Yade DEM Documentation

Yade Documentation
Vclav milauer, Emanuele Catalano, Bruno Chareyre, Sergei Dorofeenko, Jerome Duriez, Anton Gladky, Janek Kozicki, Chiara Modenese, Luc Scholts, Luc Sibille, Jan Strnsk, Klaus Thoeni
Release 2013-08-23.git-3aeba29, August 23, 2013
Authors
Vclav milauer University of Innsbruck Emanuele Catalano Grenoble INP, UJF, CNRS, lab. 3SR Bruno Chareyre Grenoble INP, UJF, CNRS, lab. 3SR Sergei Dorofeenko IPCP RAS, Chernogolovka Jerome Duriez Grenoble INP, UJF, CNRS, lab. 3SR Anton Gladky TU Bergakademie Freiberg Janek Kozicki Gdansk University of Technology - lab. 3SR Grenoble University Chiara Modenese University of Oxford Luc Scholts Grenoble INP, UJF, CNRS, lab. 3SR Luc Sibille University of Nantes, lab. GeM Jan Strnsk CVUT Prague Klaus Thoeni The University of Newcastle (Australia)
Citing this document
In order to let users cite Yade consistently in publications, we provide a list of bibliographic references for the dierent parts of the documentation. This way of acknowledging Yade is also a way to make developments and documentation of Yade more attractive for researchers, who are evaluated on the basis of citations of their work by others. We therefore kindly ask users to cite Yade as accurately as possible in their papers, as explained in http://yade-dem/doc/citing.html.
ii
Contents
1 Introduction 1.1 Getting started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Architecture overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Tutorial 2.1 Introduction . . . . . 2.2 Hands-on . . . . . . . 2.3 Data mining . . . . . 2.4 Towards geomechanics 2.5 Advanced & more . . 2.6 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 7 15 15 15 24 27 30 30 39 39 57 69 75 76 76 79 79 80 81 85 106 111 112 114 115
3 Users manual 3.1 Scene construction . . . . . . 3.2 Controlling simulation . . . . 3.3 Postprocessing . . . . . . . . 3.4 Python specialties and tricks 3.5 Extending Yade . . . . . . . 3.6 Troubleshooting . . . . . . .
4 Programmers manual 4.1 Build system . . . . . . . . . . 4.2 Development tools . . . . . . . 4.3 Conventions . . . . . . . . . . 4.4 Support framework . . . . . . 4.5 Simulation framework . . . . . 4.6 Runtime structure . . . . . . . 4.7 Python framework . . . . . . . 4.8 Maintaining compatibility . . . 4.9 Debian packaging instructions
5 Installation 117 5.1 Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.2 Source code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 6 DEM Background 6.1 Collision detection . . . . . . . . . . . 6.2 Creating interaction between particles 6.3 Strain evaluation . . . . . . . . . . . . 6.4 Stress evaluation (example) . . . . . . 6.5 Motion integration . . . . . . . . . . . 6.6 Periodic boundary conditions . . . . . 6.7 Computational aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 121 125 126 132 132 139 143
7 Class reference (yade.wrapper module) 147 7.1 Bodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 7.2 Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
iii
7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 7.12 7.13
Global engines . . . . . . . . . Partial engines . . . . . . . . . Bounding volume creation . . Interaction Geometry creation Interaction Physics creation . . Constitutive laws . . . . . . . Callbacks . . . . . . . . . . . . Preprocessors . . . . . . . . . . Rendering . . . . . . . . . . . Simulation data . . . . . . . . Other classes . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
212 286 310 315 330 340 355 356 367 380 389 399 399 400 402 404 407 408 418 421 424 425 426 438
8 Yade modules 8.1 yade.bodiesHandling module 8.2 yade.eudoxos module . . . . 8.3 yade.export module . . . . . 8.4 yade.geom module . . . . . . 8.5 yade.linterpolation module . 8.6 yade.pack module . . . . . . 8.7 yade.plot module . . . . . . . 8.8 yade.post2d module . . . . . 8.9 yade.qt module . . . . . . . . 8.10 yade.timing module . . . . . 8.11 yade.utils module . . . . . . 8.12 yade.ymport module . . . . .
9 Acknowledging Yade 441 9.1 Citing the Yade Project as a whole (the lazy citation method) . . . . . . . . . . . . . . . 441 9.2 Citing chapters of Yade Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 10 Publications on Yade 10.1 Citing Yade . . . . . . . 10.2 Journal articles . . . . . 10.3 Conference materials . 10.4 Master and PhD theses 11 References 12 Indices and tables Bibliography Python Module Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 443 443 443 443 445 447 449 461
iv
Yade Documentation, Release 2013-08-23.git-3aeba29
Note: Please consult changes to Yade documention with documentation manager (whoever that is), even if you have commit permissions. See older tentative contents
Contents
Contents
Chapter 1
Introduction
1.1 Getting started
Before you start moving around in Yade, you should have some prior knowledge. Basics of command line in your Linux system are necessary for running yade. Look on the web for tutorials. Python language; we recommend the ocial Python tutorial. Reading further documents on the topis, such as Dive into Python will certainly not hurt either. You are advised to try all commands described yourself. Dont be afraid to experiment.
1.1.1 Starting yade

Yade is being run primarily from terminal; the name of command is yade. from package, you might need to give specic path to the command 2 ):
1
(In case you did not install
$ yade Welcome to Yade TCP python prompt on localhost:9001, auth cookie `sdksuy' TCP info provider on localhost:21000 [[ ^L clears screen, Û kills line. F12 controller, F11 3d view, F10 both, F9 generator, F8 plot. ]] Yade [1]:
These initial lines give you some information about some information for Remote control , which you are unlikely to need now; basic help for the command-line that just appeared (Yade [1]:). Type quit(), exit() or simply press ^D to quit Yade.
1 The executable name can carry a sux, such as version number (yade-0.20), depending on compilation options. Packaged versions on Debian systems always provide the plain yade alias, by default pointing to latest stable version (or latest snapshot, if no stable version is installed). You can use update-alternatives to change this. 2 In general, Unix shell (command line) has environment variable PATH dened, which determines directories searched for executable les if you give name of the le without path. Typically, $PATH contains /usr/bin/, /usr/local/bin, /bin and others; you can inspect your PATH by typing echo $PATH in the shell (directories are separated by :). If Yade executable is not in directory contained in PATH, you have to specify it by hand, i.e. by typing the path in front of the lename, such as in /home/user/bin/yade and similar. You can also navigate to the directory itself (cd ~/bin/yade, where ~ is replaced by your home directory automatically) and type ./yade then (the . is the current directory, so ./ species that the le is to be found in the current directory). To save typing, you can add the directory where Yade is installed to your PATH, typically by editing ~/.profile (in normal cases automatically executed when shell starts up) le adding line like export PATH=/home/user/bin:$PATH. You can also dene an alias by saying alias yade="/home/users/bin/yade" in that le. Details depend on what shell you use (bash, zsh, tcsh, ) and you will nd more information in introductory material on Linux/Unix.
The command-line is ipython, python shell with enhanced interactive capabilities; it features persistent history (remembers commands from your last sessions), searching and so on. See ipythons documentation for more details. Typically, you will not type Yade commands by hand, but use scripts, python programs describing and running your simulations. Let us take the most simple script that will just print Hello world!:
print "Hello world!"
Saving such script as hello.py, it can be given as argument to yade:

$ yade hello.py Welcome to Yade TCP python prompt on localhost:9001, auth cookie àskcsu' TCP info provider on localhost:21000 Running script hello.py ## the script is being run Hello world! ## output from the script [[ ^L clears screen, Û kills line. F12 controller, F11 3d view, F10 both, F9 generator, F8 plot. ]] Yade [1]:
Yade will run the script and then drop to the command-line again. after running the script, use the -x switch:
$ yade -x script.py
If you want Yade to quit immediately
There is more command-line options than just -x, run yade -h to see all of them. Options: --version -h, --help show programs version number and exit show this help message and exit
-j THREADS, --threads=THREADS Number of OpenMP threads to run; defaults to 1. Equivalent to setting OMP_NUM_THREADS environment variable. --cores=CORES Set number of OpenMP threads (as threads) and in addition set anity of threads to the cores given. --update Update deprecated class names in given script(s) using text search & replace. Changed les will be backed up with ~ sux. Exit when done without running any simulation. Increase nice level (i.e. decrease priority) by given number. Exit when the script nishes Run without graphical interface (equivalent to unsetting the DISPLAY environment variable) Run regression test suite and exit; the exists status is 0 if all tests pass, 1 if a test fails and 2 for an unspecied exception. Run a series of user-dened check tests as described in /build/buildd/yadedaily1+3021+27~lucid1/scripts/test/checks/README Starts a test to measure the productivity Do not show backtrace when yade crashes (only eective with debug).
--nice=NICE -x -n --test
--check
--performance --no-gdb
3 Plain Python interpreter exits once it nishes running the script. The reason why Yade does the contrary is that most of the time script only sets up simulation and lets it run; since computation typically runs in background thread, the script is technically nished, but the computation is running.
Chapter 1. Introduction
1.1.2 Creating simulation

To create simulation, one can either use a specialized class of type FileGenerator to create full scene, possibly receiving some parameters. Generators are written in c++ and their role is limited to welldened scenarios. For instance, to create triaxial test scene:
Yade [5]: TriaxialTest(numberOfGrains=200).load() Yade [6]: len(O.bodies) Out[6]: 206
Generators are regular yade objects that support attribute access. It is also possible to construct the scene by a python script; this gives much more exibility and speed of development and is the recommended way to create simulation. Yade provides modules for streamlined body construction, import of geometries from les and reuse of common code. Since this topic is more involved, it is explained in the Users manual.
1.1.3 Running simulation

As explained above, the loop consists in running dened sequence of engines. Step number can be queried by O.iter and advancing by one step is done by O.step(). Every step advances virtual time by current timestep, O.dt:
Yade [8]: O.iter Out[8]: 0 Yade [9]: O.time Out[9]: 0.0 Yade [10]: O.dt=1e-4 Yade [11]: O.step() Yade [12]: O.iter Out[12]: 1 Yade [13]: O.time Out[13]: 0.0001
Normal simulations, however, are run continuously. Starting/stopping the loop is done by O.run() and O.pause(); note that O.run() returns control to Python and the simulation runs in background; if you want to wait for it nish, use O.wait(). Fixed number of steps can be run with O.run(1000), O.run(1000,True) will run and wait. To stop at absolute step number, O.stopAtIter can be set and O.run() called normally.
Yade [14]: O.run() Yade [15]: O.pause() Yade [16]: O.iter Out[16]: 533 Yade [17]: O.run(100000,True) Yade [18]: O.iter Out[18]: 100533 Yade [19]: O.stopAtIter=500000 Yade [20]: O.wait()
1.1. Getting started
Yade [21]: O.iter Out[21]: 100533
1.1.4 Saving and loading

Simulation can be saved at any point to (optionally compressed) XML le. With some limitations, it is generally possible to load the XML later and resume the simulation as if it were not interrupted. Note that since XML is merely readable dump of Yades internal objects, it might not (probably will not) open with dierent Yade version.
Yade [22]: O.save('/tmp/a.xml.bz2') Yade [23]: O.reload() Yade [25]: O.load('/tmp/another.xml.bz2')
The principal use of saving the simulation to XML is to use it as temporary in-memory storage for checkpoints in simulation, e.g. for reloading the initial state and running again with dierent parameters (think tension/compression test, where each begins from the same virgin state). The functions O.saveTmp() and O.loadTmp() can be optionally given a slot name, under which they will be found in memory:
Yade [26]: O.saveTmp() Yade [27]: O.loadTmp() Yade [28]: O.saveTmp('init') ## named memory slot Yade [29]: O.loadTmp('init')
Simulation can be reset to empty state by O.reset(). It can be sometimes useful to run dierent simulation, while the original one is temporarily suspended, e.g. when dynamically creating packing. O.switchWorld() toggles between the primary and secondary simulation.
1.1.5 Graphical interface

Yade can be optionally compiled with qt4-based graphical interface. It can be started by pressing F12 in the command-line, and also is started automatically when running a script.
The windows with buttons is called Controller (can be invoked by yade.qt.Controller() from python): 1. The Simulation tab is mostly self-explanatory, and permits basic simulation control. 2. The Display tab has various rendering-related options, which apply to all opened views (they can be zero or more, new one is opened by the New 3D button). 3. The Python tab has only a simple text entry area; it can be useful to enter python commands while the command-line is blocked by running script, for instance. 3d views can be controlled using mouse and keyboard shortcuts; help is displayed if you press the h key while in the 3d view. Note that having the 3d view open can slow down running simulation signicantly, it is meant only for quickly checking whether the simulation runs smoothly. Advanced post-processing is described in dedicated section.
1.2 Architecture overview

In the following, a high-level overview of Yade architecture will be given. As many of the features are directly represented in simulation scripts, which are written in Python, being familiar with this language will help you follow the examples. For the rest, this knowledge is not strictly necessary and you can ignore code examples.
1.2.1 Data and functions

To assure exibility of software design, yade makes clear distinction of 2 families of classes: data components and functional components. The former only store data without providing functionality, while the latter dene functions operating on the data. In programming, this is known as visitor pattern (as functional components visit the data, without being bound to them explicitly).
1.2. Architecture overview
Entire simulation, i.e. both data and functions, are stored in a single Scene object. It is accessible through the Omega class in python (a singleton), which is by default stored in the O global variable:
Yade [33]: O.bodies # some data components Out[33]: <yade.wrapper.BodyContainer at 0x15a5c5f0> Yade [34]: len(O.bodies) Out[34]: 0 Yade [35]: O.engines Out[35]: [] # there are no bodies as of yet
# functional components, empty at the moment
Data components
Bodies
Yade simulation (class Scene, but hidden inside Omega in Python) is represented by Bodies, their Interactions and resultant generalized forces (all stored internally in special containers). Each Body comprises the following: Shape represents particles geometry (neutral with regards to its spatial orientation), such as Sphere, Facet or ininite Wall; it usually does not change during simulation. Material stores characteristics pertaining to mechanical behavior, such as Youngs modulus or density, which are independent on particles shape and dimensions; usually constant, might be shared amongst multiple bodies. State contains state variable variables, in particular spatial position and orientation, linear and angular velocity, linear and angular accelerator; it is updated by the integrator at every step. Derived classes can hold additional data, e.g. averaged damage. Bound is used for approximate (pass 1) contact detection; updated as necessary following bodys motion. Currently, Aabb is used most often as Bound. Some bodies may have no Bound, in which case they are exempt from contact detection. (In addition to these 4 components, bodies have several more minor data associated, such as Body::id or Body::mask.) All these four properties can be of dierent types, derived from their respective base types. Yade frequently makes decisions about computation based on those types: Sphere + Sphere collision has to be treated dierently than Facet + Sphere collision. Objects making those decisions are called Dispatchers and are essential to understand Yades functioning; they are discussed below. Explicitly assigning all 4 properties to each particle by hand would be not practical; there are utility functions dened to create them with all necessary ingredients. For example, we can create sphere particle using utils.sphere:
Yade [36]: s=utils.sphere(center=[0,0,0],radius=1) Yade [37]: s.shape, s.state, s.mat, s.bound Out[37]: (<Sphere instance at 0xfb2ea20>, <State instance at 0x145d4420>, <FrictMat instance at 0x139d3150>, None) Yade [38]: s.state.pos Out[38]: Vector3(0,0,0) Yade [39]: s.shape.radius Out[39]: 1.0
We see that a sphere with material of type FrictMat (default, unless you provide another Material) and bounding volume of type Aabb (axis-aligned bounding box) was created. Its position is at origin and its radius is 1.0. Finally, this object can be inserted into the simulation; and we can insert yet one sphere as well.
Yade [40]: O.bodies.append(s) Out[40]: 0 Yade [41]: O.bodies.append(utils.sphere([0,0,2],.5)) Out[41]: 1
In each case, return value is Body.id of the body inserted. Since till now the simulation was empty, its id is 0 for the rst sphere and 1 for the second one. Saving the id value is not necessary, unless you want access this particular body later; it is remembered internally in Body itself. You can address bodies by their id:
Yade [42]: O.bodies[1].state.pos Out[42]: Vector3(0,0,2) Yade [43]: O.bodies[100] --------------------------------------------------------------------------IndexError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 O.bodies[100] IndexError: Body id out of range.
Adding the same body twice is, for reasons of the id uniqueness, not allowed:
Yade [44]: O.bodies.append(s) --------------------------------------------------------------------------IndexError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 O.bodies.append(s) IndexError: Body already has id 0 set; appending such body (for the second time) is not allowed.
Bodies can be iterated over using standard python iteration syntax:

Yade [45]: for b in O.bodies: ....: print b.id,b.shape.radius ....: 0 1.0 1 0.5
Interactions
Interactions are always between pair of bodies; usually, they are created by the collider based on spatial proximity; they can, however, be created explicitly and exist independently of distance. Each interaction has 2 components: IGeom holding geometrical conguration of the two particles in collision; it is updated automatically as the particles in question move and can be queried for various geometrical characteristics, such as penetration distance or shear strain. Based on combination of types of Shapes of the particles, there might be dierent storage requirements; for that reason, a number of derived classes exists, e.g. for representing geometry of contact between Sphere+Sphere, Facet+Sphere etc. IPhys representing non-geometrical features of the interaction; some are computed from Materials of the particles in contact using some averaging algorithm (such as contact stiness from Youngs moduli of particles), others might be internal variables like damage.
Suppose now interactions have been already created. We can access them by the id pair:
Yade [48]: O.interactions[0,1] Out[48]: <Interaction instance at 0xf64c9f0> Yade [49]: O.interactions[1,0] # order of ids is not important Out[49]: <Interaction instance at 0xf64c9f0> Yade [50]: i=O.interactions[0,1] Yade [51]: i.id1,i.id2 Out[51]: (0, 1) Yade [52]: i.geom Out[52]: <ScGeom instance at 0x694aa30> Yade [53]: i.phys Out[53]: <FrictPhys instance at 0x1759c2d0> Yade [54]: O.interactions[100,10111] --------------------------------------------------------------------------IndexError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 O.interactions[100,10111] IndexError: No such interaction
Generalized forces
Generalized forces include force, torque and forced displacement and rotation; they are stored only temporarliy, during one computation step, and reset to zero afterwards. For reasons of parallel computation, they work as accumulators, i.e. only can be added to, read and reset.
Yade [55]: O.forces.f(0) Out[55]: Vector3(0,0,0) Yade [56]: O.forces.addF(0,Vector3(1,2,3)) Yade [57]: O.forces.f(0) Out[57]: Vector3(1,2,3)
You will only rarely modify forces from Python; it is usually done in c++ code and relevant documentation can be found in the Programmers manual. Function components In a typical DEM simulation, the following sequence is run repeatedly: reset forces on bodies from previous step approximate collision detection (pass 1) detect exact collisions of bodies, update interactions as necessary solve interactions, applying forces on bodies apply other external conditions (gravity, for instance). change position of bodies based on forces, by integrating motion equations. Each of these actions is represented by an Engine, functional element of simulation. The sequence of engines is called simulation loop.
10
bodies
increment time by t Shape Material State Bound
reset forces
update bounds
collision detection pass 1
miscillaneous engines (recorders, ...) position update velocity update forces acceleration
simulation loop
interactions
geometry collision detection pass 2 strain evaluation physics properties of new interactions
forces
other forces (gravity, BC, ...)
constitutive law compute forces from strains
(generalized)
Figure 1.1: Typical simulation loop; each step begins at body-cented bit at 11 oclock, continues with interaction bit, force application bit, miscillanea and ends with time update.
Engines
Simulation loop, shown at img. img-yade-iter-loop, can be described as follows in Python (details will be explained later); each of the O.engine items is instance of a type deriving from Engine:
O.engines=[ # reset forces ForceResetter(), # approximate collision detection, create interactions InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()]), # handle interactions InteractionLoop( [Ig2_Sphere_Sphere_ScGeom(),Ig2_Facet_Sphere_ScGeom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_ScGeom_FrictPhys_CundallStrack()], ), # apply other conditions GravityEngine(gravity=(0,0,-9.81)), # update positions using Newton's equations NewtonIntegrator() ]
There are 3 fundamental types of Engines: GlobalEngines operating on the whole simulation (e.g. GravityEngine looping over all bodies and applying force based on their mass) PartialEngine operating only on some pre-selected bodies (e.g. ForceEngine applying constant force to some bodies) Dispatchers do not perform any computation themselves; they merely call other functions, represented by function objects, Functors. Each functor is specialized, able to handle certain object types, and will be dispatched if such obejct is treated by the dispatcher.
11
Dispatchers and functors
For approximate collision detection (pass 1), we want to compute bounds for all bodies in the simulation; suppose we want bound of type axis-aligned bounding box. Since the exact algorithm is dierent depending on particular shape, we need to provide functors for handling all specic cases. The line:
InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()])
creates InsertionSortCollider (it internally uses BoundDispatcher, but that is a detail). It traverses all bodies and will, based on shape type of each body, dispatch one of the functors to create/update bound for that particular body. In the case shown, it has 2 functors, one handling spheres, another facets. The name is composed from several parts: Bo (functor creating Bound), which accepts 1 type Sphere and creates an Aabb (axis-aligned bounding box; it is derived from Bound). The Aabb objects are used by InsertionSortCollider itself. All Bo1 functors derive from BoundFunctor. The next part, reading
InteractionLoop( [Ig2_Sphere_Sphere_ScGeom(),Ig2_Facet_Sphere_ScGeom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_ScGeom_FrictPhys_CundallStrack()], ),
hides 3 internal dispatchers within the InteractionLoop engine; they all operate on interactions and are, for performance reasons, put together: IGeomDispatcher uses the rst set of functors (Ig2), which are dispatched based on combination of 2 Shapes objects. Dispatched functor resolves exact collision conguration and creates IGeom (whence Ig in the name) associated with the interaction, if there is collision. The functor might as well fail on approximate interactions, indicating there is no real contact between the bodies, even if they did overlap in the approximate collision detection. 1. The rst functor, Ig2_Sphere_Sphere_ScGeom, is called on interaction of 2 Spheres and creates ScGeom instance, if appropriate. 2. The second functor, Ig2_Facet_Sphere_ScGeom, is called for interaction of Facet with Sphere and might create (again) a ScGeom instance. All Ig2 functors derive from IGeomFunctor (they are documented at the same place). IPhysDispatcher dispatches to the second set of functors based on combination of 2 Materials; these functors return return IPhys instance (the Ip prex). In our case, there is only 1 functor used, Ip2_FrictMat_FrictMat_FrictPhys, which create FrictPhys from 2 FrictMats. Ip2 functors are derived from IPhysFunctor. LawDispatcher dispatches to the third set of functors, based on combinations of IGeom and IPhys (wherefore 2 in their name again) of each particular interaction, created by preceding functors. The Law2 functors represent constitutive law; they resolve the interaction by computing forces on the interacting bodies (repulsion, attraction, shear forces, ) or otherwise update interaction state variables. Law2 functors all inherit from LawFunctor. There is chain of types produced by earlier functors and accepted by later ones; the user is responsible to satisfy type requirement (see img. img-dispatch-loop). An exception (with explanation) is raised in the contrary case. Note: When Yade starts, O.engines is lled with a reasonable default list, so that it is not strictly necessary to redene it when trying simple things. The default scene will handle spheres, boxes, and facets with frictional properties correctly, and adjusts the timestep dynamically. You can nd an example in simple-scene-default-engines.py.
12
Body
Interaction
Ig2_Sphere_S phere_Dem 3DofGeom
eom em3DofG phere_D
Sphere+Sphere Facet+Sphere
Ig2_Face
t_S
Interaction Geometry Interaction Physics
Shape
Dem3DofGeom
Law2_Dem3DofGeom_FrictPhys_Basic
Material
FrictMat+FrictMat
Ip2_Frict
Mat_FrictM
at_FrictP
hys
FrictPhys
Figure 1.2: Chain of functors producing and accepting certain types. In the case shown, the Ig2 functors produce ScfGeom instances from all handled Shape combinations; the Ig2 functor produces FrictMat. The constitutive law functor Law2 accepts the combination of types produced. Note that the types are stated in the functors class names.
13
14
Chapter 2
Tutorial
This tutorial originated as handout for a course held at Technische Universitt Dresden / Fakultt Bauingenieurwesen / Institut fr Geotechnik in Jaunary 2011. The focus was to give quick and rather practical introduction to people without prior modeling experience, but with knowledge of mechanics. Some computer literacy was assumed, though basics are reviewed in the Hands-on section. The course did not in reality follow this document, but was based on interactive writing and commenting simple Examples, which were mostly suggested by participants; many thanks to them for their ideas and suggestions. A few minor bugs were discovered during the course. They were all xed in rev. 2640 of Yade which is therefore the minimum recommended version to run the examples (notably, 0.60 will not work).
2.1 Introduction
Slides Yade: past, present and future (updated version)
2.2 Hands-on
2.2.1 Shell basics
Directory tree Directory tree is hierarchical way to organize les in operating systems. A typical (reduced) tree looks like this:
/ Root --boot System startup --bin Low-level programs --lib Low-level libraries --dev Hardware access --sbin Administration programs --proc System information --var Files modified by system services --root Root (administrator) home directory --etc Configuration files --media External drives --tmp Temporary files --usr Everything for normal operation (usr = UNIX system resources) | --bin User programs | --sbin Administration programs | --include Header files for c/c++
15
| --lib | --local | --doc --home --user --user1
Libraries Locally installed software Documentation Contains the user's home directories Home directory for user Home directory for user1
Note that there is a single root /; all other disks (such as USB sticks) attach to some point in the tree (e.g. in /media). Shell navigation Shell is the UNIX command-line, interface for conversation with the machine. Dont be afraid.
Moving around
The shell is always operated by some user, at some concrete machine; these two are constant. We can move in the directory structure, and the current place where we are is current directory. By default, it is the home directory which contains all les belonging to the respective user:
user@machine:~$ user@machine:~$ ls . user@machine:~$ ls foo user@machine:~$ ls /tmp user@machine:~$ cd foo user@machine:~/foo$ ls ~ user@machine:~/foo$ cd bar user@machine:~/foo/bar$ cd ../../foo2 user@machine:~/foo2$ cd user@machine:~$ # # # # # # # # #
user operating at machine, in the directory ~ (= user's home directory list contents of the current directory list contents of directory foo, relative to the dcurrent directory ~ ( list contents of /tmp change directory to foo list home directory (= ls /home/user) change to bar (= cd ~/foo/bar) go to the parent directory twice, then to foo2 (cd ~/foo/bar/../../foo go to the home directory (= ls ~ = ls /home/user)
Users typically have only permissions to write (i.e. modify les) only in their home directory (abbreviated ~, usually is /home/user) and /tmp, and permissions to read les in most other parts of the system:
user@machine:~$ ls /root # see what files the administrator has ls: cannot open directory /root: Permission denied
Keys
Useful keys on the command-line are: <tab> ^C (=Ctrl+C) ^D ^C ^\ Shift-PgUp Shift-PgDown

Running programs
show possible completions of what is being typed (use abundantly!) delete current line exit the shell move up and down in the command history interrupt currently running program kill currently running program scroll the screen up (show part output) scroll the screen down (show future output; works only on quantum computers)
When a program is being run (without giving its full path), several directories are searched for program of that name; those directories are given by $PATH:
user@machine:~$ echo $PATH # show the value of $PATH /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games user@machine:~$ which ls # say what is the real path of ls
16
Chapter 2. Tutorial
The rst part of the command-line is the program to be run (which), the remaining parts are arguments (ls in this case). It is upt to the program which arguments it understands. Many programs can take special arguments called options starting with - (followed by a single letter) or -- (followed by words); one of the common options is -h or --help, which displays how to use the program (try ls --help). Full documentation for each program usually exists as manual page (or man page), which can be shown using e.g. man ls (q to exit) Starting yade If yade is installed on the machine, it can be (roughly speaking) run as any other program; without any arguments, it runs in the dialog mode, where a command-line is presented:
user@machine:~$ yade Welcome to Yade bzr2616 TCP python prompt on localhost:9002, auth cookie àdcusk' XMLRPC info provider on http://localhost:21002 [[ ^L clears screen, Û kills line. F12 controller, F11 3d view, F10 both, F9 generator, F8 plot. ]] Yade [1]: #### hit ^D to exit Do you really want to exit ([y]/n)? Yade: normal exit.
The command-line is in fact python, enriched with some yade-specic features. (Pure python interpreter can be run with python or ipython commands). Instead of typing commands on-by-one on the command line, they can be be written in a le (with the .py extension) and given as argument to Yade:
user@machine:~$ yade simulation.py
For a complete help, see man yade

Exercises
1. Open the terminal, navigate to your home directory 2. Create a new empty le and save it in ~/first.py 3. Change directory to /tmp; delete the le ~/first.py 4. Run program xeyes 5. Look at the help of Yade. 6. Look at the manual page of Yade 7. Run Yade, exit and run it again.
2.2.2 Python basics

We assume the reader is familar with Python tutorial and only briey review some of the basic capabilities. The following will run in pure-python interpreter (python or ipython), but also inside Yade, which is a super-set of Python. Numerical operations and modules:
Yade [1]: (1+3*4)**2 Out[1]: 169 Yade [2]: import math # usual rules for operator precedence, ** is exponentiation
# gain access to "module" of functions
Yade [3]: math.sqrt(2) # use a function from that module Out[3]: 1.4142135623730951
2.2. Hands-on
17
Yade [4]: import math as m Yade [5]: m.cos(m.pi) Out[5]: -1.0 Yade [6]: from math import * Yade [7]: cos(pi) Out[7]: -1.0
# use the module under a different name
# import everything so that it can be used without module name
Variables:
Yade [1]: a=1; b,c=2,3 Yade [2]: a+b+c Out[2]: 6 # multiple commands separated with ;, multiple assignment
Sequences
Lists
Lists are variable-length sequences, which can be modied; they are written with braces [...], and their elements are accessed with numerical indices:
Yade [1]: a=[1,2,3] Yade [2]: a[0] Out[2]: 1 Yade [3]: a[-1] Out[3]: 3 # list of numbers # first element has index 0
# negative counts from the end
Yade [4]: a[3] # error --------------------------------------------------------------------------IndexError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 a[3] # error IndexError: list index out of range Yade [5]: len(a) Out[5]: 3 Yade [6]: a[1:] Out[6]: [2, 3] Yade [7]: a+=[4,5] # number of elements
# from second element to the end
# extend the list
Yade [8]: a+=[6]; a.append(7) # extend with single value, both have the same effect Yade [9]: 9 in a Out[9]: False # test presence of an element
Lists can be created in various ways:

Yade [1]: range(10) Out[1]: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9] Yade [2]: range(10)[-1] Out[2]: 9
List of squares of even number smaller than 20, i.e.
{ 2 } a a {0, , 19} 2a (note the similarity):
18
Chapter 2. Tutorial
Yade [1]: [a**2 for a in range(20) if a%2==0] Out[1]: [0, 4, 16, 36, 64, 100, 144, 196, 256, 324]
Tuples
Tuples are constant sequences:

Yade [1]: b=(1,2,3) Yade [2]: b[0] Out[2]: 1 Yade [3]: b[0]=4 # error --------------------------------------------------------------------------TypeError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 b[0]=4 # error TypeError: 'tuple' object does not support item assignment
Dictionaries
Mapping from keys to values:

Yade [1]: czde={'jedna':'ein','dva':'zwei','tri':'drei'} Yade [2]: de={1:'ein',2:'zwei',3:'drei'}; cz={1:'jedna',2:'dva',3:'tri'} Yade [3]: czde['jedna'] Out[3]: 'ein' Yade [4]: de[1], cz[2] Out[4]: ('ein', 'dva') ## access values
Functions, conditionals
Yade [1]: 4==5 Out[1]: False Yade [2]: a=3.1 Yade [3]: if a<pi: b=0 # conditional statement ...: else: b=1 ...: IndentationError: unexpected indent
Yade [5]: c=0 if a<1 else 1
# conditional expression # define a new function
Yade [6]: def square(x): return x**2 ...: Yade [7]: square(2) Out[7]: 4
# and call that function
Exercises
1. Read the following code and say what wil be the values of a and b: 2.2. Hands-on 19
a=range(5) b=[(aa**2 if aa%2==0 else -aa**2) for aa in a]
2.2.3 Yade basics

Yade objects are constructed in the following manner (this process is also called instantiation, since we create concrete instances of abstract classes: one individual sphere is an instance of the abstract Sphere, like Socrates is an instance of man):
Yade [2]: Sphere # try also Sphere? Out[2]: yade.wrapper.Sphere Yade [3]: s=Sphere() Yade [4]: s.radius Out[4]: nan Yade [5]: s.radius=2 Yade [6]: s.radius Out[6]: 2.0 Yade [7]: ss=Sphere(radius=3) Yade [8]: s.radius, ss.radius Out[8]: (2.0, 3.0) # create Sphere, giving radius directly # also try typing s.<tab> to see defined attributes # create a Sphere, without specifying any attributes # 'nan' is a special value meaning "not a number" (i.e. not defined)
# set radius of an existing object
Particles Particles are the data component of simulation; they are the objects that will undergo some processes, though do not dene those processes yet.
Singles
There is a number of pre-dened functions to create particles of certain type; in order to create a sphere, one has to (see the source of utils.sphere for instance): 1. Create Body 2. Set Body.shape to be an instance of Sphere with some given radius 3. Set Body.material (last-dened material is used, otherwise a default material is created) 4. Set position and orientation in Body.state, compute mass and moment of inertia based on Material and Shape In order to avoid such tasks, shorthand functions are dened in the utils module; to mention a few of them, they are utils.sphere, utils.facet, utils.wall.
Yade [2]: s=utils.sphere((0,0,0),radius=1) # create sphere particle centered at (0,0,0) with radius=1
Yade [3]: s.shape # s.shape describes the geometry of the particle Out[3]: <Sphere instance at 0x5d7e420> Yade [4]: s.shape.radius Out[4]: 1.0 # we already know the Sphere class
Yade [5]: s.state.mass, s.state.inertia # inertia is computed from density and geometry Out[5]: (4188.790204786391, Vector3(1675.5160819145563,1675.5160819145563,1675.5160819145563))
20
Chapter 2. Tutorial
Yade [6]: s.state.pos Out[6]: Vector3(0,0,0)
# position is the one we prescribed
Yade [7]: s2=utils.sphere((-2,0,0),radius=1,fixed=True)
# explanation below
In the last example, the particle was xed in space by the fixed=True parameter to utils.sphere; such a particle will not move, creating a primitive boundary condition. A particle object is not yet part of the simulation; in order to do so, a special function is called:
Yade [1]: O.bodies.append(s) Out[1]: 12
# adds particle s to the simulation; returns id of the particle(s) added
Packs
There are functions to generate a specic arrangement of particles in the pack module; for instance, cloud (random loose packing) of spheres can be generated with the pack.SpherePack class:
Yade [1]: from yade import pack Yade [2]: sp=pack.SpherePack()
# create an empty cloud; SpherePack contains only geometrical i
Yade [3]: sp.makeCloud((1,1,1),(2,2,2),rMean=.2) # put spheres with defined radius inside box given by corners ( Out[3]: 7
Yade [4]: for c,r in sp: print c,r # print center and radius of all particles (SpherePack is a seq ...: Vector3(1.4959850951991838,1.4965371932802136,1.7468636503879573) 0.2 Vector3(1.4983272328025916,1.7538613948541335,1.2433920407140553) 0.2 Vector3(1.2054697190462329,1.228821444258859,1.6399486582166969) 0.2 Vector3(1.6429499076148029,1.2349940040474703,1.3955820754129271) 0.2 Vector3(1.2174189850850208,1.4318421340844054,1.2516626635116177) 0.2 Vector3(1.221763762479428,1.75859164200592,1.5771521984386743) 0.2 Vector3(1.7931107450212451,1.648773075313096,1.5251218375983853) 0.2 Yade [5]: sp.toSimulation() Out[5]: [13, 14, 15, 16, 17, 18, 19] # create particles and add them to the simulation
Boundaries
utils.facet (triangle Facet) and utils.wall (innite axes-aligned plane Wall) geometries are typically used to dene boundaries. For instance, a oor for the simulation can be created like this:
Yade [1]: O.bodies.append(utils.wall(-1,axis=2)) Out[1]: 20
There are other conveinence functions (like utils.facetBox for creating closed or open rectangular box, or family of ymport functions) Look inside The simulation can be inspected in several ways. All data can be accessed from python directly:
Yade [1]: len(O.bodies) Out[1]: 21
Yade [2]: O.bodies[1].shape.radius # radius of body #1 (will give error if not sphere, since only spheres have --------------------------------------------------------------------------AttributeError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>()
2.2. Hands-on
21
----> 1 O.bodies[1].shape.radius
# radius of body #1 (will give error if not sphere, since only spheres have r
AttributeError: 'NoneType' object has no attribute 'radius' Yade [3]: O.bodies[2].state.pos Out[3]: Vector3(0,0,0) # position of body #2
Besides that, Yade says this at startup (the line preceding the command-line):
[[ ^L clears screen, Û kills line. F12 controller, F11 3d view, F10 both, F9 generator, F8 plot. ]]
Controller Pressing F12 brings up a window for controlling the simulation. Although typically no human intervention is done in large simulations (which run headless, without any graphical interaction), it can be handy in small examples. There are basic information on the simulation (will be used later). 3d view The 3d view can be opened with F11 (or by clicking on button in the Controller see below). There is a number of keyboard shortcuts to manipulate it (press h to get basic help), and it can be moved, rotated and zoomed using mouse. Display-related settings can be set in the Display tab of the controller (such as whether particles are drawn). Inspector Inspector is opened by clicking on the appropriate button in the Controller. It shows (and updated) internal data of the current simulation. In particular, one can have a look at engines, particles (Bodies) and interactions (Interactions). Clicking at each of the attribute names links to the appropriate section in the documentation.
Exercises
1. What is this code going to do?

Yade [1]: O.bodies.append([utils.sphere((2*i,0,0),1) for i in range(1,20)]) Out[1]: [21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]
2. Create a simple simulation with cloud of spheres enclosed in the box (0,0,0) and (1,1,1) with mean radius .1. (hint: pack.SpherePack.makeCloud) 3. Enclose the cloud created above in box with corners (0,0,0) and (1,1,1); keep the top of the box open. (hint: utils.facetBox; type utils.facetBox? or utils.facetBox?? to get help on the command line) 4. Open the 3D view, try zooming in/out; position axes so that z is upwards, y goes to the right and x towards you. Engines Engines dene processes undertaken by particles. As we know from the theoretical introduction, the sequence of engines is called simulation loop. Let us dene a simple interaction loop:
Yade [2]: O.engines=[ # newlines and indentations are not important until the brace is closed ...: ForceResetter(), ...: InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Wall_Aabb()]), ...: InteractionLoop( # dtto for the parenthesis here ...: [Ig2_Sphere_Sphere_L3Geom(),Ig2_Wall_Sphere_L3Geom()], ...: [Ip2_FrictMat_FrictMat_FrictPhys()], ...: [Law2_L3Geom_FrictPhys_ElPerfPl()] ...: ), ...: NewtonIntegrator(damping=.2,label='newton') # define a name under which we can access this engi ...: ] ...: Yade [3]: O.engines Out[3]:
22
Chapter 2. Tutorial
[<ForceResetter instance at 0x6575f40>, <InsertionSortCollider instance at 0x5892090>, <InteractionLoop instance at 0x6fa5ac0>, <NewtonIntegrator instance at 0xf5c7a10>] Yade [4]: O.engines[-1]==newton Out[4]: True Yade [5]: newton.damping Out[5]: 0.2 # is it the same object?
Instead of typing everything into the command-line, one can describe simulation in a le (script ) and then run yade with that le as an argument. We will therefore no longer show the command-line unless necessary; instead, only the script part will be shown. Like this:
O.engines=[ # newlines and indentations are not important until the brace is closed ForceResetter(), InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Wall_Aabb()]), InteractionLoop( # dtto for the parenthesis here [Ig2_Sphere_Sphere_L3Geom_Inc(),Ig2_Wall_Sphere_L3Geom_Inc()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_L3Geom_FrictPhys_ElPerfPl()] ), GravityEngine(gravity=(0,0,-9.81)), # 9.81 is the gravity acceleration, and we say that NewtonIntegrator(damping=.2,label='newton') # define a name under which we can access this engine e ]
Besides engines being run, it is likewise important to dene how often they will run. Some engines can run only sometimes (we will see this later), while most of them will run always; the time between two successive runs of engines is timestep (t). There is a mathematical limit on the timestep value, called critical timestep, which is computed from properties of particles. Since there is a function for that, we can just set timestep using utils.PWaveTimeStep:
O.dt=utils.PWaveTimeStep()
Each time when the simulation loop nishes, time O.time is advanced by the timestep O.dt:
Yade [1]: O.dt=0.01 Yade [2]: O.time Out[2]: 0.0 Yade [3]: O.step() Yade [4]: O.time Out[4]: 0.01
For experimenting with a single simulations, it is handy to save it to memory; this can be achieved, once everything is dened, with:
O.saveTmp()
Exercises
1. Dene engines as in the above example, run the Inspector and click through the engines to see their sequence. 2. Write a simple script which will (a) dene particles as in the previous exercise (cloud of spheres inside a box open from the top) (b) dene a simple simulation loop, as the one given above (c) set t equal to the critical P-Wave t 2.2. Hands-on 23
(d) save the initial simulation state to memory 3. Run the previously-dened simulation multiple times, while changing the value of timestep (use the button to reload the initial conguration). (a) See what happens as you increase t above the P-Wave value. (b) Try changing the gravity parameter, before running the simulation. (c) Try changing damping 4. Reload the simulation, open the 3d view, open the Inspector, select a particle in the 3d view (shiftclick). Then run the simulation and watch how forces on that particle change; pause the simulation somewhere in the middle, look at interactions of this particle. 5. At which point can we say that the deposition is done, so that the simulation can be stopped? See Also: The Bouncing sphere example shows a basic simulation.
2.3 Data mining

2.3.1 Read
Local data All data of the simulation are accessible from python; when you open the Inspector, blue labels of various data can be clicked left button for getting to the documentation, middle click to copy the name of the object (use Ctrl-V or middle-click to paste elsewhere). The interesting objects are among others (see Omega for a full list): 1. O.engines Engines are accessed by their index (position) in the simulation loop:
O.engines[0] O.engines[-1] # first engine # last engine
Note: The index can change if O.engines is modied. Labeling introduced below is a better solution for reliable access to a particular engine. 2. O.bodies Bodies are identied by their id, which is guaranteed to not change during the whole simulation:
O.bodies[0] [b.shape.radius in O.bodies if isinstance(b.shape,Sphere)] sum([b.state.mass for b in O.bodies]) # first body # list of radii of all spherical bodies # sum of masses of all bodies
Note: Uniqueness of Body.id is not guaranteed, since newly created bodies might recycle ids of deleted ones. 3. O.force Generalized forces (forces, torques) acting on each particle. They are (usually) reset at the beginning of each step with ForceResetter, subsequently forces from individual interactions are accumulated in InteractionLoop. To access the data, use:
O.forces.f(0) O.forces.t(1) # force on #0 # torque on #1
24
Chapter 2. Tutorial
4. O.interactions Interactions are identied by ids of the respective interacting particles (they are created and deleted automatically during the simulation):
O.interactions[0,1] O.interactions[1,0] O.bodies[0].intrs # interactions of #0 with #1 # the same object # all interactions of body #0
Labels
Engines and functors can be labeled, which means that python variable of that name is automatically created.
Yade [2]: O.engines=[ ...: NewtonIntegrator(damping=.2,label='newton') ...: ] ...: Yade [3]: newton.damping=.4 Yade [4]: O.engines[0].damping Out[4]: 0.4 # O.engines[0] and newton are the same objects
Exercises
1. Find meaning of this expression:

max([b.state.vel.norm() for b in O.bodies])
2. Run the gravity deposition script, pause after a few seconds of simulation. Write expressions that compute 1 2 (a) kinetic energy 2 mi |vi | (b) average mass (hint: use numpy.average) (c) maximum z-coordinate of all particles (d) number of interactions of body #1 Global data Useful measures of what happens in the simulation globally: unbalanced force ratio of maximum contact force and maximum per-body force; measure of staticity, computed with utils.unbalancedForce. porosity ratio of void volume and total volume; computed with utils.porosity. coordination number average number of interactions per particle, utils.avgNumInteractions stress tensor (periodic boundary conditions) averaged force in interactions, computed with utils.normalShearStressTensor and utils.stressTensorOfPeriodicCell fabric tensor distribution of contacts in space (not yet implemented); can be visualized with utils.plotDirections
Energies
Evaluating energy data for all components in the simulation (such as gravity work, kinetic energy, plastic dissipation, damping dissipation) can be enabled with
2.3. Data mining
25
O.trackEnergy=True
Subsequently, energy values are accessible in the O.energy; it is a dictionary where its entries can be retrived with keys() and their values with O.energy[key].
2.3.2 Save
PyRunner To save data that we just learned to access, we need to call Python from within the simulation loop. PyRunner is created just for that; it inherits periodicy control from PeriodicEngine and takes the code to run as text (must be quoted, i.e. inside '...') attributed called command. For instance, adding this to O.engines will print the current step number every second:
O.engines=O.engines+[ PyRunner(command='print O.iter',realPeriod=1) ]
Writing complicated code inside command is awkward; in such case, we dene a function that will be called:
def myFunction(): '''Print step number, and pause the simulation is unbalanced force is smaller than 0.05.''' print O.iter if utils.unbalancedForce()<0.05: print 'Unbalanced force is smaller than 0.05, pausing.' O.pause() O.engines=[ # ... PyRunner(command='myFunction()',iterPeriod=100) # call myFunction every 100 steps ]
Exercises
1. Run the gravity deposition simulation, but change it such that: (a) utils.unbalancedForce is printed every 2 seconds. (b) check every 1000 steps the value of unbalanced force if smaller than 0.2, set damping to 0.8 (hint: use labels) if smaller than 0.1, pause the simulation Keeping history Yade provides the plot module used for storing and plotting variables (plotting itself will be discussed later). Periodic storing of data is done with PyRunner and the plot.addData function, for instance:
from yade import plot O.engines=[ # ..., PyRunner(command='addPlotData()',realPeriod=2) # call the addPlotData function every 2 s ] def addPlotData(): # this function adds current values to the history of data, under the names specified plot.addData(i=O.iter,t=O.time,Ek=utils.kineticEnergy(),coordNum=utils.avgNumInteractions(),unForce=util
History is stored in plot.data, and can be accessed using the variable name, e.g. plot.data['Ek'], and saved to text le (for post-processing outside yade) with plot.saveTxt.
26
Chapter 2. Tutorial
2.3.3 Plot
plot provides facilities for plotting history saved with plot.addData as 2d plots. Data to be plotted are specied using dictionary plot.plots
plot.plots={'t':('coordNum','unForce',None,'Ek')}
History of all values is given as the name used for plot.addData; keys of the dictionary are x-axis values, and values are sequence of data on the y axis; the None separates data on the left and right axes (they are scaled independently). The plot itself is created with
plot.plot() # on the command line, F8 can be used as shorthand
While the plot is open, it will be updated periodically, so that simulation evolution can be seen in real-time. Energy plots Plotting all energy contributions would be dicult, since names of all energies might not be known in advance. Fortunately, there is a way to handle that in Yade. It consists in two parts: 1. plot.addData is given all the energies that are currently dened:
plot.addData(i=O.iter,total=O.energy.total(),**O.energy)
The O.energy.total functions, which sums all energies together. The **O.energy is special python syntax for converting dictionary (remember that O.energy is a dictionary) to named functions arguments, so that the following two commands are identical:
function(a=3,b=34) function(**{'a':3,'b':34}) # give arguments as arguments # create arguments from dictionary
2. Data to plot are specied using a function that gives names of data to plot, rather than providing the data names directly:
plot.plots={'i':['total',O.energy.keys()]}
where total is the name we gave to O.energy.total() above, while O.energy.keys() will always return list of currently dened energies.
Exercises
1. Run the gravity deposition script, plotting unbalanced force and kinetic energy. 2. While the script is running, try changing the NewtonIntegrator.damping parameter (do it from both Inspector and from the command-line). What inuence does it have on the evolution of unbalanced force and kinetic energy? 3. Think about and write down all energy sources (input); write down also all energy sinks (dissipation). 4. Simulate gravity deposition and plot all energies as they evolve during the simulation. See Also: Most Examples use plotting facilities of Yade, some of them also track energy of the simulation.
2.4 Towards geomechanics

See Also: Examples Gravity deposition, Oedometric test , Periodic simple shear , Periodic triaxial test deal with topics discussed here. 2.4. Towards geomechanics 27
2.4.1 Parametric studies

Input parameters of the simulation (such as size distribution, damping, various contact parameters, ) inuence the results, but frequently an analytical relationship is not known. To study such inuence, similar simulations diering only in a few parameters can be run and results compared. Yade can be run in batch mode, where one simulation script is used in conjunction with parameter table, which species parameter values for each run of the script. Batch simulation are run non-interactively, i.e. without user intervention; the user must therefore start and stop the simulation explicitly. Suppose we want to study the inuence of damping on the evolution of kinetic energy. The script has to be adapted at several places: 1. We have to make sure the script reads relevant parameters from the parameter table. This is done using utils.readParamsFromTable; the parameters which are read are created as variables in the yade.params.table module:
utils.readParamsFromTable(damping=.2) from yade.params import table
# yade.params.table.damping variable will be created # typing table.damping is easier than yade.params.table.damping
Note that utils.readParamsFromTable takes default values of its parameters, which are used if the script is not run in non-batch mode. 2. Parameters from the table are used at appropriate places:
NewtonIntegrator(damping=table.damping),
3. The simulation is run non-interactively; we must therefore specify at which point it should stop:
O.engines+=[PyRunner(iterPeriod=1000,command='checkUnbalancedForce()')] def checkUnbalancedForce(): if utils.unbalancedForce<0.05: utils.saveDataTxt(O.tags['d.id']+'.data.bz2') import sys sys.exit(0)
# call our function defined bel
# exit Yade if unbalanced force drops below 0.05 # save all data into a unique file before exiting # exit the program
4. Finally, we must start the simulation at the very end of the script:
O.run() utils.waitIfBatch()
# run forever, until stopped by checkUnbalancedForce() # do not finish the script until the simulation ends; does nothing in non-batch mode
The parameter table is a simple text-le, where each line species a simulation to run:
# comments start with # as in python damping # first non-comment line is variable name .2 .4 .6
Finally, the simulation is run using the special batch command:

user@machine:~$ yade-batch parameters.table simulation.py
Exercises
1. Run the gravity deposition script in batch mode, varying damping to take values of .2, .4, .6. See the http://localhost:9080 overview page while the batch is running.
2.4.2 Boundary
Particles moving in innite space usually need some constraints to make the simulation meaningful.
28
Chapter 2. Tutorial
Supports So far, supports (unmovable particles) were providing necessary boundary: in the gravity deposition script, utils.facetBox is by internally composed of facets (triangulation elements), which is fixed in space; facets are also used for arbitrary triangulated surfaces (see relevant sections of the Users manual ). Another frequently used boundary is utils.wall (innite axis-aligned plane). Periodic Periodic boundary is a boundary created by using periodic (rather than innite) space. Such boundary is activated by O.periodic=True , and the space conguration is decribed by O.cell . It is well suited for studying bulk material behavior, as boundary eects are avoided, leading to smaller number of particles. On the other hand, it might not be suitable for studying localization, as any cell-level eects (such as shear bands) have to satisfy periodicity as well. The periodic cell is described by its reference size of box aligned with global axes, and current transformation, which can capture stretch, shear and rotation. Deformation is prescribed via velocity gradient, which updates the transformation before the next step. Homothetic deformation can smear velocity gradient accross the cell, making the boundary dissolve in the whole cell. Stress and strains can be controlled with PeriTriaxController; it is possible to prescribe mixed strain/stress goal state using PeriTriaxController.stressMask. The following creates periodic cloud of spheres and compresses to achieve x =-10 kPa, y =-10kPa and z =-0.1. Since stress is specied for y and z, stressMask is 0b011 (x1, y2, z4, in decimal 1+2=3).
Yade [2]: sp=pack.SpherePack() Yade [3]: sp.makeCloud((1,1,1),(2,2,2),rMean=.2,periodic=True) Out[3]: 9 Yade [4]: sp.toSimulation() Out[4]: [3, 4, 5, 6, 7, 8, 9, 10, 11]
# implicitly sets O.periodic=True, and O.cell.refSize to the packing per
Yade [5]: O.engines+=[PeriTriaxController(goal=(-1e4,-1e4,-.1),stressMask=0b011,maxUnbalanced=.2,doneHook='funct
When the simulation runs, PeriTriaxController takes over the control and calls doneHook when goal is reached. A full simulation with PeriTriaxController might look like the following:
from yade import pack,plot sp=pack.SpherePack() rMean=.05 sp.makeCloud((0,0,0),(1,1,1),rMean=rMean,periodic=True) sp.toSimulation() O.engines=[ ForceResetter(), InsertionSortCollider([Bo1_Sphere_Aabb()],verletDist=.05*rMean), InteractionLoop([Ig2_Sphere_Sphere_L3Geom()],[Ip2_FrictMat_FrictMat_FrictPhys()],[Law2_L3Geom_FrictPhys_ElPer NewtonIntegrator(damping=.6), PeriTriaxController(goal=(-1e6,-1e6,-.1),stressMask=0b011,maxUnbalanced=.2,doneHook='goalReached()',label='tr PyRunner(iterPeriod=100,command='addPlotData()') ] O.dt=.5*utils.PWaveTimeStep() O.trackEnergy=True def goalReached(): print 'Goal reached, strain',triax.strain,' stress',triax.stress O.pause() def addPlotData(): plot.addData(sx=triax.stress[0],sy=triax.stress[1],sz=triax.stress[2],ex=triax.strain[0],ey=triax.strain[1],e i=O.iter,unbalanced=utils.unbalancedForce(), totalEnergy=O.energy.total(),**O.energy # plot all energies ) plot.plots={'i':(('unbalanced','go'),None,'kinetic'),' i':('ex','ey','ez',None,'sx','sy','sz'),'i ':(O.energy.ke
2.4. Towards geomechanics
29
plot.plot() O.saveTmp() O.run()
2.5 Advanced & more

2.5.1 Particle size distribution
See Periodic triaxial test
2.5.2 Clumps
Clump; see Periodic triaxial test
2.5.3 Testing laws

LawTester, scripts/test/law-test.py
2.5.4 New law 2.5.5 Visualization

See the example 3d postprocessing VTKRecorder & Paraview qt.SnapshotEngine
2.6 Examples
2.6.1 Bouncing sphere
# basic simulation showing sphere falling ball gravity, # bouncing against another sphere representing the support # DATA COMPONENTS # add 2 particles to the simulation # they the default material (utils.defaultMat) O.bodies.append([ # fixed: particle's position in space will not change (support) utils.sphere(center=(0,0,0),radius=.5,fixed=True), # this particles is free, subject to dynamics utils.sphere((0,0,2),.5) ]) # FUNCTIONAL COMPONENTS # simulation loop -- see presentation for the explanation O.engines=[ ForceResetter(), InsertionSortCollider([Bo1_Sphere_Aabb()]), InteractionLoop( # collision geometry [Ig2_Sphere_Sphere_L3Geom()], [Ip2_FrictMat_FrictMat_FrictPhys()], # collision "physics"
30
Chapter 2. Tutorial
[Law2_L3Geom_FrictPhys_ElPerfPl()] # contact law -- apply forces ), # apply gravity force to particles GravityEngine(gravity=(0,0,-9.81)), # damping: numerical dissipation of energy NewtonIntegrator(damping=0.1) ] # set timestep to a fraction of the critical timestep # the fraction is very small, so that the simulation is not too fast # and the motion can be observed O.dt=.5e-4*utils.PWaveTimeStep() # save the simulation, so that it can be reloaded later, for experimentation O.saveTmp()
2.6.2 Gravity deposition

# gravity deposition in box, showing how to plot and save history of data, # and how to control the simulation while it is running by calling # python functions from within the simulation loop # import yade modules that we will use below from yade import pack, plot # create rectangular box from facets O.bodies.append(utils.geom.facetBox((.5,.5,.5),(.5,.5,.5),wallMask=31)) # create empty sphere packing # sphere packing is not equivalent to particles in simulation, it contains only the pure geometry sp=pack.SpherePack() # generate randomly spheres with uniform radius distribution sp.makeCloud((0,0,0),(1,1,1),rMean=.05,rRelFuzz=.5) # add the sphere pack to the simulation sp.toSimulation() O.engines=[ ForceResetter(), InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()]), InteractionLoop( # handle sphere+sphere and facet+sphere collisions [Ig2_Sphere_Sphere_L3Geom(),Ig2_Facet_Sphere_L3Geom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_L3Geom_FrictPhys_ElPerfPl()] ), GravityEngine(gravity=(0,0,-9.81)), NewtonIntegrator(damping=0.4), # call the checkUnbalanced function (defined below) every 2 seconds PyRunner(command='checkUnbalanced()',realPeriod=2), # call the addPlotData function every 200 steps PyRunner(command='addPlotData()',iterPeriod=100) ] O.dt=.5*utils.PWaveTimeStep() # enable energy tracking; any simulation parts supporting it # can create and update arbitrary energy types, which can be # accessed as O.energy['energyName'] subsequently O.trackEnergy=True # if the unbalanced forces goes below .05, the packing # is considered stabilized, therefore we stop collected
2.6. Examples
31
# data history and stop def checkUnbalanced(): if utils.unbalancedForce()<.05: O.pause() plot.saveDataTxt('bbb.txt.bz2') # plot.saveGnuplot('bbb') is also possible # collect history of data which will be plotted def addPlotData(): # each item is given a names, by which it can be the unsed in plot.plots # the **O.energy converts dictionary-like O.energy to plot.addData arguments plot.addData(i=O.iter,unbalanced=utils.unbalancedForce(),**O.energy) # define how to plot data: 'i' (step number) on the x-axis, unbalanced force # on the left y-axis, all energies on the right y-axis # (O.energy.keys is function which will be called to get all defined energies) # None separates left and right y-axis plot.plots={'i':('unbalanced',None,O.energy.keys)} # show the plot on the screen, and update while the simulation runs plot.plot() O.saveTmp()
2.6.3 Oedometric test

# gravity deposition, continuing with oedometric test after stabilization # shows also how to run parametric studies with yade-batch # # # # # # # The components of the batch are: 1. table with parameters, one set of parameters per line (ccc.table) 2. utils.readParamsFromTable which reads respective line from the parameter file 3. the simulation muse be run using yade-batch, not yade $ yade-batch --job-threads=1 03-oedometric-test.table 03-oedometric-test.py
# load parameters from file if run in batch # default values are used if not run from batch utils.readParamsFromTable(rMean=.05,rRelFuzz=.3,maxLoad=1e6,minLoad=1e4) # make rMean, rRelFuzz, maxLoad accessible directly as variables later from yade.params.table import * # create box with free top, and ceate loose packing inside the box from yade import pack, plot O.bodies.append(utils.geom.facetBox((.5,.5,.5),(.5,.5,.5),wallMask=31)) sp=pack.SpherePack() sp.makeCloud((0,0,0),(1,1,1),rMean=rMean,rRelFuzz=rRelFuzz) sp.toSimulation() O.engines=[ ForceResetter(), # sphere, facet, wall InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb(),Bo1_Wall_Aabb()]), InteractionLoop( # the loading plate is a wall, we need to handle sphere+sphere, sphere+facet, sphere+wall [Ig2_Sphere_Sphere_L3Geom(),Ig2_Facet_Sphere_L3Geom(),Ig2_Wall_Sphere_L3Geom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_L3Geom_FrictPhys_ElPerfPl()] ), GravityEngine(gravity=(0,0,-9.81)),
32
Chapter 2. Tutorial
NewtonIntegrator(damping=0.5), # the label creates an automatic variable referring to this engine # we use it below to change its attributes from the functions called PyRunner(command='checkUnbalanced()',realPeriod=2,label='checker'), ] O.dt=.5*utils.PWaveTimeStep() # # # # the following checkUnbalanced, unloadPlate and stopUnloading functions are all called by the 'checker' (the last engine) one after another; this sequence defines progression of different stages of the simulation, as each of the functions, when the condition is satisfied, updates 'checker' to call the next function when it is run from within the simulation next time
# check whether the gravity deposition has already finished # if so, add wall on the top of the packing and start the oedometric test def checkUnbalanced(): # at the very start, unbalanced force can be low as there is only few contacts, but it does not mean the pack if O.iter<5000: return # the rest will be run only if unbalanced is < .1 (stabilized packing) if utils.unbalancedForce()>.1: return # add plate at the position on the top of the packing # the maximum finds the z-coordinate of the top of the topmost particle O.bodies.append(utils.wall(max([b.state.pos[2]+b.shape.radius for b in O.bodies if isinstance(b.shape,Sphere) global plate # without this line, the plate variable would only exist inside this function plate=O.bodies[-1] # the last particles is the plate # Wall objects are "fixed" by default, i.e. not subject to forces # prescribing a velocity will therefore make it move at constant velocity (downwards) plate.state.vel=(0,0,-.1) # start plotting the data now, it was not interesting before O.engines=O.engines+[PyRunner(command='addPlotData()',iterPeriod=200)] # next time, do not call this function anymore, but the next one (unloadPlate) instead checker.command='unloadPlate()' def unloadPlate(): # if the force on plate exceeds maximum load, start unloading if abs(O.forces.f(plate.id)[2])>maxLoad: plate.state.vel*=-1 # next time, do not call this function anymore, but the next one (stopUnloading) instead checker.command='stopUnloading()' def stopUnloading(): if abs(O.forces.f(plate.id)[2])<minLoad: # O.tags can be used to retrieve unique identifiers of the simulation # if running in batch, subsequent simulation would overwrite each other's output files otherwise # d (or description) is simulation description (composed of parameter values) # while the id is composed of time and process number plot.saveDataTxt(O.tags['d.id']+'.txt') O.pause() def addPlotData(): if not isinstance(O.bodies[-1].shape,Wall): plot.addData(); return Fz=O.forces.f(plate.id)[2] plot.addData(Fz=Fz,w=plate.state.pos[2]-plate.state.refPos[2],unbalanced=utils.unbalancedForce(),i=O.iter) # besides unbalanced force evolution, also plot the displacement-force diagram plot.plots={'i':('unbalanced',),'w':('Fz',)} plot.plot() O.run() # when running with yade-batch, the script must not finish until the simulation is done fully # this command will wait for that (has no influence in the non-batch mode) utils.waitIfBatch()
2.6. Examples
33
Batch table
rMean rRelFuzz maxLoad .05 .1 1e6 .05 .2 1e6 .05 .3 1e6
2.6.4 Periodic simple shear

# encoding: utf-8 # # # # # # # # # # script for periodic simple shear test, with periodic boundary first compresses to attain some isotropic stress (checkStress), then loads in shear (checkDistorsion) the initial packing is either regular (hexagonal), with empty bands along the boundary, or periodic random cloud of spheres material friction angle is initially set to zero, so that the resulting packing is dense (sphere rearrangement is easier if there is no friction)
# setup the periodic boundary O.periodic=True O.cell.refSize=(2,2,2) from yade import pack,plot # the "if 0:" block will be never executed, therefore the "else:" block will be # to use cloud instead of regular packing, change to "if 1:" or something similar if 0: # create cloud of spheres and insert them into the simulation # we give corners, mean radius, radius variation sp=pack.SpherePack() sp.makeCloud((0,0,0),(2,2,2),rMean=.1,rRelFuzz=.6,periodic=True) # insert the packing into the simulation sp.toSimulation(color=(0,0,1)) # pure blue else: # in this case, add dense packing O.bodies.append( pack.regularHexa(pack.inAlignedBox((0,0,0),(2,2,2)),radius=.1,gap=0,color=(0,0,1)) ) # create "dense" packing by setting friction to zero initially O.materials[0].frictionAngle=0 # simulation loop (will be run at every step) O.engines=[ ForceResetter(), InsertionSortCollider([Bo1_Sphere_Aabb()]), InteractionLoop( [Ig2_Sphere_Sphere_L3Geom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_L3Geom_FrictPhys_ElPerfPl()] ), NewtonIntegrator(damping=.4), # run checkStress function (defined below) every second # the label is arbitrary, and is used later to refer to this engine PyRunner(command='checkStress()',realPeriod=1,label='checker'), # record data for plotting every 100 steps; addData function is defined below
34
Chapter 2. Tutorial
PyRunner(command='addData()',iterPeriod=100) ] # set the integration timestep to be 1/2 of the "critical" timestep O.dt=.5*utils.PWaveTimeStep() # prescribe isotropic normal deformation (constant strain rate) # of the periodic cell O.cell.velGrad=Matrix3(-.1,0,0, 0,-.1,0, 0,0,-.1) # when to stop the isotropic compression (used inside checkStress) limitMeanStress=-5e5 # called every second by the PyRunner engine def checkStress(): # stress tensor as the sum of normal and shear contributions # Matrix3.Zero is the intial value for sum(...) stress=sum(utils.normalShearStressTensors(),Matrix3.Zero) print 'mean stress',stress.trace()/3. # if mean stress is below (bigger in absolute value) limitMeanStress, start shearing if stress.trace()/3.<limitMeanStress: # apply constant-rate distorsion on the periodic cell O.cell.velGrad=Matrix3(0,0,.1, 0,0,0, 0,0,0) # change the function called by the checker engine # (checkStress will not be called anymore) checker.command='checkDistorsion()' # block rotations of particles to increase tanPhi, if desired # disabled by default if 0: for b in O.bodies: # block X,Y,Z rotations, translations are free b.state.blockedDOFs='XYZ' # stop rotations if any, as blockedDOFs block accelerations really b.state.angVel=(0,0,0) # set friction angle back to non-zero value # tangensOfFrictionAngle is computed by the Ip2_* functor from material # for future contacts change material (there is only one material for all particles) O.materials[0].frictionAngle=.5 # radians # for existing contacts, set contact friction directly for i in O.interactions: i.phys.tangensOfFrictionAngle=tan(.5) # called from the 'checker' engine periodically, during def checkDistorsion(): # if the distorsion value is >.3, exit; otherwise do if abs(O.cell.trsf[0,2])>.5: # save data from addData(...) before exiting into # use O.tags['id'] to distinguish individual runs plot.saveDataTxt(O.tags['id']+'.txt') # exit the program #import sys #sys.exit(0) # no error (0) O.pause() the shear phase nothing file of the same simulation
# called periodically to store data history def addData(): # get the stress tensor (as 3x3 matrix) stress=sum(utils.normalShearStressTensors(),Matrix3.Zero) # give names to values we are interested in and save them plot.addData(exz=O.cell.trsf[0,2],szz=stress[2,2],sxz=stress[0,2],tanPhi=stress[0,2]/stress[2,2],i=O.iter) # color particles based on rotation amount for b in O.bodies: # rot() gives rotation vector between reference and current position b.shape.color=utils.scalarOnColorScale(b.state.rot().norm(),0,pi/2.)
2.6. Examples
35
# define what to plot (3 plots in total) ## exz(i), [left y axis, separate by None:] szz(i), sxz(i) ## szz(exz), sxz(exz) ## tanPhi(i) # note the space in 'i ' so that it does not overwrite the 'i' entry plot.plots={'i':('exz',None,'szz','sxz'),'exz':('szz','sxz'),'i ':('tanPhi',)} # better show rotation of particles Gl1_Sphere.stripes=True # open the plot on the screen plot.plot() O.saveTmp()
2.6.5 3d postprocessing
# demonstrate 3d postprocessing with yade # # 1. qt.SnapshotEngine saves images of the 3d view as it appears on the screen periodically # utils.makeVideo is then used to make real movie from those images # 2. VTKRecorder saves data in files which can be opened with Paraview # see the User's manual for an intro to Paraview # generate loose packing from yade import pack, qt sp=pack.SpherePack() sp.makeCloud((0,0,0),(2,2,2),rMean=.1,rRelFuzz=.6,periodic=True) # add to scene, make it periodic sp.toSimulation() O.engines=[ ForceResetter(), InsertionSortCollider([Bo1_Sphere_Aabb()]), InteractionLoop( [Ig2_Sphere_Sphere_L3Geom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_L3Geom_FrictPhys_ElPerfPl()] ), NewtonIntegrator(damping=.4), # save data for Paraview VTKRecorder(fileName='3d-vtk-',recorders=['all'],iterPeriod=1000), # save data from Yade's own 3d view qt.SnapshotEngine(fileBase='3d-',iterPeriod=200,label='snapshot'), # this engine will be called after 20000 steps, only once PyRunner(command='finish()',iterPeriod=20000) ] O.dt=.5*utils.PWaveTimeStep() # prescribe constant-strain deformation of the cell O.cell.velGrad=Matrix3(-.1,0,0, 0,-.1,0, 0,0,-.1) # we must open the view explicitly (limitation of the qt.SnapshotEngine) qt.View() # this function is called when the simulation is finished def finish(): # snapshot is label of qt.SnapshotEngine # the 'snapshots' attribute contains list of all saved files utils.makeVideo(snapshot.snapshots,'3d.mpeg',fps=10,bps=10000) O.pause()
36
Chapter 2. Tutorial
# set parameters of the renderer, to show network chains rather than particles # these settings are accessible from the Controller window, on the second tab ("Display") as well rr=yade.qt.Renderer() rr.shape=False rr.intrPhys=True
2.6.6 Periodic triaxial test

# encoding: utf-8 # # # # # # # # # # # # # # # # # # # # # periodic triaxial test simulation The initial packing is either 1. random cloud with uniform distribution, or 2. cloud with specified granulometry (radii and percentages), or 3. cloud of clumps, i.e. rigid aggregates of several particles The triaxial consists of 2 stages: 1. isotropic compaction, until sigmaIso is reached in all directions; this stage is ended by calling compactionFinished() 2. constant-strain deformation along the z-axis, while maintaining constant stress (sigmaIso) laterally; this stage is ended by calling triaxFinished() Controlling of strain and stresses is performed via PeriTriaxController, of which parameters determine type of control and also stability condition (maxUnbalanced) so that the packing is considered stabilized and the stage is done.
sigmaIso=-1e5 #import matplotlib #matplotlib.use('Agg') # generate loose packing from yade import pack, qt, plot O.periodic=True sp=pack.SpherePack() if 0: ## uniform distribution sp.makeCloud((0,0,0),(2,2,2),rMean=.1,rRelFuzz=.3,periodic=True) elif 0: ## per-fraction distribution ## passing: cummulative percentage sp.particleSD2(radii=[.09,.1,.2],passing=[40,80,100],periodic=True,numSph=1000) else: ## create packing from clumps # configuration of one clump c1=pack.SpherePack([((0,0,0),.03333),((.03,0,0),.017),((0,.03,0),.017)]) # make cloud using the configuration c1 (there could c2, c3, ...; selection between them would be random) sp.makeClumpCloud((0,0,0),(2,2,2),[c1],periodic=True,num=500) # setup periodic boundary, insert the packing sp.toSimulation() O.engines=[ ForceResetter(),
2.6. Examples
37
InsertionSortCollider([Bo1_Sphere_Aabb()]), InteractionLoop( [Ig2_Sphere_Sphere_ScGeom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_ScGeom_FrictPhys_CundallStrack()] ), PeriTriaxController(label='triax', # specify target values and whether they are strains or stresses goal=(sigmaIso,sigmaIso,sigmaIso),stressMask=7, # type of servo-control dynCell=True,maxStrainRate=(10,10,10), # wait until the unbalanced force goes below this value maxUnbalanced=.1,relStressTol=1e-3, # call this function when goal is reached and the packing is stable doneHook='compactionFinished()' ), NewtonIntegrator(damping=.2), PyRunner(command='addPlotData()',iterPeriod=100), ] O.dt=.5*utils.PWaveTimeStep() def addPlotData(): plot.addData(unbalanced=utils.unbalancedForce(),i=O.iter, sxx=triax.stress[0],syy=triax.stress[1],szz=triax.stress[2], exx=triax.strain[0],eyy=triax.strain[1],ezz=triax.strain[2], # save all available energy data Etot=O.energy.total(),**O.energy ) # enable energy tracking in the code O.trackEnergy=True # define what to plot plot.plots={'i':('unbalanced',),'i ':('sxx','syy','szz'),' i':('exx','eyy','ezz'), # energy plot ' i ':(O.energy.keys,None,'Etot'), } # show the plot plot.plot() def compactionFinished(): # set the current cell configuration to be the reference one O.cell.trsf=Matrix3.Identity # change control type: keep constant confinement in x,y, 20% compression in z triax.goal=(sigmaIso,sigmaIso,-.3) triax.stressMask=3 # allow faster deformation along x,y to better maintain stresses triax.maxStrainRate=(1.,1.,.1) # next time, call triaxFinished instead of compactionFinished triax.doneHook='triaxFinished()' # do not wait for stabilization before calling triaxFinished triax.maxUnbalanced=10 def triaxFinished(): print 'Finished' O.pause()
38
Chapter 2. Tutorial
Chapter 3
Users manual
3.1 Scene construction
3.1.1 Adding particles
The BodyContainer holds Body objects in the simulation; it is accessible as O.bodies. Creating Body objects Body objects are only rarely constructed by hand by their components (Shape, Bound, State, Material); instead, convenience functions utils.sphere, utils.facet and utils.wall are used to create them. Using these functions also ensures better future compatibility, if internals of Body change in some way. These functions receive geometry of the particle and several other characteristics. See their documentation for details. If the same Material is used for several (or many) bodies, it can be shared by adding it in O.materials, as explained below. Dening materials The O.materials object (instance of Omega.materials) holds dened shared materials for bodies. It only supports addition, and will typically hold only a few instance (though there is no limit). label given to each material is optional, but can be passed to utils.sphere and other functions forconstructing body. The value returned by O.materials.append is an id of the material, which can be also passed to utils.sphere it is a little bit faster than using label, though not noticeable for small number of particles and perhaps less convenient. If no Material is specied when calling utils.sphere, the last dened material is used; that is a convenient default. If no material is dened yet (hence there is no last material), a default material will be created using utils.defaultMaterial; this should not happen for serious simulations, but is handy in simple scripts, where exact material properties are more or less irrelevant.
Yade [2]: len(O.materials) Out[2]: 0 Yade [3]: idConcrete=O.materials.append(FrictMat(young=30e9,poisson=.2,frictionAngle=.6,label="concrete")) Yade [4]: O.materials[idConcrete] Out[4]: <FrictMat instance at 0x177be690> # uses the last defined material Yade [5]: O.bodies.append(utils.sphere(center=(0,0,0),radius=1)) Out[5]: 0
39
# material given by id Yade [6]: O.bodies.append(utils.sphere((0,0,2),1,material=idConcrete)) Out[6]: 1 # material given by label Yade [7]: O.bodies.append(utils.sphere((0,2,0),1,material="concrete")) Out[7]: 2 Yade [8]: idSteel=O.materials.append(FrictMat(young=210e9,poisson=.25,frictionAngle=.8,label="steel")) Yade [9]: len(O.materials) Out[9]: 2 # implicitly uses "steel" material, as it is the last one now Yade [10]: O.bodies.append(utils.facet([(1,0,0),(0,1,0),(-1,-1,0)])) Out[10]: 3
Adding multiple particles As shown above, bodies are added one by one or several at the same time using the append method:
Yade [2]: O.bodies.append(utils.sphere((0,10,0),1)) Out[2]: 0 Yade [3]: O.bodies.append(utils.sphere((0,0,2),1)) Out[3]: 1 # this is the same, but in one function call Yade [4]: O.bodies.append([ ...: utils.sphere((0,0,0),1), ...: utils.sphere((0,0,2),1) ...: ]) ...: Out[4]: [2, 3]
Many functions introduced in next sections return list of bodies which can be readily added to the simulation, including packing generators, such as pack.randomDensePack, pack.regularHexa surface function pack.gtsSurface2Facets import functions ymport.gmsh, ymport.stl, As those functions use utils.sphere and utils.facet internally, they accept additional argument passed to those function. In particular, material for each body is selected following the rules above (last one if not specied, by label, by index, etc.). Clumping particles together In some cases, you might want to create rigid aggregate of individual particles (i.e. particles will retain their mutual position during simulation). This we call a clump. A clump is internally represented by a special body, referenced by clumpId of its members (see also isClump, isClumpMember and isStandalone). Like every body a clump has a position, which is the balance point between all members. A clump body itself has no interactions with other bodies. Interactions between clumps is internally represented by interactions between clump members. There are also no interactions between clump members with same clumpId. YADE supports dierent ways of creating clumps: Create clumps and spheres (clump members) directly with one command:
40
Chapter 3. Users manual
The function appendClumped() is designed for this task. For instance, we might add 2 spheres tied together:
Yade [2]: O.bodies.appendClumped([ ...: utils.sphere([0,0,0],1), ...: utils.sphere([0,0,2],1) ...: ]) ...: Out[2]: (2, [0, 1]) Yade [3]: len(O.bodies) Out[3]: 3 Yade [4]: O.bodies[1].isClumpMember, O.bodies[2].clumpId Out[4]: (True, 2) Yade [5]: O.bodies[2].isClump, O.bodies[2].clumpId Out[5]: (True, 2)
-> appendClumped() returns a tuple of ids (clumpId,[memberId1,memberId2,...]) Use existing spheres and clump them together: For this case the function clump() can be used. One way to do this is to create a list of bodies, that should be clumped before using the clump() command:
Yade [2]: bodyList = [] Yade [3]: for ii in range(0,5): ...: bodyList.append(O.bodies.append(utils.sphere([ii,0,1],.5)))#create a "chain" of 5 spheres ...: Yade [4]: print bodyList [0, 1, 2, 3, 4] Yade [5]: idClump=O.bodies.clump(bodyList)
-> clump() returns clumpId Another option is to replace standalone spheres from a given packing (see SpherePack and makeCloud) using clump templates. This is done by a function called replaceByClumps(). This function takes a list of clumpTemplates() and a list of amounts and replaces spheres by clumps. The volume of a new clump will be the same as the volume of the sphere, that was replaced (clump volume/mass/inertia is accounting for overlaps assuming that there are only pair overlaps). It is also possible to add bodies to a clump and release bodies from a clump. Also you can erase the clump (clump members will get standalone spheres). Note: Have a look at examples/clumps/ folder. There you will nd some examples, that show usage of dierent functions for clumps.
-> replaceByClumps() returns a list of tuples: [(clumpId1,[memberId1,memberId2,...]),(clumpId2,[memberId1,memb
3.1.2 Sphere packings

Representing a solid of an arbitrary shape by arrangement of spheres presents the problem of sphere packing, i.e. spatial arrangement of sphere such that given solid is approximately lled with them. For the purposes of DEM simulation, there can be several requirements. 1. Distribution of spheres radii. Arbitrary volume can be lled completely with spheres provided there are no restrictions on their radius; in such case, number of spheres can be innite and their radii approach zero. Since both number of particles and minimum sphere radius (via critical 3.1. Scene construction 41
timestep) determine computation cost, radius distribution has to be given mandatorily. The most typical distribution is uniform: meandispersion; if dispersion is zero, all spheres will have the same radius. 2. Smooth boundary. Some algorithms treat boundaries in such way that spheres are aligned on them, making them smoother as surface. 3. Packing density, or the ratio of spheres volume and solid size. It is closely related to radius distribution. 4. Coordination number, (average) number of contacts per sphere. 5. Isotropy (related to regularity/irregularity); packings with preferred directions are usually not desirable, unless the modeled solid also has such preference. 6. Permissible Spheres overlap; some algorithms might create packing where spheres slightly overlap; since overlap usually causes forces in DEM, overlap-free packings are sometimes called stress-free. Volume representation There are 2 methods for representing exact volume of the solid in question in Yade: boundary representation and constructive solid geometry. Despite their fundamental dierences, they are abstracted in Yade in the Predicate class. Predicate provides the following functionality: 1. denes axis-aligned bounding box for the associated solid (optionally denes oriented bounding box); 2. can decide whether given point is inside or outside the solid; most predicates can also (exactly or approximately) tell whether the point is inside and satises some given padding distance from the represented solid boundary (so that sphere of that volume doesnt stick out of the solid).
Constructive Solid Geometry (CSG)
CSG approach describes volume by geometric primitives or primitive solids (sphere, cylinder, box, cone, ) and boolean operations on them. Primitives dened in Yade include inCylinder, inSphere, inEllipsoid, inHyperboloid, notInNotch. For instance, hyperboloid (dogbone) specimen for tension-compression test can be constructed in this way (shown at img. img-hyperboloid):
from yade import pack ## construct the predicate first pred=pack.inHyperboloid(centerBottom=(0,0,-.1),centerTop=(0,0,.1),radius=.05,skirt=.03) ## alternatively: pack.inHyperboloid((0,0,-.1),(0,0,.1),.05,.03) ## pack the predicate with spheres (will be explained later) spheres=pack.randomDensePack(pred,spheresInCell=2000,radius=3.5e-3) ## add spheres to simulation O.bodies.append(spheres)
Boundary representation (BREP)
Representing a solid by its boundary is much more exible than CSG volumes, but is mostly only approximate. Yade interfaces to GNU Triangulated Surface Library (GTS) to import surfaces readable by GTS, but also to construct them explicitly from within simulation scripts. This makes possible parametric construction of rather complicated shapes; there are functions to create set of 3d polylines from 2d polyline (pack.revolutionSurfaceMeridians), to triangulate surface between such set of 3d polylines (pack.sweptPolylines2gtsSurface). For example, we can construct a simple funnel (examples/funnel.py, shown at img-funnel):
42
Figure 3.1: Specimen pack.randomDensePack.

from numpy import linspace from yade import pack
constructed
with
the
pack.inHyperboloid
predicate,
packed
with
# angles for points on circles thetas=linspace(0,2*pi,num=16,endpoint=True) # creates list of polylines in 3d from list of 2d projections # turned from 0 to meridians=pack.revolutionSurfaceMeridians( [[(3+rad*sin(th),10*rad+rad*cos(th)) for th in thetas] for rad in linspace(1,2,num=10)], linspace(0,pi,num=10) ) # create surface surf=pack.sweptPolylines2gtsSurface( meridians+ +[[Vector3(5*sin(-th),-10+5*cos(-th),30) for th in thetas]] ) # add to simulation O.bodies.append(pack.gtsSurface2Facets(surf))
# add funnel top
Figure 3.2: Triangulated funnel, constructed with the examples/funnel.py script. GTS surface objects can be used for 2 things: 3.1. Scene construction 43
1. pack.gtsSurface2Facets function can create the triangulated surface (from Facet particles) in the simulation itself, as shown in the funnel example. (Triangulated surface can also be imported directly from a STL le using ymport.stl.) 2. pack.inGtsSurface predicate can be created, using the surface as boundary representation of the enclosed volume. The scripts/test/gts-horse.py (img. img-horse) shows both possibilities; rst, a GTS surface is imported:
import gts surf=gts.read(open('horse.coarse.gts'))
That surface object is used as predicate for packing:

pred=pack.inGtsSurface(surf) O.bodies.append(pack.regularHexa(pred,radius=radius,gap=radius/4.))
and then, after being translated, as base for triangulated surface in the simulation itself:
surf.translate(0,0,-(aabb[1][2]-aabb[0][2])) O.bodies.append(pack.gtsSurface2Facets(surf,wire=True))
Figure 3.3: Imported GTS surface (horse) used as packing predicate (top) and surface constructed from facets (bottom). See http://www.youtube.com/watch?v=PZVruIlUX1A for movie of this simulation.
Boolean operations on predicates
Boolean operations on pair of predicates (noted A and B) are dened: intersection A & B (conjunction): point must be in both predicates involved. union A | B (disjunction): point must be in the rst or in the second predicate. dierence A - B (conjunction with second predicate negated): the point must be in the rst predicate and not in the second one. symmetric dierence A ^ B (exclusive disjunction): point must be in exactly one of the two predicates.
44
Composed predicates also properly dene their bounding box. For example, we can take box and remove cylinder from inside, using the A - B operation (img. img-predicate-dierence):
pred=pack.inAlignedBox((-2,-2,-2),(2,2,2))-pack.inCylinder((0,-2,0),(0,2,0),1) spheres=pack.randomDensePack(pred,spheresInCell=2000,radius=.1,rRelFuzz=.4)
Figure 3.4: Box with cylinder removed from inside, using dierence of these two predicates.
Packing algorithms Algorithms presented below operate on geometric spheres, dened by their center and radius. With a few exception documented below, the procedure is as follows: 1. Sphere positions and radii are computed (some functions use volume predicate for this, some do not) 2. utils.sphere is called for each position and radius computed; it receives extra keyword arguments of the packing function (i.e. arguments that the packing function doesnt specify in its denition; they are noted **kw). Each utils.sphere call creates actual Body objects with Sphere shape. List of Body objects is returned. 3. List returned from the packing function can be added to simulation using O.bodies.append. Taking the example of pierced box:
pred=pack.inAlignedBox((-2,-2,-2),(2,2,2))-pack.inCylinder((0,-2,0),(0,2,0),1) spheres=pack.randomDensePack(pred,spheresInCell=2000,radius=.1,rRelFuzz=.4,wire=True,color=(0,0,1),material=1)
Keyword arguments wire, color and material are not declared in pack.randomDensePack, therefore will be passed to utils.sphere, where they are also documented. spheres is now list of Body objects, which we add to the simulation:
O.bodies.append(spheres)
Packing algorithms described below produce dense packings. If one needs loose packing, pack.SpherePack class provides functions for generating loose packing, via its pack.SpherePack.makeCloud method. It is used internally for generating initial conguration in dynamic algorithms. For instance:
from yade import pack sp=pack.SpherePack() sp.makeCloud(minCorner=(0,0,0),maxCorner=(3,3,3),rMean=.2,rRelFuzz=.5)
3.1. Scene construction
45
will ll given box with spheres, until no more spheres can be placed. The object can be used to add spheres to simulation:
for c,r in sp: O.bodies.append(utils.sphere(c,r))
or, in a more pythonic way, with one single O.bodies.append call:

O.bodies.append([utils.sphere(c,r) for c,r in sp])
Geometric
Geometric algorithms compute packing without performing dynamic simulation; among their advantages are speed; spheres touch exactly, there are no overlaps (what some people call stress-free packing); their chief disadvantage is that radius distribution cannot be prescribed exactly, save in specic cases (regular packings); sphere radii are given by the algorithm, which already makes the system determined. If exact radius distribution is important for your problem, consider dynamic algorithms instead. Regular Yade denes packing generators for spheres with constant radii, which can be used with volume predicates as described above. They are dense orthogonal packing (pack.regularOrtho) and dense hexagonal packing (pack.regularHexa). The latter creates so-called hexagonal close packing, which achieves maximum density (http://en.wikipedia.org/wiki/Close-packing_of_spheres). Clear disadvantage of regular packings is that they have very strong directional preferences, which might not be an issue in some cases. Irregular Random geometric algorithms do not integrate at all with volume predicates described above; rather, they take their own boundary/volume denition, which is used during sphere positioning. On the other hand, this makes it possible for them to respect boundary in the sense of making spheres touch it at appropriate places, rather than leaving empty space in-between. _packSpherePadder.SpherePadder constructs dense sphere packing based on pre-computed tetrahedron mesh; it is documented in _packSpherePadder.SpherePadder documentation; sample script is in scripts/test/SpherePadder.py. _packSpherePadder.SpherePadder does not return Body list as other algorithms, but a pack.SpherePack object; it can be iterated over, adding spheres to the simulation, as shown in its documentation. GenGeo is library (python module) for packing generation developed with ESyS-Particle. It creates packing by random insertion of spheres with given radius range. Inserted spheres touch each other exactly and, more importantly, they also touch the boundary, if in its neighbourhood. Boundary is represented as special object of the GenGeo library (Sphere, cylinder, box, convex polyhedron, ). Therefore, GenGeo cannot be used with volume represented by yade predicates as explained above. Packings generated by this module can be imported directly via ymport.gengeo, or from saved le via ymport.gengeoFile. There is an example script scripts/test/genCylLSM.py. Full documentation for GenGeo can be found at ESyS documentation website. To our knowledge, the GenGeo library is not currently packaged. It can be downloaded from current subversion repository
svn checkout https://svn.esscc.uq.edu.au/svn/esys3/lsm/contrib/LSMGenGeo
then following instruction in the INSTALL le.
46
Dynamic
The most versatile algorithm for random dense packing is provided by pack.randomDensePack. Initial loose packing of non-overlapping spheres is generated by randomly placing them in cuboid volume, with radii given by requested (currently only uniform) radius distribution. When no more spheres can be inserted, the packing is compressed and then uncompressed (see py/pack/pack.py for exact values of these stresses) by running a DEM simulation; Omega.switchScene is used to not aect existing simulation). Finally, resulting packing is clipped using provided predicate, as explained above. By its nature, this method might take relatively long; and there are 2 provisions to make the computation time shorter: If number of spheres using the spheresInCell parameter is specied, only smaller specimen with periodic boundary is created and then repeated as to ll the predicate. This can provide highquality packing with low regularity, depending on the spheresInCell parameter (value of several thousands is recommended). Providing memoizeDb parameter will make pack.randomDensePack rst look into provided le (SQLite database) for packings with similar parameters. On success, the packing is simply read from database and returned. If there is no similar pre-existent packing, normal procedure is run, and the result is saved in the database before being returned, so that subsequent calls with same parameters will return quickly. If you need to obtain full periodic packing (rather than packing clipped by predicate), you can use pack.randomPeriPack. In case of specic needs, you can create packing yourself, by hand. For instance, packing boundary can be constructed from facets, letting randomly positioned spheres in space fall down under gravity.
3.1.3 Triangulated surfaces

Yade integrates with the the GNU Triangulated Surface library, exposed in python via the 3rd party gts module. GTS provides variety of functions for surface manipulation (coarsening, tesselation, simplication, import), to be found in its documentation. GTS surfaces are geometrical objects, which can be inserted into simulation as set of particles whose Body.shape is of type Facet single triangulation elements. pack.gtsSurface2Facets can be used to convert GTS surface triangulation into list of bodies ready to be inserted into simulation via O.bodies.append. Facet particles are created by default as non-Body.dynamic (they have zero inertial mass). That means that they are xed in space and will not move if subject to forces. You can however prescribe arbitrary movement to facets using a PartialEngine (such as TranslationEngine or RotationEngine); assign explicitly mass and inertia to that particle; make that particle part of a clump and assign mass and inertia of the clump itself (described below). Note: Facets can only (currently) interact with spheres, not with other facets, even if they are dynamic. Collision of 2 facets will not create interaction, therefore no forces on facets.
Import Yade currently oers 3 formats for importing triangulated surfaces from external les, in the ymport module: ymport.gts text le in native GTS format. ymport.stl STereoLitography format, in either text or binary form; exported from Blender, but from many CAD systems as well.
47
ymport.gmsh. text le in native format for GMSH, popular open-source meshing program. If you need to manipulate surfaces before creating list of facets, you can study the py/ymport.py le where the import functions are dened. They are rather simple in most cases. Parametric construction The gts module provides convenient way of creating surface by vertices, edges and triangles. Frequently, though, the surface can be conveniently described as surface between polylines in space. For instance, cylinder is surface between two polygons (closed polylines). The pack.sweptPolylines2gtsSurface oers the functionality of connecting several polylines with triangulation. Note: The implementation of pack.sweptPolylines2gtsSurface is rather simplistic: all polylines must be of the same length, and they are connected with triangles between points following their indices within each polyline (not by distance). On the other hand, points can be co-incident, if the threshold parameter is positive: degenerate triangles with vertices closer that threshold are automatically eliminated. Manipulating lists eciently (in terms of code length) requires being familiar with list comprehensions in python. Another examples can be found in examples/mill.py (fully parametrized) or examples/funnel.py (with hardcoded numbers).
3.1.4 Creating interactions

In typical cases, interactions are created during simulations as particles collide. This is done by a Collider detecting approximate contact between particles and then an IGeomFunctor detecting exact collision. Some material models (such as the concrete model) rely on initial interaction network which is denser than geometrical contact of spheres: spheres radii as enlarged by a dimensionless factor called interaction radius (or interaction ratio) to create this initial network. This is done typically in this way (see examples/concrete/uniax.py for an example): 1. Approximate collision detection is adjusted so that approximate contacts are detected also between particles within the interaction radius. This consists in setting value of Bo1_Sphere_Aabb.aabbEnlargeFactor to the interaction radius value. 2. The geometry functor (Ig2) would normally say that there is no contact if given 2 spheres that are not in contact. Therefore, the same value as for Bo1_Sphere_Aabb.aabbEnlargeFactor must be given to it. (Either Ig2_Sphere_Sphere_Dem3DofGeom.distFactor or Ig2_Sphere_Sphere_ScGeom.interactionDetectionFactor, depending on the functor that is in use. Note that only Sphere + Sphere interactions are supported; there is no parameter analogous to distFactor in Ig2_Facet_Sphere_ScGeom. This is on purpose, since the interaction radius is meaningful in bulk material represented by sphere packing, whereas facets usually represent boundary conditions which should be exempt from this dense interaction network. 3. Run one single step of the simulation so that the initial network is created. 4. Reset interaction radius in both Bo1 and Ig2 functors to their default value again. 5. Continue the simulation; interactions that are already established will not be deleted (the Law2 functor in usepermitting). In code, such scenario might look similar to this one (labeling is explained in Labeling things):
intRadius=1.5 O.engines=[ ForceResetter(), InsertionSortCollider([ # enlarge here Bo1_Sphere_Aabb(aabbEnlargeFactor=intRadius,label='bo1s'),
48
Bo1_Facet_Aabb(), ]), InteractionLoop( [ # enlarge here Ig2_Sphere_Sphere_ScGeom(interactionDetectionFactor=intRadius,label='ig2ss'), Ig2_Facet_Sphere_ScGeom(), ], [Ip2_CpmMat_CpmMat_CpmPhys()], [Law2_ScGeom_CpmPhys_Cpm(epsSoft=0)], # deactivated ), NewtonIntegrator(damping=damping,label='damper'), ] # run one single step O.step() # reset interaction radius to the default value bo1s.aabbEnlargeFactor=1.0 ig2ss.interactionDetectionFactor=1.0 # now continue simulation O.run()
Individual interactions on demand It is possible to create an interaction between a pair of particles independently of collision detection using utils.createInteraction. This function looks for and uses matching Ig2 and Ip2 functors. Interaction will be created regardless of distance between given particles (by passing a special parameter to the Ig2 functor to force creation of the interaction even without any geometrical contact). Appropriate constitutive law should be used to avoid deletion of the interaction at the next simulation step.
Yade [2]: O.materials.append(FrictMat(young=3e10,poisson=.2,density=1000)) Out[2]: 0 Yade [3]: O.bodies.append([ ...: utils.sphere([0,0,0],1), ...: utils.sphere([0,0,1000],1) ...: ]) ...: Out[3]: [0, 1] # only add InteractionLoop, no other engines are needed now Yade [4]: O.engines=[ ...: InteractionLoop( ...: [Ig2_Sphere_Sphere_ScGeom(),], ...: [Ip2_FrictMat_FrictMat_FrictPhys()], ...: [] # not needed now ...: ) ...: ] ...: Yade [5]: i=utils.createInteraction(0,1) # created by functors in InteractionLoop Yade [6]: i.geom, i.phys Out[6]: (<ScGeom instance at 0x65ca6f0>, <FrictPhys instance at 0x139d31b0>)
This method will be rather slow if many interaction are to be created (the functor lookup will be repeated for each of them). In such case, ask on [email protected] to have the utils.createInteraction function accept list of pairs ids as well.
49
3.1.5 Base engines

A typical DEM simulation in Yade does at least the following at each step (see Function components for details): 1. Reset forces from previous step 2. Detect new collisions 3. Handle interactions 4. Apply forces and update positions of particles Each of these points corresponds to one or several engines:
O.engines=[ ForceResetter(), # reset forces InsertionSortCollider([...]), # approximate collision detection InteractionLoop([...],[...],[...]) # handle interactions NewtonIntegrator() # apply forces and update positions ]
The order of engines is important. In majority of cases, you will put any additional engine after InteractionLoop: if it apply force, it should come before NewtonIntegrator, otherwise the force will never be eective. if it makes use of bodies positions, it should also come before NewtonIntegrator, otherwise, positions at the next step will be used (this might not be critical in many cases, such as output for visualization with VTKRecorder). The O.engines sequence must be always assigned at once (the reason is in the fact that although engines themselves are passed by reference, the sequence is copied from c++ to Python or from Python to c++). This includes modifying an existing O.engines; therefore
O.engines.append(SomeEngine()) # wrong
will not work;

O.engines=O.engines+[SomeEngine()] # ok
must be used instead. For inserting an engine after position #2 (for example), use python slice notation:
O.engines=O.engines[:2]+[SomeEngine()]+O.engines[2:]
Note: When Yade starts, O.engines is lled with a reasonable default list, so that it is not strictly necessary to redene it when trying simple things. The default scene will handle spheres, boxes, and facets with frictional properties correctly, and adjusts the timestep dynamically. You can nd an example in simple-scene-default-engines.py.
Functors choice In the above example, we omited functors, only writing ellipses ... instead. As explained in Dispatchers and functors, there are 4 kinds of functors and associated dispatchers. User can choose which ones to use, though the choice must be consistent.
Bo1 functors
Bo1 functors must be chosen depending on the collider in use; they are given directly to the collider (which internally uses BoundDispatcher). At this moment (September 2010), the most common choice is InsertionSortCollider, which uses Aabb; functors creating Aabb must be used in that case. Depending on particle shapes in your simulation, choose appropriate functors: 50 Chapter 3. Users manual
O.engines=[..., InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()]), ... ]
Using more functors than necessary (such as Bo1_Facet_Aabb if there are no facets in the simulation) has no performance penalty. On the other hand, missing functors for existing shapes will cause those bodies to not collider with other bodies (they will freely interpenetrate). There are other colliders as well, though their usage is only experimental: SpatialQuickSortCollider is correctness-reference collider operating on Aabb; it is signicantly slower than InsertionSortCollider. PersistentTriangulationCollider only works on spheres; it does not use a BoundDispatcher, as it operates on spheres directly. FlatGridCollider is proof-of-concept grid-based collider, which computes grid positions internally (no BoundDispatcher either)
Ig2 functors
Ig2 functor choice (all of the derive from IGeomFunctor) depends on 1. shape combinations that should collide; for instance:
InteractionLoop([Ig2_Sphere_Sphere_ScGeom()],[],[])
will handle collisions for Sphere + Sphere, but not for Facet + Sphere if that is desired, an additional functor must be used:
InteractionLoop([ Ig2_Sphere_Sphere_ScGeom(), Ig2_Facet_Sphere_ScGeom() ],[],[])
Again, missing combination will cause given shape combinations to freely interpenetrate one another. 2. IGeom type accepted by the Law2 functor (below); it is the rst part of functors name after Law2 (for instance, Law2_ScGeom_CpmPhys_Cpm accepts ScGeom). This is (for most cases) either Dem3DofGeom (total shear formulation) or ScGeom (incremental shear formulation). For Dem3DofGeom, the above example would simply change to:
InteractionLoop([ Ig2_Sphere_Sphere_Dem3DofGeom(), Ig2_Facet_Sphere_Dem3DofGeom() ],[],[])
Ip2 functors
Ip2 functors (deriving from IPhysFunctor) must be chosen depending on 1. Material combinations within the simulation. In most cases, Ip2 functors handle 2 instances of the same Material class (such as Ip2_FrictMat_FrictMat_FrictPhys for 2 bodies with FrictMat) 2. IPhys accepted by the constitutive law (Law2 functor), which is the second part of the Law2 functors name (e.g. Law2_ScGeom_FrictPhys_CundallStrack accepts FrictPhys) Note: Unlike with Bo1 and Ig2 functors, unhandled combination of Materials is an error condition signaled by an exception.
51
Law2 functor(s)
Law2 functor was the ultimate criterion for the choice of Ig2 and Ip2 functors; there are no restrictions on its choice in itself, as it only applies forces without creating new objects. In most simulations, only one Law2 functor will be in use; it is possible, though, to have several of them, dispatched based on combination of IGeom and IPhys produced previously by Ig2 and Ip2 functors respectively (in turn based on combination of Shapes and Materials). Note: As in the case of Ip2 functors, receiving a combination of IGeom and IPhys which is not handled by any Law2 functor is an error. Warning: Many Law2 exist in Yade, and new ones can appear at any time. In some cases dierent functors are only dierent implementations of the same contact law (e.g. Law2_ScGeom_FrictPhys_CundallStrack and Law2_L3Geom_FrictPhys_ElPerfPl). Also, sometimes, the peculiarity of one functor may be reproduced as a special case of a more general one. Therefore, for a given constitutive behavior, the user may have the choice between dierent functors. It is strongly recommended to favor the most used and most validated implementation when facing such choice. A list of available functors classied from mature to unmaintained is updated here to guide this choice.
Examples
Let us give several example of the chain of created and accepted types.
Basic DEM model
Suppose we want to use the Law2_ScGeom_FrictPhys_CundallStrack constitutive law. We see that 1. the Ig2 functors must create ScGeom. If we have for instance spheres and boxes in the simulation, we will need functors accepting Sphere + Sphere and Box + Sphere combinations. We dont want interactions between boxes themselves (as a matter of fact, there is no such functor anyway). That gives us Ig2_Sphere_Sphere_ScGeom and Ig2_Box_Sphere_ScGeom. 2. the Ip2 functors should create FrictPhys. Looking at InteractionPhysicsFunctors, there is only Ip2_FrictMat_FrictMat_FrictPhys. That obliges us to use FrictMat for particles. The result will be therefore:
InteractionLoop( [Ig2_Sphere_Sphere_ScGeom(),Ig2_Box_Sphere_ScGeom()], [Ip2_FrictMat_FrictMat_FrictPhys()], [Law2_ScGeom_FrictPhys_CundallStrack()] )
Concrete model
In this case, our goal is to use the Law2_ScGeom_CpmPhys_Cpm constitutive law. We use spheres and facets in the simulation, which selects Ig2 functors accepting those types and producing ScGeom: Ig2_Sphere_Sphere_ScGeom and Ig2_Facet_Sphere_ScGeom. We have to use Material which can be used for creating CpmPhys. We nd that CpmPhys is only created by Ip2_CpmMat_CpmMat_CpmPhys, which determines the choice of CpmMat for all particles. Therefore, we will use:
52
InteractionLoop( [Ig2_Sphere_Sphere_ScGeom(),Ig2_Facet_Sphere_ScGeom()], [Ip2_CpmMat_CpmMat_CpmPhys()], [Law2_ScGeom_CpmPhys_Cpm()] )
3.1.6 Imposing conditions

In most simulations, it is not desired that all particles oat freely in space. There are several ways of imposing boundary conditions that block movement of all or some particles with regard to global space. Motion constraints Body.dynamic determines whether a body will be moved by NewtonIntegrator; it is mandatory for bodies with zero mass, where applying non-zero force would result in innite displacement. Facets are case in the point: utils.facet makes them non-dynamic by default, as they have zero volume and zero mass (this can be changed, by passing dynamic=True to utils.facet or setting Body.dynamic; setting State.mass to a non-zero value must be done as well). The same is true for utils.wall. Making sphere non-dynamic is achieved simply by:
utils.sphere([x,y,z],radius,dynamic=False)
State.blockedDOFs permits selective blocking of any of 6 degrees of freedom in global space. For instance, a sphere can be made to move only in the xy plane by saying:
Yade [2]: O.bodies.append(utils.sphere((0,0,0),1)) Out[2]: 0 Yade [3]: O.bodies[0].state.blockedDOFs='zXY'
In contrast to Body.dynamic, blockedDOFs will only block forces (and acceleration) in that direction being eective; if you prescribed linear or angular velocity, they will be applied regardless of blockedDOFs. It also implies that if the velocity is not zero when degrees of freedom are blocked the body will keep moving at the velocity it has at the time of blocking. It might be desirable to constrain motion of some particles constructed from a generated sphere packing, following some condition, such as being at the bottom of a specimen; this can be done by looping over all bodies with a conditional:
for b in O.bodies: # block all particles with z coord below .5: if b.state.pos[2]<.5: b.dynamic=False
Arbitrary spatial predicates introduced above can be expoited here as well:

from yade import pack pred=pack.inAlignedBox(lowerCorner,upperCorner) for b in O.bodies: if b.shape.name!=Sphere: continue # skip non-spheres # ask the predicate if we are inside if pred(b.state.pos,b.shape.radius): b.dynamic=False
Imposing motion and forces If a degree of freedom is blocked and a velocity is assigned along that direction (translational or rotational velocity), then the body will move at constant velocity. This is the simpler and recommended method to impose the motion of a body. This, for instance, will result in a constant velocity along x 3.1. Scene construction 53
O.bodies.append(utils.sphere((0,0,0),1)) O.bodies[0].state.blockedDOFs='x' O.bodies[0].state.vel=(10,0,0)
Conversely, modifying the position directly is likely to break Yades algorithms, especially those related to collision detection and contact laws, as they are based oon bodies velocities. Therefore, unless you really know what you are doing, dont do that for imposing a motion:
O.bodies.append(utils.sphere((0,0,0),1)) O.bodies[0].state.blockedDOFs='x' O.bodies[0].state.pos=10*O.dt #REALLY BAD! Don't assign position
Applying a force or a torque on a body is done via functions of the ForceContainer. It is as simple as this:
O.forces.addF(0,(1,0,0)) #applies for one step
By default, the force applies for one time step only, and is resetted at the beginning of each step. For this reason, imposing a force at the begining of one step will have no eect at all, since it will be immediatly resetted. The only way is to place a PyRunner inside the simulation loop. Applying the force permanently is possible with an optional argument (in this case it does not matter if the command comes at the begining of the time step):
O.forces.addF(0,(1,0,0),permanent=True) #applies permanently
The force will persist across iterations, until it is overwritten by another call to O.forces.addF(id,f,True) or erased by O.forces.reset(resetAll=True). The permanent force on a body can be checked with O.forces.permF(id). Boundary controllers Engines deriving from BoundaryController impose boundary conditions during simulation, either directly, or by inuencing several bodies. You are referred to their individual documentation for details, though you might nd interesting in particular UniaxialStrainer for applying strain along one axis at constant rate; useful for plotting strain-stress diagrams for uniaxial loading case. See examples/concrete/uniax.py for an example. TriaxialStressController which applies prescribed stress/strain along 3 perpendicular axes on cuboid-shaped packing using 6 walls (Box objects) (ThreeDTriaxialEngine is generalized such that it allows independent value of stress along each axis) PeriTriaxController for applying stress/strain along 3 axes independently, for simulations using periodic boundary conditions (Cell) Field appliers Engines deriving from FieldApplier acting on all particles. The one most used is GravityEngine applying uniform acceleration eld. Partial engines Engines deriving from PartialEngine dene the ids attribute determining bodies which will be aected. Several of them warrant explicit mention here: TranslationEngine and RotationEngine for applying constant speed linear and rotational motion on subscribers. ForceEngine and TorqueEngine applying given values of force/torque on subscribed bodies at every step.
54
StepDisplacer for applying generalized displacement delta at every timestep; designed for precise control of motion when testing constitutive laws on 2 particles. The real value of partial engines is if you need to prescribe complex types of force or displacement elds. For moving a body at constant velocity or for imposing a single force, the methods explained in Imposing motion and forces are much simpler. There are several interpolating engines (InterpolatingDirectedForceEngine for applying force with varying magnitude, InterpolatingSpiralEngine for applying spiral displacement with varying angular velocity and possibly others); writing a new interpolating engine is rather simple using examples of those that already exist.
3.1.7 Convenience features

Labeling things Engines and functors can dene that label attribute. Whenever the O.engines sequence is modied, python variables of those names are created/update; since it happens in the __builtins__ namespaces, these names are immediately accessible from anywhere. This was used in Creating interactions to change interaction radius in multiple functors at once. Warning: Make sure you do not use label that will overwrite (or shadow) an object that you already use under that variable name. Take care not to use syntactically wrong names, such as er*452 or my engine; only variable names permissible in Python can be used.
Simulation tags Omega.tags is a dictionary (it behaves like a dictionary, although the implementation in c++ is dierent) mapping keys to labels. Contrary to regular python dictionaries that you could create, O.tags is saved and loaded with simulation; O.tags has some values pre-initialized. After Yade startup, O.tags contains the following:
Yade [2]: dict(O.tags) # convert to real dictionary Out[2]: {'author': 'BuildBot~system~user~(buildbot@r0calcul5)', 'd.id': '20130823T161357p31978', 'id': '20130823T161357p31978', 'id.d': '20130823T161357p31978', 'isoTime': '20130823T161357'}
author Real name, username and machine as obtained from your system at simulation creation id Unique identier of this Yade instance (or of the instance which created a loaded simulation). It is composed of date, time and process number. Useful if you run simulations in parallel and want to avoid overwriting each others outputs; embed O.tags['id'] in output lenames (either as directory name, or as part of the les name itself) to avoid it. This is explained in Separating output les from jobs in detail. isoTime Time when simulation was created (with second resolution). d.id, id.d Simulation description and id joined by period (and vice-versa). Description is used in batch jobs; in non-batch jobs, these tags are identical to id. You can add your own tags by simply assigning value, with the restriction that the left-hand side object must be a string and must not contain =.
Yade [1]: O.tags['anythingThat I lik3']='whatever' Yade [2]: O.tags['anythingThat I lik3'] Out[2]: 'whatever'
55
Saving python variables Python variable lifetime is limited; in particular, if you save simulation, variables will be lost after reloading. Yade provides limited support for data persistence for this reason (internally, it uses special values of O.tags). The functions in question are utils.saveVars and utils.loadVars. utils.saveVars takes dictionary (variable names and their values) and a mark (identication string for the variable set); it saves the dictionary inside the simulation. These variables can be re-created (after the simulation was loaded from a XML le, for instance) in the yade.params.mark namespace by calling utils.loadVars with the same identication mark :
Yade [1]: a=45; b=pi/3 Yade [2]: utils.saveVars('ab',a=a,b=b) # save simulation (we could save to disk just as well) Yade [2]: O.saveTmp() Yade [4]: O.loadTmp() Yade [5]: utils.loadVars('ab') Yade [6]: yade.params.ab.a Out[6]: 45 # import like this Yade [7]: from yade.params import ab Yade [8]: ab.a, ab.b Out[8]: (45, 1.0471975511965976) # also possible Yade [9]: from yade.params import * Yade [10]: ab.a, ab.b Out[10]: (45, 1.0471975511965976)
Enumeration of variables can be tedious if they are many; creating local scope (which is a function denition in Python, for instance) can help:
def setGeomVars(): radius=a*4 thickness=22 p_t=4/3*pi dim=Vector3(1.23,2.2,3) # # define as much as you want here # it all appears in locals() (and nothing else does) # utils.saveVars('geom',loadNow=True,**locals()) setGeomVars() from yade.params.geom import * # use the variables now
Note: Only types that can be pickled can be passed to utils.saveVars.
56
3.2 Controlling simulation

3.2.1 Tracking variables
Running python code A special engine PyRunner can be used to periodically call python code, specied via the command parameter. Periodicity can be controlled by specifying computation time (realPeriod), virutal time (virtPeriod) or iteration number (iterPeriod). For instance, to print kinetic energy (using utils.kineticEnergy) every 5 seconds, this engine will be put PyRunner(command=print kinetic energy,utils.kineticEnergy(),realPeriod=5) For running more complex commands, it is convenient to dene an external function and only call it from within the engine. Since the command is run in the scripts namespace, functions dened within scripts can be called. Let us print information on interaction between bodies 0 and 1 periodically:
def intrInfo(id1,id2): try: i=O.interactions[id1,id2] # assuming it is a CpmPhys instance print id1,id2,i.phys.sigmaN except: # in case the interaction doesn't exist (yet?) print "No interaction between",id1,id2 O.engines=[..., PyRunner(command="intrInfo(0,1)",realPeriod=5) ]
More useful examples will be given below. The plot module provides simple interface and storage for tracking various data. Although originally conceived for plotting only, it is widely used for tracking variables in general. The data are in plot.data dictionary, which maps variable names to list of their values; the plot.addData function is used to add them.
Yade [2]: from yade import plot Yade [3]: plot.data Out[3]: {'eps': [0.0001, 0.001, nan], 'force': [nan, nan, 1000.0], 'sigma': [12, nan, nan]} Yade [4]: plot.addData(sigma=12,eps=1e-4) # not adding sigma will add a NaN automatically # this assures all variables have the same number of records Yade [5]: plot.addData(eps=1e-3) # adds NaNs to already existing sigma and eps columns Yade [6]: plot.addData(force=1e3) Yade [7]: plot.data Out[7]: {'eps': [0.0001, 0.001, nan, 0.0001, 0.001, nan], 'force': [nan, nan, 1000.0, nan, nan, 1000.0], 'sigma': [12, nan, nan, 12, nan, nan]} # retrieve only one column Yade [8]: plot.data['eps'] Out[8]: [0.0001, 0.001, nan, 0.0001, 0.001, nan]
3.2. Controlling simulation
57
# get maximum eps Yade [9]: max(plot.data['eps']) Out[9]: 0.001
New record is added to all columns at every time plot.addData is called; this assures that lines in dierent columns always match. The special value nan or NaN (Not a Number) is inserted to mark the record invalid. Note: It is not possible to have two columns with the same name, since data are stored as a dictionary. To record data periodically, use PyRunner. This will record the z coordinate and velocity of body #1, iteration number and simulation time (every 20 iterations):
O.engines=O.engines+[PyRunner(command='myAddData()', iterPeriod=20)] from yade import plot def myAddData(): b=O.bodies[1] plot.addData(z1=b.state.pos[2], v1=b.state.vel.norm(), i=O.iter, t=O.time)
Note: Arbitrary string can be used as column label for plot.data. If it cannot be used as keyword name for plot.addData (since it is a python keyword (for), or has spaces inside (my funny column), you can pass dictionary to plot.addData instead:
plot.addData(z=b.state.pos[2],**{'my funny column':b.state.vel.norm()})
An exception are columns having leading of trailing whitespaces. They are handled specially in plot.plots and should not be used (see below). Labels can be conveniently used to access engines in the myAddData function:
O.engines=[..., UniaxialStrainer(...,label='strainer') ] def myAddData(): plot.addData(sigma=strainer.stress,eps=strainer.strain)
In that case, naturally, the labeled object must dene attributes which are used (UniaxialStrainer.strain and UniaxialStrainer.stress in this case). Plotting variables Above, we explained how to track variables by storing them using plot.addData. These data can be readily used for plotting. Yade provides a simple, quick to use, plotting in the plot module. Naturally, since direct access to underlying data is possible via plot.data, these data can be processed in any way. The plot.plots dictionary is a simple specication of plots. Keys are x-axis variable, and values are tuple of y-axis variables, given as strings that were used for plot.addData; each entry in the dictionary represents a separate gure:
plot.plots={ 'i':('t',), # plot t(i) 't':('z1','v1') # z1(t) and v1(t) }
Actual plot using data in plot.data and plot specication of plot.plots can be triggered by invoking the plot.plot function.
58
Live updates of plots
Yade features live-updates of gures during calculations. It is controlled by following settings: plot.live - By setting yade.plot.live=True you can watch the plot being updated while the calculations run. Set to False otherwise. plot.liveInterval - This is the interval in seconds between the plot updates. plot.autozoom - When set to True the plot will be automatically rezoomed.
Controlling line properties
In this subsection let us use a basic complete script like examples/simple-scene/simple-scene-plot.py, which we will later modify to make the plots prettier. Line of interest from that le is, and generates a picture presented below:
plot.plots={'i':('t'),'t':('z_sph',None,('v_sph','go-'),'z_sph_half')}
2.0 1.9 1.8 z_sph 1.7 1.6 1.5 1.4 0.0
z_sph
v_sph z_sph_half
3.0 2.5 2.0 1.5 1.0 0.5
0.5
1.0
1.5
0.0 2.0
Figure 3.5: Figure generated by examples/simple-scene/simple-scene-plot.py. The line plots take an optional second string argument composed of a line color (eg. 'r', 'g' or 'b'), a line style (eg. '-', '-' or ':') and a line marker ('o', 's' or 'd'). A red dotted line with circle markers is created with ro: argument. For a listing of all options please have a look at http://matplotlib.sourceforge.net/api/pyplot_api.html#matplotlib.pyplot.plot For example using following plot.plots() command, will produce a following graph:
plot.plots={'i':(('t','xr:'),),'t':(('z_sph','r:'),None,('v_sph','g--'),('z_sph_half','b-.'))}
And this one will produce a following graph: 3.2. Controlling simulation 59
v_sph,z_sph_half
2.0 1.9 1.8 z_sph 1.7 1.6 1.5 1.4 0.0
z_sph
v_sph z_sph_half
3.0 2.5 2.0 1.5 1.0 0.5
0.5
1.0
1.5
0.0 2.0
Figure 3.6: Figure generated by changing parameters to plot.plots as above.
60
v_sph,z_sph_half
plot.plots={'i':(('t','xr:'),),'t':(('z_sph','Hr:'),None,('v_sph','+g--'),('z_sph_half','*b-.'))}
2.0 1.9 1.8 z_sph 1.7 1.6 1.5 1.4 0.0
z_sph
v_sph z_sph_half
3.0 2.5 2.0 1.5 1.0 0.5
0.5
1.0
1.5
0.0 2.0
Figure 3.7: Figure generated by changing parameters to plot.plots as above.
Note: You can learn more in matplotlib tutorial http://matplotlib.sourceforge.net/users/pyplot_tutorial.html and documentation http://matplotlib.sourceforge.net/users/pyplot_tutorial.html#controllingline-properties
Note: Please note that there is an extra , in 'i':(('t','xr:'),), otherwise the 'xr:' wouldnt be recognized as a line style parameter, but would be treated as an extra data to plot.
Controlling text labels
It is possible to use TeX syntax in plot labels. For example using following two lines in examples/simplescene/simple-scene-plot.py, will produce a following picture:
plot.plots={'i':(('t','xr:'),),'t':(('z_sph','r:'),None,('v_sph','g--'),('z_sph_half','b-.'))} plot.labels={'z_sph':'$z_{sph}$' , 'v_sph':'$v_{sph}$' , 'z_sph_half':'$z_{sph}/2$'}
Greek letters are simply a '$\alpha$', '$\beta$' etc. in those labels. To change the font style a following command could be used:
yade.plot.matplotlib.rc('mathtext', fontset='stixsans')
But this is not part of yade, but a part of matplotlib, and if you want something more complex you really should have a look at matplotlib users manual http://matplotlib.sourceforge.net/users/index.html 3.2. Controlling simulation 61
v_sph,z_sph_half
2.0 1.9 1.8

zsph
zsph
vsph zsph/2
3.0 2.5 2.0 1.5 1.0 0.5

vsph,zsph/2
1.7 1.6 1.5 1.4 0.0 0.5 1.0 1.5
0.0 2.0
Figure 3.8: Figure generated by examples/simple-scene/simple-scene-plot.py, with TeX labels.
62
Multiple gures
Since plot.plots is a dictionary, multiple entries with the same key (x-axis variable) would not be possible, since they overwrite each other:
Yade [1]: plot.plots={ ...: 'i':('t',), ...: 'i':('z1','v1') ...: } ...: Yade [2]: plot.plots Out[2]: {'i': ('z1', 'v1')}
You can, however, distinguish them by prepending/appending space to the x-axis variable, which will be removed automatically when looking for the variable in plot.data both x-axes will use the i column:
Yade [1]: plot.plots={ ...: 'i':('t',), ...: 'i ':('z1','v1') # note the space in 'i ' ...: } ...: Yade [2]: plot.plots Out[2]: {'i': ('t',), 'i ': ('z1', 'v1')}
Split y1 y2 axes
To avoid big range dierences on the y axis, it is possible to have left and right y axes separate (like axes x1y2 in gnuplot). This is achieved by inserting None to the plot specier; variables coming before will be plot normally (on the left y-axis), while those after will appear on the right:
plot.plots={'i':('z1',None,'v1')}
Exporting
Plots can be exported to external les for later post-processing via that plot.saveGnuplot function. Note that all data you added via plot.addData is saved - even data that you dont plot live during simulation. By editing the generated .gnuplot le you can plot any of the added Data afterwards. Data le is saved (compressed using bzip2) separately from the gnuplot le, so any other programs can be used to process them. In particular, the numpy.genfromtxt (documented here) can be useful to import those data back to python; the decompression happens automatically. The gnuplot le can be run through gnuplot to produce the gure; see plot.saveGnuplot documentation for details.
3.2.2 Stop conditions

For simulations with pre-determined number of steps, number of steps can be prescribed: # absolute iteration number O.stopAtIter=35466 O.run() O.wait() or
# number of iterations to run from now O.run(35466,True) # wait=True
causes the simulation to run 35466 iterations, then stopping.
63
Frequently, decisions have to be made based on evolution of the simulation itself, which is not yet known. In such case, a function checking some specic condition is called periodically; if the condition is satised, O.pause or other functions can be called to stop the stimulation. See documentation for Omega.run, Omega.pause, Omega.step, Omega.stopAtIter for details. For simulations that seek static equilibrium, the _utils.unbalancedForce can provide a useful metrics (see its documentation for details); for a desired value of 1e-2 or less, for instance, we can use:
def checkUnbalanced(): if utils.unbalancedForce<1e-2: O.pause() O.engines=O.engines+[PyRunner(command="checkUnbalanced()",iterPeriod=100)] # this would work as well, without the function defined apart: # PyRunner(command="if utils.unablancedForce<1e-2: O.pause()",iterPeriod=100) O.run(); O.wait() # will continue after O.pause() will have been called
Arbitrary functions can be periodically checked, and they can also use history of variables tracked via plot.addData. For example, this is a simplied version of damage control in examples/concrete/uniax.py; it stops when current stress is lower than half of the peak stress:
O.engines=[..., UniaxialStrainer=(...,label='strainer'), PyRunner(command='myAddData()',iterPeriod=100), PyRunner(command='stopIfDamaged()',iterPeriod=100) ] def myAddData(): plot.addData(t=O.time,eps=strainer.strain,sigma=strainer.stress) def stopIfDamaged(): currSig=plot.data['sigma'][-1] # last sigma value maxSig=max(plot.data['sigma']) # maximum sigma value # print something in any case, so that we know what is happening print plot.data['eps'][-1],currSig if currSig<.5*maxSig: print "Damaged, stopping" print 'gnuplot',plot.saveGnuplot(O.tags['id']) import sys sys.exit(0) O.run(); O.wait() # this place is never reached, since we call sys.exit(0) directly
Checkpoints Occasionally, it is useful to revert to simulation at some past point and continue from it with dierent parameters. For instance, tension/compression test will use the same initial state but load it in 2 dierent directions. Two functions, Omega.saveTmp and Omega.loadTmp are provided for this purpose; memory is used as storage medium, which means that saving is faster, and also that the simulation will disappear when Yade nishes.
O.saveTmp() # do something O.saveTmp('foo') O.loadTmp() # loads the first state O.loadTmp('foo') # loads the second state
64
Warning: O.loadTmp cannot be called from inside an engine, since before loading a simulation, the old one must nish the current iteration; it would lead to deadlock, since O.loadTmp would wait for the current iteration to nish, while the current iteration would be blocked on O.loadTmp. A special trick must be used: a separate function to be run after the current iteration is dened and is invoked from an independent thread launched only for that purpose:
O.engines=[...,PyRunner('myFunc()',iterPeriod=345)] def myFunc(): if someCondition: import thread # the () are arguments passed to the function thread.start_new_thread(afterIterFunc,()) def afterIterFunc(): O.pause(); O.wait() # wait till the iteration really finishes O.loadTmp() O.saveTmp() O.run()
3.2.3 Remote control

Yade can be controlled remotely over network. At yade startup, the following lines appear, among other messages:
TCP python prompt on localhost:9000, auth cookie `dcekyu' TCP info provider on localhost:21000
They inform about 2 ports on which connection of 2 dierent kind is accepted. Python prompt TCP python prompt is telnet server with authenticated connection, providing full python command-line. It listens on port 9000, or higher if already occupied (by another yade instance, for example). Using the authentication cookie, connection can be made using telnet:
$ telnet localhost 9000 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. Enter auth cookie: dcekyu __ __ ____ __ _____ ____ ____ \ \ / /_ _| _ \ ___ ___ / / |_ _/ ___| _ \ \ V / _` | | | |/ _ \ / _ \ / / | || | | |_) | | | (_| | |_| | __/ | (_) / / | || |___| __/ |_|\__,_|____/ \___| \___/_/ |_| \____|_| (connected from 127.0.0.1:40372) >>>
The python pseudo-prompt >>> lets you write commands to manipulate simulation in variety of ways as usual. Two things to notice: 1. The new python interpreter (>>>) lives in a namespace separate from Yade [1]: command-line. For your convenience, from yade import * is run in the new python instance rst, but local and global variables are not accessible (only builtins are). 2. The (fake) >>> interpreter does not have rich interactive feature of IPython, which handles the usual command-line Yade [1]:; therefore, you will have no command history, ? help and so on.
65
Note: By giving access to python interpreter, full control of the system (including reading users les) is possible. For this reason, connection are only allowed from localhost, not over network remotely. Of course you can log into the system via SSH over network to get remote access. Warning: Authentication cookie is trivial to crack via bruteforce attack. Although the listener stalls for 5 seconds after every failed login attempt (and disconnects), the cookie could be guessed by trial-and-error during very long simulations on a shared computer.
Info provider TCP Info provider listens at port 21000 (or higher) and returns some basic information about current simulation upon connection; the connection terminates immediately afterwards. The information is python dictionary represented as string (serialized) using standard pickle module. This functionality is used by the batch system (described below) to be informed about individual simulation progress and estimated times. If you want to access this information yourself, you can study core/main/yade-multi.in for details.
3.2.4 Batch queuing and execution (yade-batch)

Yade features light-weight system for running one simulation with dierent parameters; it handles assignment of parameter values to python variables in simulation script, scheduling jobs based on number of available and required cores and more. The whole batch consists of 2 les: simulation script regular Yade script, which calls utils.readParamsFromTable to obtain parameters from parameter table. In order to make the script runnable outside the batch, utils.readParamsFromTable takes default values of parameters, which might be overridden from the parameter table. utils.readParamsFromTable knows which parameter le and which line to read by inspecting the PARAM_TABLE environment variable, set by the batch system. parameter table simple text le, each line representing one parameter set. This le is read by utils.readParamsFromTable (using utils.TableParamReader class), called from simulation script, as explained above. For better reading of the text le you can make use of tabulators, these will be ignored by utils.readParamsFromTable. Parameters are not restricted to numerical values. You can also make use of strings by quoting them ( may also be used instead of ). This can be useful for nominal parameters. The batch can be run as
yade-batch parameters.table simulation.py
and it will intelligently run one simulation for each parameter table line. A minimal example is found in examples/test/batch/params.table and examples/test/batch/sim.py, another example follows. Example Suppose we want to study inuence of parameters density and initialVelocity on position of a sphere falling on xed box. We create parameter table like this:
description density initialVelocity # first non-empty line are column headings reference 2400 10 hi_v = 20 # = to use value from previous line lo_v = 5 # comments are allowed hi_rho 5000 10 # blank lines as well:
66
hi_rho_v hi_rh0_lo_v
= =
20 5
Each line give one combination of these 2 parameters and assigns (which is optional) a description of this simulation. In the simulation le, we read parameters from table, at the beginning of the script; each parameter has default value, which is used if not specied in the parameters le:
from yade import utils utils.readParamsFromTable( gravity=-9.81, density=2400, initialVelocity=20, noTableOk=True # use default values if not run in batch ) from yade.params.table import * print gravity, density, initialVelocity
after the call to utils.readParamsFromTable, corresponding python variables are created in the yade.params.table module and can be readily used in the script, e.g.
GravityEngine(gravity=(0,0,gravity))
Let us see what happens when running the batch:
$ yade-batch batch.table batch.py Will run `/usr/local/bin/yade-trunk' on `batch.py' with nice value 10, output redirected to `[email protected]', 4 job Will use table `batch.table', with available lines 2, 3, 4, 5, 6, 7. Will use lines 2 (reference), 3 (hi_v), 4 (lo_v), 5 (hi_rho), 6 (hi_rho_v), 7 (hi_rh0_lo_v). Master process pid 7030
These lines inform us about general batch information: nice level, log le names, how many cores will be used (4); table name, and line numbers that contain parameters; nally, which lines will be used; master PID is useful for killing (stopping) the whole batch with the kill command.
Job summary: #0 (reference/4): PARAM_TABLE=batch.table:2 DISPLAY= /usr/local/bin/yade-trunk --threads=4 --nice=10 -x batc #1 (hi_v/4): PARAM_TABLE=batch.table:3 DISPLAY= /usr/local/bin/yade-trunk --threads=4 --nice=10 -x batch.py #2 (lo_v/4): PARAM_TABLE=batch.table:4 DISPLAY= /usr/local/bin/yade-trunk --threads=4 --nice=10 -x batch.py #3 (hi_rho/4): PARAM_TABLE=batch.table:5 DISPLAY= /usr/local/bin/yade-trunk --threads=4 --nice=10 -x batch.p #4 (hi_rho_v/4): PARAM_TABLE=batch.table:6 DISPLAY= /usr/local/bin/yade-trunk --threads=4 --nice=10 -x batch #5 (hi_rh0_lo_v/4): PARAM_TABLE=batch.table:7 DISPLAY= /usr/local/bin/yade-trunk --threads=4 --nice=10 -x ba
displays all jobs with command-lines that will be run for each of them. At this moment, the batch starts to be run.
#0 #0 #1 #1 #2 #2 #3 #3 #4 #4 #5 #5 (reference/4) started on Tue Apr 13 13:59:32 2010 (reference/4) done (exit status 0), duration 00:00:01, log batch.reference.log (hi_v/4) started on Tue Apr 13 13:59:34 2010 (hi_v/4) done (exit status 0), duration 00:00:01, log batch.hi_v.log (lo_v/4) started on Tue Apr 13 13:59:35 2010 (lo_v/4) done (exit status 0), duration 00:00:01, log batch.lo_v.log (hi_rho/4) started on Tue Apr 13 13:59:37 2010 (hi_rho/4) done (exit status 0), duration 00:00:01, log batch.hi_rho.log (hi_rho_v/4) started on Tue Apr 13 13:59:38 2010 (hi_rho_v/4) done (exit status 0), duration 00:00:01, log batch.hi_rho_v.log (hi_rh0_lo_v/4) started on Tue Apr 13 13:59:40 2010 (hi_rh0_lo_v/4) done (exit status 0), duration 00:00:01, log batch.hi_rh0_lo_v.log
information about job status changes is being printed, until:

All jobs finished, total time Log files: 00:00:08
67
batch.reference.log batch.hi_v.log batch.lo_v.log batch.hi_rho.log batch.hi_rho_v.log batch.hi_rh0_lo_v.log Bye.
Separating output les from jobs As one might output data to external les during simulation (using classes such as VTKRecorder, it is important to name les in such way that they are not overwritten by next (or concurrent) job in the same batch. A special tag O.tags['id'] is provided for such purposes: it is comprised of date, time and PID, which makes it always unique (e.g. 20100413T144723p7625); additional advantage is that alphabetical order of the id tag is also chronological. To add the used parameterset or if set the description of the job you could add O.tags[params] to the lename. For smaller simulations, prepending all output le names with O.tags['id'] can be sucient:
utils.saveGnuplot(O.tags['id'])
For larger simulations, it is advisable to create separate directory of that name rst, putting all les inside afterwards:
os.mkdir(O.tags['id']) O.engines=[ # VTKRecorder(fileName=O.tags['id']+'/'+'vtk'), # ] # O.saveGnuplot(O.tags['id']+'/'+'graph1')
Controlling parallel compuation Default total number of available cores is determined from /proc/cpuinfo (provided by Linux kernel); in addition, if OMP_NUM_THREADS environment variable is set, minimum of these two is taken. The -j/--jobs option can be used to override this number. By default, each job uses all available cores for itself, which causes jobs to be eectively run in parallel. Number of cores per job can be globally changed via the --job-threads option. Table column named !OMP_NUM_THREADS (! prepended to column generally means to assign environment variable, rather than python variable) controls number of threads for each job separately, if it exists. If number of cores for a job exceeds total number of cores, warning is issued and only the total number of cores is used instead. Merging gnuplot from individual jobs Frequently, it is desirable to obtain single gure for all jobs in the batch, for comparison purposes. Somewhat heiristic way for this functionality is provided by the batch system. yade-batch must be run with the --gnuplot option, specifying some le name that will be used for the merged gure:
yade-trunk --gnuplot merged.gnuplot batch.table batch.py
Data are collected in usual way during the simulation (using plot.addData) and saved to gnuplot le via plot.saveGnuplot (it creates 2 les: gnuplot command le and compressed data le). The batch system scans, once the job is nished, log le for line of the form gnuplot [something]. Therefore, in order to print this magic line we put:
print 'gnuplot',plot.saveGnuplot(O.tags['id'])
and the end of the script (even after utils.waitIfBatch()) , which prints:
68
gnuplot 20100413T144723p7625.gnuplot
to the output (redirected to log le). This le itself contains single graph:
Figure 3.9: Figure from single job in the batch. At the end, the batch system knows about all gnuplot les and tries to merge them together, by assembling the merged.gnuplot le. HTTP overview While job is running, the batch system presents progress via simple HTTP server running at port 9080, which can be acessed from regular web browser by requesting the http://localhost:9080 URL. This page can be accessed remotely over network as well.
3.3 Postprocessing
3.3.1 3d rendering & videos
There are multiple ways to produce a video of simulation: 1. Capture screen output (the 3d rendering window) during the simulation there are tools available for that (such as Istanbul or RecordMyDesktop, which are also packaged for most Linux distributions). The output is what you see is what you get, with all the advantages and disadvantages. 2. Periodic frame snapshot using SnapshotEngine (see examples/bulldozer.py for a full example):
3.3. Postprocessing
69
Figure 3.10: Merged gure from all jobs in the batch. Note that labels are prepended by job description to make lines distinguishable.
70
Figure 3.11: Summary page available at port 9080 as batch is processed (updates every 5 seconds automatically). Possible job statuses are pending, running, done, failed.
3.3. Postprocessing
71
O.engines=[ #... SnapshotEngine(iterPeriod=100,fileBase='/tmp/bulldozer-',viewNo=0,label='snapshooter') ]
which will save numbered les like /tmp/bulldozer-0000.png. These les can be processed externally (with mencoder and similar tools) or directly with the utils.makeVideo:
utils.makeVideo(frameSpec,out,renameNotOverwrite=True,fps=24,kbps=6000,bps=None)
The video is encoded using the default mencoder codec (mpeg4). 3. Specialized post-processing tools, notably Paraview. This is described in more detail in the following section. Paraview
Saving data during the simulation
Paraview is based on the Visualization Toolkit, which denes formats for saving various types of data. One of them (with the .vtu extension) can be written by a special engine VTKRecorder. It is added to the simulation loop:
O.engines=[ # ... VTKRecorder(iterPeriod=100,recorders=['spheres','facets','colors'],fileName='/tmp/p1-') ]
iterPeriod determines how often to save simulation data (besides iterPeriod, you can also use virtPeriod or realPeriod). If the period is too high (and data are saved only few times), the video will have few frames. leName is the prex for les being saved. In this case, output les will be named /tmp/p1-spheres.0.vtu and /tmp/p1-facets.0.vtu, where the number is the number of iteration; many les are created, putting them in a separate directory is advisable. recorders determines what data to save (see the documentation) VTKExporter plays a similar role, with the dierence that it is more exible. It will save any user dened variable associated to the bodies.
Loading data into Paraview
All sets of les (spheres, facets, ) must be opened one-by-one in Paraview. The open dialogue automatically collapses numbered les in one, making it easy to select all of them: Click on the Apply button in the Object inspector sub-window to make loaded objects visible. You can see tree of displayed objects in the Pipeline browser: Rendering spherical particles. Glyphs Spheres will only appear as points. To make them look
as spheres, you have to add glyph to the p1-spheres.* item in the pipeline using the Then set (in the Object inspector) Glyph type to Sphere Radius to 1
icon.
Scale mode to Scalar (Scalar is set above to be the radii value saved in the le, therefore spheres with radius 1 will be scaled by their true radius) Set scale factor to 1 72 Chapter 3. Users manual
optionally uncheck Mask points and Random mode (they make some particles not to be rendered for performance reasons, controlled by the Maximum Number of Points) After clicking Apply, spheres will appear. They will be rendered over the original white points, which you can disable by clicking on the eye icon next to p1-spheres.* in the Pipeline browser. Rendering spherical particles. PointSprite Another opportunity to display spheres is an using PointSprite plugin. This technique requires much less RAM in comparison to Glyphs. Tools -> Manage Plugins PointSprite_Plugin -> Load selected -> Close Load VTU-les Representation -> Point Sprite Point Sprite -> Scale By -> radii Edit Radius Transfer Function -> Proportional -> Multiplier = 1.0 -> Close Facet transparency If you want to make facet objects transparent, select p1-facets.* in the Pipeline browser, then go to the Object inspector on the Display tab. Under Style, you can set the Opacity value to something smaller than 1. Animation You can move between frames (snapshots that were saved) via the Animation menu. After setting the view angle, zoom etc to your satisfaction, the animation can be saved with File/Save animation.
3.3.2 Micro-stress and micro-strain

It is sometimes useful to visualize a DEM simulation through equivalent strain elds or stress elds. This is possible with TesselationWrapper. This class handles the triangulation of spheres in a scene, build tesselation on request, and give access to computed quantities: volume, porosity and local deformation for each sphere. The denition of microstrain and microstress is at the scale of particle-centered subdomains shown below, as explained in [Catalano2013a] . Below is an output of the defToVtk function visualized with paraview (in this case Yades TesselationWrapper was used to process experimental data obtained on sand by Edward Ando at Grenoble University, 3SR lab.). The output is visualized with paraview, as explained in the previous section. Similar results can be generated from simulations:
3.3. Postprocessing
73
74
tt=TriaxialTest() tt.generate("test.yade") O.load("test.yade") O.run(100,True) TW=TesselationWrapper() TW.triangulate() #compute regular Delaunay triangulation, dont construct tesselation TW.computeVolumes() #will silently tesselate the packing, then compute volume of each Voronoi cell TW.volume(10) #get volume associated to sphere of id 10 TW.setState(0) #store current positions internaly for later use as the "0" state O.run(100,True) #make particles move a little (let's hope they will!) TW.setState(1) #store current positions internaly in the "1" (deformed) state #Now we can define strain by comparing states 0 and 1, and average them at the particles scale TW.defToVtk("strain.vtk")
Stress elds can be generated by combining the volume returned by TesselationWrapper to per-particle stress given by bodyStressTensors. Since the stress from bodyStressTensor implies a division by the volume Vb of the solid particle, one has to re-normalize it in order to obtain the micro-stress as dened k k k in [Catalano2013a] (equation 39 therein), i.e. k = k Vb /V where V is the volume assigned to particle k in the tesselation. For instance:
#"b" being a body TW=TesselationWrapper() TW.computeVolumes() s=bodyStressTensors() stress = s[b.id]**4.*pi/3.*b.shape.radius**3/TW.volume(b.id)
As any other value, the stress can be exported to a vtk le for display in Paraview using VTKExporter.
3.4 Python specialties and tricks

3.4.1 Importing Yade in other Python applications
Yade can be imported in other Python applications. To do so, you need somehow to make yade executable .py extended. The easiest way is to create a symbolic link, i.e. (suppose your Yade executable le is called yade-trunk and you want make it yadeimport.py):
3.4. Python specialties and tricks
75
$ cd /path/where/you/want/yadeimport $ ln -s /path/to/yade/executable/yade-trunk yadeimport.py
Then you need to make your yadeimport.py ndable by Python. You can export PYTHONPATH environment variable, or simply use sys.path directly in Python script:
import sys sys.path.append('/path/where/you/want/yadeimport') from yadeimport import * print Matrix3(1,2,3, 4,5,6, 7,8,9) print O.bodies # any other Yade code
3.5 Extending Yade

new particle shape new constitutive law
3.6 Troubleshooting
3.6.1 Crashes
It is possible that you encounter crash of Yade, i.e. Yade terminates with error message such as
76
Segmentation fault (core dumped)
without further explanation. Frequent causes of such conditions are program error in Yade itself; fatal condition in your particular simulation (such as impossible dispatch); problem with graphics card driver. Try to reproduce the error (run the same script) with debug-enabled version of Yade. Debugger will be automatically launched at crash, showing backtrace of the code (in this case, we triggered crash by hand):
Yade [1]: import os,signal Yade [2]: os.kill(os.getpid(),signal.SIGSEGV) SIGSEGV/SIGABRT handler called; gdb batch file is `/tmp/yade-YwtfRY/tmp-0' GNU gdb (GDB) 7.1-ubuntu Copyright (C) 2010 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-linux-gnu". For bug reporting instructions, please see: <http://www.gnu.org/software/gdb/bugs/>. [Thread debugging using libthread_db enabled] [New Thread 0x7f0fb1268710 (LWP 16471)] [New Thread 0x7f0fb29f2710 (LWP 16470)] [New Thread 0x7f0fb31f3710 (LWP 16469)]
What looks as cryptic message is valuable information for developers to locate source of the bug. In particular, there is (usually) line <signal handler called>; lines below it are source of the bug (at least very likely so):
Thread 1 (Thread 0x7f0fcee53700 (LWP 16465)): #0 0x00007f0fcd8f4f7d in __libc_waitpid (pid=16497, stat_loc=<value optimized out>, options=0) at ../sysdeps/un #1 0x00007f0fcd88c7e9 in do_system (line=<value optimized out>) at ../sysdeps/posix/system.c:149 #2 0x00007f0fcd88cb20 in __libc_system (line=<value optimized out>) at ../sysdeps/posix/system.c:190 #3 0x00007f0fcd0b4b23 in crashHandler (sig=11) at core/main/pyboot.cpp:45 #4 <signal handler called> #5 0x00007f0fcd87ed57 in kill () at ../sysdeps/unix/syscall-template.S:82 #6 0x000000000051336d in posix_kill (self=<value optimized out>, args=<value optimized out>) at ../Modules/posi #7 0x00000000004a7c5e in call_function (f=Frame 0x1c54620, for file <ipython console>, line 1, in <module> (), #8 PyEval_EvalFrameEx (f=Frame 0x1c54620, for file <ipython console>, line 1, in <module> (), throwflag=<value
If you think this might be error in Yade, le a bug report as explained below. Do not forget to attach full yade output from terminal, including startup messages and debugger output select with right mouse button, with middle button paste the bugreport to a le and attach it. Attach your simulation script as well.
3.6.2 Reporting bugs

Bugs are general name for defects (functionality shortcomings, misdocumentation, crashes) or feature requests. They are tracked at http://bugs.launchpad.net/yade. When reporting a new bug, be as specic as possible; state version of yade you use, system version and so on, as explained in the above section on crashes.
3.6. Troubleshooting
77
3.6.3 Getting help

Mailing lists Yade has two mailing-lists. Both are hosted at http://www.launchpad.net and before posting, you must register to Launchpad and subscribe to the list by adding yourself to team of the same name running the list. [email protected] is general help list for Yade users. Add yourself to yade-users team so that you can post messages. List archive is available. [email protected] is for discussions about Yade development; you must be member of yade-dev team to post. This list is archived as well. Read How To Ask Questions The Smart Way before posting. Do not forget to state what version of yade you use (shown when you start yade), what operating system (such as Ubuntu 10.04), and if you have done any local modications to source code. Questions and answers Launchpad provides interface for giving questions at https://answers.launchpad.net/yade/ which you can use instead of mailing lists; at the moment, it functionality somewhat overlaps with yade-users, but has the advantage of tracking whether a particular question has already been answered. Wiki http://www.yade-dem.org/wiki/ Private and/or paid support You might contact developers by their private mail (rather than by mailing list) if you do not want to disclose details on the mailing list. This is also a suitable method for proposing nancial reward for implementation of a substantial feature that is not yet in Yade typically, though, we will request this feature to be part of the public codebase once completed, so that the rest of the community can benet from it as well.
78
Chapter 4
Programmers manual
4.1 Build system
Yade uses cmake the cross-platform, open-source build system for managing the build process. It takes care of conguration, compilation and installation. CMake is used to control the software compilation process using simple platform and compiler independent conguration les. CMake generates native makeles and workspaces that can be used in the compiler environment of your choice.
4.1.1 Building
Yade source tree has the following structure (omiting, doc, examples and scripts which dont participate in the build process); we shall call each top-level component module:
core/ extra/ gui/ qt4/ py/ lib/ pkg/ common/ dem/ py/ ## core simulation building blocks ## miscillanea ## user interfaces ## graphical user interface based on qt3 and OpenGL ## python console interface (phased out) ## support libraries, not specific to simulations ## simulation-specific files ## generally useful classes ## classes for Discrete Element Method ## python modules
Header installation To allow exibility in source layout, CMAKE will copy (symlink) all headers into attened structure within the build directory. First 2 components of the original directory are joind by dash, deeper levels are discarded (in case of core and extra, only 1 level is used). The following table makes gives a few examples: Original header location core/Scene.hpp lib/base/Logging.hpp lib/serialization/Serializable.hpp pkg/dem/DataClass/SpherePack.hpp gui/qt3/QtGUI.hpp Included as <yade/core/Scene.hpp> <yade/lib-base/Logging.hpp> <yade/lib-serialization/Serializable.hpp> <yade/pkg-dem/SpherePack.hpp> <yade/gui-qt3/QtGUI.hpp> style of inclusion rather than
It is advised to use #include<yade/module/Class.hpp> #include"Class.hpp even if you are in the same directory.
79
Automatic compilation
In the pkg/ directory, situation is dierent. In order to maximally ease addition of modules to yade, all *.cpp les are automatically scanned by CMAKE and considered for compilation. Each le may contain multiple lines that declare features that are necessary for this le to be compiled:
YADE_REQUIRE_FEATURE(vtk); YADE_REQUIRE_FEATURE(gts);
This le will be compiled only if both vtk and gts features are enabled. Depending on current feature set, only selection of plugins will be compiled. It is possible to disable compilation of a le by requiring any non-existent feature, such as:
YADE_REQUIRE_FEATURE(temporarily disabled 345uiysdijkn);
The YADE_REQUIRE_FEATURE macro expands to nothing during actual compilation. Linking The order in which modules might depend on each other is given as follows: module lib core pkg gui py resulting shared library libyade-support.so libcore.so libplugins.so libQtGUI.so, libPythonUI.so (many les) dependencies can depend on external libraries, may not depend on any other part of Yade. yade-support; may depend on external libraries. core, yade-support lib, core, pkg lib, core, pkg, external
4.2 Development tools

4.2.1 Integrated Development Environment and other tools
A frequently used IDE is Kdevelop. We recommend using this software for navigating in the sources, compiling and debugging. Other useful tools for debugging and proling are Valgrind and KCachegrind. A series a wiki pages are dedicated to these tools in the development section of the wiki.
4.2.2 Hosting and versioning

The Yade project is kindly hosted at launchpad, which is used for source code, bug tracking, planning, package downloads and more. The versioning software used is GIT, for which a short tutorial can be found in a Yades wiki pages. GIT is a distributed revision control system. It is available packaged for all major linux distributions. The source code is hosted on GitHub , which is periodically imported to Launchpad for building PPApackages. The repository can be http-browsed.
4.2.3 Build robot

A build robot hosted at 3SR lab. is tracking souce code changes. Each time a change in the source code is commited to the main development branch via GIT, the buildbot downloads and compiles the new version, and start a series of tests.
80
Chapter 4. Programmers manual
If a compilation error has been introduced, it will be notied to the yade-dev mailing list and to the commiter, thus helping to x problems quickly. If the compilation is successfull, the buildbot starts unit regression tests and check tests (see below) and report the results. If all tests are passed, a new version of the documentation is generated and uploaded to the website in html and pdf formats. As a consequence, those two links always point to the documentation (the one you are reading now) of the last successfull build, and the delay between commits and documentation updates are very short (minutes). The buildbot activity and logs can be browsed online.
4.2.4 Regression tests

Yade contains two types of regression tests, some are unit tests while others are testing more complex simulations. Altough both types can be considered regression tests, the usage is that we name the rst simply regression tests, while the latest are called check tests. Both series of tests can be ran at yade startup by passing the options test or check
yade --test yade --check
Unit regression tests Unit regression tests are testing the output of individual functors and engines in well dened conditions. They are dened in the folder py/tests/. The purpose of unit testing is to make sure that the behaviour of the most important classes remains correct during code development. Since they test classes one by one, unit tests cant detect problems coming from the interaction between dierent engines in a typical simulation. That is why check tests have been introduced. Check tests Check tests perform comparisons of simulation results between dierent versions of yade, as discussed here. They dier with regression tests in the sense that they simulate more complex situations and combinations of dierent engines, and usually dont have a mathematical proof (though there is no restriction on the latest). They compare the values obtained in version N with values obtained in a previous version or any other expected results. The reference values must be hardcoded in the script itself or in data les provided with the script. Check tests are based on regular yade scripts, so that users can easily commit their own scripts to trunk in order to get some automatized testing after commits from other developers. Since the check tests history will be mostly based on standard output generated by yade check, a meaningfull checkTest should include some print command telling if something went wrong. If the script itself fails for some reason and cant generate an output, the log will contain scriptName failure. If the script denes dierences on obtained and awaited data, it should print some useful information about the problem and increase the value of global variable resultStatus. After this occurs, the automatic test will stop the execution with error message. An example check test can be found in checkTestTriax.py. It shows results comparison, output, and how to dene the path to data les using checksPath. Users are encouraged to add their own scripts into the scripts/test/checks/ folder. Discussion of some specic checktests design in users question is welcome. Note that re-compiling is required before that added scripts can be launched by yade check (or direct changes have to be performed in lib subfolders). A check test should never need more than a few seconds to run. If your typical script needs more, try and reduce the number of element or the number of steps.
4.3 Conventions
The following rules that should be respected; documentation is treated separately. general 4.3. Conventions 81
C++ source les have .hpp and .cpp extensions (for headers and implementation, respectively). All header les should have the #pragma once multiple-inclusion guard. Try to avoid using namespace in header les. Use tabs for indentation. While this is merely visual in c++, it has semantic meaning in python; inadverently mixing tabs and spaces can result in syntax errors. capitalization style Types should be always capitalized. Use CamelCase for composed names (GlobalEngine). Underscores should be used only in special cases, such as functor names. Class data members and methods must not be capitalized, composed names should use use lowercased camelCase (glutSlices). The same applies for functions in python modules. Preprocessor macros are uppercase, separated by underscores; those that are used outside the core take (with exceptions) the form YADE_*, such as YADE_CLASS_BASE_DOC_* macro family . programming style Be defensive, if it has no signicant performance impact. Use assertions abundantly: they dont aect performance (in the optimized build) and make spotting error conditions much easier. Use YADE_CAST and YADE_PTR_CAST where you want type-check during debug builds, but fast casting in optimized build. Initialize all class variables in the default constructor. This avoids bugs that may manifest randomly and are dicult to x. Initializing with NaNs will help you nd otherwise unitialized variable. (This is taken care of by YADE_CLASS_BASE_DOC_* macro family macros for user classes)
4.3.1 Class naming

Although for historial reasons the naming scheme is not completely consistent, these rules should be obeyed especially when adding a new class. GlobalEngines and PartialEngines GlobalEngines should be named in a way suggesting that it is a performer of certain action (like ForceResetter, InsertionSortCollider, Recorder); if this is not appropriate, append the Engine to the characteristics (GravityEngine). PartialEngines have no special naming convention dierent from GlobalEngines. Dispatchers Names of all dispatchers end in Dispatcher. The name is composed of type it creates or, in case it doesnt create any objects, its main characteristics. Currently, the following dispatchers 1 are dened: dispatcher arity dispatch created functor functor types type type prex BoundDis1 Shape Bound BoundFunc- Bo1 patcher tor IGeomDis2 (symet2 Shape IGeom IGeomIg2 patcher ric) Functor IPhysDis2 (symet2 Material IPhys IPhysFuncIp2 patcher ric) tor LawDis2 (asyIGeom IPhys (none) LawFunctor Law2 patcher metric) Respective abstract functors for each dispatchers are BoundFunctor, IGeomFunctor, IPhysFunctor and LawFunctor. Functors Functor name is composed of 3 parts, separated by underscore.
1
Not considering OpenGL dispatchers, which might be replaced by regular virtual functions in the future.
82
1. prex, composed of abbreviated functor type and arity (see table above) 2. Types entering the dispatcher logic (1 for unary and 2 for binary functors) 3. Return type for functors that create instances, simple characteristics for functors that dont create instances. To give a few examples: Bo1_Sphere_Aabb is a BoundFunctor which is called for Sphere, creating an instance of Aabb. Ig2_Facet_Sphere_ScGeom is binary functor called for Facet and Sphere, creating and instace of ScGeom. Law2_ScGeom_CpmPhys_Cpm is binary functor (LawFunctor) called for types ScGeom (Geom) and CpmPhys.
4.3.2 Documentation
Documenting code properly is one of the most important aspects of sustained development. Read it again. Most code in research software like Yade is not only used, but also read, by developers or even by regular users. Therefore, when adding new class, always mention the following in the documentation: purpose details of the functionality, unless obvious (algorithms, internal logic) limitations (by design, by implementation), bugs bibliographical reference, if using non-trivial published algorithms (see below) references to other related classes hyperlinks to bugs, blueprints, wiki or mailing list about this particular feature. As much as it is meaningful, you should also update any other documentation aected provide a simple python script demonstrating the new functionality in scripts/test. Sphinx documentation Most c++ classes are wrapped in Python, which provides good introspection and interactive documentation (try writing Material? in the ipython prompt; or help(CpmState)). Syntax of documentation is ReST <http://docutils.sourceforge.net/rst.html__ (reStructuredText, see reStructuredText Primer). It is the same for c++ and python code. Documentation of c++ classes exposed to python is given as 3rd argument to YADE_CLASS_BASE_DOC_* macro family introduced below. Python classes/functions are documented using regular python docstrings. Besides explaining functionality, meaning and types of all arguments should also be documented. Short pieces of code might be very helpful. See the utils module for an example. In addition to standard ReST syntax, yade provides several shorthand macros: :yref: creates hyperlink to referenced term, for instance:
:yref:`CpmMat`
becomes CpmMat; link name and target can be dierent:

:yref:`Material used in the CPM model<CpmMat>`
yielding Material used in the CPM model. 4.3. Conventions 83
:ysrc: creates hyperlink to le within the source tree (to its latest version in the repository), for instance core/Cell.hpp. Just like with :yref:, alternate text can be used with
:ysrc:`Link text<target/file>`
like this. |ycomp| is used in attribute description for those that should not be provided by the used, but are auto-computed instead; |ycomp| expands to (auto-computed). |yupdate| marks attributes that are periodically update, being subset of the previous. |yupdate| expands to (auto-updated). $...$ delimits inline math expressions; they will be replaced by:
:math:`...`
and rendered via LaTeX. To write a single dollar sign, escape it with backslash \$. Displayed mathematics (standalone equations) can be inserted as explained in Math support in Sphinx. Bibliographical references As in any scientic documentation, references to publications are very important. To cite an article, add it to BibTeX le in doc/references.bib, using the BibTeX format. Please adhere to the following conventions: 1. Keep entries in the form Author2008 (Author is the rst author), Author2008b etc if multiple articles from one author; 2. Try to ll mandatory elds for given type of citation; 3. Do not use \'{i} funny escapes for accents, since they will not work with the HTML output; put everything in straight utf-8. In your docstring, the Author2008 article can be cited by [Author2008]_; for example:
According to [Allen1989]_, the integration scheme
will be rendered as According to [Allen1989], the integration scheme Separate class/function documentation Some c++ might have long or content-rich documentation, which is rather inconvenient to type in the c++ source itself as string literals. Yade provides a way to write documentation separately in py/_extraDocs.py le: it is executed after loading c++ plugins and can set __doc__ attribute of any object directly, overwriting docstring from c++. In such (exceptional) cases: 1. Provide at least a brief description of the class in the c++ code nevertheless, for people only reading the code. 2. Add notice saying This class is documented in detail in the py/_extraDocs.py le. 3. Add documentation to py/_etraDocs.py in this way:
module.YourClass.__doc__=''' This is the docstring for YourClass. Class, methods and functions can be documented this way. .. note:: It can use any syntax features you like. '''
84
Note: Boost::python embeds function signatures in the docstring (before the one provided by the user). Therefore, before creating separate documentation of your function, have a look at its __doc__ attribute and copy the rst line (and the blank lie afterwards) in the separate docstring. The rst line is then used to create the function signature (arguments and return value).
Internal c++ documentation doxygen was used for automatic generation of c++ code. Since user-visible classes are dened with sphinx now, it is not meaningful to use doxygen to generate overall documentation. However, take care to document well internal parts of code using regular comments, including public and private data members.
4.4 Support framework

Besides the framework provided by the c++ standard library (including STL), boost and other dependencies, yade provides its own specic services.
4.4.1 Pointers
Shared pointers Yade makes extensive use of shared pointers shared_ptr. 2 Although it probably has some performance impacts, it greatly simplies memory management, ownership management of c++ objects in python and so forth. To obtain raw pointer from a shared_ptr, use its get() method; raw pointers should be used in case the object will be used only for short time (during a function call, for instance) and not stored anywhere. Python denes thin wrappers for most c++ Yade classes (for all those registered with YADE_CLASS_BASE_DOC_* macro family and several others), which can be constructed from shared_ptr; in this way, Python reference counting blends with the shared_ptr reference counting model, preventing crashes due to python objects pointing to c++ objects that were destructed in the meantime. Typecasting Frequently, pointers have to be typecast; there is choice between static and dynamic casting. dynamic_cast (dynamic_pointer_cast for a shared_ptr) assures cast admissibility by checking runtime type of its argument and returns NULL if the cast is invalid; such check obviously costs time. Invalid cast is easily caught by checking whether the pointer is NULL or not; even if such check (e.g. assert) is absent, dereferencing NULL pointer is easily spotted from the stacktrace (debugger output) after crash. Moreover, shared_ptr checks that the pointer is non-NULL before dereferencing in debug build and aborts with Assertion px!=0 failed. if the check fails. static_cast is fast but potentially dangerous (static_pointer_cast for shared_ptr). Static cast will return non-NULL pointer even if types dont allow the cast (such as casting from State* to Material*); the consequence of such cast is interpreting garbage data as instance of the class cast to, leading very likely to invalid memory access (segmentation fault, crash for short). To have both speed and safety, Yade provides 2 macros: YADE_CAST expands to static_cast in optimized builds and to dynamic_cast in debug builds. YADE_PTR_CAST expands to static_pointer_cast in optimized builds and to dynamic_pointer_cast in debug builds.
2 Either boost::shared_ptr or tr1::shared_ptr is used, but it is always imported with the using statement so that unqualied shared_ptr can be used.
4.4. Support framework
85
4.4.2 Basic numerics

The oating point type to use in Yade Real, which is by default typedef for double.
3
Yade uses the Eigen library for computations. It provides classes for 2d and 3d vectors, quaternions and 3x3 matrices templated by number type; their specialization for the Real type are typedefed with the r sux, and occasionally useful integer types with the i sux: Vector2r, Vector2i Vector3r, Vector3i Quaternionr Matrix3r Yade additionally denes a class named Se3r, which contains spatial position (Vector3r Se3r::position) and orientation (Quaternionr Se3r::orientation), since they are frequently used one with another, and it is convenient to pass them as single parameter to functions. Eigen provides full rich linear algebra functionality. Some code rther uses the [cgal] library for computational geometry. In Python, basic numeric types are wrapped and imported from the minieigen module; the types drop the r type qualier at the end, the syntax is otherwise similar. Se3r is not wrapped at all, only converted automatically, rarely as it is needed, from/to a (Vector3,Quaternion) tuple/list.
# cross product Yade [2]: Vector3(1,2,3).cross(Vector3(0,0,1)) Out[2]: Vector3(2,-1,0) # construct quaternion from axis and angle Yade [3]: Quaternion(Vector3(0,0,1),pi/2) Out[3]: Quaternion((0,0,1),1.5707963267948966)
Note: Quaternions are internally stored as 4 numbers. Their usual human-readable representation is, however, (normalized) axis and angle of rotation around that axis, and it is also how they are input/output in Python. Raw internal values can be accessed using the [0] [3] element access (or .W(), .X(), .Y() and .Z() methods), in both c++ and Python.
4.4.3 Run-time type identication (RTTI)

Since serialization and dispatchers need extended type and inheritance information, which is not suciently provided by standard RTTI. Each yade class is therefore derived from Factorable and it must use macro to override its virtual functions providing this extended RTTI: YADE_CLASS_BASE_DOC(Foo,Bar Baz,"Docstring) creates the following virtual methods (mediated via the REGISTER_CLASS_AND_BASE macro, which is not user-visible and should not be used directly): std::string getClassName() returning class name (Foo) as string. (There is the typeid(instanceOrType).name() standard c++ construct, but the name returned is compilerdependent.) unsigned getBaseClassNumber() returning number of base classes (in this case, 2). std::string getBaseClassName(unsigned i=0) returning name of i -th base class (here, Bar for i=0 and Baz for i=1).
3 Historically, it was thought that Yade could be also run with single precision based on build-time parameter; it turned out however that the impact on numerical stability was such disastrous that this option is not available now. There is, however, QUAD_PRECISION parameter to scons, which will make Real a typedef for long double (extended precision; quad precision in the proper sense on IA64 processors); this option is experimental and is unlikely to be used in near future, though.
86
Warning: RTTI relies on virtual functions; in order for virtual functions to work, at least one virtual method must be present in the implementation (.cpp) le. Otherwise, virtual method table (vtable) will not be generated for this class by the compiler, preventing virtual methods from functioning properly. Some RTTI information can be accessed from python:
Yade [2]: yade.system.childClasses('Shape') Out[2]: set(['Box', 'Facet', 'Cylinder', 'ChainedCylinder', 'GridConnection', 'Sphere', 'Wall', 'Tetra', 'Clump', 'GridNode']) Yade [3]: Sphere().name ## getClassName() --------------------------------------------------------------------------AttributeError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 Sphere().name ## getClassName() AttributeError: 'Sphere' object has no attribute 'name'
4.4.4 Serialization
Serialization serves to save simulation to le and restore it later. This process has several necessary conditions: classes know which attributes (data members) they have and what are their names (as strings); creating class instances based solely on its name; knowing what classes are dened inside a particular shared library (plugin). This functionality is provided by 3 macros and 4 optional methods; details are provided below. Serializable::preLoad, Serializable::preSave, Serializable::postLoad, Serializable::postSave Prepare attributes before serialization (saving) or deserialization (loading) or process them after serialization or deserialization. See Attribute registration. YADE_CLASS_BASE_DOC_* Inside the class declaration (i.e. in the .hpp le within the class Foo { /* */}; block). See Attribute registration. Enumerate class attributes that should be saved and loaded; associate each attribute with its literal name, which can be used to retrieve it. See YADE_CLASS_BASE_DOC_* macro family . Additionally documents the class in python, adds methods for attribute access from python, and documents each attribute. REGISTER_SERIALIZABLE In header le, but after the class declaration block. See Class factory . Associate literal name of the class with functions that will create its new instance (ClassFactory). YADE_PLUGIN In the implementation .cpp le. See Plugin registration. Declare what classes are declared inside a particular plugin at time the plugin is being loaded (yade startup).
87
Attribute registration All (serializable) types in Yade are one of the following: Type deriving from Serializable, which provide information on how to serialize themselves via overriding the Serializable::registerAttributes method; it declares data members that should be serialzed along with their literal names, by which they are identied. This method then invokes registerAttributes of its base class (until Serializable itself is reached); in this way, derived classes properly serialize data of their base classes. This funcionality is hidden behind the macro YADE_CLASS_BASE_DOC_* macro family used in class declaration body (header le), which takes base class and list of attributes:
YADE_CLASS_BASE_DOC_ATTRS(ThisClass,BaseClass,"class documentation",((type1,attribute1,initValue1,,"Docume
Note that attributes are encodes in double parentheses, not separated by commas. Empty attribute list can be given simply by YADE_CLASS_BASE_DOC_ATTRS(ThisClass,BaseClass,"documentation",) (the last comma is mandatory), or by omiting ATTRS from macro name and last parameter altogether. Fundamental type: strings, various number types, booleans, Vector3r and others. Their handlers (serializers and deserializers) are dened in lib/serialization. Standard container of any serializable objects. Shared pointer to serializable object. Yade uses the excellent boost::serialization library internally for serialization of data. Note: YADE_CLASS_BASE_DOC_ATTRS also generates code for attribute access from python; this will be discussed later. Since this macro serves both purposes, the consequence is that attributes that are serialized can always be accessed from python. Yade also provides callback for before/after (de) serialization, virtual functions Serializable::preProcessAttributes and Serializable::postProcessAttributes, which receive one bool deserializing argument (true when deserializing, false when serializing). Their default implementation in Serializable doesnt do anything, but their typical use is: converting some non-serializable internal data structure of the class (such as multi-dimensional array, hash table, array of pointers) into a serializable one (pre-processing) and ll this nonserializable structure back after deserialization (post-processing); for instance, InteractionContainer uses these hooks to ask its concrete implementation to store its contents to a unied storage (vector<shared_ptr<Interaction> >) before serialization and to restore from it after deserialization. precomputing non-serialized attributes from the serialized values; e.g. Facet computes its (local) edge normals and edge lengths from vertices coordinates. Class factory Each serializable class must use REGISTER_SERIALIZABLE, which denes function to create that class by ClassFactory. ClassFactory is able to instantiate a class given its name (as string), which is necessary for deserialization. Although mostly used internally by the serialization framework, programmer can ask for a class instantiation using shared_ptr<Factorable> f=ClassFactory::instance().createShared("ClassName");, casting the returned shared_ptr<Factorable> to desired type afterwards. Serializable itself derives from Factorable, i.e. all serializable types are also factorable (It is possible that dierent mechanism will be in place if boost::serialization is used, though.)
88
Plugin registration Yade loads dynamic libraries containing all its functionality at startup. ClassFactory must be taught about classes each particular le provides. YADE_PLUGIN serves this purpose and, contrary to YADE_CLASS_BASE_DOC_* macro family, must be place in the implementation (.cpp) le. It simple enumerates classes that are provided by this le:
YADE_PLUGIN((ClassFoo)(ClassBar));
Note: You must use parentheses around the class name even if there is only one (preprocessor limitation): YADE_PLUGIN((classFoo));. If there is no class in this le, do not use this macro at all. Internally, this macro creates function registerThisPluginClasses_ declared specially as __attribute__((constructor)) (see GCC Function Attributes); this attributes makes the function being executed when the plugin is loaded via dlopen from ClassFactory::load(...). It registers all factorable classes from that le in the Class factory . Note: Classes that do not derive from Factorable, such as Shop or SpherePack, are not declared with YADE_PLUGIN.
This is an example of a serializable class header:

/*! Homogeneous gravity field; applies gravitymass force on all bodies. */ class GravityEngine: public GlobalEngine{ public: virtual void action(); // registering class and its base for the RTTI system YADE_CLASS_BASE_DOC_ATTRS(GravityEngine,GlobalEngine, // documentation visible from python and generated reference documentation "Homogeneous gravity field; applies gravitymass force on all bodies.", // enumerating attributes here, include documentation ((Vector3r,gravity,Vector3r::ZERO,"acceleration, zero by default [kgms]")) ); }; // registration function for ClassFactory REGISTER_SERIALIZABLE(GravityEngine);
and this is the implementation:

#include<yade/ pkg-common/ GravityEngine.hpp> #include<yade/ core/ Scene.hpp> // registering the plugin YADE_PLUGIN((GravityEngine)); void GravityEngine::action(){ /* do the work here */ }
We can create a mini-simulation (with only one GravityEngine):

Yade [1]: O.engines=[GravityEngine(gravity=Vector3(0,0,-9.81))] Yade [2]: O.save('abc.xml')
and the XML looks like this:

<?xml version="1.0" encoding="UTF-8" standalone="yes" ?> <!DOCTYPE boost_serialization> <boost_serialization signature="serialization::archive" version="9">
89
<scene class_id="0" tracking_level="0" version="1"> <px class_id="1" tracking_level="1" version="0" object_id="_0"> <Serializable class_id="2" tracking_level="1" version="0" object_id="_1"></Serializable> <dt>1e-08</dt> <iter>0</iter> <subStepping>0</subStepping> <subStep>-1</subStep> <time>0</time> <speed>0</speed> <stopAtIter>0</stopAtIter> <stopAtTime>0</stopAtTime> <isPeriodic>0</isPeriodic> <trackEnergy>0</trackEnergy> <doSort>0</doSort> <runInternalConsistencyChecks>1</runInternalConsistencyChecks> <selectedBody>-1</selectedBody> <flags>0</flags> <tags class_id="3" tracking_level="0" version="0"> <count>5</count> <item_version>0</item_version> <item>author=BuildBot~system~user~(buildbot@r0calcul5)</item> <item>isoTime=20130823T161353</item> <item>id=20130823T161353p31978</item> <item>d.id=20130823T161353p31978</item> <item>id.d=20130823T161353p31978</item> </tags> <engines class_id="4" tracking_level="0" version="0"> <count>1</count> <item_version>1</item_version> <item class_id="5" tracking_level="0" version="1"> <px class_id="7" class_name="GravityEngine" tracking_level="1" version="0" objec <FieldApplier class_id="8" tracking_level="1" version="0" object_id="_3" <GlobalEngine class_id="9" tracking_level="1" version="0" object <Engine class_id="6" tracking_level="1" version="0" obje <Serializable object_id="_6"></Serializable> <dead>0</dead> <ompThreads>-1</ompThreads> <label></label> </Engine> </GlobalEngine> </FieldApplier> <gravity class_id="10" tracking_level="0" version="0"> <x>0</x> <y>0</y> <z>-9.8100000000000005</z> </gravity> <mask>0</mask> <warnOnce>1</warnOnce> </px> </item> </engines> <_nextEngines> <count>0</count> <item_version>1</item_version> </_nextEngines> <bodies class_id="11" tracking_level="0" version="1"> <px class_id="12" tracking_level="1" version="0" object_id="_7"> <Serializable object_id="_8"></Serializable> <body class_id="13" tracking_level="0" version="0"> <count>0</count> <item_version>1</item_version> </body> </px>
90
</bodies> <interactions class_id="14" tracking_level="0" version="1"> <px class_id="15" tracking_level="1" version="0" object_id="_9"> <Serializable object_id="_10"></Serializable> <interaction class_id="16" tracking_level="0" version="0"> <count>0</count> <item_version>1</item_version> </interaction> <serializeSorted>0</serializeSorted> <dirty>1</dirty> </px> </interactions> <energy class_id="17" tracking_level="0" version="1"> <px class_id="18" tracking_level="1" version="0" object_id="_11"> <Serializable object_id="_12"></Serializable> <energies class_id="19" tracking_level="0" version="0"> <size>0</size> </energies> <names class_id="20" tracking_level="0" version="0"> <count>0</count> <item_version>0</item_version> </names> <resetStep> <count>0</count> </resetStep> </px> </energy> <materials class_id="22" tracking_level="0" version="0"> <count>0</count> <item_version>1</item_version> </materials> <bound class_id="23" tracking_level="0" version="1"> <px class_id="-1"></px> </bound> <cell class_id="25" tracking_level="0" version="1"> <px class_id="26" tracking_level="1" version="0" object_id="_13"> <Serializable object_id="_14"></Serializable> <trsf class_id="27" tracking_level="0" version="0"> <m00>1</m00> <m01>0</m01> <m02>0</m02> <m10>0</m10> <m11>1</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>1</m22> </trsf> <refHSize> <m00>1</m00> <m01>0</m01> <m02>0</m02> <m10>0</m10> <m11>1</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>1</m22> </refHSize> <hSize> <m00>1</m00> <m01>0</m01> <m02>0</m02>
91
<m10>0</m10> <m11>1</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>1</m22> </hSize> <prevHSize> <m00>1</m00> <m01>0</m01> <m02>0</m02> <m10>0</m10> <m11>1</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>1</m22> </prevHSize> <velGrad> <m00>0</m00> <m01>0</m01> <m02>0</m02> <m10>0</m10> <m11>0</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>0</m22> </velGrad> <nextVelGrad> <m00>0</m00> <m01>0</m01> <m02>0</m02> <m10>0</m10> <m11>0</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>0</m22> </nextVelGrad> <prevVelGrad> <m00>0</m00> <m01>0</m01> <m02>0</m02> <m10>0</m10> <m11>0</m11> <m12>0</m12> <m20>0</m20> <m21>0</m21> <m22>0</m22> </prevVelGrad> <homoDeform>1</homoDeform> <velGradChanged>0</velGradChanged> </px> </cell> <miscParams class_id="28" tracking_level="0" version="0"> <count>0</count> <item_version>1</item_version> </miscParams> <dispParams class_id="29" tracking_level="0" version="0"> <count>0</count> <item_version>1</item_version> </dispParams>
92
</px> </scene> </boost_serialization>
Warning: Since XML les closely reect implementation details of Yade, they will not be compatible between dierent versions. Use them only for short-term saving of scenes. Python is the high-level description Yade uses.
Python attribute access The macro YADE_CLASS_BASE_DOC_* macro family introduced above is (behind the scenes) also used to create functions for accessing attributes from Python. As already noted, set of serialized attributes and set of attributes accessible from Python are identical. Besides attribute access, these wrapper classes imitate also some functionality of regular python dictionaries:
Yade [1]: s=Sphere() Yade [2]: s.radius Out[2]: nan Yade [3]: s.radius=4. ## read-access
## write access
Yade [4]: s.dict().keys() ## show all available keys Out[4]: ['color', 'highlight', 'wire', 'radius'] Yade [5]: for k in s.dict().keys(): print s.dict()[k] ...: Vector3(1,1,1) False False 4.0 Yade [6]: s.dict()['radius'] Out[6]: 4.0 ## iterate over keys, print their values
## same as: 'radius' in s.keys()
Yade [7]: s.dict() ## show dictionary of both attributes and values Out[7]: {'color': Vector3(1,1,1), 'highlight': False, 'radius': 4.0, 'wire': False}
4.4.5 YADE_CLASS_BASE_DOC_* macro family

There is several macros that hide behind them the functionality of Sphinx documentation, Run-time type identication (RTTI), Attribute registration, Python attribute access, plus automatic attribute initialization and documentation. They are all dened as shorthands for base macro YADE_CLASS_BASE_DOC_ATTRS_INIT_CTOR_PY with some arguments left out. They must be placed in class declarations body (.hpp le):
#define YADE_CLASS_BASE_DOC(klass,base,doc) \ YADE_CLASS_BASE_DOC_ATTRS(klass,base,doc,) #define YADE_CLASS_BASE_DOC_ATTRS(klass,base,doc,attrs) \ YADE_CLASS_BASE_DOC_ATTRS_CTOR(klass,base,doc,attrs,) #define YADE_CLASS_BASE_DOC_ATTRS_CTOR(klass,base,doc,attrs,ctor) \ YADE_CLASS_BASE_DOC_ATTRS_CTOR_PY(klass,base,doc,attrs,ctor,) #define YADE_CLASS_BASE_DOC_ATTRS_CTOR_PY(klass,base,doc,attrs,ctor,py) \ YADE_CLASS_BASE_DOC_ATTRS_INIT_CTOR_PY(klass,base,doc,attrs,,ctor,py) #define YADE_CLASS_BASE_DOC_ATTRS_INIT_CTOR_PY(klass,base,doc,attrs,init,ctor,py) \ YADE_CLASS_BASE_DOC_ATTRS_DEPREC_INIT_CTOR_PY(klass,base,doc,attrs,,init,ctor,py)
Expected parameters are indicated by macro name components separated with underscores. Their meaning is as follows: 4.4. Support framework 93
klass (unquoted) name of this class (used for RTTI and python) base (unquoted) name of the base class (used for RTTI and python) doc docstring of this class, written in the ReST syntax. This docstring will appear in generated documentation (such as CpmMat). It can be as long as necessary, but sequences interpreted by c++ compiler must be properly escaped (therefore some backslashes must be doubled, like in = E:
":math:`\\sigma=\\epsilon E"
Use \n and \t for indentation inside the docstring. Hyperlink the documentation abundantly with yref (all references to other classes should be hyperlinks). See Sphinx documentation for syntax details. attrs Attribute must be written in the form of parethesized list:
((type1,attr1,initValue1,attrFlags,"Attribute 1 documentation")) ((type2,attr2,,,"Attribute 2 documentation")) // initValue and attrFlags unspecified
This will expand to 1. data members declaration in c++ (note that all attributes are public):
public: type1 attr1; type2 attr2;
2. Initializers of the default (argument-less) constructor, for attributes that have non-empty initValue:
Klass(): attr1(initValue1), attr2() { /* constructor body */ }
No initial value will be assigned for attribute of which initial value is left empty (as is for attr2 in the above example). Note that you still have to write the commas. 3. Registration of the attribute in the serialization system (unless disabled by attrFlags see below)
4. Registration of the attribute in python (unless disabled by attrFlags), so that it can be accesse The attribute is read-write by default, see attrFlags to change that. This attribute will carry the docstring provided, along with knowledge of the initial value. You can add text description to the default value using the comma operator of c++ and casting the char* to (void):
((Real,dmgTau,((void)"deactivated if negative",-1),,"Characteristic time for normal viscosity. [s]
leading to CpmMat::dmgTau. The attribute is registered via boost::python::add_property specifying return_by_value policy rather than return_internal_reference, which is the default when using def_readwrite. The reason is that we need to honor custom converters for those values; see note in Custom converters for details. Attribute ags By default, an attribute will be serialized and will be read-write from python. There is a number of ags that can be passed as the 4th argument (empty by default) to change that: Attr::noSave avoids serialization of the attribute (while still keeping its accessibility from Python) Attr::readonly makes the attribute read-only from Python Attr::triggerPostLoad will trigger call to postLoad function to handle attribute change after its value is set from Python; this is to ensure consistency of other precomputed data which depend on this value (such as Cell.trsf and such)
94
Attr::hidden will not expose the attribute to Python at all Attr::noResize will not permit changing size of the array from Python [not yet used] Flags can be combined as usual using bitwise disjunction | (such as Attr::noSave | Attr::readonly), though in such case the value should be parenthesized to avoid a warning with some compilers (g++ specically), i.e. (Attr::noSave | Attr::readonly). Currently, the ags logic handled at runtime; that means that even for attributes with Attr::noSave, their serialization template must be dened (although it will never be used). In the future, the implementation might be template-based, avoiding this necessity. deprec List of deprecated attribute names. The syntax is
((oldName1,newName1,"Explanation why renamed etc.")) ((oldName2,newName2,"! Explanation why removed and what to do instaed."))
This will make accessing oldName1 attribute from Python return value of newName, but displaying warning message about the attribute name change, displaying provided explanation. This happens whether the access is read or write. If the explanations rst character is ! (bang), the message will be displayed upon attribute access, but exception will be thrown immediately. Use this in cases where attribute is no longer meaningful or was not straightforwardsly replaced by another, but more complex adaptation of users script is needed. You still have to give newName2, although its value will never be used you can use any variable you like, but something must be given for syntax reasons). Warning: Due to compiler limitations, this feature only works if Yade is compiled with gcc >= 4.4. In the contrary case, deprecated attribute functionality is disabled, even if such attributes are declared. init Parethesized list of the form:
((attr3,value3)) ((attr4,value4))
which will be expanded to initializers in the default ctor:
Klass(): /* attributes declared with the attrs argument */ attr4(value4), attr5(value5) { /* constructor b
The purpose of this argument is to make it possible to initialize constants and references (which are not declared as attributes using this macro themselves, but separately), as that cannot be done in constructor body. This argument is rarely used, though. ctor will be put directly into the generated constructors body. Mostly used for calling createIndex(); in the constructor. Note: The code must not contain commas ouside parentheses (since preprocessor uses commas to separate macro arguments). If you need complex things at construction time, create a separate init() function and call it from the constructor instead. py will be appeneded directly after generated python code that registers the class and all its attributes. You can use it to access class methods from python, for instance, to override an existing attribute with the same name etc:
.def_readonly("omega",&CpmPhys::omega,"Damage internal variable") .def_readonly("Fn",&CpmPhys::Fn,"Magnitude of normal force.")
def_readonly will not work for custom types (such as std::vector), as it bypasses conversion registry; see Custom converters for details.
95
Special python constructors The Python wrapper automatically create constructor that takes keyword (named) arguments corresponding to instance attributes; those attributes are set to values provided in the constructor. In some cases, more exibility is desired (such as InteractionLoop, which takes 3 lists of functors). For such cases, you can override the function Serializable::pyHandleCustomCtorArgs, which can arbitrarily modify the new (already existing) instance. It should modify in-place arguments given to it, as they will be passed further down to the routine which sets attribute values. In such cases, you should document the constructor:
.. admonition:: Special constructor Constructs from lists of
which then appears in the documentation similar to InteractionLoop. Static attributes Some classes (such as OpenGL functors) are instantiated automatically; since we want their attributes to be persistent throughout the session, they are static. To expose class with static attributes, use the YADE_CLASS_BASE_DOC_STATICATTRS macro. Attribute syntax is the same as for YADE_CLASS_BASE_DOC_ATTRS:
class SomeClass: public BaseClass{ YADE_CLASS_BASE_DOC_STATICATTRS(SomeClass,BaseClass,"Documentation of SomeClass", ((Type1,attr1,default1,"doc for attr1")) ((Type2,attr2,default2,"doc for attr2")) ); };
additionally, you have to allocate memory for static data members in the .cpp le (otherwise, error about undened symbol will appear when the plugin is loaded): There is no way to expose class that has both static and non-static attributes using YADE_CLASS_BASE_* macros. You have to expose non-static attributes normally and wrap static attributes separately in the py parameter. Returning attribute by value or by reference When attribute is passed from c++ to python, it can be passed either as value: new python object representing the original c++ object is constructed, but not bound to it; changing the python object doesnt modify the c++ object, unless explicitly assigned back to it, where inverse conversion takes place and the c++ object is replaced. reference: only reference to the underlying c++ object is given back to python; modifying python object will make the c++ object modied automatically. The way of passing attributes given to YADE_CLASS_BASE_DOC_ATTRS in the attrs parameter is determined automatically in the following manner:
Vector3, Vector3i, Vector2, Vector2i, Matrix3 and Quaternion objects are passed by reference. For ins O.bodies[0].state.pos[0]=1.33 will assign correct value to x component of position, without changing the other ones.
Yade classes (all that use shared_ptr when declared in python: all classes deriving from Serializable O.engines[4].damping=.3 will change damping parameter on the original engine object, not on its copy.
All other types are passed by value. This includes, most importantly, sequence types declared in Cu O.engines[4]=NewtonIntegrator()
96
will not work as expected; it will replace 5th element of a copy of the sequence, and this change will not propagate back to c++.
4.4.6 Multiple dispatch

Multiple dispatch is generalization of virtual methods: a Dispatcher decides based on type(s) of its argument(s) which of its Functors to call. Numer of arguments (currently 1 or 2) determines arity of the dispatcher (and of the functor): unary or binary. For example:
InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()])
creates InsertionSortCollider, which internally contains Collider.boundDispatcher, a BoundDispatcher (a Dispatcher), with 2 functors; they receive Sphere or Facet instances and create Aabb. This code would look like this in c++:
shared_ptr<InsertionSortCollider> collider=(new InsertionSortCollider); collider->boundDispatcher->add(new Bo1_Sphere_Aabb()); collider->boundDispatcher->add(new Bo1_Facet_Aabb());
There are currenly 4 predened dispatchers (see dispatcher-names) and corresponding functor types. They are inherit from template instantiations of Dispatcher1D or Dispatcher2D (for functors, Functor1D or Functor2D). These templates themselves derive from DynlibDispatcher (for dispatchers) and FunctorWrapper (for functors). Example: IGeomDispatcher Lets take (the most complicated perhaps) IGeomDispatcher. IGeomFunctor, which is dispatched based on types of 2 Shape instances (a Functor), takes a number of arguments and returns bool. The functor call is always provided by its overridden Functor::go method; it always receives the dispatched instances as rst argument(s) (2 const shared_ptr<Shape>&) and a number of other arguments it needs:
class IGeomFunctor: public Functor2D< bool, TYPELIST_7(const shared_ptr<Shape>&, const shared_ptr<Shape>&, const State&, const State&, const Vector3r&, const bool&, const shared_ptr<Interaction>& ) > //return type // 1st class for dispatch // 2nd class for dispatch // other arguments passed to ::go // // // //
The dispatcher is declared as follows:

class IGeomDispatcher: public Dispatcher2D< Shape, // 1st class for dispatch Shape, // 2nd class for dispatch IGeomFunctor, // functor type bool, // return type of the functor // follow argument types for functor call // they must be exactly the same as types // given to the IGeomFunctor above. TYPELIST_7(const shared_ptr<Shape>&, const shared_ptr<Shape>&, const State&, const State&, const Vector3r&, const bool &,
97
const shared_ptr<Interaction>& ), // handle symetry automatically // (if the dispatcher receives Sphere+Facet, // the dispatcher might call functor for Facet+Sphere, // reversing the arguments) false > { /* */ }
Functor derived from IGeomFunctor must then override the ::go method with appropriate arguments (they must match exactly types given to TYPELIST_* macro); declare what types they should be dispatched for, and in what order if they are not the same.
class Ig2_Facet_Sphere_Dem3DofGeom: public IGeomFunctor{ public: // override the IGeomFunctor::go // (it is really inherited from FunctorWrapper template, // therefore not declare explicitly in the // IGeomFunctor declaration as such) // since dispatcher dispatches only for declared types // (or types derived from them), we can do // static_cast<Facet>(shape1) and static_cast<Sphere>(shape2) // in the ::go body, without worrying about types being wrong. virtual bool go( // objects for dispatch const shared_ptr<Shape>& shape1, const shared_ptr<Shape>& shape2, // other arguments const State& state1, const State& state2, const Vector3r& shift2, const bool& force, const shared_ptr<Interaction>& c ); /* */ // this declares the type we want to be dispatched for, matching // first 2 arguments to ::go and first 2 classes in TYPELIST_7 above // shape1 is a Facet and shape2 is a Sphere // (or vice versa, see lines below) FUNCTOR2D(Facet,Sphere); // declare how to swap the arguments // so that we can receive those as well DEFINE_FUNCTOR_ORDER_2D(Facet,Sphere); /* */ };
Dispatch resolution The dispatcher doesnt always have functors that exactly match the actual types it receives. In the same way as virtual methods, it tries to nd the closest match in such way that: 1. the actual instances are derived types of those the functor accepts, or exactly the accepted types; 2. sum of distances from actual to accepted types is sharp-minimized (each step up in the class hierarchy counts as 1) If no functor is able to accept given types (rst condition violated) or multiple functors have the same distance (in condition 2), an exception is thrown.
98
This resolution mechanism makes it possible, for instance, to have a hierarchy of Dem3DofGeom classes (for dierent combination of shapes: Dem3DofGeom_SphereSphere, Dem3DofGeom_FacetSphere, Dem3DofGeom_WallSphere), but only provide a LawFunctor accepting Dem3DofGeom, rather than having dierent laws for each shape combination. Note: Performance implications of dispatch resolution are relatively low. The dispatcher lookup is only done once, and uses fast lookup matrix (1D or 2D); then, the functor found for this type(s) is cached within the Interaction (or Body) instance. Thus, regular functor call costs the same as dereferencing pointer and calling virtual method. There is blueprint to avoid virtual function call as well.
Note: At the beginning, the dispatch matrix contains just entries exactly matching given functors. Only when necessary (by passing other types), appropriate entries are lled in as well.
Indexing dispatch types Classes entering the dispatch mechanism must provide for fast identication of themselves and of their parent class. 4 This is called class indexing and all such classes derive from Indexable. There are top-level Indexables (types that the dispatchers accept) and each derived class registers its index related to this top-level Indexable. Currently, there are: Top-level Indexable Shape Material IPhys IGeom used by BoundFunctor, IGeomDispatcher IPhysDispatcher LawDispatcher LawDispatcher
The top-level Indexable must use the REGISTER_INDEX_COUNTER macro, which sets up the machinery for identifying types of derived classes; they must then use the REGISTER_CLASS_INDEX macro and call createIndex() in their constructor. For instance, taking the Shape class (which is a top-level Indexable):
// derive from Indexable class Shape: public Serializable, public Indexable { // never call createIndex() in the top-level Indexable ctor! /* */ // allow index registration for classes deriving from ``Shape`` REGISTER_INDEX_COUNTER(Shape); };
Now, all derived classes (such as Sphere or Facet) use this:

class Sphere: public Shape{ /* */ YADE_CLASS_BASE_DOC_ATTRS_CTOR(Sphere,Shape,"docstring", ((Type1,attr1,default1,"docstring1")) /* */ , // this is the CTOR argument // important; assigns index to the class at runtime createIndex(); ); // register index for this class, and give name of the immediate parent class // (i.e. if there were a class deriving from Sphere, it would use // REGISTER_CLASS_INDEX(SpecialSphere,Sphere), // not REGISTER_CLASS_INDEX(SpecialSphere,Shape)!) REGISTER_CLASS_INDEX(Sphere,Shape); };
4 The functionality described in Run-time type identication (RTTI) serves a dierent purpose (serialization) and would hurt the performance here. For this reason, classes provide numbers (indices) in addition to strings.
99
At runtime, each class within the top-level Indexable hierarchy has its own unique numerical index. These indices serve to build the dispatch matrix for each dispatcher. Inspecting dispatch in python If there is a need to debug/study multiple dispatch, python provides convenient interface for this low-level functionality. We can inspect indices with the dispIndex property (note that the top-level indexable Shape has negative (invalid) class index; we purposively didnt call createIndex in its constructor):
Yade [1]: Sphere().dispIndex, Facet().dispIndex, Wall().dispIndex Out[1]: (1, 5, 9) Yade [2]: Shape().dispIndex Out[2]: -1 # top-level indexable
Dispatch hierarchy for a particular class can be shown with the dispHierarchy() function, returning list of class names: 0th element is the instance itself, last element is the top-level indexable (again, with invalid index); for instance:
Yade [1]: ScGeom().dispHierarchy() # parent class of all other ScGeom_ classes Out[1]: ['ScGeom', 'GenericSpheresContact', 'IGeom'] Yade [2]: ScGridCoGeom().dispHierarchy(), ScGeom6D().dispHierarchy(), CylScGeom().dispHierarchy() Out[2]: (['ScGridCoGeom', 'ScGeom6D', 'ScGeom', 'GenericSpheresContact', 'IGeom'], ['ScGeom6D', 'ScGeom', 'GenericSpheresContact', 'IGeom'], ['CylScGeom', 'ScGeom', 'GenericSpheresContact', 'IGeom']) Yade [3]: CylScGeom().dispHierarchy(names=False) Out[3]: [4, 1, 0, -1] # show numeric indices instead
Dispatchers can also be inspected, using the .dispMatrix() method:

Yade [1]: ig=IGeomDispatcher([ ...: Ig2_Sphere_Sphere_ScGeom(), ...: Ig2_Facet_Sphere_ScGeom(), ...: Ig2_Wall_Sphere_ScGeom() ...: ]) ...: Yade [2]: ig.dispMatrix() Out[2]: {('Facet', 'Sphere'): 'Ig2_Facet_Sphere_ScGeom', ('Sphere', 'Facet'): 'Ig2_Facet_Sphere_ScGeom', ('Sphere', 'Sphere'): 'Ig2_Sphere_Sphere_ScGeom', ('Sphere', 'Wall'): 'Ig2_Wall_Sphere_ScGeom', ('Wall', 'Sphere'): 'Ig2_Wall_Sphere_ScGeom'} Yade [3]: ig.dispMatrix(False) Out[3]: {(1, 1): 'Ig2_Sphere_Sphere_ScGeom', (1, 5): 'Ig2_Facet_Sphere_ScGeom', (1, 9): 'Ig2_Wall_Sphere_ScGeom', (5, 1): 'Ig2_Facet_Sphere_ScGeom', (9, 1): 'Ig2_Wall_Sphere_ScGeom'} # don't convert to class names
We can see that functors make use of symmetry (i.e. that Sphere+Wall are dispatched to the same functor as Wall+Sphere). Finally, dispatcher can be asked to return functor suitable for given argument(s):
100
Yade [1]: ld=LawDispatcher([Law2_ScGeom_CpmPhys_Cpm()]) Yade [2]: ld.dispMatrix() Out[2]: {('GenericSpheresContact', 'CpmPhys'): 'Law2_ScGeom_CpmPhys_Cpm'} # see how the entry for ScGridCoGeom will be filled after this request Yade [3]: ld.dispFunctor(ScGridCoGeom(),CpmPhys()) Out[3]: <Law2_ScGeom_CpmPhys_Cpm instance at 0x5da9930> Yade [4]: ld.dispMatrix() Out[4]: {('GenericSpheresContact', 'CpmPhys'): 'Law2_ScGeom_CpmPhys_Cpm', ('ScGridCoGeom', 'CpmPhys'): 'Law2_ScGeom_CpmPhys_Cpm'}
OpenGL functors OpenGL rendering is being done also by 1D functors (dispatched for the type to be rendered). Since it is sucient to have exactly one class for each rendered type, the functors are found automatically. Their base functor types are GlShapeFunctor, GlBoundFunctor, GlIGeomFunctor and so on. These classes register the type they render using the RENDERS macro:
class Gl1_Sphere: public GlShapeFunctor { public : virtual void go(const shared_ptr<Shape>&, const shared_ptr<State>&, bool wire, const GLViewInfo& ); RENDERS(Sphere); YADE_CLASS_BASE_DOC_STATICATTRS(Gl1_Sphere,GlShapeFunctor,"docstring", ((Type1,staticAttr1,informativeDefault,"docstring")) /* */ ); }; REGISTER_SERIALIZABLE(Gl1_Sphere);
You can list available functors of a particular type by querying child classes of the base functor:
Yade [2]: yade.system.childClasses('GlShapeFunctor') Out[2]: set(['Gl1_Sphere', 'Gl1_Tetra', 'Gl1_Wall', 'Gl1_GridConnection', 'Gl1_Box', 'Gl1_ChainedCylinder', 'Gl1_Cylinder', 'Gl1_Facet'])
Note: OpenGL functors may disappear in the future, being replaced by virtual functions of each class that can be rendered.
4.4.7 Parallel execution

Yade was originally not designed with parallel computation in mind, but rather with maximum exibility (for good or for bad). Parallel execution was added later; in order to not have to rewrite whole Yade from scratch, relatively non-instrusive way of parallelizing was used: OpenMP. OpenMP is standartized shared-memory parallel execution environment, where parallel sections are marked by special #pragma
101
in the code (which means that they can compile with compiler that doesnt support OpenMP) and a few functions to query/manipulate OpenMP runtime if necessary. There is parallelism at 3 levels: Computation, interaction (python, GUI) and rendering threads are separate. This is done via regular threads (boost::threads) and is not related to OpenMP. ParallelEngine can run multiple engine groups (which are themselves run serially) in parallel; it rarely nds use in regular simulations, but it could be used for example when coupling with an independent expensive computation:
ParallelEngine([ [Engine1(),Engine2()], [Engine3()] ])
# Engine1 will run before Engine2 # Engine3() will run in parallel with the group [Engine1(),Engine2()] # arbitrary number of groups can be used
Engine2 will be run after Engine1, but in parallel with Engine3. Warning: It is your reponsibility to avoid concurrent access to data when using ParallelEngine. Make sure you understand very well what the engines run in parallel do. Parallelism inside Engines. Some loops over bodies or interactions are parallelized (notably InteractionLoop and NewtonIntegrator, which are treated in detail later (FIXME: link)):
#pragma omp parallel for for(long id=0; id<size; id++){ const shared_ptr<Body>& b(scene->bodies[id]); /* */ }
Note: OpenMP requires loops over contiguous range of integers (OpenMP 3 also accepts containers with random-access iterators). If you consider running parallelized loop in your engine, always evalue its benets. OpenMP has some overhead fo creating threads and distributing workload, which is proportionally more expensive if the loop body execution is fast. The results are highly hardware-dependent (CPU caches, RAM controller). Maximum number of OpenMP threads is determined by the OMP_NUM_THREADS environment variable and is constant throughout the program run. Yade main program also sets this variable (before loading OpenMP libraries) if you use the -j/--threads option. It can be queried at runtime with the omp_get_max_threads function. At places which are susceptible of being accessed concurrently from multiple threads, Yade provides some mutual exclusion mechanisms, discussed elsewhere (FIXME): simultaneously writeable container for ForceContainer , mutex for Body::state.
4.4.8 Timing
Yade provides 2 services for measuring time spent in dierent pars of the code. One has the granularity of engine and can be enabled at runtime. The other one is ner, but requires adjusting and recompiling the code being measured. Per-engine timing The coarser timing works by merely accumulating numebr of invocations and time (with the precision of the clock_gettime function) spent in each engine, which can be then post-processed by associated 102 Chapter 4. Programmers manual
Python module yade.timing. There is a static bool variable controlling whether such measurements take place (disabled by default), which you can change
TimingInfo::enabled=True; O.timingEnabled=True // in c++ ## in python
After running the simulation, yade.timing.stats() function will show table with the results and percentages:
Yade [1]: TriaxialTest(numberOfGrains=100).load() Yade [2]: O.engines[0].label='firstEngine' Yade [3]: import yade.timing; Yade [4]: O.timingEnabled=True Yade [5]: yade.timing.reset() Yade [6]: O.run(50); O.wait() Yade [7]: yade.timing.stats() Name Count Time Rel. time ------------------------------------------------------------------------------------------------------"firstEngine" 50 87us 0.80% InsertionSortCollider 26 3069us 27.94% InteractionLoop 50 5603us 51.01% GlobalStiffnessTimeStepper 2 43us 0.40% TriaxialCompressionEngine 50 541us 4.93% TriaxialStateRecorder 3 191us 1.74% NewtonIntegrator 50 1447us 13.17% TOTAL 10984us 100.00% ## not necessary if used for the first time ## labeled engines will show by labels in the stats table
Exec count and time can be accessed and manipulated through Engine::timingInfo from c++ or Engine().execCount and Engine().execTime properties in Python. In-engine and in-functor timing Timing within engines (and functors) is based on TimingDeltas class. It is made for timing loops (functors loop is in their respective dispatcher) and stores cummulatively time dierences between checkpoints. Note: Fine timing with TimingDeltas will only work if timing is enabled globally (see previous section). The code would still run, but giving zero times and exec counts. 1. Engine::timingDeltas must point to an instance of TimingDeltas (preerably instantiate TimingDeltas in the constructor):
// header file class Law2_ScGeom_CpmPhys_Cpm: public LawFunctor { /* */ YADE_CLASS_BASE_DOC_ATTRS_CTOR(Law2_ScGeom_CpmPhys_Cpm,LawFunctor,"docstring", /* attrs */ , /* constructor */ timingDeltas=shared_ptr<TimingDeltas>(new TimingDeltas); // timingDeltas object is automatically ); // ... };
2. Inside the loop, start the timing by calling timingDeltas->start();
103
3. At places of interest, call timingDeltas->checkpoint("label"). The label is used only for postprocessing, data are stored based on the checkpoint position, not the label. Warning: Checkpoints must be always reached in the same order, otherwise the timing data will be garbage. Your code can still branch, but you have to put checkpoints to places which are in common.
void Law2_ScGeom_CpmPhys_Cpm::go(shared_ptr<IGeom>& _geom, shared_ptr<IPhys>& _phys, Interaction* I, Scene* scene) { timingDeltas->start(); // the point at which the first timing starts // prepare some variables etc here timingDeltas->checkpoint("setup"); // find geometrical data (deformations) here timingDeltas->checkpoint("geom"); // compute forces here timingDeltas->checkpoint("material"); // apply forces, cleanup here timingDeltas->checkpoint("rest"); }
4. Alternatively, you can compile Yade using -DCMAKE_CXX_FLAGS=-DUSE_TIMING_DELTA

void Law2_ScGeom_CpmPhys_Cpm::go(shared_ptr<IGeom>& _geom, shared_ptr<IPhys>& _phys, Interaction* I, Scene* scene) { TIMING_DELTAS_START(); // prepare some variables etc here TIMING_DELTAS_CHECKPOINT("setup") // find geometrical data (deformations) here TIMING_DELTAS_CHECKPOINT("geom") // compute forces here TIMING_DELTAS_CHECKPOINT("material") // apply forces, cleanup here TIMING_DELTAS_CHECKPOINT("rest") }
The output might look like this (note that functors are nested inside dispatchers and TimingDeltas inside their engine/functor):
Name Count Time Rel. time ------------------------------------------------------------------------------------ForceReseter 400 9449s 0.01% BoundDispatcher 400 1171770s 1.15% InsertionSortCollider 400 9433093s 9.24% IGeomDispatcher 400 15177607s 14.87% IPhysDispatcher 400 9518738s 9.33% LawDispatcher 400 64810867s 63.49% Law2_ScGeom_CpmPhys_Cpm setup 4926145 7649131s 15.25% geom 4926145 23216292s 46.28% material 4926145 8595686s 17.14% rest 4926145 10700007s 21.33% TOTAL 50161117s 100.00% NewtonIntegrator 400 1866816s 1.83% "strainer" 400 21589s 0.02% "plotDataCollector" 160 64284s 0.06% "damageChecker" 9 3272s 0.00%
104
TOTAL
102077490s
100.00%
Warning: Do not use TimingDeltas in parallel sections, results might not be meaningful. In particular, avoid timing functors inside InteractionLoop when running with multiple OpenMP threads. TimingDeltas data are accessible from Python as list of (label,*time*,*count*) tuples, one tuple representing each checkpoint:
deltas=someEngineOrFunctor.timingDeltas.data() deltas[0][0] # 0th checkpoint label deltas[0][1] # 0th checkpoint time in nanoseconds deltas[0][2] # 0th checkpoint execution count deltas[1][0] # 1st checkpoint label # deltas.reset()
Timing overhead The overhead of the coarser, per-engine timing, is very small. For simulations with at least several hundreds of elements, they are below the usual time variance (a few percent). The ner TimingDeltas timing can have major performance impact and should be only used during debugging and performance-tuning phase. The parts that are le-timed will take disproportionally longer time that the rest of engine; in the output presented above, LawDispatcher takes almost of total simulation time in average, but the number would be twice of thrice lower typically (note that each checkpoint was timed almost 5 million times in this particular case).
4.4.9 OpenGL Rendering

Yade provides 3d rendering based on QGLViewer. It is not meant to be full-featured rendering and post-processing, but rather a way to quickly check that scene is as intended or that simulation behaves sanely. Note: Although 3d rendering runs in a separate thread, it has performance impact on the computation itself, since interaction container requires mutual exclusion for interaction creation/deletion. The InteractionContainer::drawloopmutex is either held by the renderer (OpenGLRenderingEngine) or by the insertion/deletion routine. Warning: There are 2 possible causes of crash, which are not prevented because of serious performance penalty that would result: 1. access to BodyContainer, in particular deleting bodies from simulation; this is a rare operation, though. 2. deleting Interaction::phys or Interaction::geom. Renderable entities (Shape, State, Bound, IGeom, IPhys) have their associated OpenGL functors. An entity is rendered if 1. Rendering such entities is enabled by appropriate attribute in OpenGLRenderingEngine 2. Functor for that particular entity type is found via the dispatch mechanism. Gl1_* functors operating on Bodys attributes (Shape, State, Bound) are called with the OpenGL context translated and rotated according to State::pos and State::ori. Interaction functors work in global coordinates.
105
4.5 Simulation framework

Besides the support framework mentioned in the previous section, some functionality pertaining to simulation itself is also provided. There are special containers for storing bodies, interactions and (generalized) forces. Their internal functioning is normally opaque to the programmer, but should be understood as it can inuence performance.
4.5.1 Scene
Scene is the object containing the whole simulation. Although multiple scenes can be present in the memory, only one of them is active. Saving and loading (serializing and deserializing) the Scene object should make the simulation run from the point where it left o. Note: All Engines and functors have interally a Scene* scene pointer which is updated regularly by engine/functor callers; this ensures that the current scene can be accessed from within user code. For outside functions (such as those called from python, or static functions in Shop), you can use Omega::instance().getScene() to retrieve a shared_ptr<Scene> of the current scene.
4.5.2 Body container

Body container is linear storage of bodies. Each body in the simulation has its unique id, under which it must be found in the BodyContainer. Body that is not yet part of the simulation typically has id equal to invalid value Body::ID_NONE, and will have its id assigned upon insertion into the container. The requirements on BodyContainer are O(1) access to elements, linear-addressability (0n indexability), store shared_ptr, not objects themselves, no mutual exclusion for insertion/removal (this must be assured by the called, if desired), intelligent allocation of id for new bodies (tracking removed bodies), easy iteration over all bodies. Note: Currently, there is abstract class BodyContainer, from which derive concrete implementations; the initial idea was the ability to select at runtime which implementation to use (to nd one that performs the best for given simulation). This incurs the penalty of many virtual function calls, and will probably change in the future. All implementations of BodyContainer were removed in the meantime, except BodyVector (internally a vector<shared_ptr<Body> > plus a few methods around), which is the fastest.
Insertion/deletion Body insertion is typically used in FileGenerators:

shared_ptr<Body> body(new Body); // (body setup) scene->bodies->insert(body); // assigns the id
Bodies are deleted only rarely:

scene->bodies->erase(id);
Warning: Since mutual exclusion is not assured, never insert/erase bodies from parallel sections, unless you explicitly assure there will be no concurrent access.
106
Iteration The container can be iterated over using FOREACH macro (shorthand for BOOST_FOREACH):
FOREACH(const shared_ptr<Body>& b, *scene->bodies){ if(!b) continue; // skip deleted bodies /* do something here */ }
Note a few important things: 1. Always use const shared_ptr<Body>& (const reference); that avoids incrementing and decrementing the reference count on each shared_ptr. 2. Take care to skip NULL bodies (if(!b) continue): deleted bodies are deallocated from the container, but since body ids must be persistent, their place is simply held by an empty shared_ptr<Body>() object, which is implicitly convertible to false. In python, the BodyContainer wrapper also has iteration capabilities; for convenience (which is dierent from the c++ iterator), NULL bodies as silently skipped:
Yade [2]: O.bodies.append([Body(),Body(),Body()]) Out[2]: [0, 1, 2] Yade [3]: O.bodies.erase(1) Out[3]: True Yade [4]: [b.id for b in O.bodies] Out[4]: [0, 2]
In loops parallelized using OpenMP, the loop must traverse integer interval (rather than using iterators):
const long size=(long)bodies.size(); // store this value, since it doesn't change during the loop #pragma omp parallel for for(long _id=0; _id<size; _id++){ const shared_ptr<Body>& b(bodies[_id]); if(!b) continue; /* */ }
4.5.3 InteractionContainer
Interactions are stored in special container, and each interaction must be uniquely identied by pair of ids (id1,id2). O(1) access to elements, linear-addressability (0n indexability), store shared_ptr, not objects themselves, mutual exclusion for insertion/removal, easy iteration over all interactions, addressing symmetry, i.e. interaction(id1,id2)interaction(id2,id1) Note: As with BodyContainer, there is abstract class InteractionContainer, and then its concrete implementations. Currently, only InteractionVecMap implementation is used and all the other were removed. Therefore, the abstract InteractionContainer class may disappear in the future, to avoid unnecessary virtual calls. Further, there is a blueprint for storing interactions inside bodies, as that would give extra advantage of quickly getting all interactions of one particular body (currently, this necessitates loop over all interactions); in that case, InteractionContainer would disappear.
4.5. Simulation framework
107
Insert/erase Creating new interactions and deleting them is delicate topic, since many eleents of simulation must be synchronized; the exact workow is described in Handling interactions. You will almost certainly never need to insert/delete an interaction manually from the container; if you do, consider designing your code dierently.
// both insertion and erase are internally protected by a mutex, // and can be done from parallel sections safely scene->interactions->insert(shared_ptr<Interaction>(new Interactions(id1,id2))); scene->interactions->erase(id1,id2);
Iteration As with BodyContainer, iteration over interactions should use the FOREACH macro:
FOREACH(const shared_ptr<Interaction>& i, *scene->interactions){ if(!i->isReal()) continue; /* */ }
Again, note the usage const reference for i. The check if(!i->isReal()) lters away interactions that exist only potentially, i.e. there is only Bound overlap of the two bodies, but not (yet) overlap of bodies themselves. The i->isReal() function is equivalent to i->geom && i->phys. Details are again explained in Handling interactions. In some cases, such as OpenMP-loops requiring integral index (OpenMP >= 3.0 allows parallelization using random-access iterator as well), you need to iterate over interaction indices instead:
inr nIntr=(int)scene->interactions->size(); // hoist container size #pragma omp parallel for for(int j=0; j<nIntr, j++){ const shared_ptr<Interaction>& i(scene->interactions[j]); if(!->isReal()) continue; /* */ }
4.5.4 ForceContainer
ForceContainer holds generalized forces, i.e. forces, torques, (explicit) dispalcements and rotations for each body. During each computation step, there are typically 3 phases pertaining to forces: 1. Resetting forces to zero (usually done by the ForceResetter engine) 2. Incrementing forces from parallel sections (solving interactions from LawFunctor) 3. Reading absolute force values sequentially for each body: forces applied from dierent interactions are summed together to give overall force applied on that body (NewtonIntegrator, but also various other engine that read forces) This scenario leads to special design, which allows fast parallel write access: each thread has its own storage (zeroed upon request), and only writes to its own storage; this avoids concurrency issues. Each thread identies itself by the omp_get_thread_num() function provided by the OpenMP runtime. before reading absolute values, the container must be synchronized, i.e. values from all threads are summed up and stored separately. This is a relatively slow operation and we provide ForceContainer::syncCount that you might check to nd cummulative number of synchronizations and compare it against number of steps. Ideally, ForceContainer is only synchronized once at each step.
108
the container is resized whenever an element outside the current range is read/written to (the read returns zero in that case); this avoids the necessity of tracking number of bodies, but also is potential danger (such as scene->forces.getForce(1000000000), which will probably exhaust your RAM). Unlike c++, Python does check given id against number of bodies.
// resetting forces (inside ForceResetter) scene->forces.reset() // in a parallel section scene->forces.addForce(id,force); // add force // container is not synced after we wrote to it, sync before reading scene->forces.sync(); const Vector3r& f=scene->forces.getForce(id);
Synchronization is handled automatically if values are read from python:

Yade [1]: O.bodies.append(Body()) Out[1]: 1 Yade [2]: O.forces.addF(0,Vector3(1,2,3)) Yade [3]: O.forces.f(0) Out[3]: Vector3(1,2,3) Yade [4]: O.forces.f(100) --------------------------------------------------------------------------IndexError Traceback (most recent call last) /home/buildbot/yade/yade-full/install/lib/x86_64-linux-gnu/yade-buildbot/py/yade/__init__.pyc in <module>() ----> 1 O.forces.f(100) IndexError: Body id out of range.
4.5.5 Handling interactions

Creating and removing interactions is a rather delicate topic and number of components must cooperate so that the whole behaves as expected. Terminologically, we distinguish potential interactions, having neither geometry nor physics. Interaction.isReal can be used to query the status (Interaction::isReal() in c++). real interactions, having both geometry and physics. Below, we shall discuss the possibility of interactions that only have geometry but no physics. During each step in the simulation, the following operations are performed on interactions in a typical simulation: 1. Collider creates potential interactions based on spatial proximity. Not all pairs of bodies are susceptible of entering interaction; the decision is done in Collider::mayCollide: clumps may not enter interactions (only their members can) clump members may not interact if they belong to the same clump bitwise AND on both bodies masks must be non-zero (i.e. there must be at least one bit set in common) 2. Collider erases interactions that were requested for being erased (see below). 3. InteractionLoop (via IGeomDispatcher) calls appropriate IGeomFunctor based on Shape combination of both bodies, if such functor exists. For real interactions, the functor updates associated IGeom. For potential interactions, the functor returns
4.5. Simulation framework
109
false if there is no geometrical overlap, and the interaction will stillremain potentialonly true if there is geometrical overlap; the functor will have created an IGeom in such case. Note: For real interactions, the functor must return true, even if there is no more spatial overlap between bodies. If you wish to delete an interaction without geometrical overlap, you have to do this in the LawFunctor. This behavior is deliberate, since dierent laws have dierent requirements, though ideally using relatively small number of generally useful geometry functors.
Note: If there is no functor suitable to handle given combination of shapes, the interaction will be left in potential state, without raising any error. 4. For real interactions (already existing or just created in last step), InteractionLoop (via IPhysDispatcher) calls appropriate IPhysFunctor based on Material combination of both bodies. The functor must update (or create, if it doesnt exist yet) associated IPhys instance. It is an error if no suitable functor is found, and an exception will be thrown. 5. For real interactions, InteractionLoop (via LawDispatcher) calls appropriate LawFunctor based on combination of IGeom and IPhys of the interaction. Again, it is an error if no functor capable of handling it is found. 6. LawDispatcher can decide that an interaction should be removed (such as if bodies get too far apart for non-cohesive laws; or in case of complete damage for damage models). This is done by calling
InteractionContainer::requestErase(id1,id2)
Such interaction will not be deleted immediately, but will be reset to potential state. At next step, the collider will call InteractionContainer::erasePending, which will only completely erase interactions the collider indicates; the rest will be kept in potential state. Creating interactions explicitly Interactions may still be created explicitly with utils.createInteraction, without any spatial requirements. This function searches current engines for dispatchers and uses them. IGeomFunctor is called with the force parameter, obliging it to return true even if there is no spatial overlap.
4.5.6 Associating Material and State types

Some models keep extra state information in the Body.state object, therefore requiring strict association of a Material with a certain State (for instance, CpmMat is associated to CpmState and this combination is supposed by engines such as CpmStateUpdater). If a Material has such a requirement, it must override 2 virtual methods: 1. Material.newAssocState, which returns a new State object of the corresponding type. The default implementation returns State itself. 2. Material.stateTypeOk, which checks whether a given State object is of the corresponding type (this check is run at the beginning of the simulation for all particles). In c++, the code looks like this (for CpmMat):
class CpmMat: public FrictMat { public: virtual shared_ptr<State> newAssocState() const { return shared_ptr<State>(new CpmState); } virtual bool stateTypeOk(State* s) const { return (bool)dynamic_cast<CpmState*>(s); }
110
/* ... */ };
This allows one to construct Body objects from functions such as utils.sphere only by knowing the requires Material type, enforcing the expectation of the model implementor.
4.6 Runtime structure

4.6.1 Startup sequence
Yades main program is python script in core/main/main.py.in; the build system replaces a few ${variables} in that le before copying it to its install location. It does the following: 1. Process command-line options, set environment variables based on those options. 2. Import main yade module (import yade), residing in py/__init__.py.in. This module locates plugins (recursive search for les lib*.so in the lib installation directory). yade.boot module is used to setup temporary directory, and, most importantly, loads plugins. 3. Manage further actions, such as running scripts given at command line, opening qt.Controller (if desired), launching the ipython prompt.
4.6.2 Singletons
There are several global variables that are always accessible from c++ code; properly speaking, they are Singletons, classes of which exactly one instance always exists. The interest is to have some general functionality acessible from anywhere in the code, without the necessity of passing pointers to such objects everywhere. The instance is created at startup and can be always retrieved (as non-const reference) using the instance() static method (e.g. Omega::instance().getScene()). There are 3 singletons: SerializableSingleton Handles serialization/deserialization; it is not used anywhere except for the serialization code proper. ClassFactory Registers classes from plugins and able to factor instance of a class given its name as string (the class must derive from Factorable). Not exposed to python. Omega Access to simulation(s); deserves separate section due to its importance. Omega The Omega class handles all simulation-related functionality: loading/saving, running, pausing. In python, the wrapper class to the singleton is instantiated 5 as global variable O. For convenience, Omega is used as proxy for scenes attribute: although multiple Scene objects may be instantiated in c++, it is always the current scene that Omega represents. The correspondence of data is literal: Omega.materials corresponds to Scene::materials of the current scene; likewise for materials, bodies, interactions, tags, cell, engines, initializers, miscParams. To give an overview of (some) variables: Python Omega.iter Omega.dt Omega.time Omega.realtime Omega.stopAtIter
5
c++ Scene::iter Scene::dt Scene::time Omega::getRealTime() Scene::stopAtIter
It is understood that instantiating Omega() in python only instantiates the wrapper class, not the singleton itself.
4.6. Runtime structure
111
Omega in c++ contains pointer to the current scene (Omega::scene, retrieved by Omega::instance().getScene()). Using Omega.switchScene, it is possible to swap this pointer with Omega::sceneAnother, a completely independent simulation. This can be useful for example (and this motivated this functionality) if while constructing simulation, another simulation has to be run to dynamically generate (i.e. by running simulation) packing of spheres.
4.6.3 Engine loop

Running simulation consists in looping over Engines and calling them in sequence. This loop is dened in Scene::moveToNextTimeStep function in core/Scene.cpp. Before the loop starts, O.initializers are called; they are only run once. The engine loop does the following in each iteration over O.engines: 1. set Engine::scene pointer to point to the current Scene. 2. Call Engine::isActivated(); if it returns false, the engine is skipped. 3. Call Engine::action() 4. If O.timingEnabled, increment Engine::execTime by the dierence from the last time reading (either after the previous engine was run, or immediately before the loop started, if this engine comes rst). Increment Engine::execCount by 1. After engines are processed, virtual time is incremented by timestep and iteration number is incremented by 1. Background execution The engine loop is (normally) executed in background thread (handled by SimulationFlow class), leaving foreground thread free to manage user interaction or running python script. The background thread is managed by O.run() and O.pause() commands. Foreground thread can be blocked until the loop nishes using O.wait(). Single iteration can be run without spawning additional thread using O.step().
4.7 Python framework

4.7.1 Wrapping c++ classes
Each class deriving from Serializable is automatically exposed to python, with access to its (registered) attributes. This is achieved via YADE_CLASS_BASE_DOC_* macro family. All classes registered in class factory are default-constructed in Omega::buildDynlibDatabase. Then, each serializable class calls Serializable::pyRegisterClass virtual method, which injects the class wrapper into (initially empty) yade.wrapper module. pyRegisterClass is dened by YADE_CLASS_BASE_DOC and knows about class, base class, docstring, attributes, which subsequently all appear in boost::python class denition. Wrapped classes dene special constructor taking keyword arguments corresponding to class attributes; therefore, it is the same to write:
Yade [1]: f1=ForceEngine() Yade [2]: f1.ids=[0,4,5] Yade [3]: f1.force=Vector3(0,-1,-2)
and
Yade [1]: f2=ForceEngine(ids=[0,4,5],force=Vector3(0,-1,-2)) Yade [2]: print f1.dict() {'ompThreads': -1, 'force': Vector3(0,-1,-2), 'ids': [0, 4, 5], 'dead': False, 'label': ''}
112
Yade [3]: print f2.dict() {'ompThreads': -1, 'force': Vector3(0,-1,-2), 'ids': [0, 4, 5], 'dead': False, 'label': ''}
Wrapped classes also inherit from Serializable several special virtual methods: dict() returning all registered class attributes as dictionary (shown above), clone() returning copy of instance (by copying attribute values), updateAttrs() and updateExistingAttrs() assigning attributes from given dictionary (the former thrown for unknown attribute, the latter doesnt). Read-only property name wraps c++ method getClassName() returning class name as string. (Since c++ class and the wrapper class always have the same name, getting python type using __class__ and its property __name__ will give the same value).
Yade [1]: s=Sphere() Yade [2]: s.__class__.__name__ Out[2]: 'Sphere'
4.7.2 Subclassing c++ types in python

In some (rare) cases, it can be useful to derive new class from wrapped c++ type in pure python. This is done in the yade.pack module module: Predicate is c++ base class; from this class, several c++ classes are derived (such as inGtsSurface), but also python classes (such as the trivial inSpace predicate). inSpace derives from python class Predicate; it is, however, not direct wrapper of the c++ Predicate class, since virtual methods would not work. boost::python provides special boost::python::wrapper template for such cases, where each overridable virtual method has to be declared explicitly, requesting python override of that method, if present. See Overridable virtual functions for more details.
4.7.3 Reference counting

Python internally uses reference counting on all its objects, which is not visible to casual user. It has to be handled explicitly if using pure Python/C API with Py_INCREF and similar functions. boost::python used in Yade fortunately handles reference counting internally. Additionally, it automatically integrates reference counting for shared_ptr and python objects, if class A is wrapped as boost::python::class_<A,shared_ptr<A>>. Since all Yade classes wrapped using YADE_CLASS_BASE_DOC_* macro family are wrapped in this way, returning shared_ptr<> objects from is the preferred way of passing objects from c++ to python. Returning shared_ptr is much more ecient, since only one pointer is returned and reference count internally incremented. Modifying the object from python will modify the (same) object in c++ and vice versa. It also makes sure that the c++ object will not be deleted as long as it is used somewhere in python, preventing (important) source of crashes.
4.7.4 Custom converters

When an object is passed from c++ to python or vice versa, then either 1. the type is basic type which is transparently passed between c++ and python (int, bool, std::string etc) 2. the type is wrapped by boost::python (such as Yade classes, Vector3 and so on), in which case wrapped object is returned; 6
6 Wrapped classes are automatically registered when the class wrapper is created. If wrapped class derives from another wrapped class (and if this dependency is declared with the boost::python::bases template, which Yades classes do automatically), parent class must be registered before derived class, however. (This is handled via loop in Omega::buildDynlibDatabase, which reiterates over classes, skipping failures, until they all successfully register) Math classes (Vector3, Matrix3, Quaternion) are wrapped by hand, to be found in py/mathWrap/miniEigen.cpp; this module is imported at startup. On systems, where minieigen is available as a separate package, the Yades miniEigen is skipped.
4.7. Python framework
113
Other classes, including template containers such as std::vector must have their custom converters written separately. Some of them are provided in py/wrapper/customConverters.cpp, notably converters between python (homogeneous, i.e. with all elements of the same type) sequences and c++ std::vector of corresponding type; look in that source le to add your own converter or for inspiration. When an object is crossing c++/python boundary, boost::pythons global converters registry is searched for class that can perform conversion between corresponding c++ and python types. The converters registry is common for the whole program instance: there is no need to register converters in each script (by importing _customConverters, for instance), as that is done by yade at startup already. Note: Custom converters only work for value that are passed by value to python (not by reference): some attributes dened using YADE_CLASS_BASE_DOC_* macro family are passed by value, but if you dene your own, make sure that you read and understand Why is my automatic to-python conversion not being found?. In short, the default for def_readwrite and def_readonly is to return references to underlying c++ objects, which avoids performing conversion on them. For that reason, return value policy must be set to return_by_value explicitly, using slighly more complicated add_property syntax, as explained at the page referenced.
4.8 Maintaining compatibility

In Yade development, we identied compatibility to be very strong desire of users. Compatibility concerns python scripts, not simulations saved in XML or old c++ code.
4.8.1 Renaming class

Script scripts/rename-class.py should be used to rename class in c++ code. It takes 2 parameters (old name and new name) and must be run from top-level source directory:
$ scripts/rename-class.py OldClassName NewClassName Replaced 4 occurences, moved 0 files and 0 directories Update python scripts (if wanted) by running: perl -pi -e 's/\bOldClassName\b/NewClassName/g' `ls **/*.py |grep
This has the following eects: 1. If le or directory has basename OldClassName (plus extension), it will be renamed using bzr. 2. All occurences of whole word OldClassName will be replaced by NewClassName in c++ sources. 3. An extry is added to py/system.py, which contains map of deprecated class names. At yade startup, proxy class with OldClassName will be created, which issues a DeprecationWarning when being instantiated, informing you of the new name you should use; it creates an instance of NewClassName, hence not disruting your scripts functioning:
Yade [3]: SimpleViscoelasticMat() /usr/local/lib/yade-trunk/py/yade/__init__.py:1: DeprecationWarning: Class `SimpleViscoelasticMat' was ren -> [3]: <ViscElMat instance at 0x2d06770>
As you have just been informed, you can run yade --update to all old names with their new names in scripts you provide:
$ yade-trunk --update script1.py some/where/script2.py
This gives you enough freedom to make your class name descriptive and intuitive.
114
4.8.2 Renaming class attribute

Renaming class attribute is handled from c++ code. You have the choice of merely warning at accessing old attribute (giving the new name), or of throwing exception in addition, both with provided explanation. See deprec parameter to YADE_CLASS_BASE_DOC_* macro family for details.
4.9 Debian packaging instructions

In order to make parallel installation of several Yade version possible, we adopted similar strategy as e.g. gcc packagers in Debian did: 1. Real Yade packages are named yade-0.30 (for stable versions) or yade-bzr2341 (for snapshots). 2. They provide yade or yade-snapshot virtual packages respectively. 3. Each source package creates several installable packages (using bzr2341 as example version): (a) yade-bzr2341 with the optimized binaries; (yade-bzr2341-multi, ) the executable binary is yade-bzr2341
(b) yade-bzr2341-dbg with debug binaries (debugging symbols, non-optimized, and with crash handlers); the executable binary is yade-bzr2341-dbg (c) yade-bzr2341-doc with sample scripts and some documentation (see bug #398176 however) (d) (future?) yade-bzr2341-reference with reference documentation (see bug #401004) 4. Using Debian alternatives, the highest installed package provides additionally commands without the version specication like yade, yade-multi, as aliases to that versions binaries. (yade-dbg, for the debuggin packages). The exact rule is: (a) Stable releases have always higher priority than snapshots (b) Higher versions/revisions have higher pripority than lower versions/revisions.
4.9.1 Prepare source package

Debian packaging les are located in debian/ directory. They contain build recipe debian/rules, dependecy and package declarations debian/control and maintainer scripts. Some of those les are only provided as templates, where some variables (such as version number) are replaced by special script. The script scripts/debian-prep processes templates in debian/ and creates les which can be used by debian packaging system. Before running this script: 1. If you are releasing stable version, make sure there is le named RELEASE containing single line with version number (such as 0.30). This will make scripts/debian-prep create release packages. In absence of this le, snapshots packaging will be created instead. Release or revision number (as detected by running bzr revno in the source tree) is stored in VERSION le, where it is picked up during package build and embedded in the binary. 2. Find out for which debian/ubuntu series your package will be built. This is the name that will appear on the top of (newly created) debian/changelog le. This name will be usually unstable, testing or stable for debian and karmic, lucid etc for ubuntu. WHen package is uploaded to Launchpads build service, the package will be built for this specied release. Then run the script from the top-level directory, giving series name as its rst (only) argument:
$ scripts/debian-prep lucid
After this, signed debian source package can be created:

$ debuild -S -sa -k62A21250 -I -Iattic
(-k gives GPG key identier, -I skips .bzr and similar directories, -Iattic will skip the useless attic directory).
4.9. Debian packaging instructions
115
4.9.2 Create binary package

Local in-tree build Once les in debian/ are prepared, packages can be build by issuing:: $ fakeroot debian/rules binary Clean system build Using pbuilder system, package can be built in a chroot containing clean debian/ubuntu system, as if freshly installed. Package dependencies are automatically installed and package build attempted. This is a good way of testing packaging before having the package built remotely at Launchpad. Details are provided at wiki page. Launchpad build service Launchpad provides service to compile package for dierent ubuntu releases (series), for all supported architectures, and host archive of those packages for download via APT. Having appropriate permissions at Launchpad (veried GPG key), source package can be uploaded to yades archive by:
$ dput ppa:yade-users/ppa ../yade-bzr2341_1_source.changes
After several hours (depending on load of Launchpad build farm), new binary packages will be published at https://launchpad.net/~yade-users/+archive/ppa. This process is well documented at https://help.launchpad.net/Packaging/PPA.
116
Chapter 5
Installation
Yade can be installed from packages (precompiled binaries) or source code. The choice depends on what you need: if you dont plan to modify Yade itself, package installation is easier. In the contrary case, you must download and install source code.
5.1 Packages
Packages from Launchpad PPA service (package personal archive) are provided for all currently supported Ubuntu versions for stable and daily releases. yade-daily is a automatically daily (if there were some commtis during the previous days) geenrated package, which includes all the newly added features. To install version from PPA, run the following: For stable releases:
sudo sudo sudo sudo add-apt-repository ppa:yade-pkg/stable add-apt-repository ppa:yade-users/external apt-get update apt-get install yade-stable # for stable releases # optional (updates of other packages)
For latest builds from trunk:

sudo add-apt-repository ppa:yade-pkg/snapshots sudo apt-get update sudo apt-get install yade-daily # for latest daily (mostly) releases
After you added whether stable or snapshot PPAs, you will get automatically the updates of the package, when they arrive the PPA. More detailed instructions are available at the corresponding pages of ppas (links above). Since 2011 all Ubuntu versions (starting from 11.10, Oneiric) and Debian (starting from Wheezy) are having already Yade in their main repositories. There are only stable releases are placed. To install the program, run the following:
sudo apt-get install yade
To check, what version of Yade is in specic Distribution, check the links for Ubuntu and Debian.
5.2 Source code

Installation from source code is reasonable, when you want to add or modify constitutive laws, engines or functions... Installing the latest trunk version allows one to use newly added features, which are not yet available in packaged versions.
117
5.2.1 Download
If you want to install from source, you can install either a release (numbered version, which is frozen) or the current developement version (updated by the developers frequently). You should download the development version (called trunk) if you want to modify the source code, as you might encounter problems that will be xed by the developers. Release version will not be modied (except for updates due to critical and easy-to-x bugs), but they are in a more stabilized state that trunk generally. 1. Releases can be downloaded from the download page, as compressed archive. Uncompressing the archive gives you a directory with the sources. 2. developement version (trunk) can be obtained from the code repository at github. We use GIT (the git command) for code management (install the git package in your distribution and create a GitHub account):
git clone [email protected]:yade/trunk.git
will download the whole code repository of trunk. Check out Yade on github wiki page for more. Alternatively, a read-only checkout is possible via https without a GitHub account (easier if you dont want to modify the main Yade branch):
git clone https://github.com/yade/trunk.git
For those behind rewall, you can download any revision of the repository as compressed archive. Release and trunk sources are compiled in the same way.
5.2.2 Prerequisities
Yade relies on a number of external software to run; its installation is checked before the compilation starts. cmake build system gcc compiler (g++); other compilers will not work; you need g++>=4.2 for openMP support boost 1.35 or later qt4 library freeglut3 libQGLViewer python, numpy, ipython matplotlib eigen3 algebra library gdb debugger sqlite3 database engine Loki library VTK library (optional but recommended) Most of the list above is very likely already packaged for your distribution. The following commands have to be executed in command line of corresponding distributions. Just copy&paste to the terminal. To perform commands you should have root privileges Ubuntu, Debian and their derivatives:
sudo apt-get install cmake git freeglut3-dev libloki-dev \ libboost-date-time-dev libboost-filesystem-dev libboost-thread-dev \ libboost-program-options-dev \ libboost-regex-dev fakeroot dpkg-dev build-essential g++ \ libboost-iostreams-dev python-dev libboost-python-dev ipython \
118
Chapter 5. Installation
python-matplotlib libsqlite3-dev python-numpy python-tk gnuplot \ libgts-dev python-pygraphviz libvtk5-dev python-scientific libeigen3-dev \ binutils-gold python-xlib python-qt4 pyqt4-dev-tools \ gtk2-engines-pixbuf python-argparse \ libqglviewer-qt4-dev python-imaging libjs-jquery python-sphinx python-git python-bibtex \ libxmu-dev libxi-dev libgmp3-dev libcgal-dev help2man libsuitesparse-dev \ libopenblas-dev libmetis-dev
Fedora:
yum install cmake qt3-devel freeglut-devel boost-devel boost-date-time \ boost-filesystem boost-thread boost-regex fakeroot gcc gcc-c++ boost-iostreams \ python-devel boost-python ipython python-matplotlib \ sqlite-devel python-numeric ScientificPython-tk gnuplot doxygen gts-devel \ graphviz-python vtk-devel ScientificPython eigen2-devel libQGLViewer-devel \ loki-lib-devel python-xlib PyQt4 PyQt4-devel python-imaging python-sphinx python-bibtex
5.2.3 Compilation
You should create a separate build-place-folder, where Yade will be congured and where the source code will be compiled. Then inside this build-directory you should start cmake to congure the compilation process:
cmake -DINSTALL_PREFIX=/path/to/installfolder /path/to/sources
Additional options can be congured in the same line with the following syntax:
cmake -DOPTION1=VALUE1 -DOPTION2=VALUE2
The following options are available: INSTALL_PREFIX: path where Yade should be installed (/usr/local by default) LIBRARY_OUTPUT_PATH: path to install libraries (lib by default) DEBUG: compile in debug-mode (OFF by default) CMAKE_VERBOSE_MAKEFILE: output additional information during compiling (OFF by default) SUFFIX: sux, added after binary-names (version number by default) NOSUFFIX: do not add a sux after binary-name (OFF by default) YADE_VERSION: explicitely set version number (is dened from git-directory by default) ENABLE_GUI: enable GUI option (ON by default) ENABLE_CGAL: enable CGAL option (ON by default) ENABLE_VTK: enable VTK-export option (ON by default) ENABLE_OPENMP: enable OpenMP-parallelizing option (ON by default) ENABLE_GTS: enable GTS-option (ON by default) ENABLE_GL2PS: enable GL2PS-option (ON by default) runtimePREFIX: used for packaging, when install directory is not the same is runtime directory (/usr/local by default) CHUNKSIZE: used, if you want several sources to be compiled at once. Increases compilation speed and RAM-consumption during it (1 by default). For using an extended parameters of cmake, please, follow the corresponding documentation on cmakewebpage. If the compilation is nished without errors, you will see all enabled and disabled options. Then start the standard the compilation process:
5.2. Source code
119
make
Installing performs with the following command:

make install
The install command will in fact also recompile if source les have been modied. Hence there is no absolute need to type the two commands separately. The compilation process can take a long time, be patient. An additional parameter on many cores systems -j can be added to decrease compilation time and split the compilation on many cores. For example, on 4-core machines it would be reasonable to set the parameter -j4. Note, the Yade requires approximately 2GB/core for compilation, otherwise the swap-le will be used and a compilation time dramatically increases. For building the documentation you should at rst execute the command make install and then make doc to build it. The generated les will be stored in your current build directory/doc/sphinx/_build. make manpage command generates and moves manpages in a standard place. make check command executes standard test to check the functionality of compiled program. Yade can be compiled not only by GCC-compiler, but also by CLANG front-end for the LLVM compiler. For that you set the environment variables CC and CXX upon detecting the C and C++ compiler to use:
export CC=/usr/bin/clang export CXX=/usr/bin/clang++ cmake -DOPTION1=VALUE1 -DOPTION2=VALUE2
Clang does not support OpenMP-parallelizing for the moment, that is why the feature will be disabled.
120
Chapter 5. Installation
Chapter 6
DEM Background
In this chapter, we mathematically describe general features of explicit DEM simulations, with some reference to Yade implementation of these algorithms. They are given roughly in the order as they appear in simulation; rst, two particles might establish a new interaction, which consists in 1. detecting collision between particles; 2. creating new interaction and determining its properties (such as stiness); they are either precomputed or derived from properties of both particles; Then, for already existing interactions, the following is performed: 1. strain evaluation; 2. stress computation based on strains; 3. force application to particles in interaction. This simplied description serves only to give meaning to the ordering of sections within this chapter. A more detailed description of this simulation loop is given later.
6.1 Collision detection

6.1.1 Generalities
Exact computation of collision conguration between two particles can be relatively expensive (for instance between Sphere and Facet). Taking a general pair of bodies i and j and their exact (In the sense of precision admissible by numerical implementation.) spatial predicates (called Shape in Yade) represented by point sets Pi , Pj the detection generally proceeds in 2 passes: i and P j ; they are pre-constructed in such a 1. fast collision detection using approximate predicate P way as to abstract away individual features of Pi and Pj and satisfy the condition i x R3 : x Pi x P (6.1)
(likewise for Pj ). The approximate predicate is called bounding volume (Bound in Yade) since it bounds any particles volume from outside (by virtue of the implication). It follows that (Pi Pj ) = i P j ) = and, by applying modus tollens, (P ( ) ( ) i P j = Pi Pj = (6.2) P
which is a candidate exclusion rule in the proper sense.
121
2. By ltering away impossible collisions in (6.2), a more expensive, exact collision detection algo rithms ( can )be run on possible interactions, ltering out remaining spurious couples (Pi Pj ) = Pi Pj = . These algorithms operate on Pi and Pj and have to be able to handle all possible combinations of shape types. It is only the rst step we are concerned with here.
6.1.2 Algorithms
Collision evaluation algorithms have been the subject of extensive research in elds such as robotics, computer graphics and simulations. They can be roughly divided in two groups: Hierarchical algorithms which recursively subdivide space and restrict the number of approximate checks in the rst pass, knowing that lower-level bounding volumes can intersect only if they are part of the same higher-level bounding volume. Hierarchy elements are bounding volumes of dierent kinds: octrees [Jung1997], bounding spheres [Hubbard1996], k-DOPs [Klosowski1998]. Flat algorithms work directly with bounding volumes without grouping them in hierarchies rst; let us only mention two kinds commonly used in particle simulations: Sweep and prune algorithm operates on axis-aligned bounding boxes, which overlap if and only if they overlap along all axes. These algorithms have roughly O(n log n) complexity, where n is number of particles as long as they exploit temporal coherence of the simulation. Grid algorithms represent continuous R3 space by a nite set of regularly spaced points, leading to very fast neighbor search; they can reach the O(n) complexity [Munjiza1998] and recent research suggests ways to overcome one of the major drawbacks of this method, which is the necessity to adjust grid cell size to the largest particle in the simulation ([Munjiza2006], the multistep extension). Temporal coherence expresses the fact that motion of particles in simulation is not arbitrary but governed by physical laws. This knowledge can be exploited to optimize performance. Numerical stability of integrating motion equations dictates an upper limit on t (sect. Stability considerations) and, by consequence, on displacement of particles during one step. This consideration is taken into account in [Munjiza2006], implying that any particle may not move further than to a neighboring grid cell during one step allowing the O(n) complexity; it is also explored in the periodic variant of the sweep and prune algorithm described below. i predicates in such a way that they satisfy the (6.1) condition On a ner level, it is common to enlarge P during several timesteps; the rst collision detection pass might then be run with stride, speeding up the simulation considerably. The original publication of this optimization by Verlet [Verlet1967] used enlarged list of neighbors, giving this technique the name Verlet list. In general cases, however, where neighbor lists are not necessarily used, the term Verlet distance is employed.
6.1.3 Sweep and prune

Let us describe in detail the sweep and prune algorithm used for collision detection in Yade (class i ; each Aabb is given by lower InsertionSortCollider). Axis-aligned bounding boxes (Aabb) are used as P 3 x0 x1 i along the and upper corner R (in the following, Pi , Pi are minimum/maximum coordinates of P x-axis and so on). Construction of Aabb from various particle Shapes (such as Sphere, Facet, Wall) is straightforward, handled by appropriate classes deriving form BoundFunctor (Bo1_Sphere_Aabb, Bo1_Facet_Aabb, ). Presence of overlap of two Aabbs can be determined from conjunction of separate overlaps of intervals along each axis (g-sweep-and-prune): ( ) ) ( )) ] [(( i P j = w0 , P w1 P w0 , P w1 P P = i i j j
w{x,y,z}
122
Chapter 6. DEM Background
y1 P 3 y1 P 2
+y
2 P
3 P
P2
y1 P 1 y0 P 2 1 P
P3
P1
y0 P 3
y0 P 1 x0 P 1 x0 P 2 x0 P 3 x1 P 1 x1 P 2 x1 P 3
+x
Figure 6.1: Sweep and prune algorithm (shown in 2D), where Aabb of each sphere is represented by minimum and maximum value along each axis. Spatial overlap of Aabbs is present if they overlap along 1 P 2 = (but note that P1 P2 = ) and P 2 P 3 = .} all axes. In this case, P where (a, b) denotes interval in R. The collider keeps 3 separate lists (arrays) Lw for each axis w {x, y, z} } { w1 w0 , P Lw = P i i
i
where i traverses all particles. Lw arrays (sorted sets) contain respective coordinates of minimum and maximum corners for each Aabb (we call these coordinates bound in the following); besides bound, each of list elements further carries id referring to particle it belongs to, and a ag whether it is lower or upper bound. In the initial step, all lists are sorted (using quicksort, average O(n log n)) and one axis is used to create initial interactions: the range between lower and upper bound for each body is traversed, while bounds in-between indicate potential Aabb overlaps which must be checked on the remaining axes as well. At each successive step, lists are already pre-sorted. Inversions occur where a particles coordinate has just crossed another particles coordinate; this number is limited by numerical stability of simulation and its physical meaning (giving spatio-temporal coherence to the algorithm). The insertion sort algorithm swaps neighboring elements if they are inverted, and has complexity between bigO{n} and bigO{n^2}, for pre-sorted and unsorted lists respectively. For our purposes, we need only to handle inversions, which by nature of the sort algorithm are detected inside the sort loop. An inversion might signify: overlap along the current axis, if an upper bound inverts (swaps) with a lower bound (i.e. that the upper bound with a higher coordinate was out of order in coming before the lower bound with a lower coordinate). Overlap along the other 2 axes is checked and if there is overlap along all axes, a new potential interaction is created. End of overlap along the current axis, if lower bound inverts (swaps) with an upper bound. If there is only potential interaction between the two particles in question, it is deleted. Nothing if both bounds are upper or both lower.
6.1. Collision detection
123
Aperiodic insertion sort Let us show the sort algorithm on a sample sequence of numbers: || 3 7 2 4 ||
Elements are traversed from left to right; each of them keeps inverting (swapping) with neighbors to the left, moving left itself, until any of the following conditions is satised: () (||) the sorting order with the left neighbor is correct, or the element is at the beginning of the sequence.
We start at the leftmost element (the current element is marked i ) || 3 7 2 4 ||.
It obviously immediately satises (||), and we move to the next element: || 3 g
4 ||.
Condition () holds, therefore we move to the right. The 2 is not in order (violating ()) and two inversions take place; after that, (||) holds: || || || 3 3h
7h
2 7 7
4 ||, 4 ||, 4 ||.
2 3
The last element 4 rst violates (), but satises it after one inversion || 2 || 2 3 3g
7h
4 7
||, ||.
All elements having been traversed, the sequence is now sorted. It is obvious that if the initial sequence were sorted, elements only would have to be traversed without any inversion to handle (that happens in O(n) time). For each inversion during the sort in simulation, the function that investigates change in Aabb overlap is invoked, creating or deleting interactions. The periodic variant of the sort algorithm is described in Periodic insertion sort algorithm, along with other periodic-boundary related topics. Optimization with Verlet distances As noted above, [Verlet1967] explored the possibility of running the collision detection only sparsely by i . enlarging predicates P In Yade, this is achieved by enlarging Aabb of particles by xed relative length (or Verlets distance) in all dimensions L (InsertionSortCollider.sweepLength). Suppose the collider run last time at step m and the
124
current step is n. NewtonIntegrator tracks the cummulated distance traversed by each particle between m and n by comparing the current position with the reference position from time n (Bound::refPos), Lmn = |Xn Xm | triggering the collider re-run as soon as one particle gives: Lmn > L. (6.4) (6.3)
InsertionSortCollider.targetInterv is used to adjust L independently for each particle. Larger L will be assigned to the fastest ones, so that all particles would ideally reach the edge of their bounds after this target number of iterations. Results of using Verlet distance depend highly on the nature of simulation and choice of InsertionSortCollider.targetInterv. Adjusting the sizes independently for each particle is especially ecient if some parts of a problem have high-speed particles will others are not moving. If it is not the case, no signicant gain should be expected as compared to targetInterv=0 (assigning the same L to all particles). The number of particles and the number of available threads is also to be considered for choosing an appropriate Verlets distance. A larger distance will result in less time spent in the collider (which runs single-threaded) and more time in computing interactions (multi-threaded). Typically, large L will be used for large simulations with more than 105 particles on multi-core computers. On the other hand simulations with less than 104 particles on single processor will probably benet from smaller L. Users benchmarks may be found on Yades wiki (see e.g. https://yade-dem.org/wiki/Colliders_performace).
6.2 Creating interaction between particles

Collision detection described above is only approximate. Exact collision detection depends on the geometry of individual particles and is handled separately. In Yade terminology, the Collider creates only potential interactions; potential interactions are evaluated exactly using specialized algorithms for collision of two spheres or other combinations. Exact collision detection must be run at every timestep since it is at every step that particles can change their mutual position (the collider is only run sometimes if the Verlet distance optimization is in use). Some exact collision detection algorithms are described in Strain evaluation; in Yade, they are implemented in classes deriving from IGeomFunctor (prexed with Ig2). Besides detection of geometrical overlap (which corresponds to IGeom in Yade), there are also nongeometrical properties of the interaction to be determined (IPhys). In Yade, they are computed for every new interaction by calling a functor deriving from IPhysFunctor (prexed with Ip2) which accepts the given combination of Material types of both particles.
6.2.1 Stinesses
Basic DEM interaction denes two stinesses: normal stiness KN and shear (tangent) stiness KT . It is desirable that KN be related to ctitious Youngs modulus of the particles material, while KT is typically determined as a given fraction of computed KN . The KT /KN ratio determines macroscopic Poissons ratio of the arrangement, which can be shown by dimensional analysis: elastic continuum has two parameters (E and ) and basic DEM model also has 2 parameters with the same dimensions KN and KT /KN ; macroscopic Poissons ratio is therefore determined solely by KT /KN and macroscopic Youngs modulus is then proportional to KN and aected by KT /KN . Naturally, such analysis is highly simplifying and does not account for particle radius distribution, packing conguration and other possible parameters such as the interaction radius introduced later. Normal stiness The algorithm commonly used in Yade computes normal interaction stiness as stiness of two springs in serial conguration with lengths equal to the sphere radii (g-spheres-contact-stiness). 6.2. Creating interaction between particles 125
l 1 = r1
l 2 = r2
E1
l = l1 + l2
E2
Figure 6.2: Series of 2 springs representing normal stiness of contact between 2 spheres. Let us dene distance l = l1 + l2 , where li are distances between contact point and sphere centers, which are initially (roughly speaking) equal to sphere radii. Change of distance between the sphere centers l is distributed onto deformations of both spheres l = l1 + l2 proportionally to their compliances. Displacement change li generates force Fi = Ki li , where Ki assures proportionality and has physical meaning and dimension of stiness; Ki is related to the sphere material modulus Ei and some length li proportional to ri . l = l1 + l2 Ki = Ei li KN l = F = F1 = F2 KN (l1 + l2 ) = F ( ) F F KN + =F K1 K2
1 1 1 K1 + K 2 = KN K1 K2 KN = K1 + K2 E1 l1 E2 l2 KN = E1 l1 + E2 l2
The most used class computing interaction properties Ip2_FrictMat_FrictMat_FrictPhys uses li = 2ri . Some formulations dene an equivalent cross-section Aeq , which in that case appears in the li term as A Ki = Ei li = Ei eq . Such is the case for the concrete model (Ip2_CpmMat_CpmMat_CpmPhys), where Aeq = min(r1 , r2 ). For reasons given above, no pretense about equality of particle-level Ei and macroscopic modulus E should be made. Some formulations, such as [Hentz2003], introduce parameters to match them numerically. This is not appropriate, in our opinion, since it binds those values to particular features of the sphere arrangement that was used for calibration.
li
6.2.2 Other parameters

Non-elastic parameters dier for various material models. Usually, though, they are averaged from the particles material properties, if it makes sense. For instance, Ip2_CpmMat_CpmMat_CpmPhys averages most quantities, while Ip2_FrictMat_FrictMat_FrictPhys computes internal friction angle as = min(1 , 2 ) to avoid friction with bodies that are frictionless.
6.3 Strain evaluation

In the general case, mutual conguration of two particles has 6 degrees of freedom (DoFs) just like a beam in 3D space: both particles have 6 DoFs each, but the interaction itself is free to move and rotate in space (with both spheres) having 6 DoFs itself; then 12 6 = 6. They are shown at g-spheres-dofs.
126
initial conguration
normal straining (1DoF)
shearing (2 DoFs)
twisting (1DoF)
bending (2 DoFs)
Figure 6.3: Degrees of freedom of conguration of two spheres. Normal strain appears if there is a dierence of linear velocity along the interaction axis (n); shearing originates from the dierence of linear velocities perpendicular to n and from the part of 1 + 2 perpendicular to n; twisting is caused by the part of 1 2 parallel with n; bending comes from the part of 1 2 perpendicular to n. We will only describe normal and shear components of strain in the following, leaving torsion and bending aside. The reason is that most constitutive laws for contacts do not use the latter two.
6.3.1 Normal strain

Constants 1 , C 2 and radii r1 , r2 that enter into contact. The Let us consider two spheres with initial centers C order of spheres within the contact is arbitrary and has no inuence on the behavior. Then we dene lengths 2 C 1| d0 = |C d0 r1 r2 d1 = r1 + , 2
d2 = d0 d1 .
These quantities are constant throughout the life of the interaction and are computed only once when the interaction is established. The distance d0 is the reference distance and is used for the conversion of absolute displacements to dimensionless strain, for instance. It is also the distance where (for usual contact laws) there is neither repulsive nor attractive force between the spheres, whence the name equilibrium distance.
r2 r1 1 C d0 = d1 + d2 2 C C d1 d2
Figure 6.4: Geometry of the initial contact of 2 spheres; this case pictures spheres which already overlap when the contact is created (which can be the case at the beginning of a simulation) for the sake of is in the middle of the overlap zone. generality. The initial contact point C
6.3. Strain evaluation
127
Distances d1 and d2 dene reduced (or expanded) radii of spheres; geometrical radii r1 and r2 are used only for collision detection and may not be the same as d1 and d2 , as shown in g. g-sphere-sphere. This dierence is exploited in cases where the average number of contacts between spheres should be increased, e.g. to inuence the response in compression or to stabilize the packing. In such case, interactions will be created also for spheres that do not geometrically overlap based on the interaction radius RI , a dimensionless parameter determining non-locality of contact detection. For RI = 1, only spheres that touch are considered in contact; the general condition reads d0 RI (r1 + r2 ). (6.5)
The value of RI directly inuences the average number of interactions per sphere (percolation), which i predicates for some models is necessary in order to achieve realistic results. In such cases, Aabb (or P in general) must be enlarged accordingly (Bo1_Sphere_Aabb.aabbEnlargeFactor).
Contact cross-section
Some constitutive laws are formulated with strains and stresses (Law2_ScGeom_CpmPhys_Cpm, the concrete model described later, for instance); in that case, equivalent cross-section of the contact must be introduced for the sake of dimensionality. The exact denition is rather arbitrary; the CPM model (Ip2_CpmMat_CpmMat_CpmPhys) uses the relation Aeq = min(r1 , r2 )2 (6.6)
which will be used to convert stresses to forces, if the constitutive law used is formulated in terms of stresses and strains. Note that other values than can be used; it will merely scale macroscopic packing stiness; it is only for the intuitive notion of a truss-like element between the particle centers that we choose Aeq representing the circle area. Besides that, another function than min(r1 , r2 ) can be used, although the result should depend linearly on r1 and r2 so that the equation gives consistent results if the particle dimensions are scaled. Variables The following state variables are updated as spheres undergo motion during the simulation (as C 1 and C change): 2 n = and ( ) d0 |C 2 C1 | C = C + d n. 1 1 2 (6.8)
C 2 C1 C2 C1 |C C | 2 1
(6.7)
The contact point C is always in the middle of the spheres overlap zone (even if the overlap is negative, when it is in the middle of the empty space between the spheres). The contact plane is always perpendicular to the contact plane normal n and passes through C . Normal displacement and strain can be dened as
uN = |C 2 C1 | d0 ,
N =
|C C uN 1| = 2 1. d0 d0
Since uN is always aligned with n, it can be stored as a scalar value multiplied by n if necessary. For massively compressive simulations, it might be benecial to use the logarithmic strain, such that the strain tends to (rather than 1) as centers of both spheres approach. Otherwise, repulsive force 128 Chapter 6. DEM Background
would remain nite and the spheres could penetrate through each other. Therefore, we can adjust the denition of normal strain as follows: { ( |C C | ) 2 1 log if |C 2 C1 | < d0 N = |C C |d0 2 1 1 otherwise. d0 Such denition, however, has the disadvantage of eectively increasing rigidity (up to innity) of contacts, requiring t to be adjusted, lest the simulation becomes unstable. Such dynamic adjustment is possible using a stiness-based time-stepper (GlobalStinessTimeStepper in Yade).
6.3.2 Shear strain

In order to keep uT consistent (e.g. that uT must be constant if two spheres retain mutually constant conguration but move arbitrarily in space), then either uT must track spheres spatial motion or must (somehow) rely on sphere-local data exclusively. These two possibilities lead to two algorithms of computing shear strains. They should give the same results (disregarding numerical imprecision), but there is a trade-o between computational cost of the incremental method and robustness of the total one. Geometrical meaning of shear strain is shown in g-shear-2d.
uT n C
Figure 6.5: Evolution of shear displacement uT due to mutual motion of spheres, both linear and rotational. Left conguration is the initial contact, right conguration is after displacement and rotation of one particle.
Incremental algorithm The incremental algorithm is widely used in DEM codes and is described frequently ([Luding2008], [Alonso2004]). Yade implements this algorithm in the ScGeom class. At each step, shear displacement uT is updated; the update increment can be decomposed in 2 parts: motion of the interaction (i.e. C and n) in global space and mutual motion of spheres. 1. Contact moves dues to changes of the spheres positions C1 and C2 , which updates current C and n as per (6.8) and (6.7). u T is perpendicular to the contact plane at the previous step n and must be updated so that uT + (uT ) = uT n ; this is done by perpendicular projection to the plane rst (which might decrease |uT |) and adding what corresponds to spatial rotation of the interaction instead:
(uT )1 = u T (n n ) ) ( t n (1 + 2 ) n (uT )2 = uT 2
129
2. Mutual movement of spheres, using only its part perpendicular to n ; v12 denotes mutual velocity of spheres at the contact point: ( ) ( ) v12 = v 2 + 2 (d2 n ) v1 + 1 (d1 n )
v 12 = v12 (n v12 )n
(uT )3 = tv 12 Finally, we compute

u T = uT + (uT )1 + (uT )2 + (uT )3 .
Total algorithm The following algorithm, aiming at stabilization of response even with large rotation speeds or t approaching stability limit, was designed in [Smilauer2010b]. (A similar algorithm based on total formulation, which covers additionally bending and torsion, was proposed in [Wang2009].) It is based on tracking original contact points (with zero shear) in the particle-local frame. In this section, variable symbols implicitly denote their current values unless explicitly stated otherwise. Shear strain may have two sources: mutual rotation of spheres or transversal displacement of one sphere with respect to the other. Shear strain does not change if both spheres move or rotate but are not in linear or angular motion mutually. To accurately and reliably model this situation, for every new contact is mapped into local sphere coordinates (p , p ). As we want to determine the initial contact point C 01 02 the distance between both points (i.e. how long the trajectory in on both spheres surfaces together), the shortest path from current C to the initial locally mapped point on the spheres surface is unrolled to the contact plane (p01 , p02 ); then we can measure their linear distance uT and dene shear strain T = uT /d0 (g. g-shear-displacement). i, q i for the sphere initial positions and orientations (as quaterions) in global More formally, taking C ) is rememcoordinates, the initial sphere-local contact point orientation (relative to sphere-local axis x bered: = C1 C2 , n q 01 = Align(x , q q 1n 1 ), )q 02 = Align(x , q 2 (n q 2 ). After some spheres motion, the original point can be unrolled to the current contact plane:
p01
01 x (q1 q 01 ) ) q = Align(n, q1 q = q d1 (qu n)
(auxiliary)
where qu , q are axis and angle components of q and p01 is the unrolled point. Similarly,
02 x (q2 q 02 ) ) q = Align(n, q2 q
p02
= q d1 (qu (n)).
Shear displacement and strain are then computed easily:

uT = p02 p01 uT T = d0
When using material law with plasticity in shear, it may be necessary to limit maximum shear strain, in which case the mapped points are moved closer together to the requested distance (without changing T ). This allows us to remember the previous strain direction and also avoids summation of increments u of plastic strain at every step (g-shear-slip). This algorithm is straightforwardly modied to facet-sphere interactions. In Yade, it is implemented by Dem3DofGeom and related classes. 130 Chapter 6. DEM Background
contact plane
C1 p02 p02 p01 p01 uT C2
Figure 6.6: Shear displacement computation for two spheres in relative motion.
contact plane
C1 old p02 p02 old p02 p02 old p01 p01 uT p01 C2
old p01
requested max |uT |
Figure 6.7: Shear plastic slip for two spheres.
131
6.4 Stress evaluation (example)

Once strain on a contact is computed, it can be used to compute stresses/forces acting on both spheres. The constitutive law presented here is the most usual DEM formulation, originally proposed by Cundall. While the strain evaluation will be similar to algorithms described in the previous section regardless of stress evaluation, stress evaluation itself depends on the nature of the material being modeled. The constitutive law presented here is the most simple non-cohesive elastic case with dry friction, which Yade implements in Law2_ScGeom_FrictPhys_CundallStrack (all constitutive laws derive from base class LawFunctor). In DEM generally, some constitutive laws are expressed using strains and stresses while others prefer displacement/force formulation. The law described here falls in the latter category. When new contact is established (discussed in Engines) it has its properties (IPhys) computed from Materials associated with both particles. In the simple case of frictional material FrictMat, Ip2_FrictMat_FrictMat_FrictPhys creates a new FrictPhys instance, which denes normal stiness KN , shear stiness KT and friction angle . At each step, given normal and shear displacements uN , uT , normal and shear forces are computed (if uN > 0, the contact is deleted without generating any forces): FN = KN uN n, Ft T = KT uT where FN is normal force and FT is trial shear force. A simple non-associated stress return algorithm is applied to compute nal shear force { t |F | tan if |FT | > |FN | tan , FT N Ft T FT = t FT otherwise. Summary force F = FN + FT is then applied to both particles each particle accumulates forces and torques acting on it in the course of each step. Because the force computed acts at contact point C, which is dierence from spheres centers, torque generated by F must also be considered. F1 + = F T 1 + = d1 (n) F F2 + = F T 2 + = d2 n F.
6.5 Motion integration

Each particle accumulates generalized forces (forces and torques) from the contacts in which it participates. These generalized forces are then used to integrate motion equations for each particle separately; therefore, we omit i indices denoting the i-th particle in this section. The customary leapfrog scheme (also known as the Verlet scheme) is used, with some adjustments for rotation of non-spherical particles, as explained below. The leapfrog name comes from the fact that even derivatives of position/orientation are known at on-step points, whereas odd derivatives are known at mid-step points. Let us recall that we use a , a , a+ for on-step values of a at t t, t and t + t respectively; and a , a for mid-step values of a at t t/2, t + t/2. Described integration algorithms are implemented in the NewtonIntegrator class in Yade.
6.5.1 Position
on a particle to update its position from the Integrating motion consists in using current acceleration u current value u to its value at the next timestep u+ . Computation of acceleration, knowing current forces F acting on the particle in question and its mass m, is simply = F/m. u
132
Using the 2nd order nite dierence with step t, we obtain

+ u 2u + u = u t2
from which we express t2 = u+ = 2 u u + u ( ) u u t . = u + t +u t

()
Typically, u is already not known (only u is); we notice, however, that u u u , t
i.e. the mean velocity during the previous step, which is known. Plugging this approximate into the () term, we also notice that mean velocity during the current step can be approximated as u +u t, u which is (); we arrive nally at
( ) +u t . u+ = u + t u
which we need to The algorithm can then be written down by rst computing current mean velocity u now), then computing the position for the next store for the next step (just as we use its old value u time step u+ : = u +u t u t. u+ = u + u Positions are known at times it (if t is constant) while velocities are known at it + t 2 . The facet that they interleave (jump over each other) in such way gave rise to the colloquial name leapfrog scheme.
6.5.2 Orientation (spherical)

Updating particle orientation q proceeds in an analogous way to position update. First, we compute from known current torque T . For spherical particles where the inertia current angular acceleration tensor is diagonal in any orientation (therefore also in current global orientation), satisfying I11 = I22 = I33 , we can write i = T i /I11 , We use the same approximation scheme, obtaining an equation analogous to (??) . = + t The quaternion q representing rotation vector t is constructed, i.e. such that (q) = | |, (q)u = Finally, we compute the next orientation q+ by rotation composition q+ = qq .
6.5. Motion integration
133
6.5.3 Orientation (aspherical)

Integrating rotation of aspherical particles is considerably more complicated than their position, as their local reference frame is not inertial. Rotation of rigid body in the local frame, where inertia matrix I is diagonal, is described in the continuous form by Eulers equations (i {1, 2, 3} and i, j, k are subsequent indices): i + (Ikk Ijj )j k . T i = Iii , they cannot be solved using the standard Due to the presence of the current values of both and leapfrog algorithm (that was the case for translational motion and also for the spherical bodies rotation ). where this equation reduced to T = I The algorithm presented here is described by [Allen1989] (pg. 8489) and was designed by Fincham for molecular dynamics problems; it is based on extending the leapfrog algorithm by mid-step/on-step estimators of quantities known at on-step/mid-step points in the basic formulation. Although it has received criticism and more precise algorithms are known ([Omelyan1999], [Neto2006], [Johnson2008]), this one is currently implemented in Yade for its relative simplicity. Each body has its local coordinate system based on the principal axes of inertia for that body. We use to denote vectors in local coordinates. The orientation of the local system is given by the current particles orientation q as a quaternion; this quaternion can be expressed as the (current) rotation matrix A. Therefore, every vector a is transformed as a = qaq = Aa. Since A is a rotation (orthogonal) matrix, the inverse rotation A1 = AT . For given particle in question, we know I (constant) inertia matrix; diagonal, since in local, principal coordinates, T external torque, q current orientation (and its equivalent rotation matrix A), mid-step angular velocity, L mid-step angular momentum; this is an auxiliary variable that must be tracked in addition for use in this algorithm. It will be zero in the initial step. Our goal is to compute new values of the latter three, that is L , q+ , . We rst estimate current angular momentum and compute current local angular velocity: L = L + T
1
t , 2
L = AL , L = AL ,
L = L + T t, =I =I

L , L .
, using q and : Then we compute q qw w q q q 1 x x = q y 2 qy q q z z
q x q w q z q y t q =q +q . 2
q y q z q w q x
q 0 z q y x , y qx q z w
from q and in the same way as in (??) but shifted by t/2 ahead. Then we can We evaluate q nally compute the desired values t, q+ = q + q = A1 134
6.5.4 Clumps (rigid aggregates)

DEM simulations frequently make use of rigid aggregates of particles to model complex shapes [Price2007] called clumps, typically composed of many spheres. Dynamic properties of clumps are computed from the properties of its members: the clumps mass mc is summed over members, the inertia tensor Ic with respect to the clumps centroid is computed using the parallel axes theorem; local axes are oriented such that they are principal and inertia tensor is diagonal and clumps orientation is changed to compensate rotation of the local system, as to not change the clump members positions in global space. Initial positions and orientations of all clump members in local coordinate system are stored. In Yade (class Clump), clump members behave as stand-alone particles during simulation for purposes of collision detection and contact resolution, except that they have no contacts created among themselves within one clump. It is at the stage of motion integration that they are treated specially. Instead of integrating each of them separately, forces/torques on those particles Fi , T i are converted to forces/torques on the clump itself. Let us denote ri relative position of each particle with regards to clumps centroid, in global orientation. Then summary force and torque on the clump are Fc = Fi , Tc = ri Fi + Ti . Motion of the clump is then integrated, using aspherical rotation integration. Afterwards, clump members are displaced in global space, to keep their initial positions and orientations in the clumps local coordinate system. In such a way, relative positions of clump members are always the same, resulting in the behavior of a rigid aggregate.
6.5.5 Numerical damping

In simulations of quasi-static phenomena, it it desirable to dissipate kinetic energy of particles. Since most constitutive laws (including Law_ScGeom_FrictPhys_Basic shown above, Stress evaluation (example)) do not include velocity-based damping (such as one in [Addetta2001]), it is possible to use articial numerical damping. The formulation is described in [Pfc3dManual30], although our version is slightly adapted. The basic idea is to decrease forces which increase the particle velocities and vice versa by (F)d , comparing the current acceleration sense and particle velocity sense. This is done by component, which makes the damping scheme clearly non-physical, as it is not invariant with respect to coordinate system rotation; on the other hand, it is very easy to compute. Cundall proposed the form (we omit particle indices i since it applies to all of them separately): (F)dw = d sgn(Fw u w ), Fw w {x, y, z}
where d is the damping coecient. This formulation has several advantages [Hentz2003]: it acts on forces (accelerations), not constraining uniform motion; it is independent of eigenfrequencies of particles, they will be all damped equally; it needs only the dimensionless parameter d which does not have to be scaled. In Yade, we use the adapted form ( ) (F)dw u w t w + = d sgn Fw u , Fw 2
u w
(6.9)
by its on-step estimate in parentheses. This is to where we replaced the previous mid-step velocity u avoid locked-in forces that appear if the velocity changes its sign due to force application at each step, i.e. when the particle in question oscillates around the position of equilibrium with 2t period. In Yade, damping (6.9) is implemented in the NewtonIntegrator engine; the damping coecient d is NewtonIntegrator.damping. 6.5. Motion integration 135
6.5.6 Stability considerations

Critical timestep In order to ensure stability for the explicit integration sceheme, an upper limit is imposed on t: tcr = 2 max (6.10)
where max is the highest eigenfrequency within the system.

Single mass-spring system
Single 1D mass-spring system with mass m and stiness K is governed by the equation = Kx mx
where x is displacement from the mean (equilibrium) position. The solution of harmonic oscillation is x(t) = A cos(t + ) where phase and amplitude A are determined by initial conditions. The angular frequency K (1 ) (6.11) = m does not depend on initial conditions. Since there is one single mass, max = (1) . Plugging (6.11) into (6.10), we obtain 1) (1) t( cr = 2/max = 2 m/K
(1)
for a single oscillator.

General mass-spring system
In a general mass-spring system, the highest frequency occurs if two connected masses mi , mj are in opposite motion; let us suppose they have equal velocities (which is conservative) and they are connected by a spring with stiness Ki : displacement xi of mi will be accompained by xj = xi of mj , giving (2) Fi = Ki (xi (xi )) = 2Ki xi . That results in apparent stiness Ki = 2Ki , giving maximum eigenfrequency of the whole system (2) max = max Ki /mi .
i
The overall critical timestep is then tcr = 2 max = min 2

i
mi
(2) Ki
= min 2
i
mi = min 2 i 2Ki
mi . Ki
(6.12)
This equation can be used for all 6 degrees of freedom (DOF) in translation and rotation, by considering 1 i . generalized mass and stiness matrices M and K, and replacing fractions m Ki by eigen values of M.K The critical timestep is then associated to the eigen mode with highest frequency : tcr = min tcr k , k {1, ..., 6}. (6.13)
136
DEM simulations
In DEM simulations, per-particle stiness Kij is determined from the stinesses of contacts in which it participates [Chareyre2005]. Suppose each contact has normal stiness KNk , shear stiness KTk = KNk and is oriented by normal nk . A translational stiness matrix Kij can be dened as the sum of contributions of all contacts in which it participates (indices k), as Kij = (KNk KTk )ni nj + KTk = KNk ((1 )ni nj + ) (6.14)
k j
with i and j {x, y, z}. Equations (6.13) and (6.14) determine tcr in a simulation. A similar approach generalized to all 6 DOFs is implemented by the GlobalStinessTimeStepper engine in Yade. The derivation of generalized stiness including rotational terms is very similar but not developped here, for simplicity. For full reference, see PFC3D - Theoretical Background. Note that for computation eciency reasons, eigenvalues of the stiness matrices are not computed. They are only approximated assuming than DOFs are uncoupled, and using diagonal terms of K.M1 . They give good approximates in typical mechanical systems. There is one important condition that max > 0: if there are no contacts between particles and max = 0, we would obtain value tcr = . While formally correct, this value is numerically erroneous: we were silently supposing that stiness remains constant during each timestep, which is not true if contacts are created as particles collide. In case of no contact, therefore, stiness must be pre-estimated based on future interactions, as shown in the next section. Estimation of tcr by wave propagation speed Estimating timestep in absence of interactions is based on the connection between interaction stinesses and the particles properties. Note that in this section, symbols E and refer exceptionally to Youngs modulus and density of particles, not of macroscopic arrangement. In Yade, particles have associated Material which denes density (Material.density), and also may dene (in ElastMat and derived classes) particles Youngs modulus E (ElastMat.young). is used when particles mass m is initially computed from its , while E is taken in account when creating new interaction between particles, aecting stiness KN . Knowing m and KN , we can estimate (6.14) for each particle; we obviously neglect number of interactions per particle Ni ; for a reasonable radius distribution, however, there is a geometrically imposed upper limit (6 for a packing of spheres with equal radii, for instance); the exact relationship the between particles rigidities Ei , Ej , supposing only that KN is somehow proportional to them. By dening E and , particles have continuum-like quantities. Explicit integration schemes for continuum equations impose a critical timestep based on sonic speed E/; the elastic wave must not propagate farther than the minimum distance of integration points lmin during one step. Since E, are parameters of the elastic continuum and lmin is xed beforehand, we obtain c) t( = l . min cr E For our purposes, we dene E and for each particle separately; lmin can be replaced by the spheres radius Ri ; technically, lmin = 2Ri could be used, but because of possible interactions of spheres and facets (which have zero thickness), we consider lmin = Ri instead. Then i p) t( = min R . i cr i Ei This algorithm is implemented in the utils.PWaveTimeStep function. Let us compare this result to (6.12); this necessitates making several simplifying hypotheses: 6.5. Motion integration 137
all particles are spherical and have the same radius R; the spheres material has the same E and the average number of contacts per sphere is N; the contacts have suciently uniform spatial distribution around each particle; the = KN /KT ratio is constant for all interactions; contact stiness KN is computed from E using a formula of the form KN = E R , (6.15)
where is some constant depending on the algorithm in usefootnote{For example, = /2 in the concrete particle model (Ip2_CpmMat_CpmMat_CpmPhys), while = 2 in the classical DEM model (Ip2_FrictMat_FrictMat_FrictPhys) as implemented in Yade.} and R is half-distance between spheres in contact, equal to R for the case of interaction radius RI = 1. If RI = 1 (and R R by consequence), all interactions will have the same stiness KN . In other cases, we will consider KN as the average stiness computed from average R (see below). As all particles have the same parameters, we drop the i index in the following formulas. We try to express the average per-particle stiness from (6.14). It is a sum over all interactions where KN and are scalars that will not rotate with interaction, while nw is w-th component of unit interaction 2 normal n. Since we supposed uniform spatial distribution, we can replace n2 w by its average value nw . 2 Recognizing components of n as direction cosines, the average values of nw is 1/3. %we nd the average value by integrating over all possible orientations, which are uniformly distributed in space: Moreover, since all directions are equal, we can write the per-body stiness as K = Kw for all w {x, y, z}. We obtain ( ) 1 1 2 K= KN (1 ) + = KN 3 3
and can put constant terms (everything) in front of the summation. per sphere, i.e. N. Arriving at K = NKN 1 2 , 3
1 equals the number of contacts
we substitute K into (6.12) using (6.15): tcr = 2 m = 2 K

4 3 3 R 2 NE R 1 3
=R
tcr
2 E
/ . N(1 2)
(p)
The ratio of timestep tcr predicted by the p-wave velocity and numerically stable timestep tcr is the inverse value of the last (dimensionless) term: (p) N(1 + ) tcr =2 . tcr /
(p)
Actual values of this ratio depend on characteristics of packing N, KN /KT = ratio and the way of computing contact stiness from particle rigidity. Let us show it for two models in Yade: Concrete particle model computes contact stiness from the equivalent area Aeq rst (6.6), Aeq = R2 KN = Aeq E . d0 Chapter 6. DEM Background
138
d0 is the initial contact length, which will be, for interaction radius (6.5) RI > 1, in average larger than 2R. For RI = 1.5 (sect.~ref{sect-calibration-elastic-properties}), we can roughly estimate d0 = 1.25 2R = 5 2 R, getting ( ) 2 KN = E R 5
where
2 5
= by comparison with (6.15).
Interaction radius RI = 1.5 leads to average N 12 interactions per sphere for dense packing of spheres with the same radius R. = 0.2 is calibrated (sect.~ref{sect-calibration-elastic-properties}) to match the desired macroscopic Poissons ratio = 0.2. Finally, we obtain the ratio tcr =2 tcr
(p)
12(1 2 0.2)
(2/5)
= 3.39,
showing signicant overestimation by the p-wave algorithm. Non-cohesive dry friction model is the basic model proposed by Cundall explained in ref{sectformulation-stress-cundall}. Supposing almost-constant sphere radius R and rather dense packing, each sphere will have N = 6 interactions on average (that corresponds to maximally dense packing of spheres with a constant radius). If we use the Ip2_FrictMat_FrictMat_FrictPhys class, we have = 2, as KN = E2R; we again use = 0.2 (for lack of a more signicant value). In this case, we obtain the result (p) tcr 6(1 2 0.2) =2 = 3.02 tcr /2
which again overestimates the numerical critical timestep. To conclude, p-wave timestep gives estimate proportional to the real tcr , but in the cases shown, the (p) value of about t = 0.3tcr should be used to guarantee stable simulation. Non-elastic t constraints Let us note at this place that not only tcr assuring numerical stability of motion integration is a constraint. In systems where particles move at relatively high velocities, position change during one timestep can lead to non-elastic irreversible eects such as damage. The t needed for reasonable result can be lower tcr . We have no rigorously derived rules for such cases.
6.6 Periodic boundary conditions

While most DEM simulations happen in R3 space, it is frequently useful to avoid boundary eects by using periodic space instead. In order to satisfy periodicity conditions, periodic space is created by repetition of parallelepiped-shaped cell. In Yade, periodic space is implemented in the Cell class. The geometry of the cell in the reference coordinates system is dened by three edges of the parallepiped. The corresponding base vectors are stored in the columns of matrix H (Cell.hSize). The initial H can be explicitly dened as a 3x3 matrix at the beginning of the simulation. There are no restricitions on the possible shapes: any parallelepiped is accepted as the initial cell. If the base vectors are axis-aligned, dening only their sizes can be more convenient than dening the full H matrix; in that case it is enough to dene the norms of columns in H (see Cell.size).
6.6. Periodic boundary conditions
139
After the denition of the initial cells geometry, H should generally not be modied by direct assignment. Instead, its deformation rate will be dened via the velocity gradient Cell.velGrad described below. It is the only variable that let the period deformation be correctly accounted for in constitutive laws and Newton integrator (NewtonIntegrator).
6.6.1 Deformations handling

The deformation of the cell over time is dened via a matrix representing the gradient of an homogeneous velocity eld v (Cell.velGrad). This gradient represents arbitrary combinations of rotations and stretches. It can be imposed externaly or updated by boundary controllers (see PeriTriaxController or Peri3dController) in order to reach target strain values or to maintain some prescribed stress. The velocity gradient is integrated automatically over time, and the cumulated transformation is reected in the transformation matrix F (Cell.trsf) and the current shape of the cell H. The per-step transformation update reads (it is similar for H), with I the identity matrix: F+ = (I + vt)F .
F can be set back to identity at any point in simulations, in order to dene the current state as reference for strains denition in boundary controllers. It will have no eect on H. Along with the automatic integration of cell transformation, there is an option to homothetically displace all particles so that v is applied over the whole simulation (enabled via Cell.homoDeform). This avoids all boundary eects coming from change of the velocity gradient.
6.6.2 Collision detection in periodic cell

In usual implementations, particle positions are forced to be inside the cell by wrapping their positions if they get over the boundary (so that they appear on the other side). As we wanted to avoid abrupt changes of position (it would make particles velocity inconsistent with step displacement change), a dierent method was chosen. Approximate collision detection Pass 1 collision detection (based on sweep and prune algorithm, sect.~ref{sect-sweep-and-prune}) operates on axis-aligned bounding boxes (Aabb) of particles. During the collision detection phase, bounds of all Aabbs are wrapped inside the cell in the rst step. At subsequent runs, every bound remembers by how many cells it was initially shifted from coordinate given by the Aabb and uses this oset repeatedly as it is being updated from Aabb during particles motion. Bounds are sorted using the periodic insertion sort algorithm (sect.~ref{sect-periodic-insertion-sort}), which tracks periodic cell boundary ||. Upon inversion of two Aabbs, their collision along all three axes is checked, wrapping real coordinates inside the cell for that purpose. This algorithm detects collisions as if all particles were inside the cell but without the need of constructing ghost particles (to represent periodic image of a particle which enters the cell from the other side) or changing the particles positions. It is required by the implementation (and partly by the algorithm itself) that particles do not span more than half of the current cell size along any axis; the reason is that otherwise two (or more) contacts between both particles could appear, on each side. Since Yade identies contacts by Body.id of both bodies, they would not be distinguishable. In presence of shear, the sweep-and-prune collider could not sort bounds independently along three axes: collision along x axis depends on the mutual position of particles on the y axis. Therefore, bounding boxes are expressed in transformed coordinates which are perpendicular in the sense of collision detection. This requires some extra computation: Aabb of sphere in transformed coordinates will no longer be cube, but cuboid, as the sphere itself will appear as ellipsoid after transformation. Inversely, the sphere in simulation space will have a parallelepiped bounding box, which is cuboid around the ellipsoid in 140 Chapter 6. DEM Background
transformed axes (the Aabb has axes aligned with transformed cell basis). This is shown in g. g-cellshear-aabb. The restriction of a single particle not spanning more than half of the transformed axis becomes stringent as Aabb is enlarged due to shear. Considering Aabb of a sphere with radius r in the cell where x x, z z, but (y, y ) = , the x-span of the Aabb will be multiplied by 1/ cos . For the innite shear /2, which can be desirable to simulate, we have 1/ cos . Fortunately, this limitation can be easily circumvented by realizing the quasi-identity of all periodic cells which, if repeated in space, create the same grid with their corners: the periodic cell can be ipped, keeping all particle interactions intact, as shown in g. g-cell-ip. It only necessitates adjusting the Interaction.cellDist of interactions and re-initialization of the collider (Collider::invalidatePersistentData). Cell ipping is implemented in the utils.ipCell function.
y 1 y1 y2 2 y
x1
x2 x1
Figure 6.8: Flipping cell (utils.ipCell) to avoid innite stretch of the bounding boxes spans with growing . Cell ip does not aect interactions from the point of view of the simulation. The periodic arrangement on the left is the same as the one on the right, only the cell is situated dierently between identical grid points of repetition; at the same time |2 | < |1 | and sphere bounding boxs x-span stretched by 1/ cos becomes smaller. Flipping can be repeated, making eective innite shear possible. This algorithm is implemented in InsertionSortCollider and is used whenever simulation is periodic (Omega.isPeriodic); individual BoundFunctors are responsible for computing sheared Aabbs; currently it is implemented for spheres and facets (in Bo1_Sphere_Aabb and Bo1_Facet_Aabb respectively).
yy y y
x xx x
Figure 6.9: Constructing axis-aligned bounding box (Aabb) of a sphere in simulation space coordinates (without periodic cell left) and transformed cell coordinates (right), where collision detection axes x , y are not identical with simulation space axes x, y. Bounds projection to axes is shown by orange lines.
Exact collision detection When the collider detects approximate contact (on the Aabb level) and the contact does not yet exist, it creates potential contact, which is subsequently checked by exact collision algorithms (depending on the combination of Shapes). Since particles can interact over many periodic cells (recall we never change their positions in simulation space), the collider embeds the relative cell coordinate of particles
6.6. Periodic boundary conditions
141
in the interaction itself (Interaction.cellDist) as an integer vector c. Multiplying current cell size T s by c component-wise, we obtain particle oset x in aperiodic R3 ; this value is passed (from InteractionLoop) to the functor computing exact collision (IGeomFunctor), which adds it to the position of the particle Interaction.id2. By storing the integral oset c, x automatically updates as cell parameters change. Periodic insertion sort algorithm The extension of sweep and prune algorithm (described in Sweep and prune) to periodic boundary conditions is non-trivial. Its cornerstone is a periodic variant of the insertion sort algorithm, which involves keeping track of the period of each boundary; e.g. taking period 0, 10), then 81 22 < 22 (subscript indicating period). Doing so eciently (without shuing data in memory around as bound wraps from one period to another) requires moving period boundary rather than bounds themselves and making the comparison work transparently at the edge of the container. This algorithm was also extended to handle non-orthogonal periodic Cell boundaries by working in transformed rather than Cartesian coordinates; this modies computation of Aabb from Cartesian coordinates in which bodies are positioned (treated in detail in Approximate collision detection). The sort algorithm is tracking Aabb extrema along all axes. At the colliders initialization, each value is assigned an integral period, i.e. its distance from the cells interior expressed in the cells dimension along its respective axis, and is wrapped to a value inside the cell. We put the period number in subscript. Let us give an example of coordinate sequence along x axis (in a real case, the number of elements would be even, as there is maximum and minimum value couple for each particle; this demonstration only shows the sorting algorithm, however.) 41 122 || 12 24 50
with cell x-size sx = 10. The 41 value then means that the real coordinate xi of this extremum is xi + 1 10 = 4, i.e. xi = 4. The || symbol denotes the periodic cell boundary. Sorting starts from the rst element in the cell, i.e. right of ||, and inverts elements as in the aperiodic variant. The rules are, however, more complicated due to the presence of the boundary ||: () stop inverting if neighbors are ordered; (||) current element left of || is below 0 (lower period boundary); in this case, decrement elements period, decrease its coordinate by sx and move || right; (||) current element right of || is above sx (upper period boundary); increment elements period, increase its coordinate by sx and move || left; ( < || ) inversion across || must subtract sx from the left coordinate during comparison. If the elements are not in order, they are swapped, but they must have their periods changed as they traverse ||. Apply (||) if necessary; (||) if after ( < || ) the element that is now right of || has xi < sx , decrease its coordinate by sx and decrement its period. Do not move ||. In the rst step, (||) is applied, and inversion with 122 happens; then we stop because of (): 41 41 41 i 122 122 j 91
||
12 91
24 || 24 || 24
50 , 50 , 50 .
122
142
We move to next element 24 ; rst, we apply (||), then invert until (): 41 41 41 41 i 91 91 91 i 83
122 122 j 83
||
24 83
50 , || 50 , || 50 , || 50 .
122 122
91
The next element is 50 ; we satisfy ( < || ), therefore instead of comparing 122 > 50 , we must do (122 sx ) = 23 5; we adjust periods when swapping over || and apply (||), turning 122 into 23 ; then we keep inverting, until (): 41 41 41 41 i 83 83 83 j
91 91 j
122 j || 51 91 91
50 , 23 , 23 , 23 .
|| || ||
51 83
51
We move (wrapping around) to 41 , which is ordered: 41 51 83
91 || 8 23
and so is the last element 41 i 51
83
91 || 23 .
6.7 Computational aspects

6.7.1 Cost
The DEM computation using an explicit integration scheme demands a relatively high number of steps during simulation, compared to implicit scehemes. The total computation time Z of simulation spanning T seconds (of simulated time), containing N particles in volume V depends on: linearly, the number of steps i = T/(st tcr ), where st is timestep safety factor; tcr can be estimated (p) by p-wave velocity using E and (sect.~ref{sect-dt-pwave}) as tcr = r E . Therefore T E . i= st r
the number of particles N; for xed value of simulated domain volume V and particle radius r N=p4 V , 3 r 3
6.7. Computational aspects
143
where p is packing porosity, roughly
1 2
for dense irregular packings of spheres of similar radius.
The dependency is not strictly linear (which would be the best case), as some algorithms do not scale linearly; a case in point is the sweep and prune collision detection algorithm introduced in sect.~ref{sect-sweep-and-prune}, with scaling roughly O(N log N). The number of interactions scales with N, as long as packing characteristics are the same. the number of computational cores ncpu ; in the ideal case, the dependency would be inverse-linear were all algorithms parallelized (in Yade, collision detection is not). Let us suppose linear scaling. Additionally, let us suppose that the material to be simulated (E, ) and the simulation setup (V , T ) are given in advance. Finally, dimensionless constants st , p and ncpu will have a xed value. This leaves us with one last degree of freedom, r. We may write 1 T V 1 1 E 1 1 Z iN = p4 3 = 4. 3 ncpu st r 3 r ncpu rr r
This (rather trivial) result is essential to realize DEM scaling; if we want to have ner results, rening the mesh by halving r, the computation time will grow 24 = 16 times. For very crude estimates, one can use a known simulation to obtain a machine constant = Z Ni
with the meaning of time per particle and per timestep (in the order of 106 s for current machines). will be only useful if simulation characteristics are similar and non-linearities in scaling do not have major inuence, i.e. N should be in the same order of magnitude as in the reference case.
6.7.2 Result indeterminism

It is naturally expected that running the same simulation several times will give exactly the same results: although the computation is done with nite precision, round-o errors would be deterministically the same at every run. While this is true for single-threaded computation where exact order of all operations is given by the simulation itself, it is not true anymore in multi-threaded computation which is described in detail in later sections. The straight-forward manner of parallel processing in explicit DEM is given by the possibility of treating interactions in arbitrary order. Strain and stress is evaluated for each interaction independently, but forces from interactions have to be summed up. If summation order is also arbitrary (in Yade, forces are accumulated for each thread in the order interactions are processed, then summed together), then the results can be slightly dierent. For instance
(1/10.)+(1/13.)+(1/17.)=0.23574660633484162 (1/17.)+(1/13.)+(1/10.)=0.23574660633484165
As forces generated by interactions are assigned to bodies in quasi-random order, summary force Fi on the body can be dierent between single-threaded and multi-threaded computations, but also between dierent runs of multi-threaded computation with exactly the same parameters. Exact thread scheduling by the kernel is not predictable since it depends on asynchronous events (hardware interrupts) and other unrelated tasks running on the system; and it is thread scheduling that ultimately determines summation order of force contributions from interactions.
144
Numerical damping inuence The eect of summation order can be signicantly amplied by the usage of a discontinuous damping function in NewtonIntegrator given in (6.9) as ( ) (F)dw u w t w + = d sgn Fw u . Fw 2
If the sgn argument is close to zero then the least signicant nite precision artifact can determine whether the equation (relative increment of Fw ) is +d or d . Given commonly used values of d = 0.4, it means that such artifact propagates from least signicant place to the most signicant one at once.
6.7. Computational aspects
145
146
Chapter 7
Class reference (yade.wrapper module)
7.1 Bodies
7.1.1 Body
class yade.wrapper.Body(inherits Serializable) A particle, basic element of simulation; interacts with other bodies. aspherical(=false) Whether this body has dierent inertia along principal axes; NewtonIntegrator makes use of this ag to call rotation integration routine for aspherical bodies, which is more expensive. bound(=uninitalized ) Bound, approximating volume for the purposes of collision detection. bounded(=true) Whether this body should have Body.bound created. Note that bodies without a bound do not participate in collision detection. (In c++, use Body::isBounded/Body::setBounded) chain Returns Id of chain to which the body belongs. clumpId Id of clump this body makes part of; invalid number if not part of clump; see Body::isStandalone, Body::isClump, Body::isClumpMember properties. Not meant to be modied directly from Python, use O.bodies.appendClumped instead. dict() dict Return dictionary of attributes. dynamic(=true) Whether this body will be moved Body::isDynamic/Body::setDynamic) by forces. (In c++, use
flags(=FLAG_BOUNDED) Bits of various body-related ags. Do not access directly. In c++, use isDynamic/setDynamic, isBounded/setBounded, isAspherical/setAspherical. In python, use Body.dynamic, Body.bounded, Body.aspherical. groupMask(=1) Bitmask for determining interactions.
147
id(=Body::ID_NONE ) Unique id of this body. intrs() list Return all interactions in which this body participates. isClump True if this body is clump itself, false otherwise. isClumpMember True if this body is clump member, false otherwise. isStandalone True if this body is neither clump, nor clump member; false otherwise. iterBorn Returns step number at which the body was added to simulation. mask Shorthand for Body::groupMask mat Shorthand for Body::material material(=uninitalized ) Material instance associated with this body. shape(=uninitalized ) Geometrical Shape. state(=new State) Physical state. timeBorn Returns time at which the body was added to simulation. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.1.2 Shape
Box
Facet
Cylinder
ChainedCylinder
S phere S hape W all
GridConnection
GridNode
Tetra
Clump
class yade.wrapper.Shape(inherits Serializable) Geometry of a body
148
Chapter 7. Class reference (yade.wrapper module)
color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Box(inherits Shape Serializable) Box (cuboid) particle geometry. (Avoid using in new code, prefer Facet instead. color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. extents(=uninitalized ) Half-size of the cuboid highlight(=false) Whether this Shape will be highlighted when rendered. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.ChainedCylinder(inherits Cylinder Sphere Shape Serializable) Geometry of a deformable chained cylinder, using geometry Cylinder. chainedOrientation(=Quaternionr::Identity()) Deviation of node1 orientation from node-to-node vector color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself,
7.1. Bodies
149
top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. initLength(=0 ) tensile-free length, used as reference for tensile strain length(=NaN ) Length [m] radius(=NaN ) Radius [m] segment(=Vector3r::Zero()) Length vector updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Clump(inherits Shape Serializable) Rigid aggregate of bodies color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. members Return clump members as {id1:(relPos,relOri),...} updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Cylinder(inherits Sphere Shape Serializable) Geometry of a cylinder, as Minkowski sum of line and sphere. color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself,
150
top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. length(=NaN ) Length [m] radius(=NaN ) Radius [m] segment(=Vector3r::Zero()) Length vector updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Facet(inherits Shape Serializable) Facet (triangular particle) geometry. area(=NaN ) Facets area color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. normal(=Vector3r(NaN, NaN, NaN)) Facets normal updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vertices(=vector<Vector3r>(3, Vector3r(NaN, NaN, NaN))) Vertex positions in local coordinates. wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.GridConnection(inherits Sphere Shape Serializable) GridConnection shape. Component of a grid designed to link two GridNodes. Its highly recommended to use utils.gridConnection(...) to generate correct GridConnections. cellDist(=Vector3i(0, 0, 0)) missing doc :( color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB).
7.1. Bodies
151
dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. node1(=uninitalized ) First Body the GridConnection is connected to. node2(=uninitalized ) Second Body the GridConnection is connected to. periodic(=false) true if two nodes from dierent periods are connected. radius(=NaN ) Radius [m] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.GridNode(inherits Sphere Shape Serializable) GridNode shape, component of a grid. To create a Grid, place the nodes rst, they will dene the spacial discretisation of it. Its highly recommended to use utils.gridNode(...) to generate correct GridNodes. Note that the GridNodes should only be in an Interaction with other GridNodes. The Sphere-Grid contact is only handled by the GridConnections. ConnList(=uninitalized ) List of GridConnections the GridNode is connected to. addConnection((Body)Body) None Add a GridConnection to the GridNode. color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. radius(=NaN ) Radius [m] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
152
wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Sphere(inherits Shape Serializable) Geometry of spherical particle. color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. radius(=NaN ) Radius [m] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Tetra(inherits Shape Serializable) Tetrahedron geometry. color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary v(=std::vector<Vector3r>(4)) Tetrahedron vertices in global coordinate system. wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer). class yade.wrapper.Wall(inherits Shape Serializable) Object representing innite plane aligned with the coordinate system (axis-aligned wall).
7.1. Bodies
153
axis(=0 ) Axis of the normal; can be 0,1,2 for +x, +y, +z respectively (Bodys orientation is disregarded for walls) color(=Vector3r(1, 1, 1)) Color for rendering (normalized RGB). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. highlight(=false) Whether this Shape will be highlighted when rendered. sense(=0 ) Which side of the wall interacts: -1 for negative only, 0 for both, +1 for positive only updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Whether this Shape is rendered using color surfaces, or only wireframe (can still be overridden by global cong of the renderer).
7.1.3 State
JCFpmS tate
RpmS tate
ChainedS tate S tate CpmS tate
CFpmS tate
WireS tate
class yade.wrapper.State(inherits Serializable) State of a body (spatial conguration, internal variables). angMom(=Vector3r::Zero()) Current angular momentum angVel(=Vector3r::Zero()) Current angular velocity
154
blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system. isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body ori Current orientation. pos Current position. refOri(=Quaternionr::Identity()) Reference orientation refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity. class yade.wrapper.CFpmState(inherits State Serializable) CFpm state information about each body. None of that is used for computation (at least not now), only for post-processing. angMom(=Vector3r::Zero()) Current angular momentum angVel(=Vector3r::Zero()) Current angular velocity blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to
7.1. Bodies
155
an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system. isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body numBrokenCohesive(=0 ) Number of broken cohesive links. [-] ori Current orientation. pos Current position. refOri(=Quaternionr::Identity()) Reference orientation refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity. class yade.wrapper.ChainedState(inherits State Serializable) State of a chained bodies, containing information on connectivity in order to track contacts jumping over contiguous elements. Chains are 1D lists from which id of chained bodies are retrieved via rank and chainNumber. addToChain((int)bodyId ) None Add body to current active chain angMom(=Vector3r::Zero()) Current angular momentum
156
angVel(=Vector3r::Zero()) Current angular velocity bId(=-1) id of the body containing - for postLoad operations only. blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). chainNumber(=0 ) chain id. currentChain = 0 densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system. isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body ori Current orientation. pos Current position. rank(=0 ) rank in the chain. refOri(=Quaternionr::Identity()) Reference orientation refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity.
7.1. Bodies
157
class yade.wrapper.CpmState(inherits State Serializable) State information about body use by cpm-model. None of that is used for computation (at least not now), only for post-processing. angMom(=Vector3r::Zero()) Current angular momentum angVel(=Vector3r::Zero()) Current angular velocity blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). damageTensor(=Matrix3r::Zero()) Damage tensor computed with microplane theory averaging. state.damageTensor.trace() = state.normDmg densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) epsVolumetric(=0 ) Volumetric strain around this body (unused for now) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system. isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body normDmg(=0 ) Average damage including already deleted contacts (it is really not damage, but 1relResidualStrength now) numBrokenCohesive(=0 ) Number of (cohesive) contacts that damaged completely numContacts(=0 ) Number of contacts with this body ori Current orientation. pos Current position. refOri(=Quaternionr::Identity()) Reference orientation
158
refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. stress(=Matrix3r::Zero()) Stress tensor of the spherical particle (under assumption that particle volume = pi*r*r*r*4/3.) for packing fraction 0.62 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity. class yade.wrapper.JCFpmState(inherits State Serializable) JCFpm state information about each body. None of that is used for computation (at least not now), only for post-processing. angMom(=Vector3r::Zero()) Current angular momentum angVel(=Vector3r::Zero()) Current angular velocity blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system. isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body ori Current orientation. pos Current position.
7.1. Bodies
159
refOri(=Quaternionr::Identity()) Reference orientation refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. shearBreak(=0 ) number of shear breakages. [-] tensBreak(=0 ) number of tensile breakages. [-] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity. class yade.wrapper.RpmState(inherits State Serializable) State information about Rpm body. angMom(=Vector3r::Zero()) Current angular momentum angVel(=Vector3r::Zero()) Current angular velocity blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system. isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body ori Current orientation.
160
pos Current position. refOri(=Quaternionr::Identity()) Reference orientation refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. specimenMass(=0 ) Indicates the mass of the whole stone, which owns the particle. specimenMaxDiam(=0 ) Indicates the maximal diametr of the specimen. specimenNumber(=0 ) The variable is used for particle size distribution analyze. Indicates, to which part of specimen belongs para of particles. specimenVol(=0 ) Indicates the mass of the whole stone, which owns the particle. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity. class yade.wrapper.WireState(inherits State Serializable) Wire state information of each body. None of that is used for computation (at least not now), only for post-processing. angMom(=Vector3r::Zero()) Current angular momentum angVel(=Vector3r::Zero()) Current angular velocity blockedDOFs Degress of freedom where linear/angular velocity will be always constant (equal to zero, or to an user-dened value), regardless of applied force/torque. String that may contain xyzXYZ (translations and rotations). densityScaling(=1) (auto-updated) see GlobalStinessTimeStepper::targetDt. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displ() Vector3 Displacement from reference position (pos - refPos) inertia(=Vector3r::Zero()) Inertia of associated body, in local coordinate system.
7.1. Bodies
161
isDamped(=true) Damping in Newtonintegrator can be deactivated for individual particles by setting this variable to FALSE. E.g. damping is inappropriate for particles in free ight under gravity but it might still be applicable to other particles in the same simulation. mass(=0 ) Mass of this body numBrokenLinks(=0 ) Number of broken links (e.g. number of wires connected to the body which are broken). [-] ori Current orientation. pos Current position. refOri(=Quaternionr::Identity()) Reference orientation refPos(=Vector3r::Zero()) Reference position rot() Vector3 Rotation from reference orientation (as rotation vector) se3(=Se3r(Vector3r::Zero(), Quaternionr::Identity())) Position and orientation as one object. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vel(=Vector3r::Zero()) Current linear velocity.
7.1.4 Material
RpmMat
CFpmMat
CpmMat
WireMat Elas tMat FrictMat CohFrictMat LudingMat Material BubbleMat NormalInelas ticMat Vis cElMat JCFpmMat Inelas tCohFrictMat
162
class yade.wrapper.Material(inherits Serializable) Material properties of a body. density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.BubbleMat(inherits Material Serializable) material for bubble interactions, for use with other Bubble classes density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created
7.1. Bodies
163
will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). surfaceTension(=0.07197 ) The surface tension in the uid surrounding the bubbles. The default value is that of water at 25 degrees Celcius. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CFpmMat(inherits FrictMat ElastMat Material Serializable) cohesive frictional material, for use with other CFpm classes density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. type(=0 ) Type of the particle. If particles of two dierent types interact, it will be with friction only (no cohesion).[-] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.CohFrictMat(inherits FrictMat ElastMat Material Serializable) alphaKr(=2.0 ) Dimensionless coecient used for the rolling stiness. alphaKtw(=2.0 ) Dimensionless coecient used for the twist stiness.
164
density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. etaRoll(=-1.) Dimensionless coecient used to calculate the plastic rolling moment (if negative, plasticity will not be applied). frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) isCohesive(=true) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. momentRotationLaw(=false) Use bending/twisting moment at contact. The contact will have moments only if both bodies have this ag true. See CohFrictPhys::cohesionDisablesFriction for details. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). normalCohesion(=0 ) poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. shearCohesion(=0 ) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.CpmMat(inherits FrictMat ElastMat Material Serializable) Concrete material, for use with other Cpm classes. Note: Density is initialized to 4800 kgmautomatically, which gives approximate 2800 kgm on 0.5 density packing. Concrete Particle Model (CPM) CpmMat is particle material, Ip2_CpmMat_CpmMat_CpmPhys averages two particles materials, creating CpmPhys, which is then used in interaction resultion by Law2_ScGeom_CpmPhys_Cpm.
7.1. Bodies
165
CpmState is associated to CpmMat and keeps state dened on particles rather than interactions (such as number of completely damaged interactions). The model is contained in externally dened macro CPM_MATERIAL_MODEL, which features damage in tension, plasticity in shear and compression and rate-dependence. For commercial reasons, rate-dependence and compression-plasticity is not present in reduced version of the model, used when CPM_MATERIAL_MODEL is not dened. The full model will be described in detail in my (Vclav milauer) thesis along with calibration procedures (rigidity, poissons ratio, compressive/tensile strength ratio, fracture energy, behavior under connement, rate-dependent behavior). Even the public model is useful enough to run simulation on concrete samples, such as uniaxial tension-compression test. damLaw(=1) Law for damage evolution in uniaxial tension. 0 for linear stress-strain softening branch, 1 (default) for exponential damage evolution law density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. dmgRateExp(=0 ) Exponent for normal viscosity function. [-] dmgTau(=-1, deactivated if negative) Characteristic time for normal viscosity. [s] epsCrackOnset(=NaN ) Limit elastic strain [-] frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) isoPrestress(=0 ) Isotropic prestress of the whole specimen. [Pa] label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. neverDamage(=false) If true, no damage will occur (for testing only). newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). plRateExp(=0 ) Exponent for visco-plasticity function. [-] 166 Chapter 7. Class reference (yade.wrapper module)
plTau(=-1, deactivated if negative) Characteristic time for visco-plasticity. [s] poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. relDuctility(=NaN ) relative ductility of bonds in normal direction sigmaT(=NaN ) Initial cohesion [Pa] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.ElastMat(inherits Material Serializable) Purely elastic material. The material parameters may have dierent meanings depending on the IPhysFunctor used : true Young and Poisson in Ip2_FrictMat_FrictMat_MindlinPhys, or contact stinesses in Ip2_FrictMat_FrictMat_FrictPhys. density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.FrictMat(inherits ElastMat Material Serializable) Elastic material with contact friction. See also ElastMat.
7.1. Bodies
167
density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.InelastCohFrictMat(inherits FrictMat ElastMat Material Serializable) alphaKr(=2.0 ) Dimensionless coecient used for the rolling stiness. alphaKtw(=2.0 ) Dimensionless coecient used for the twist stiness. compressionModulus(=0.0 ) Compresion elasticity modulus creepBending(=0.0 ) Bending creeping coecient. Usual values between 0 and 1. creepTension(=0.0 ) Tension/compression creeping coecient. Usual values between 0 and 1. creepTwist(=0.0 ) Twist creeping coecient. Usual values between 0 and 1. density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. 168 Chapter 7. Class reference (yade.wrapper module)
dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. epsilonMaxCompression(=0.0 ) Maximal plastic strain compression epsilonMaxTension(=0.0 ) Maximal plastic strain tension etaMaxBending(=0.0 ) Maximal plastic bending strain etaMaxTwist(=0.0 ) Maximal plastic twist strain frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). nuBending(=0.0 ) Bending elastic stress limit nuTwist(=0.0 ) Twist elastic stress limit poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. shearCohesion(=0.0 ) Shear elastic stress limit shearModulus(=0.0 ) shear elasticity modulus sigmaCompression(=0.0 ) Compression elastic stress limit sigmaTension(=0.0 ) Tension elastic stress limit tensionModulus(=0.0 ) Tension elasticity modulus unloadBending(=0.0 ) Bending plastic unload coecient. Usual values between 0 and +innity. unloadTension(=0.0 ) Tension/compression plastic unload coecient. Usual values between 0 and +innity.
7.1. Bodies
169
unloadTwist(=0.0 ) Twist plastic unload coecient. Usual values between 0 and +innity. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.JCFpmMat(inherits FrictMat ElastMat Material Serializable) possibly jointed cohesive frictional material, for use with other JCFpm classes density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) joint(=0 ) indicates the number of joint surfaces to which the particle belongs (0-> no joint, 1->1 joint, etc..). [-] jointNormal1(=Vector3r::Zero()) species the normal direction to the joint plane 1. Rk: the ideal here would be to create a vector of vector wich size is dened by the joint integer (as much joint normals as joints). However, it needs to make the pushback function works with python since joint detection is done through a python script. lines 272 to 312 of cpp le should therefore be adapted. [-] jointNormal2(=Vector3r::Zero()) species the normal direction to the joint plane 2. [-] jointNormal3(=Vector3r::Zero()) species the normal direction to the joint plane 3. [-] label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). onJoint(=false) identies if the particle is on a joint surface. [-] poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor.
170
type(=0 ) if particles of two dierent types interact, it will be with friction only (no cohesion).[-] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.LudingMat(inherits Material Serializable) Material for simple Ludnings model of contact. G0(=NaN ) Viscous damping PhiF(=NaN ) Dimensionless plasticity depth density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=NaN ) Friction angle [rad] id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) k1(=NaN ) Slope of loading plastic branch kc(=NaN ) Slope of irreversible, tensile adhesive branch kp(=NaN ) Slope of unloading and reloading limit elastic branch label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.NormalInelasticMat(inherits FrictMat ElastMat Material Serializable) Material class for particles whose contact obey to a normal inelasticity (governed by this coe_dech).
7.1. Bodies
171
coeff_dech(=1.0 ) =kn(unload) / kn(load) density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.RpmMat(inherits FrictMat ElastMat Material Serializable) Rock material, for use with other Rpm classes. G_over_E(=1) Ratio of normal/shear stiness at interaction level. [-] Zeta(=0 ) Damping Ratio, http://en.wikipedia.org/wiki/Damping_ratio [-] density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance.
172
exampleNumber(=0 ) Number of the specimen. This value is equal for all particles of one specimen. [-] frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) initCohesive(=false) The ag shows, whether particles of this material can be cohesive. [-] label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. stressCompressMax(=0 ) Maximal strength for compression. The main destruction parameter. [Pa] //(Needs to be reworked) stressShearMax(=0 ) Maximal strength for shearing. [Pa] stressStretchingMax(=0 ) Maximal strength for stretching. [Pa] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor. class yade.wrapper.ViscElMat(inherits Material Serializable) Material for simple viscoelastic model of contact. Note: Shop::getViscoelasticFromSpheresInteraction (and utils.getViscoelasticFromSpheresInteraction in python) compute kn, cn, ks, cs from analytical solution of a pair spheres interaction problem. Capillar(=false) True, if capillar forces need to be added. CapillarType(= ) Dierent types of capillar interaction: Willett_numeric, Willett_analytic [Willett2000] , Weigert [Weigert1999] , Rabinovich [Rabinov2005] Vb(=NaN ) Liquid bridge volume [m^3] cn(=NaN ) Normal viscous constant cs(=NaN ) Shear viscous constant 7.1. Bodies 173
density(=1000 ) Density of the material [kg/m] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=NaN ) Friction angle [rad] gamma(=NaN ) Surface tension [N/m] id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) kn(=NaN ) Normal elastic stiness ks(=NaN ) Shear elastic stiness label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. mR(=0.0 ) Rolling resistance, see [Zhou1999536]. mRtype(=1) Rolling resistance type, see [Zhou1999536]. mRtype=1 - equation (3) in [Zhou1999536]; mRtype=2 - equation (4) in [Zhou1999536]. massMultiply(=true) Stiness and viscosity are multiplied by the reduced mass. If massMultiply=false, these parameter are set explicitly without mass multiplication newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). theta(=NaN ) Contact angle [] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.WireMat(inherits FrictMat ElastMat Material Serializable) Material for use with the Wire classes as(=0.) Cross-section area of a single wire used to transform stress into force. [m] density(=1000 ) Density of the material [kg/m]
174
diameter(=0.0027 ) Diameter of the single wire in [m] (the diameter is used to compute the cross-section area of the wire). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=.5 ) Contact friction angle (in radians). Hint : use radians(degreesValue) in python scripts. id(=-1, not shared ) Numeric id of this material; is non-negative only if this Material is shared (i.e. in O.materials), -1 otherwise. This value is set automatically when the material is inserted to the simulation via O.materials.append. (This id was necessary since before boost::serialization was used, shared pointers were not tracked properly; it might disappear in the future) isDoubleTwist(=false) Type of the mesh. If true two particles of the same material which body ids dier by one will be considered as double-twisted interaction. label(=uninitalized ) Textual identier for this material; can be used for shared materials lookup in MaterialContainer. lambdaEps(=0.47 ) Parameter between 0 and 1 to reduce strain at failure of a double-twisted wire (as used by [Bertrand2008]). [-] lambdaF(=1.0 ) Parameter between 0 and 1 introduced by [Thoeni2013] which denes where the shifted forcedisplacement curve intersects with the new initial stiness: F = F Felastic . [-] lambdak(=0.73 ) Parameter between 0 and 1 to compute the elastic stiness of a double-twisted wire (as used by [Bertrand2008]): kD = 2(k kh + (1 k )kS ). [-] lambdau(=0.2 ) Parameter between 0 and 1 introduced by [Thoeni2013] which denes the maximum shift of the force-displacement curve in order to take an additional initial elongation (e.g. wire distortion/imperfections, slipping, system exibility) into account: l = u l0 rnd(seed). [-] newAssocState() State Return new State instance, which is associated with this Material. Some materials have special requirement on Body::state type and calling this function when the body is created will ensure that they match. (This is done automatically if you use utils.sphere, functions from python). poisson(=.25 ) Poissons ratio or the ratio between shear and normal stiness [-]. It has dierent meanings depending on the Ip functor. seed(=12345 ) Integer used to initialize the random number generator for the calculation of the distortion. If the integer is equal to 0 a internal seed number based on the time is computed. [-] strainStressValues(=uninitalized ) Piecewise linear denition of the stress-strain curve by set of points (strain[-]>0,stress[Pa]>0) for one single wire. Tension only is considered and the point (0,0) is not needed! NOTE: Vector needs to be initialized! 7.1. Bodies 175
strainStressValuesDT(=uninitalized ) Piecewise linear denition of the stress-strain curve by set of points (strain[-]>0,stress[Pa]>0) for the double twist. Tension only is considered and the point (0,0) is not needed! If this value is given the calculation will be based on two dierent stress-strain curves without considering the parameter introduced by [Bertrand2008] (see [Thoeni2013]). type Three dierent types are considered: 0 1 2 Corresponds to Bertrands approach (see [Bertrand2008]): only one stress-strain curve is used New approach: two separate stress-strain curves can be used (see [Thoeni2013]) New approach with stochastically distorted contact model: two separate stress-strain curves with changed initial stiness and horizontal shift (shift is random if seed 0, for more details see [Thoeni2013])
By default the type is 0. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary young(=1e9 ) elastic modulus [Pa]. It has dierent meanings depending on the Ip functor.
7.1.5 Bound
Bound
Aabb
class yade.wrapper.Bound(inherits Serializable) Object bounding part of space taken by associated body; might be larger, used to optimalize collision detection color(=Vector3r(1, 1, 1)) Color for rendering this object dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. lastUpdateIter(=0 ) record iteration of last reference position update (auto-updated) max(=Vector3r(NaN, NaN, NaN)) Lower corner of box containing this bound (and the Body as well) min(=Vector3r(NaN, NaN, NaN)) Lower corner of box containing this bound (and the Body as well) refPos(=Vector3r(NaN, NaN, NaN)) Reference position, updated at current body position each time the bound dispatcher update bounds (auto-updated) sweepLength(=0 ) The length used to increase the bounding boxe size, can be adjusted on the basis of previous displacement if BoundDispatcher::targetInterv>0. (auto-updated) 176 Chapter 7. Class reference (yade.wrapper module)
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Aabb(inherits Bound Serializable) Axis-aligned bounding box, for use with InsertionSortCollider. (This class is quasi-redundant since min,max are already contained in Bound itself. That might change at some point, though.) color(=Vector3r(1, 1, 1)) Color for rendering this object dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. lastUpdateIter(=0 ) record iteration of last reference position update (auto-updated) max(=Vector3r(NaN, NaN, NaN)) Lower corner of box containing this bound (and the Body as well) min(=Vector3r(NaN, NaN, NaN)) Lower corner of box containing this bound (and the Body as well) refPos(=Vector3r(NaN, NaN, NaN)) Reference position, updated at current body position each time the bound dispatcher update bounds (auto-updated) sweepLength(=0 ) The length used to increase the bounding boxe size, can be adjusted on the basis of previous displacement if BoundDispatcher::targetInterv>0. (auto-updated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.2 Interactions
7.2.1 Interaction
class yade.wrapper.Interaction(inherits Serializable) Interaction between pair of bodies. cellDist Distance of bodies in cell size units, if using periodic boundary conditions; id2 is shifted by this number of cells from its State::pos coordinates for this interaction to exist. Assigned by the collider. Warning: (internal) cellDist must survive Interaction::reset(), it is only initialized in ctor. Interaction that was cancelled by the constitutive law, was reset() and became only potential must have thepriod information if the geometric functor again makes it real. Good to know after few days of debugging that :-) dict() dict Return dictionary of attributes. geom(=uninitalized ) Geometry part of the interaction.
7.2. Interactions
177
id1(=0 ) Id of the rst body in this interaction. id2(=0 ) Id of the second body in this interaction. isActive True if this interaction is active. Otherwise the forces from this interaction will not be taken into account. True by default. isReal True if this interaction has both geom and phys; False otherwise. iterBorn(=-1) Step number at which the interaction was added to simulation. iterMadeReal(=-1) Step number at which the interaction was fully (in the sense of geom and phys) created. (Should be touched only by IPhysDispatcher and InteractionLoop, therefore they are made friends of Interaction phys(=uninitalized ) Physical (material) part of the interaction. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.2.2 IGeom
Dem3DofGeom_FacetS phere
Dem3DofGeom_W allS phere Dem3DofGeom Dem3DofGeom_S phereS phere CylS cGeom6D S cGeom6D GridNodeGeom6D GenericS pheres Contact IGeom TTetraGeom L3Geom L6Geom GridCoGridCoGeom S cGeom CylS cGeom ChCylGeom6D S cGridCoGeom
class yade.wrapper.IGeom(inherits Serializable) Geometrical conguration of interaction dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ChCylGeom6D(inherits ScGeom6D ScGeom GenericSpheresContact IGeom Serializable) Test
178
bending(=Vector3r::Zero()) Bending at contact as a vector dening axis of rotation and angle (angle=norm). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). initialOrientation1(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 1 one at initialisation time (auto-updated) initialOrientation2(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 2 one at initialisation time (auto-updated) normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. shearInc(=Vector3r::Zero()) Shear displacement increment in the last step twist(=0 ) Elastic twist angle of the contact. twistCreep(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Stored creep, substracted from total relative rotation for computation of elastic moment (autoupdated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CylScGeom(inherits ScGeom GenericSpheresContact IGeom Serializable) Geometry of a cylinder-sphere contact. contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. 7.2. Interactions 179
dispIndex Return class index of this instance. end(=Vector3r::Zero()) position of 2nd node (auto-updated) id3(=0 ) id of next chained cylinder (auto-updated) incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). isDuplicate(=0 ) this ag is turned true (1) automatically if the contact is shared between two chained cylinders. A duplicated interaction will be skipped once by the constitutive law, so that only one contact at a time is eective. If isDuplicate=2, it means one of the two duplicates has no longer geometric interaction, and should be erased by the constitutive laws. normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) onNode(=false) contact on node? penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. relPos(=0 ) position of the contact on the cylinder (0: node-, 1:node+) (auto-updated) shearInc(=Vector3r::Zero()) Shear displacement increment in the last step start(=Vector3r::Zero()) position of 1st node (auto-updated) trueInt(=-1) Denes the body id of the cylinder where the contact is real, when CylScGeom::isDuplicate>0. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CylScGeom6D(inherits ScGeom6D ScGeom GenericSpheresContact IGeom Serializable) Class representing geometry of two bodies in contact. The contact has 6 DOFs (normal, 2shear, twist, 2xbending) and uses ScGeom incremental algorithm for updating shear. bending(=Vector3r::Zero()) Bending at contact as a vector dening axis of rotation and angle (angle=norm). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself,
180
top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. end(=Vector3r::Zero()) position of 2nd node (auto-updated) id3(=0 ) id of next chained cylinder (auto-updated) incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). initialOrientation1(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 1 one at initialisation time (auto-updated) initialOrientation2(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 2 one at initialisation time (auto-updated) isDuplicate(=0 ) this ag is turned true (1) automatically if the contact is shared between two chained cylinders. A duplicated interaction will be skipped once by the constitutive law, so that only one contact at a time is eective. If isDuplicate=2, it means one of the two duplicates has no longer geometric interaction, and should be erased by the constitutive laws. normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) onNode(=false) contact on node? penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. relPos(=0 ) position of the contact on the cylinder (0: node-, 1:node+) (auto-updated) shearInc(=Vector3r::Zero()) Shear displacement increment in the last step start(=Vector3r::Zero()) position of 1st node (auto-updated) trueInt(=-1) Denes the body id of the cylinder where the contact is real, when CylScGeom::isDuplicate>0. twist(=0 ) Elastic twist angle of the contact. twistCreep(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Stored creep, substracted from total relative rotation for computation of elastic moment (autoupdated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.2. Interactions
181
class yade.wrapper.Dem3DofGeom(inherits GenericSpheresContact IGeom Serializable) Abstract base class for representing contact geometry of 2 elements that has 3 degrees of freedom: normal (1 component) and shear (Vector3r, but in plane perpendicular to the normal). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displacementN() oat displacementT() Vector3 logCompression(=false) make strain go to - for length going to zero (false by default). normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) refLength(=uninitalized ) some length used to convert displacements to strains. (auto-computed) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) se31(=uninitalized ) Copy of body #1 se3 (needed to compute torque from the contact, strains etc). (auto-updated) se32(=uninitalized ) Copy of body #2 se3. (auto-updated) slipToDisplacementTMax((oat)arg2 ) oat slipToStrainTMax((oat)arg2 ) oat strainN() oat strainT() Vector3 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Dem3DofGeom_FacetSphere(inherits Dem3DofGeom GenericSpheresContact IGeom Serializable) Class representing facet+sphere in contact which computes 3 degrees of freedom (normal and shear deformation). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) cp1pt(=uninitalized ) Reference contact point on the facet in facet-local coords. cp2rel(=uninitalized ) Orientation between +x and the reference contact point (on the sphere) in sphere-local coords dict() dict Return dictionary of attributes.
182
dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displacementN() oat displacementT() Vector3 effR2(=uninitalized ) Eective radius of sphere localFacetNormal(=uninitalized ) Unit normal of the facet plane in facet-local coordinates logCompression(=false) make strain go to - for length going to zero (false by default). normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) refLength(=uninitalized ) some length used to convert displacements to strains. (auto-computed) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) se31(=uninitalized ) Copy of body #1 se3 (needed to compute torque from the contact, strains etc). (auto-updated) se32(=uninitalized ) Copy of body #2 se3. (auto-updated) slipToDisplacementTMax((oat)arg2 ) oat slipToStrainTMax((oat)arg2 ) oat strainN() oat strainT() Vector3 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Dem3DofGeom_SphereSphere(inherits Dem3DofGeom GenericSpheresContact IGeom Serializable) Class representing 2 spheres in contact which computes 3 degrees of freedom (normal and shear deformation). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) cp1rel(=uninitalized ) Spheres #1 relative orientation of the contact point with regards to sphere-local +x axis (quasi-constant) cp2rel(=uninitalized ) Same as cp1rel, but for sphere #2. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself,
7.2. Interactions
183
top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displacementN() oat displacementT() Vector3 effR1(=uninitalized ) Eective radius of sphere #1; can be smaller/larger than refR1 (the actual radius), but quasiconstant throughout interaction life effR2(=uninitalized ) Same as eR1, but for sphere #2. logCompression(=false) make strain go to - for length going to zero (false by default). normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) refLength(=uninitalized ) some length used to convert displacements to strains. (auto-computed) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) se31(=uninitalized ) Copy of body #1 se3 (needed to compute torque from the contact, strains etc). (auto-updated) se32(=uninitalized ) Copy of body #2 se3. (auto-updated) slipToDisplacementTMax((oat)arg2 ) oat slipToStrainTMax((oat)arg2 ) oat strainN() oat strainT() Vector3 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Dem3DofGeom_WallSphere(inherits Dem3DofGeom GenericSpheresContact IGeom Serializable) Representation of contact between wall and sphere, based on Dem3DofGeom. contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) cp1pt(=uninitalized ) initial contact point on the wall, relative to the current contact point cp2rel(=uninitalized ) orientation between +x and the reference contact point (on the sphere) in sphere-local coords dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices.
184
dispIndex Return class index of this instance. displacementN() oat displacementT() Vector3 effR2(=uninitalized ) eective radius of sphere logCompression(=false) make strain go to - for length going to zero (false by default). normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) refLength(=uninitalized ) some length used to convert displacements to strains. (auto-computed) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) se31(=uninitalized ) Copy of body #1 se3 (needed to compute torque from the contact, strains etc). (auto-updated) se32(=uninitalized ) Copy of body #2 se3. (auto-updated) slipToDisplacementTMax((oat)arg2 ) oat slipToStrainTMax((oat)arg2 ) oat strainN() oat strainT() Vector3 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GenericSpheresContact(inherits IGeom Serializable) Class uniting ScGeom and Dem3DofGeom, for the purposes of GlobalStinessTimeStepper. (It might be removed inthe future). Do not use this class directly. contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed)
7.2. Interactions
185
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GridCoGridCoGeom(inherits ScGeom GenericSpheresContact IGeom Serializable) Geometry of a GridConnection-GridConnection contact. contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. relPos1(=0 ) position of the contact on the rst connection (0: node-, 1:node+) (auto-updated) relPos2(=0 ) position of the contact on the rst connection (0: node-, 1:node+) (auto-updated) shearInc(=Vector3r::Zero()) Shear displacement increment in the last step updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GridNodeGeom6D(inherits ScGeom6D ScGeom GenericSpheresContact IGeom Serializable) Geometry of a GridNode-GridNode contact. Inherits almost everything from ScGeom6D. bending(=Vector3r::Zero()) Bending at contact as a vector dening axis of rotation and angle (angle=norm). connectionBody(=uninitalized ) Reference to the GridNode Body who is linking the two GridNodes. contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes.
186
dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). initialOrientation1(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 1 one at initialisation time (auto-updated) initialOrientation2(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 2 one at initialisation time (auto-updated) normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. shearInc(=Vector3r::Zero()) Shear displacement increment in the last step twist(=0 ) Elastic twist angle of the contact. twistCreep(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Stored creep, substracted from total relative rotation for computation of elastic moment (autoupdated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.L3Geom(inherits GenericSpheresContact IGeom Serializable) Geometry of contact given in local coordinates with 3 degress of freedom: normal and two in shear plane. [experimental] F(=Vector3r::Zero()) Applied force in local coordinates [debugging only, will be removed] contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance.
7.2. Interactions
187
normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) trsf(=Matrix3r::Identity()) Transformation (rotation) from global to local coordinates. (the translation part is in GenericSpheresContact.contactPoint) u(=Vector3r::Zero()) Displacement components, in local coordinates. (auto-updated) u0 Zero displacement value; u0 should be always subtracted from the geometrical displacement u computed by appropriate IGeomFunctor, resulting in u. This value can be changed for instance 1.by IGeomFunctor, e.g. to take in account large shear displacement value unrepresentable by underlying geomeric algorithm based on quaternions) 2.by LawFunctor, to account for normal equilibrium position dierent from zero geometric overlap (set once, just after the interaction is created) 3.by LawFunctor to account for plastic slip. Note: Never set an absolute value of u0, only increment, since both IGeomFunctor and LawFunctor use it. If you need to keep track of plastic deformation, store it in IPhys isntead (this might be changed: have u0 for LawFunctor exclusively, and a separate value stored (when that is needed) inside classes deriving from L3Geom. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.L6Geom(inherits L3Geom GenericSpheresContact IGeom Serializable) Geometric of contact in local coordinates with 6 degrees of freedom. [experimental] F(=Vector3r::Zero()) Applied force in local coordinates [debugging only, will be removed] contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) phi(=Vector3r::Zero()) Rotation components, in local coordinates. (auto-updated) phi0(=Vector3r::Zero()) Zero rotation, should be always subtracted from phi to get the value. See L3Geom.u0. 188 Chapter 7. Class reference (yade.wrapper module)
refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) trsf(=Matrix3r::Identity()) Transformation (rotation) from global to local coordinates. (the translation part is in GenericSpheresContact.contactPoint) u(=Vector3r::Zero()) Displacement components, in local coordinates. (auto-updated) u0 Zero displacement value; u0 should be always subtracted from the geometrical displacement u computed by appropriate IGeomFunctor, resulting in u. This value can be changed for instance 1.by IGeomFunctor, e.g. to take in account large shear displacement value unrepresentable by underlying geomeric algorithm based on quaternions) 2.by LawFunctor, to account for normal equilibrium position dierent from zero geometric overlap (set once, just after the interaction is created) 3.by LawFunctor to account for plastic slip. Note: Never set an absolute value of u0, only increment, since both IGeomFunctor and LawFunctor use it. If you need to keep track of plastic deformation, store it in IPhys isntead (this might be changed: have u0 for LawFunctor exclusively, and a separate value stored (when that is needed) inside classes deriving from L3Geom. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ScGeom(inherits GenericSpheresContact IGeom Serializable) Class representing geometry of a contact point between two bodies. It is more general than spheresphere contact even though it is primarily focused on spheres interactions (reason for the Sc namming); it is also used for representing contacts of a Sphere with non-spherical bodies (Facet, Plane, Box, ChainedCylinder), or between two non-spherical bodies (ChainedCylinder). The contact has 3 DOFs (normal and 2shear) and uses incremental algorithm for updating shear. We use symbols x, v, respectively for position, linear and angular velocities (all in global coordinates) and r for particles radii; subscripted with 1 or 2 to distinguish 2 spheres in contact. Then we dene branch length and unit contact normal l = ||x2 x1 ||, n = The relative velocity of the spheres is then v12 = r1 + r2 (v2 v1 ) (r2 2 + r1 1 ) n l x2 x1 ||x2 x1 ||
where the fraction multplying translational velocities is to make the denition objective and avoid ratcheting eects (see Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting). The shear component is vs 12 = v12 (n v12 )n. Tangential displacement increment over last step then reads
s xs 12 = tv12 .
contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) 7.2. Interactions 189
dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. shearInc(=Vector3r::Zero()) Shear displacement increment in the last step updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ScGeom6D(inherits ScGeom GenericSpheresContact IGeom Serializable) Class representing geometry of two bodies in contact. The contact has 6 DOFs (normal, 2shear, twist, 2xbending) and uses ScGeom incremental algorithm for updating shear. bending(=Vector3r::Zero()) Bending at contact as a vector dening axis of rotation and angle (angle=norm). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). initialOrientation1(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 1 one at initialisation time (auto-updated) initialOrientation2(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 2 one at initialisation time (auto-updated)
190
normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. shearInc(=Vector3r::Zero()) Shear displacement increment in the last step twist(=0 ) Elastic twist angle of the contact. twistCreep(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Stored creep, substracted from total relative rotation for computation of elastic moment (autoupdated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ScGridCoGeom(inherits ScGeom6D ScGeom GenericSpheresContact IGeom Serializable) Geometry of a GridConnection-Sphere contact. bending(=Vector3r::Zero()) Bending at contact as a vector dening axis of rotation and angle (angle=norm). contactPoint(=uninitalized ) some reference point for the interaction (usually in the middle). (auto-computed) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. id3(=0 ) id of the rst GridNode. (auto-updated) id4(=0 ) id of the second GridNode. (auto-updated) incidentVel((Interaction)i [, (bool)avoidGranularRatcheting=True ]) Vector3 Return incident velocity of the interaction (see also Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting for explanation of the ratcheting argument). initialOrientation1(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 1 one at initialisation time (auto-updated) initialOrientation2(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Orientation of body 2 one at initialisation time (auto-updated) isDuplicate(=0 ) this ag is turned true (1) automatically if the contact is shared between two Connections. A duplicated interaction will be skipped once by the constitutive law, so that only one contact
7.2. Interactions
191
at a time is eective. If isDuplicate=2, it means one of the two duplicates has no longer geometric interaction, and should be erased by the constitutive laws. normal(=uninitalized ) Unit vector oriented along the interaction, from particle #1, towards particle #2. (autoupdated) penetrationDepth(=NaN ) Penetration distance of spheres (positive if overlapping) refR1(=uninitalized ) Reference radius of particle #1. (auto-computed) refR2(=uninitalized ) Reference radius of particle #2. (auto-computed) relAngVel((Interaction)i ) Vector3 Return relative angular velocity of the interaction. relPos(=0 ) position of the contact on the connection (0: node-, 1:node+) (auto-updated) shearInc(=Vector3r::Zero()) Shear displacement increment in the last step trueInt(=-1) Denes the body id of the GridConnection where the contact is real, when ScGridCoGeom::isDuplicate>0. twist(=0 ) Elastic twist angle of the contact. twistCreep(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Stored creep, substracted from total relative rotation for computation of elastic moment (autoupdated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.TTetraGeom(inherits IGeom Serializable) Geometry of interaction between 2 tetrahedra, including volumetric characteristics contactPoint(=uninitalized ) Contact point (global coords) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. equivalentCrossSection(=NaN ) Cross-section of the overlap (perpendicular to the axis of least inertia equivalentPenetrationDepth(=NaN ) ?? maxPenetrationDepthA(=NaN ) ?? maxPenetrationDepthB(=NaN ) ?? normal(=uninitalized ) Normal of the interaction, directed in the sense of least inertia of the overlap volume
192
penetrationVolume(=NaN ) Volume of overlap [m] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.2.3 IPhys
MindlinPhys MindlinCapillaryPhys
Inelas tCohFrictPhys
Vis cElPhys
LudingPhys
FrictPhys
NormalInelas ticityPhys
CFpmPhys
CapillaryPhys
NormPhys IPhys BubblePhys
NormS hearPhys
RpmPhys
WirePhys
CpmPhys
CohFrictPhys
JCFpmPhys
Vis coFrictPhys
class yade.wrapper.IPhys(inherits Serializable) Physical (material) properties of interaction. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.BubblePhys(inherits IPhys Serializable) Physics of bubble-bubble interactions, for use with BubbleMat computeForce((oat)arg1, (oat)arg2, (oat)arg3, (int)arg4, (oat)arg5 ) oat : Computes the normal force acting between the two interacting bubbles using the NewtonRhapson method dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. fN(=NaN ) Contact normal force 7.2. Interactions 193
newtonIter(=50 ) Maximum number of force iterations allowed newtonTol(=1e-6 ) Convergence criteria for force iterations normalForce(=Vector3r::Zero()) Normal force rAvg(=NaN ) Average radius of the two interacting bubbles surfaceTension(=NaN ) Surface tension of the surrounding liquid updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CFpmPhys(inherits NormShearPhys NormPhys IPhys Serializable) Representation of a single interaction of the CFpm type, storage for relevant parameters FnMax(=0 ) Denes the maximum admissible normal force in Max=tensileStrength*crossSection, with crossSection=pi*Rmin^2. [Pa] traction Fn-
FsMax(=0 ) Denes the maximum admissible tangential force in shear FsMax=cohesion*FnMax, with crossSection=pi*Rmin^2. [Pa] cumulativeRotation(=0 ) Cumulated rotation... [-] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=0 ) denes Coulomb friction. [deg] initD(=0 ) equilibrium distance for particles. Computed as the initial interparticular distance when bonded particle interact. initD=0 for non cohesive interactions. initialOrientation1(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Used for moment computation. initialOrientation2(=Quaternionr(1.0, 0.0, 0.0, 0.0)) Used for moment computation. isCohesive(=false) If false, particles interact in a frictional way. If true, particles are bonded regarding the given cohesion and tensileStrength. kn(=0 ) Normal stiness kr(=0 ) Denes the stiness to compute the resistive moment in rotation. [-] ks(=0 ) Shear stiness
194
maxBend(=0 ) Denes the maximum admissible resistive moment in rotation Mtmax=maxBend*Fn, maxBend=eta*meanRadius. [m] moment_bending(=Vector3r::Zero()) [N.m] moment_twist(=Vector3r::Zero()) [N.m] normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). prevNormal(=Vector3r::Zero()) Normal to the contact at previous time step. shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). strengthSoftening(=0 ) Denes the softening when Dtensile is reached to avoid explosion. Typically, when D > Dtensile, Fn=FnMax - (kn/strengthSoftening)*(Dtensile-D). [-] tanFrictionAngle(=0 ) Tangent of frictionAngle. [-] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CapillaryPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) Physics (of interaction) for Law2_ScGeom_CapillaryPhys_Capillarity. Delta1(=0.) Denes the surface area wetted by the meniscus on the smallest grains of radius R1 (R1<R2) Delta2(=0.) Denes the surface area wetted by the meniscus on the biggest grains of radius R2 (R1<R2) capillaryPressure(=0.) Value of the capillary pressure Uc denes as Ugas-Uliquid dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. fCap(=Vector3r::Zero()) Capillary Force produces by the presence of the meniscus fusionNumber(=0.) Indicates the number of meniscii that overlap with this one isBroken(=false) If true, capillary force is zero and liquid bridge is inactive. kn(=0 ) Normal stiness ks(=0 ) Shear stiness meniscus(=false) Presence of a meniscus if true
7.2. Interactions
195
normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vMeniscus(=0.) Volume of the menicus class yade.wrapper.CohFrictPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) cohesionBroken(=true) is cohesion active? will be set false when a fragile contact is broken cohesionDisablesFriction(=false) is shear strength the sum of friction and adhesion or only adhesion? creep_viscosity(=-1) creep viscosity [Pa.s/m]. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. fragile(=true) do cohesion disapear when contact strength is exceeded? initCohesion(=false) Initialize the cohesive behaviour with current state as equilibrium state (same as Ip2_CohFrictMat_CohFrictMat_CohFrictPhys::setCohesionNow but acting on only one interaction) kn(=0 ) Normal stiness kr(=0 ) rotational stiness [N.m/rad] ks(=0 ) Shear stiness ktw(=0 ) twist stiness [N.m/rad] maxRollPl(=0.0 ) Coecient to determine the maximum plastic rolling moment. maxTwistMoment(=Vector3r::Zero()) Maximum elastic value for the twisting moment (if zero, plasticity will not be applied). In CohFrictMat a parameter should be added to decide what value should be attributed to this threshold value. momentRotationLaw(=false) use bending/twisting moment at contacts. See CohFrictPhys::cohesionDisablesFriction for details.
196
moment_bending(=Vector3r(0, 0, 0)) Bending moment moment_twist(=Vector3r(0, 0, 0)) Twist moment normalAdhesion(=0 ) tensile strength normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). shearAdhesion(=0 ) cohesive part of the shear strength (a frictional term might be added depending on Law2_ScGeom6D_CohFrictPhys_CohesionMoment::always_use_moment_law) shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction unp(=0 ) plastic normal displacement, Phys::fragile=false. :ydefault:0 only used for tensile behaviour and if CohFrict-
unpMax(=0 ) maximum value of plastic normal displacement, after that the interaction breaks even if CohFrictPhys::fragile=false. The default value (0) means no maximum. :ydefault:0 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CpmPhys(inherits NormShearPhys NormPhys IPhys Serializable) Representation of a single interaction of the Cpm type: storage for relevant parameters. Evolution of the contact is governed by Law2_ScGeom_CpmPhys_Cpm, that includes damage eects and chages of parameters inside CpmPhys. See cpm-model for details. E(=NaN ) normal modulus (stiness / crossSection) [Pa] Fn Magnitude of normal force. Fs Magnitude of shear force G(=NaN ) shear modulus [Pa] crossSection(=NaN ) equivalent cross-section associated with this contact [m] cummBetaCount = 0 cummBetaIter = 0 damLaw(=1) Law for softening part of uniaxial tension. 0 for linear, 1 for exponential (default) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. 7.2. Interactions 197
dmgOverstress(=0 ) damage viscous overstress (at previous step or at current step) dmgRateExp(=0 ) exponent in the rate-dependent damage evolution dmgStrain(=0 ) damage strain (at previous or current step) dmgTau(=-1) characteristic time for damage (if non-positive, the law without rate-dependence is used) epsCrackOnset(=NaN ) strain at which the material starts to behave non-linearly epsFracture(=NaN ) strain at which the bond is fully broken [-] epsN Current normal strain epsNPl(=0 ) normal plastic strain (initially zero) epsT(=Vector3r::Zero()) Total shear strain (either computed from increments with ScGeom) (auto-updated) epsTPl(=Vector3r::Zero()) shear plastic strain (initially zero) epsTrans(=0 ) Transversal strain (perpendicular to the contact axis) funcG((oat)kappaD, (oat)epsCrackOnset, (oat)epsFracture[, (bool)neverDamage=False[, (int)damLaw=1 ]]) oat : Damage evolution law, evaluating the parameter. D is historically maximum strain, epsCrackOnset (0 ) = CpmPhys.epsCrackOnset, epsFracture = CpmPhys.epsFracture; if neverDamage is True, the value returned will always be 0 (no damage). TODO funcGInv((oat)omega, (oat)epsCrackOnset, (oat)epsFracture[, (bool)neverDamage=False[, (int)damLaw=1 ]]) oat : Inversion of damage evolution law, evaluating the D parameter. is damage, for other parameters see funcG function isCohesive(=false) if not cohesive, interaction is deleted when distance is greater than zero. isoPrestress(=0 ) prestress of this link (used to simulate isotropic stress) kappaD(=0 ) Up to now maximum normal strain (semi-norm), non-decreasing in time. kn(=0 ) Normal stiness ks(=0 ) Shear stiness neverDamage(=false) the damage evolution function will always return virgin state normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). omega Damage internal variable
198
plRateExp(=0 ) exponent in the rate-dependent viscoplasticity plTau(=-1) characteristic time for viscoplasticity (if non-positive, no rate-dependence for shear) refLength(=NaN ) initial length of interaction [m] refPD(=NaN ) initial penetration depth of interaction [m] (used with ScGeom) relDuctility(=NaN ) Relative ductility of bonds in normal direction relResidualStrength Relative residual strength setDamage((oat)arg2 ) None TODO setRelResidualStrength((oat)arg2 ) None TODO shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). sigmaN Current normal stress sigmaT Current shear stress tanFrictionAngle(=NaN ) tangens of internal friction angle [-] undamagedCohesion(=NaN ) virgin material cohesion [Pa] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.FrictPhys(inherits NormShearPhys NormPhys IPhys Serializable) The simple linear elastic-plastic interaction with friction angle, like in the traditional [CundallStrack1979] dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. kn(=0 ) Normal stiness ks(=0 ) Shear stiness normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction 7.2. Interactions 199
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.InelastCohFrictPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) cohesionBroken(=false) is cohesion active? will be set false when a fragile contact is broken dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. isBroken(=false) true if compression plastic fracture achieved kDam(=0 ) Damage coecient on bending, computed from maximum bending moment reached and pure creep behaviour. Its values will vary between InelastCohFrictPhys::kr and InelastCohFrictPhys::kRCrp . kRCrp(=0.0 ) Bending creep stiness kRUnld(=0.0 ) Bending plastic unload stiness kTCrp(=0.0 ) Tension/compression creep stiness kTUnld(=0.0 ) Tension/compression plastic unload stiness kTwCrp(=0.0 ) Twist creep stiness kTwUnld(=0.0 ) Twist plastic unload stiness kn(=0 ) Normal stiness knC(=0 ) compression stiness knT(=0 ) tension stiness kr(=0 ) bending stiness ks(=0 ) shear stiness ktw(=0 ) twist shear stiness maxBendMom(=0.0 ) Plastic failure bending moment. maxContract(=0.0 ) Plastic failure contraction (shrinkage).
200
maxCrpRchdB(=Vector3r(0, 0, 0)) maximal bending moment reached on plastic deformation. maxCrpRchdC(=Vector2r(0, 0)) maximal compression reached on plastic deformation. maxCrpRchdC[0] stores un and maxCrpRchdC[1] stores Fn. maxCrpRchdT(=Vector2r(0, 0)) maximal extension reached on plastic deformation. maxCrpRchdT[0] stores un and maxCrpRchdT[1] stores Fn. maxCrpRchdTw(=Vector2r(0, 0)) maximal twist reached on plastic deformation. maxCrpRchdTw[0] stores twist angle and maxCrpRchdTw[1] stores twist moment. maxElB(=0.0 ) Maximum bending elastic moment. maxElC(=0.0 ) Maximum compression elastic force. maxElT(=0.0 ) Maximum tension elastic force. maxElTw(=0.0 ) Maximum twist elastic moment. maxExten(=0.0 ) Plastic failure extension (stretching). maxTwist(=0.0 ) Plastic failure twist angle moment_bending(=Vector3r(0, 0, 0)) Bending moment moment_twist(=Vector3r(0, 0, 0)) Twist moment normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). onPlastB(=false) true if plasticity achieved on bending onPlastC(=false) true if plasticity achieved on compression onPlastT(=false) true if plasticity achieved on traction onPlastTw(=false) true if plasticity achieved on twisting pureCreep(=Vector3r(0, 0, 0)) Pure creep curve, used for comparison in calculation. shearAdhesion(=0 ) Maximum elastic shear force (cohesion). shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction twp(=0 ) plastic twist penetration depth describing the equilibrium state. unp(=0 ) plastic normal penetration depth describing the equilibrium state. 7.2. Interactions 201
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.JCFpmPhys(inherits NormShearPhys NormPhys IPhys Serializable) representation of a single interaction of the JCFpm type, storage for relevant parameters FnMax(=0 ) denes the maximum admissible Max=tensileStrength*crossSection. [N] normal force in traction Fn-
FsMax(=0 ) denes the maximum admissible tangential force in shear FsMax=cohesion*FnMax. [N] crossSection(=0 ) crossSection=pi*Rmin^2. [m2] dict() dict Return dictionary of attributes. dilation(=0 ) denes the normal displacement in the joint after sliding treshold. [m] dilationAngle(=0 ) denes the dilatancy of the joint surface. [rad] dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. frictionAngle(=0 ) denes Coulomb friction. [rad] initD(=0 ) equilibrium distance for interacting particles. Computed as the interparticular distance at rst contact detection. isCohesive(=false) If false, particles interact in a frictional way. If true, particles are bonded regarding the given cohesion and tensileStrength. [-] isOnJoint(=false) if true, particles interact as part of a joint (see Law2_ScGeom_JCFpmPhys_JointedCohesiveFrictionalPM for details). [-] jointCumulativeSliding(=0 ) sliding distance for particles interacting on a joint. Used when soothJoint is true in Law2_ScGeom_JCFpmPhys_JointedCohesiveFrictionalPM to take into account dilatancy due to shearing. [-] jointNormal(=Vector3r::Zero()) Normal direction to the joint. kn(=0 ) Normal stiness ks(=0 ) Shear stiness more(=false) species if the interaction is crossed by more than 3 joints. If true, interaction is deleted (temporary solution). [-] normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates).
202
shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tanDilationAngle(=0 ) tangent of dilationAngle. [-] tanFrictionAngle(=0 ) tangent of frictionAngle. [-] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.LudingPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) IPhys created from LudingMat, for use with Law2_ScGeom_LudingPhys_Basic. G0(=NaN ) Viscous damping PhiF(=NaN ) Dimensionless plasticity depth ThetMax(=NaN ) Maximum overlap between particles for a collision ThetNull(=NaN ) Force free overlap, plastic contact deformation ThetPMax(=NaN ) Maximum overlap between particles for the limit case dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. k1(=NaN ) Slope of loading plastic branch k2(=NaN ) Slope of unloading and reloading elastic branch kc(=NaN ) Slope of irreversible, tensile adhesive branch kn(=0 ) Normal stiness kp(=NaN ) Slope of unloading and reloading limit elastic branch ks(=0 ) Shear stiness normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.2. Interactions
203
class yade.wrapper.MindlinCapillaryPhys(inherits MindlinPhys FrictPhys NormShearPhys NormPhys IPhys Serializable) Adds capillary physics to Mindlins interaction physics. Delta1(=0.) Denes the surface area wetted by the meniscus on the smallest grains of radius R1 (R1<R2) Delta2(=0.) Denes the surface area wetted by the meniscus on the biggest grains of radius R2 (R1<R2) Fs(=Vector2r::Zero()) Shear force in local axes (computed incrementally) adhesionForce(=0.0 ) Force of adhesion as predicted by DMT alpha(=0.0 ) Constant coecient to dene contact viscous damping for non-linear elastic force-displacement relationship. betan(=0.0 ) Fraction of the viscous damping coecient (normal direction) equal to betas(=0.0 ) Fraction of the viscous damping coecient (shear direction) equal to capillaryPressure(=0.) Value of the capillary pressure Uc denes as Ugas-Uliquid dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. fCap(=Vector3r::Zero()) Capillary Force produces by the presence of the meniscus fusionNumber(=0.) Indicates the number of meniscii that overlap with this one isAdhesive(=false) bool to identify if the contact is adhesive, that is to say if the contact force is attractive isBroken(=false) If true, capillary force is zero and liquid bridge is inactive. isSliding(=false) check if the contact is sliding (useful to calculate the ratio of sliding contacts) kn(=0 ) Normal stiness kno(=0.0 ) Constant value in the formulation of the normal stiness kr(=0.0 ) Rotational stiness ks(=0 ) Shear stiness kso(=0.0 ) Constant value in the formulation of the tangential stiness
cn Cn,crit .
cs Cs,crit .
204
ktw(=0.0 ) Rotational stiness maxBendPl(=0.0 ) Coecient to determine the maximum plastic moment to apply at the contact meniscus(=false) Presence of a meniscus if true momentBend(=Vector3r::Zero()) Articial bending moment to provide rolling resistance in order to account for some degree of interlocking between particles momentTwist(=Vector3r::Zero()) Articial twisting moment (no plastic condition can be applied at the moment) normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). normalViscous(=Vector3r::Zero()) Normal viscous component prevU(=Vector3r::Zero()) Previous local displacement; only used with Law2_L3Geom_FrictPhys_HertzMindlin. radius(=NaN ) Contact radius (only computed with Law2_ScGeom_MindlinPhys_Mindlin::calcEnergy) shearElastic(=Vector3r::Zero()) Total elastic shear force shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). shearViscous(=Vector3r::Zero()) Shear viscous component tangensOfFrictionAngle(=NaN ) tan of angle of friction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary usElastic(=Vector3r::Zero()) Total elastic shear displacement (only elastic part) usTotal(=Vector3r::Zero()) Total elastic shear displacement (elastic+plastic part) vMeniscus(=0.) Volume of the menicus class yade.wrapper.MindlinPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) Representation of an interaction of the Hertz-Mindlin type. Fs(=Vector2r::Zero()) Shear force in local axes (computed incrementally) adhesionForce(=0.0 ) Force of adhesion as predicted by DMT alpha(=0.0 ) Constant coecient to dene contact viscous damping for non-linear elastic force-displacement relationship. betan(=0.0 ) Fraction of the viscous damping coecient (normal direction) equal to betas(=0.0 ) Fraction of the viscous damping coecient (shear direction) equal to
cn Cn,crit .
cs Cs,crit .
7.2. Interactions
205
dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. isAdhesive(=false) bool to identify if the contact is adhesive, that is to say if the contact force is attractive isSliding(=false) check if the contact is sliding (useful to calculate the ratio of sliding contacts) kn(=0 ) Normal stiness kno(=0.0 ) Constant value in the formulation of the normal stiness kr(=0.0 ) Rotational stiness ks(=0 ) Shear stiness kso(=0.0 ) Constant value in the formulation of the tangential stiness ktw(=0.0 ) Rotational stiness maxBendPl(=0.0 ) Coecient to determine the maximum plastic moment to apply at the contact momentBend(=Vector3r::Zero()) Articial bending moment to provide rolling resistance in order to account for some degree of interlocking between particles momentTwist(=Vector3r::Zero()) Articial twisting moment (no plastic condition can be applied at the moment) normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). normalViscous(=Vector3r::Zero()) Normal viscous component prevU(=Vector3r::Zero()) Previous local displacement; only used with Law2_L3Geom_FrictPhys_HertzMindlin. radius(=NaN ) Contact radius (only computed with Law2_ScGeom_MindlinPhys_Mindlin::calcEnergy) shearElastic(=Vector3r::Zero()) Total elastic shear force shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). shearViscous(=Vector3r::Zero()) Shear viscous component tangensOfFrictionAngle(=NaN ) tan of angle of friction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary 206 Chapter 7. Class reference (yade.wrapper module)
usElastic(=Vector3r::Zero()) Total elastic shear displacement (only elastic part) usTotal(=Vector3r::Zero()) Total elastic shear displacement (elastic+plastic part) class yade.wrapper.NormPhys(inherits IPhys Serializable) Abstract class for interactions that have normal stiness. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. kn(=0 ) Normal stiness normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.NormShearPhys(inherits NormPhys IPhys Serializable) Abstract class for interactions that have shear stinesses, in addition to normal stiness. This class is used in the PFC3d-style stiness timestepper. dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. kn(=0 ) Normal stiness ks(=0 ) Shear stiness normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.NormalInelasticityPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) Physics (of interaction) for using Law2_ScGeom6D_NormalInelasticityPhys_NormalInelasticity : with inelastic unloadings dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself,
7.2. Interactions
207
top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. forMaxMoment(=1.0 ) parameter stored for each interaction, and allowing to compute the maximum value of the exchanged torque : TorqueMax= forMaxMoment * NormalForce kn(=0 ) Normal stiness knLower(=0.0 ) the stifness corresponding to a virgin load for example kr(=0.0 ) the rolling stiness of the interaction ks(=0 ) Shear stiness moment_bending(=Vector3r(0, 0, 0)) Bending moment. Dened here, being initialized as it should be, to be used in Law2_ScGeom6D_NormalInelasticityPhys_NormalInelasticity moment_twist(=Vector3r(0, 0, 0)) Twist moment. Dened here, being initialized as it should be, to be used in Law2_ScGeom6D_NormalInelasticityPhys_NormalInelasticity normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). previousFn(=0.0 ) the value of the normal force at the last time step previousun(=0.0 ) the value of this un at the last time step shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction unMax(=0.0 ) the maximum value of penetration depth of the history of this interaction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.RpmPhys(inherits NormShearPhys NormPhys IPhys Serializable) Representation of a single interaction of the Rpm type: storage for relevant parameters. Evolution of the contact is governed by Law2_Dem3DofGeom_RpmPhys_Rpm, that includes damage eects and chages of parameters inside RpmPhys E(=NaN ) normal modulus (stiness / crossSection) [Pa] G(=NaN ) shear modulus [Pa] Kn(=0 ) Stiness in normal direction [N/m] Ks(=0 ) Stiness in shear direction [N/m] crossSection(=0 ) equivalent cross-section associated with this contact [m] 208 Chapter 7. Class reference (yade.wrapper module)
dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. epsMaxCompression(=0 ) Maximal relative penetration of particles during compression. If it is more, the interaction is deleted [m] epsMaxShear(=0 ) Maximal relative distance between particles in tangential direction. If it is more, the interaction is deleted [m] epsMaxTension(=0 ) Maximal relative distance between particles during tension. If it is more, the interaction is deleted [m] isCohesive(=false) if not cohesive, interaction is deleted when distance is greater than lengthMaxTension or less than lengthMaxCompression. kn(=0 ) Normal stiness ks(=0 ) Shear stiness normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tanFrictionAngle(=NaN ) tangens of internal friction angle [-] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ViscElPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) IPhys created from ViscElMat, for use with Law2_ScGeom_ViscElPhys_Basic. Capillar(=false) True, if capillar forces need to be added. CapillarType(= ) Dierent types of capillar interaction: Willett_numeric, Willett_analytic, Weigert, Rabinovich Vb(=NaN ) Liquid bridge volume [m^3] cn(=NaN ) Normal viscous constant cs(=NaN ) Shear viscous constant dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, 7.2. Interactions 209
top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. gamma(=NaN ) Surface tension [N/m] kn(=0 ) Normal stiness ks(=0 ) Shear stiness liqBridgeCreated(=false) Whether liquid bridge was created, only after a normal contact of spheres mR(=0.0 ) Rolling resistance, see [Zhou1999536]. mRtype(=1) Rolling resistance type, see [Zhou1999536]. mRtype=1 - equation (3) in [Zhou1999536]; mRtype=2 - equation (4) in [Zhou1999536] normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). sCrit(=false) Critical bridge length [m] shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction theta(=NaN ) Contact angle [rad] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ViscoFrictPhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) Temporary version of FrictPhys for compatibility with e.g. Law2_ScGeom6D_NormalInelasticityPhys_NormalInelasticity creepedShear(=Vector3r(0, 0, 0)) Creeped force (parallel) dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. kn(=0 ) Normal stiness ks(=0 ) Shear stiness normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates).
210
shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates). tangensOfFrictionAngle(=NaN ) tan of angle of friction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.WirePhys(inherits FrictPhys NormShearPhys NormPhys IPhys Serializable) Representation of a single interaction of the WirePM type, storage for relevant parameters dL(=0.) Additional wire length for considering the distortion for WireMat type=2 (see [Thoeni2013]). dict() dict Return dictionary of attributes. dispHierarchy([(bool)names=True ]) list Return list of dispatch classes (from down upwards), starting with the class instance itself, top-level indexable at last. If names is true (default), return class names rather than numerical indices. dispIndex Return class index of this instance. displForceValues(=uninitalized ) Denes the values for force-displacement curve. initD(=0.) Equilibrium distance for particles. Computed as the initial inter-particular distance when particle are linked. isDoubleTwist(=false) If true the properties of the interaction will be dened as a double-twisted wire. isLinked(=false) If true particles are linked and will interact. Interactions are linked automatically by the denition of the corresponding interaction radius. The value is false if the wire breaks (no more interaction). isShifted(=false) If true WireMat type=2 and the force-displacement curve will be shifted. kn(=0 ) Normal stiness ks(=0 ) Shear stiness limitFactor(=0.) This value indicates on how far from failing the wire is, e.g. actual normal displacement divided by admissible normal displacement. normalForce(=Vector3r::Zero()) Normal force after previous step (in global coordinates). plastD Plastic part of the inter-particular distance of the previous step. Note: Only elastic displacements are reversible (the elastic stiness is used for unloading) and compressive forces are inadmissible. The compressive stiness is assumed to be equal to zero. shearForce(=Vector3r::Zero()) Shear force after previous step (in global coordinates).
7.2. Interactions
211
stiffnessValues(=uninitalized ) Denes the values for the various stinesses (the elastic stiness is stored as kn). tangensOfFrictionAngle(=NaN ) tan of angle of friction updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.3 Global engines

7.3.1 GlobalEngine
CapillaryS tres s Recorder
ParticleS izeDis trbutionRPMRecorder
ForceRecorder Recorder TriaxialS tateRecorder DomainLimiter TorqueRecorder CpmS tateUpdater Collider Res etRandomPos ition PeriodicEngine PyRunner BoundaryController VTKRecorder TimeS tepper S naps hotEngine FacetTopologyAnalyzer GlobalS ti nes s TimeS tepper FieldApplier Cohes iveS tateRPMRecorder
S pheres Factory
BoxFactory
Cohes iveFrictionalContactLaw GlobalEngine NewtonIntegrator
CircularFactory
Law2_S cGeom_CapillaryPhys _Capillarity
Tes s elationWrapper
ForceRes etter
TetraVolumetricLaw
MicroMacroAnalys er
Elas ticContactLaw
InteractionLoop
class yade.wrapper.GlobalEngine(inherits Engine Serializable) Engine that will generally aect the whole simulation (contrary to PartialEngine).
212
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.BoxFactory(inherits SpheresFactory GlobalEngine Engine Serializable) Box geometry of the SpheresFactory region, given by extents and center PSDcalculateMass(=true) PSD-Input is in mass (true), otherwise the number of particles will be considered. PSDcum(=uninitalized ) PSD-dispersion, cumulative procent meanings [-] PSDsizes(=uninitalized ) PSD-dispersion, sizes of cells, Diameter [m] blockedDOFs(= ) Blocked degress of freedom center(=Vector3r(NaN, NaN, NaN)) Center of the region color(=Vector3r(-1, -1, -1)) Use the color for newly created particles, if specied dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. exactDiam(=true) If true, the particles only with the dened in PSDsizes diameters will be created. Otherwise the diameter will be randomly chosen in the range [PSDsizes[i-1]:PSDsizes[i]], in this case the length of PSDsizes should be more on 1, than the length of PSDcum. execCount Cummulative count this engine was run (only used if O.timingEnabled==True).
7.3. Global engines
213
execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). extents(=Vector3r(NaN, NaN, NaN)) Extents of the region goalMass(=0 ) Total mass that should be attained at the end of the current step. (auto-updated) ids(=uninitalized ) ids of created bodies label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=-1) groupMask to apply for newly created spheres massFlowRate(=NaN ) Mass ow rate [kg/s] materialId(=-1) Shared material id to use for newly created spheres (can be negative to count from the end) maxAttempt(=5000 ) Maximum number of attempts to position a new sphere randomly. maxMass(=-1) Maximal mass at which to stop generating new particles regardless of massFlowRate. if maxMass=-1 - this parameter is ignored. maxParticles(=100 ) The number of particles at which to stop generating new ones regardless of massFlowRate. if maxParticles=-1 - this parameter is ignored . normal(=Vector3r(NaN, NaN, NaN)) Orientation of the regions geometry, direction of particles velocites if normalVel is not set. normalVel(=Vector3r(NaN, NaN, NaN)) Direction of particles velocites. numParticles(=0 ) Cummulative number of particles produces so far (auto-updated) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rMax(=NaN ) Maximum radius of generated spheres (uniform distribution) rMin(=NaN ) Minimum radius of generated spheres (uniform distribution) silent(=false) If true no complain about excessing maxAttempt but disable the factory (by set massFlowRate=0). stopIfFailed(=true) If true, the SpheresFactory stops (sets massFlowRate=0), when maximal number of attempts to insert particle exceed.
214
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. totalMass(=0 ) Mass of spheres that was produced so far. (auto-updated) totalVolume(=0 ) Volume of spheres that was produced so far. (auto-updated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vAngle(=NaN ) Maximum angle by which the initial sphere velocity deviates from the normal. vMax(=NaN ) Maximum velocity norm of generated spheres (uniform distribution) vMin(=NaN ) Minimum velocity norm of generated spheres (uniform distribution) class yade.wrapper.CapillaryStressRecorder(inherits Recorder PeriodicEngine GlobalEngine Engine Serializable) Records information from capillary meniscii on samples submitted to triaxial compressions. -> New formalism needs to be tested!!! addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes 7.3. Global engines 215
openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.CircularFactory(inherits SpheresFactory GlobalEngine Engine Serializable) Circular geometry of the SpheresFactory region. It can be disk (given by radius and center), or cylinder (given by radius, length and center). PSDcalculateMass(=true) PSD-Input is in mass (true), otherwise the number of particles will be considered. PSDcum(=uninitalized ) PSD-dispersion, cumulative procent meanings [-] PSDsizes(=uninitalized ) PSD-dispersion, sizes of cells, Diameter [m] blockedDOFs(= ) Blocked degress of freedom center(=Vector3r(NaN, NaN, NaN)) Center of the region color(=Vector3r(-1, -1, -1)) Use the color for newly created particles, if specied dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. exactDiam(=true) If true, the particles only with the dened in PSDsizes diameters will be created. Otherwise the diameter will be randomly chosen in the range [PSDsizes[i-1]:PSDsizes[i]], in this case the length of PSDsizes should be more on 1, than the length of PSDcum. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). goalMass(=0 ) Total mass that should be attained at the end of the current step. (auto-updated)
216
ids(=uninitalized ) ids of created bodies label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. length(=0 ) Length of the cylindrical region (0 by default) mask(=-1) groupMask to apply for newly created spheres massFlowRate(=NaN ) Mass ow rate [kg/s] materialId(=-1) Shared material id to use for newly created spheres (can be negative to count from the end) maxAttempt(=5000 ) Maximum number of attempts to position a new sphere randomly. maxMass(=-1) Maximal mass at which to stop generating new particles regardless of massFlowRate. if maxMass=-1 - this parameter is ignored. maxParticles(=100 ) The number of particles at which to stop generating new ones regardless of massFlowRate. if maxParticles=-1 - this parameter is ignored . normal(=Vector3r(NaN, NaN, NaN)) Orientation of the regions geometry, direction of particles velocites if normalVel is not set. normalVel(=Vector3r(NaN, NaN, NaN)) Direction of particles velocites. numParticles(=0 ) Cummulative number of particles produces so far (auto-updated) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rMax(=NaN ) Maximum radius of generated spheres (uniform distribution) rMin(=NaN ) Minimum radius of generated spheres (uniform distribution) radius(=NaN ) Radius of the region silent(=false) If true no complain about excessing maxAttempt but disable the factory (by set massFlowRate=0). stopIfFailed(=true) If true, the SpheresFactory stops (sets massFlowRate=0), when maximal number of attempts to insert particle exceed. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True.
7.3. Global engines
217
totalMass(=0 ) Mass of spheres that was produced so far. (auto-updated) totalVolume(=0 ) Volume of spheres that was produced so far. (auto-updated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vAngle(=NaN ) Maximum angle by which the initial sphere velocity deviates from the normal. vMax(=NaN ) Maximum velocity norm of generated spheres (uniform distribution) vMin(=NaN ) Minimum velocity norm of generated spheres (uniform distribution) class yade.wrapper.CohesiveFrictionalContactLaw(inherits GlobalEngine Engine Serializable) [DEPRECATED] Loop over interactions applying Law2_ScGeom6D_CohFrictPhys_CohesionMoment on all interactions. Note: Use InteractionLoop and Law2_ScGeom6D_CohFrictPhys_CohesionMoment instead of this class for performance reasons. always_use_moment_law(=false) If true, use bending/twisting moments at all contacts. If false, compute moments only for cohesive contacts. creep_viscosity(=false) creep viscosity [Pa.s/m]. probably should be moved to Ip2_CohFrictMat_CohFrictMat_CohFrictPhys... dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. shear_creep(=false) activate creep on the shear force, using CohesiveFrictionalContactLaw::creep_viscosity.
218
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. twist_creep(=false) activate creep on the twisting moment, using CohesiveFrictionalContactLaw::creep_viscosity. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CohesiveStateRPMRecorder(inherits Recorder PeriodicEngine GlobalEngine Engine Serializable) Store number of cohesive contacts in RPM model to le. addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). numberCohesiveContacts(=0 ) Number of cohesive contacts found at last run. [-] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0)
7.3. Global engines
219
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.CpmStateUpdater(inherits PeriodicEngine GlobalEngine Engine Serializable) Update CpmState of bodies based on state variables in CpmPhys of interactions with this bod. In particular, bodies colors and CpmState::normDmg depending on average damage of their interactions and number of interactions that were already fully broken and have disappeared is updated. This engine contains its own loop (2 loops, more precisely) over all bodies and should be run periodically to update colors during the simulation, if desired. avgRelResidual(=NaN ) Average residual strength at last run. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxOmega(=NaN ) Globally maximum damage parameter at last run. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. 220 Chapter 7. Class reference (yade.wrapper module)
realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.DomainLimiter(inherits PeriodicEngine GlobalEngine Engine Serializable) Delete particles that are out of axis-aligned box given by lo and hi. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). hi(=Vector3r(0, 0, 0)) Upper corner of the domain. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. lo(=Vector3r(0, 0, 0)) Lower corner of the domain. mDeleted(=0 ) Mass of deleted particles. mask(=-1) If mask is dened, only particles with corresponding groupMask will be deleted. nDeleted(=0 ) Cummulative number of particles deleted. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated).
7.3. Global engines
221
ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vDeleted(=0 ) Volume of deleted particles. virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.ElasticContactLaw(inherits GlobalEngine Engine Serializable) [DEPRECATED] Loop over interactions applying Law2_ScGeom_FrictPhys_CundallStrack on all interactions. Note: Use InteractionLoop and Law2_ScGeom_FrictPhys_CundallStrack instead of this class for performance reasons. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level.
222
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.FacetTopologyAnalyzer(inherits GlobalEngine Engine Serializable) Initializer for lling adjacency geometry data for facets. Common vertices and common edges are identied and mutual angle between facet faces is written to Facet instances. If facets dont move with respect to each other, this must be done only at the beginng. commonEdgesFound(=0 ) how many common edges were identied during last run. (auto-updated) commonVerticesFound(=0 ) how many common vertices were identied during last run. (auto-updated) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. projectionAxis(=Vector3r::UnitX()) Axis along which to do the initial vertex sort relTolerance(=1e-4) maximum distance of identical vertices, relative to minimum facet size timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ForceRecorder(inherits Recorder PeriodicEngine GlobalEngine Engine Serializable) Engine saves the resultant force aecting to bodies, listed in ids. For instance, can be useful for dening the forces, which aects to _buldozer_ during its work. addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default)
7.3. Global engines
223
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. ids(=uninitalized ) List of bodies whose state will be measured initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. totalForce(=Vector3r::Zero()) Resultant force, returning by the function. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) 224 Chapter 7. Class reference (yade.wrapper module)
class yade.wrapper.ForceResetter(inherits GlobalEngine Engine Serializable) Reset all forces stored in Scene::forces (O.forces in python). Typically, this is the rst engine to be run at every step. In addition, reset those energies that should be reset, if energy tracing is enabled. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlobalStiffnessTimeStepper(inherits TimeStepper GlobalEngine Engine Serializable) An engine assigning the time-step as a fraction of the minimum eigen-period in the problem. The derivation is detailed in the chapter on DEM formulation active(=true) is the engine active? dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. defaultDt(=-1) used as the initial value of the timestep (especially useful in the rst steps when no contact exist). If negative, it will be dened by utils.PWaveTimeStep * GlobalStinessTimeStepper::timestepSafetyCoecient densityScaling(=false) (auto-updated) dont modify this value if you dont plan to modify the scaling factor manually for some bodies. In most cases, it is enough to set NewtonIntegrator::densityScaling and let this one be adjusted automatically. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True).
7.3. Global engines
225
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxDt(=Mathr::MAX_REAL) if positive, used as max value of the timestep whatever the computed value ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. previousDt(=1) last computed dt (auto-updated) targetDt(=1) if NewtonIntegrator::densityScaling is active, this value will be used as the simulation timestep and the scaling will use this value of dt as the target value. The value of targetDt is arbitrary and should have no eect in the result in general. However if some bodies have imposed velocities, for instance, they will move more or less per each step depending on this value. timeStepUpdateInterval(=1) dt update interval timestepSafetyCoefficient(=0.8 ) safety factor between the minimum eigen-period and the nal assigned dt (less than 1)) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.InteractionLoop(inherits GlobalEngine Engine Serializable) Unied dispatcher for handling interaction loop at every step, for parallel performance reasons. Special constructor Constructs from 3 lists of Ig2, Ip2, Law functors respectively; they will be passed to interal dispatchers, which you might retrieve. callbacks(=uninitalized ) Callbacks which will be called for every Interaction, if activated. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. eraseIntsInLoop(=false) Denes if the interaction loop should erase pending interactions, else the collider takes care of that alone (depends on what collider is used). execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). geomDispatcher(=new IGeomDispatcher ) IGeomDispatcher object that is used for dispatch. 226 Chapter 7. Class reference (yade.wrapper module)
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. lawDispatcher(=new LawDispatcher ) LawDispatcher object used for dispatch. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. physDispatcher(=new IPhysDispatcher ) IPhysDispatcher object used for dispatch. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_CapillaryPhys_Capillarity(inherits GlobalEngine Engine Serializable) This law allows one to take into account capillary forces/eects between spheres coming from the presence of interparticular liquid bridges (menisci). The control parameter is the capillary pressure (or suction) Uc = ugas - Uliquid. Liquid bridges properties (volume V, extent over interacting grains delta1 and delta2) are computed as a result of the dened capillary pressure and of the interacting geometry (spheres radii and interparticular distance). References: in english [Scholtes2009b]; more detailed, but in french [Scholtes2009d]. The law needs ascii les M(r=i) with i=R1/R2 to work (see https://yadedem.org/index.php/CapillaryTriaxialTest). These ASCII les contain a set of results from the resolution of the Laplace-Young equation for dierent congurations of the interacting geometry. In order to allow capillary forces between distant spheres, it is necessary to enlarge the bounding boxes using Bo1_Sphere_Aabb::aabbEnlargeFactor and make the Ig2 dene dene distant interactions via interactionDetectionFactor. It is also necessary to disable interactions removal by the constitutive law (Law2). The only combinations of laws supported are currently capillary law + Law2_ScGeom_FrictPhys_CundallStrack and capillary law + Law2_ScGeom_MindlinPhys_Mindlin (and the other variants of Hertz-Mindlin). See CapillaryPhys-example.py for an example script. binaryFusion(=true) If true, capillary forces are set to zero as soon as, at least, 1 overlap (menisci fusion) is detected capillaryPressure(=0.) Value of the capillary pressure Uc denes as Uc=Ugas-Uliquid createDistantMeniscii(=false) Generate meniscii between distant spheres? Else only maintain the existing ones. For modeling a wetting path this ag should always be false. For a drying path it should be true for one step (initialization) then false, as in the logic of [Scholtes2009c] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point.
7.3. Global engines
227
dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). fusionDetection(=false) If true potential menisci overlaps are checked hertzOn(=false) (auto-updated) true if hertz model is used label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.MicroMacroAnalyser(inherits GlobalEngine Engine Serializable) Compute fabric tensor, local porosity, local deformation, and other micromechanicaly dened quantities based on triangulation/tesselation of the packing. compDeformation(=false) Is the engine just saving states or also computing and outputing deformations for each increment? compIncrt(=false) Should increments of force and displacements be dened on [n,n+1]? If not, states will be saved with only positions and forces (no displacements). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). incrtNumber(=1) interval(=100 ) Number of timesteps between analyzed states. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python.
228
nonSphereAsFictious(=true) bodies that are not spheres will be used to denes bounds (else just skipped). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. outputFile(=MicroMacroAnalysis ) Base name for increment analysis output le. stateFileName(=state ) Base name of state les. stateNumber(=0 ) A number incremented and appended at the end of output les to reect increment number. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.NewtonIntegrator(inherits GlobalEngine Engine Serializable) Engine integrating newtonian motion equations. damping(=0.2 ) damping coecient for Cundalls non viscous damping (see [Chareyre2005]) [-] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. densityScaling if True, then density scaling [Pfc3dManual30] will be applied in order to have a critical timestep equal to GlobalStinessTimeStepper::targetDt for all bodies. This option makes the simulation unrealistic from a dynamic point of view, but may speedup quasistatic simulations. In rare situations, it could be useful to not set the scalling factor automatically for each body (which the time-stepper does). In such case revert GlobalStinessTimeStepper.densityScaling to False. dict() dict Return dictionary of attributes. exactAsphericalRot(=true) Enable more exact body rotation integrator for aspherical bodies only, using formulation from [Allen1989], pg. 89. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). gravity(=Vector3r::Zero()) Gravitational acceleration (eectifely replaces GravityEngine). kinSplit(=false) Whether to separately track translational and rotational kinetic energy. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python.
7.3. Global engines
229
maxVelocitySq(=NaN ) store square of max. velocity, for informative purposes; computed again at every step. (autoupdated) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. prevVelGrad(=Matrix3r::Zero()) Store previous velocity gradient (Cell::velGrad) to track acceleration. (auto-updated) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary warnNoForceReset(=true) Warn when forces were not resetted in this step by ForceResetter; this mostly points to ForceResetter being forgotten incidentally and should be disabled only with a good reason. class yade.wrapper.ParticleSizeDistrbutionRPMRecorder(inherits Recorder PeriodicEngine GlobalEngine Engine Serializable) Store number of PSD in RPM model to le. addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). 230 Chapter 7. Class reference (yade.wrapper module)
numberCohesiveContacts(=0 ) Number of cohesive contacts found at last run. [-] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.PeriodicEngine(inherits GlobalEngine Engine Serializable) Run Engine::action with given xed periodicity real time (=wall clock time, computation time), virtual time (simulation time), iteration number), by setting any of those criteria (virtPeriod, realPeriod, iterPeriod) to a positive value. They are all negative (inactive) by default. The number of times this engine is activated can be limited by setting nDo>0. If the number of activations will have been already reached, no action will be called even if an active period has elapsed. If initRun is set (false by default), the engine will run when called for the rst time; otherwise it will only start counting period (realLast etc interal variables) from that point, but without actually running, and will run only once a period has elapsed since the initial run. This class should not be used directly; rather, derive your own engine which you want to be run periodically. Derived engines should override Engine::action(), which will be called periodically. If the derived Engine overrides also Engine::isActivated, it should also take in account return value from PeriodicEngine::isActivated, since otherwise the periodicity will not be functional. Example with PyRunner, which derives from PeriodicEngine; likely to be encountered in python scripts:
PyRunner(realPeriod=5,iterPeriod=10000,command='print O.iter')
will print iteration number every 10000 iterations or every 5 seconds of wall clock time, whiever comes rst since it was last run. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes.
7.3. Global engines
231
execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.PyRunner(inherits PeriodicEngine GlobalEngine Engine Serializable) Execute a python command periodically, with dened (and adjustable) periodicity. See PeriodicEngine documentation for details. command(= ) Command to be run by python interpreter. Not run if empty. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True).
232
execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.Recorder(inherits PeriodicEngine GlobalEngine Engine Serializable) Engine periodically storing some data to (one) external le. In addition PeriodicEngine, it handles opening the le as needed. See PeriodicEngine for controlling periodicity. addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True).
7.3. Global engines
233
execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.ResetRandomPosition(inherits PeriodicEngine GlobalEngine Engine Serializable) Creates spheres during simulation, placing them at random positions. Every time called, one new sphere will be created and inserted in the simulation. angularVelocity(=Vector3r::Zero()) Mean angularVelocity of spheres. angularVelocityRange(=Vector3r::Zero()) Half size of a angularVelocity distribution interval. New sphere will have random angularVelocity within the range angularVelocityangularVelocityRange.
234
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). factoryFacets(=uninitalized ) The geometry of the section where spheres will be placed; they will be placed on facets or in volume between them depending on volumeSection ag. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxAttempts(=20 ) Max attempts to place sphere. If placing the sphere in certain random position would cause an overlap with any other physical body in the model, SpheresFactory will try to nd another position. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). normal(=Vector3r(0, 1, 0)) ?? ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. point(=Vector3r::Zero()) ?? realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) subscribedBodies(=uninitalized ) Aected bodies. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True.
7.3. Global engines
235
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary velocity(=Vector3r::Zero()) Mean velocity of spheres. velocityRange(=Vector3r::Zero()) Half size of a velocities distribution interval. New sphere will have random velocity within the range velocityvelocityRange. virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) volumeSection(=false, dene factory by facets.) Create new spheres inside factory volume rather than on its surface. class yade.wrapper.SnapshotEngine(inherits PeriodicEngine GlobalEngine Engine Serializable) Periodically save snapshots of GLView(s) as .png les. Files are named leBase + counter + .png (counter is left-padded by 0s, i.e. snap00004.png). counter(=0 ) Number that will be appended to leBase when the next snapshot is saved (incremented at every save). (auto-updated) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. deadTimeout(=3 ) Timeout for 3d operations (opening new view, saving snapshot); after timing out, throw exception (or only report error if ignoreErrors) and make myself dead. [s] dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). fileBase(= ) Basename for snapshots format(=PNG ) Format of snapshots (one of JPEG, PNG, EPS, PS, PPM, BMP) QGLViewer documentation. File extension will be lowercased format. Validity of format is not checked. ignoreErrors(=true) Only report errors instead of throwing exceptions, in case of timeouts. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. msecSleep(=0 ) number of msec to sleep after snapshot (to prevent 3d hw problems) [ms]
236
nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. plot(=uninitalized ) Name of eld in plot.imgData to which taken snapshots will be appended automatically. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) snapshots(=uninitalized ) Files that have been created so far timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.SpheresFactory(inherits GlobalEngine Engine Serializable) Engine for spitting spheres based on mass ow rate, particle size distribution etc. Initial velocity of particles is given by vMin, vMax, the massFlowRate determines how many particles to generate at each step. When goalMass is attained or positive maxParticles is reached, the engine does not produce particles anymore. Geometry of the region should be dened in a derived engine by overridden SpheresFactory::pickRandomPosition(). A sample script for this engine is in scripts/spheresFactory.py. PSDcalculateMass(=true) PSD-Input is in mass (true), otherwise the number of particles will be considered. PSDcum(=uninitalized ) PSD-dispersion, cumulative procent meanings [-] PSDsizes(=uninitalized ) PSD-dispersion, sizes of cells, Diameter [m] blockedDOFs(= ) Blocked degress of freedom color(=Vector3r(-1, -1, -1)) Use the color for newly created particles, if specied dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes.
7.3. Global engines
237
exactDiam(=true) If true, the particles only with the dened in PSDsizes diameters will be created. Otherwise the diameter will be randomly chosen in the range [PSDsizes[i-1]:PSDsizes[i]], in this case the length of PSDsizes should be more on 1, than the length of PSDcum. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). goalMass(=0 ) Total mass that should be attained at the end of the current step. (auto-updated) ids(=uninitalized ) ids of created bodies label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=-1) groupMask to apply for newly created spheres massFlowRate(=NaN ) Mass ow rate [kg/s] materialId(=-1) Shared material id to use for newly created spheres (can be negative to count from the end) maxAttempt(=5000 ) Maximum number of attempts to position a new sphere randomly. maxMass(=-1) Maximal mass at which to stop generating new particles regardless of massFlowRate. if maxMass=-1 - this parameter is ignored. maxParticles(=100 ) The number of particles at which to stop generating new ones regardless of massFlowRate. if maxParticles=-1 - this parameter is ignored . normal(=Vector3r(NaN, NaN, NaN)) Orientation of the regions geometry, direction of particles velocites if normalVel is not set. normalVel(=Vector3r(NaN, NaN, NaN)) Direction of particles velocites. numParticles(=0 ) Cummulative number of particles produces so far (auto-updated) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rMax(=NaN ) Maximum radius of generated spheres (uniform distribution) rMin(=NaN ) Minimum radius of generated spheres (uniform distribution) silent(=false) If true no complain about excessing maxAttempt but disable the factory (by set massFlowRate=0).
238
stopIfFailed(=true) If true, the SpheresFactory stops (sets massFlowRate=0), when maximal number of attempts to insert particle exceed. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. totalMass(=0 ) Mass of spheres that was produced so far. (auto-updated) totalVolume(=0 ) Volume of spheres that was produced so far. (auto-updated) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary vAngle(=NaN ) Maximum angle by which the initial sphere velocity deviates from the normal. vMax(=NaN ) Maximum velocity norm of generated spheres (uniform distribution) vMin(=NaN ) Minimum velocity norm of generated spheres (uniform distribution) class yade.wrapper.TesselationWrapper(inherits GlobalEngine Engine Serializable) Handle the triangulation of spheres in a scene, build tesselation on request, and give access to computed quantities (see also the dedicated section in user manual). The calculation of microstrain is explained in [Catalano2013a] See example usage in script example/tesselationWrapper/tesselationWrapper.py. Below is an output of the defToVtk function visualized with paraview (in this case Yades TesselationWrapper was used to process experimental data obtained on sand by Edward Ando at Grenoble University, 3SR lab.) ComputeDeformations() None Compute per-particle deformation. Get it with TesselationWrapper::deformation (id,i,j). computeVolumes() None Compute volumes of all Voronois cells. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. defToVtk([(str)outputFile=def.vtk ]) None Write local deformations in vtk format from states 0 and 1. defToVtkFromPositions([(str)input1=pos1 [, (str)input2=pos2 [, (str)outputFile=def.vtk [, (bool)bz2=False ]]]]) None Write local deformations in vtk format from positions les (one sphere per line, with x,y,z,rad separated by spaces). defToVtkFromStates([(str)input1=state1 [, (str)input2=state2 [, (str)outputFile=def.vtk [, (bool)bz2=False ]]]]) None Write local deformations in vtk format from state les (since the le format is very special, consider using defToVtkFromPositions if the input les were not generated by TesselationWrapper). deformation((int)id, (int)i, (int)j ) oat Get particle deformation dict() dict Return dictionary of attributes.
7.3. Global engines
239
execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). getVolPoroDef([(bool)deformation=False ]) dict Return a table with per-sphere computed quantities. Include deformations on the increment dened by states 0 and 1 if deformation=True (make sure to dene states 0 and 1 consistently). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. loadState((str)arg2, (bool)arg3 [, (bool)stateNumber=0 ]) None Load a le with positions to dene state 0 or 1. n_spheres(=0 ) (auto-computed) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. saveState([(str)outputFile=state [, (bool)state=0 [, (bool)bz2=True ]]]) None Save a le with positions, can be later reloaded in order to dene state 0 or 1. setState([(bool)state=0 ]) None Make the current state of the simulation the initial (0) or nal (1) conguration for the
240
denition of displacement increments, use only state=0 if you just want to get volmumes and porosity. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. triangulate([(bool)reset=True ]) None triangulate spheres of the packing updateAttrs((dict)arg2 ) None Update object attributes from given dictionary volume([(int)id=0 ]) oat Returns the volume of Voronois cell of a sphere. class yade.wrapper.TetraVolumetricLaw(inherits GlobalEngine Engine Serializable) Calculate physical response of 2 tetrahedra in interaction, based on penetration conguration given by TTetraGeom. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.TimeStepper(inherits GlobalEngine Engine Serializable) Engine dening time-step (fundamental class) active(=true) is the engine active? dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True).
7.3. Global engines
241
execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timeStepUpdateInterval(=1) dt update interval timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.TorqueRecorder(inherits Recorder PeriodicEngine GlobalEngine Engine Serializable) Engine saves the total torque according to the given axis and ZeroPoint, the force is taken from bodies, listed in ids For instance, can be useful for dening the torque, which aects on ball mill during its work. addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. ids(=uninitalized ) List of bodies whose state will be measured initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) 242 Chapter 7. Class reference (yade.wrapper module)
nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) rotationAxis(=Vector3r::UnitX()) Rotation axis timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. totalTorque(=0 ) Resultant torque, returning by the function. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) zeroPoint(=Vector3r::Zero()) Point of rotation center class yade.wrapper.TriaxialStateRecorder(inherits Recorder PeriodicEngine GlobalEngine Engine Serializable) Engine recording triaxial variables (see the variables list in the rst line of the output le). This recorder needs TriaxialCompressionEngine or ThreeDTriaxialEngine present in the simulation). addIterNum(=false) Adds an iteration number to the le name, when the le was created. Useful for creating new les at each call (false by default) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). file(=uninitalized ) Name of le to save to; must not be empty. initRun(=false) Run the rst time we are called as well.
7.3. Global engines
243
iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. porosity(=1) porosity of the packing [-] realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. truncate(=false) Whether to delete current le contents, if any, when opening (false by default) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0) class yade.wrapper.VTKRecorder(inherits PeriodicEngine GlobalEngine Engine Serializable) Engine recording snapshots of simulation into series of *.vtu les, readable by VTK-based postprocessing programs such as Paraview. Both bodies (spheres and facets) and interactions can be recorded, with various vector/scalar quantities that are dened on them. PeriodicEngine.initRun is initialized to True automatically. ascii(=false) Store data as readable text in the XML le (sets vtkXMLWriter data mode to vtkXMLWriter::Ascii, while the default is Appended compress(=false) Compress output XML les [experimental]. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point.
244
dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). fileName(= ) Base le name; it will be appended with {spheres,intrs,facets}-243100.vtu (unless multiblock is True) depending on active recorders and step number (243100 in this case). It can contain slashes, but the directory must exist already. initRun(=false) Run the rst time we are called as well. iterLast(=0 ) Tracks step number of last run (auto-updated). iterPeriod(=0, deactivated ) Periodicity criterion using step number (deactivated if <= 0) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=0 ) If mask dened, only bodies with corresponding groupMask will be exported. If 0, all bodies will be exported. multiblock(=false) Use multi-block (.vtm) les to store data, rather than separate .vtu les. nDo(=-1, deactivated ) Limit number of executions by this number (deactivated if negative) nDone(=0 ) Track number of executions (cummulative) (auto-updated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. realLast(=0 ) Tracks real time of last run (auto-updated). realPeriod(=0, deactivated ) Periodicity criterion using real (wall clock, computation, human) time (deactivated if <=0) recorders List of active recorders (as strings). all (the default value) enables all base and generic recorders. Base recorders Base recorders save the geometry (unstructured grids) on which other data is dened. They are implicitly activated by many of the other recorders. Each of them creates a new le (or a block, if multiblock is set). spheres Saves positions and radii (radii) of spherical particles. facets Save facets positions (vertices). boxes Save boxes positions (edges). 7.3. Global engines 245
intr Store interactions as lines between nodes at respective particles positions. Additionally stores magnitude of normal (forceN) and shear (absForceT) forces on interactions (the geom).
Generic recorders Generic recorders do not depend on specic model being used and save commonly useful data. id Saves ids (eld id) of spheres; active only if spheres is active. mass Saves masses (eld mass) of spheres; active only if spheres is active. clumpId Saves ids of clumps to which each sphere belongs (eld clumpId); active only if spheres is active. colors Saves colors of spheres and of facets (eld color); only active if spheres or facets are activated. mask Saves groupMasks of spheres and of facets (eld mask); only active if spheres or facets are activated. materialId Saves materialID of spheres and of facets; only active if spheres or facets are activated. velocity Saves linear and angular velocities of spherical particles as Vector3 and length(elds linVelVec, linVelLen and angVelVec, angVelLen respectively); only eective with spheres. stress Saves stresses of spheres and of facets as Vector3 and length; only active if spheres or facets are activated.
Specic recorders The following should only be activated in appropriate cases, otherwise crashes can occur due to violation of type presuppositions. cpm Saves data pertaining to the concrete model: cpmDamage (normalized residual strength averaged on particle), cpmStress (stress on particle); intr is activated automatically by cpm rpm Saves data pertaining to the rock particle model: rpmSpecNum shows dierent pieces of separated stones, only ids. rpmSpecMass shows masses of separated stones. wpm Saves data pertaining to the wire particle model: wpmForceNFactor shows the loading factor for the wire, e.g. normal force divided by threshold normal force. skipFacetIntr(=true) Skip interactions with facets, when saving interactions skipNondynamic(=false) Skip non-dynamic spheres (but not facets). timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary virtLast(=0 ) Tracks virtual time of last run (auto-updated). virtPeriod(=0, deactivated ) Periodicity criterion using virtual (simulation) time (deactivated if <= 0)
246
7.3.2 BoundaryController
Dis p2DPropLoadEngine KinemCTDEngine Peri3dController KinemCNS Engine PeriIs oCompres s or KinemCNLEngine BoundaryController KinemS impleS hearBox KinemCNDEngine TriaxialS tres s Controller TriaxialCompres s ionEngine UniaxialS trainer ThreeDTriaxialEngine PeriTriaxController
class yade.wrapper.BoundaryController(inherits GlobalEngine Engine Serializable) Base for engines controlling boundary conditions of simulations. Not to be used directly. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Disp2DPropLoadEngine(inherits BoundaryController GlobalEngine Engine Serializable) Disturbs a simple shear sample in a given displacement direction
7.3. Global engines
247
This engine allows one to apply, on a simple shear sample, a loading controlled by du/dgamma = cste, which is equivalent to du + cste * dgamma = 0 (proportionnal path loadings). To do so, the upper plate of the simple shear box is moved in a given direction (corresponding to a given du/dgamma), whereas lateral plates are moved so that the box remains closed. This engine can easily be used to perform directionnal probes, with a python script launching successivly the same .xml which contains this engine, after having modied the direction of loading (see theta attribute). Thats why this Engine contains a saveData procedure which can save data on the state of the sample at the end of the loading (in case of successive loadings - for successive directions - through a python script, each line would correspond to one direction of loading). Key(= ) string to add at the names of the saved les, and of the output le lled by saveData LOG(=false) boolean controling the output of messages on the screen dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). id_boxback(=4) the id of the wall at the back of the sample id_boxbas(=1) the id of the lower wall id_boxfront(=5 ) the id of the wall in front of the sample id_boxleft(=0 ) the id of the left wall id_boxright(=2 ) the id of the right wall id_topbox(=3 ) the id of the upper wall label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nbre_iter(=0 ) the number of iterations of loading to perform ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. theta(=0.0 ) the angle, in a (gamma,h=-u) plane from the gamma - axis to the perturbation vector (trigo wise) [degrees]
248
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary v(=0.0 ) the speed at which the perturbation is imposed. In case of samples which are more sensitive to normal loadings than tangential ones, one possibility is to take v = V_shear - | (V_shearV_comp)*sin(theta) | => v=V_shear in shear; V_comp in compression [m/s] class yade.wrapper.KinemCNDEngine(inherits KinemSimpleShearBox BoundaryController GlobalEngine Engine Serializable) To apply a Constant Normal Displacement (CND) shear for a parallelogram box This engine, designed for simulations implying a simple shear box (SimpleShear Preprocessor or scripts/simpleShear.py), allows one to perform a constant normal displacement shear, by translating horizontally the upper plate, while the lateral ones rotate so that they always keep contact with the lower and upper walls. Key(= ) string to add at the names of the saved les LOG(=false) boolean controling the output of messages on the screen alpha(=Mathr::PI/2.0 ) the angle from the lower box to the left box (trigo wise). Measured by this Engine. Has to be saved, but not to be changed by the user. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f0(=0.0 ) the (vertical) force acting on the upper plate on the very rst time step (determined by the Engine). Controls of the loadings in case of KinemCNSEngine or KinemCNLEngine will be done according to this initial value [N]. Has to be saved, but not to be changed by the user. firstRun(=true) boolean set to false as soon as the engine has done its job one time : useful to know if initial height of, and normal force sustained by, the upper box are known or not (and thus if they have to be initialized). Has to be saved, but not to be changed by the user. gamma(=0.0 ) the current value of the tangential displacement gamma_save(=uninitalized ) vector with the values of gamma at which a save of the simulation is performed [m] gammalim(=0.0 ) the value of the tangential displacement at wich the displacement is stopped [m] id_boxback(=4) the id of the wall at the back of the sample id_boxbas(=1) the id of the lower wall
7.3. Global engines
249
id_boxfront(=5 ) the id of the wall in front of the sample id_boxleft(=0 ) the id of the left wall id_boxright(=2 ) the id of the right wall id_topbox(=3 ) the id of the upper wall label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. max_vel(=1.0 ) to limit the speed of the vertical displacements done to control (CNL or CNS cases) [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. shearSpeed(=0.0 ) the speed at which the shear is performed : speed of the upper plate [m/s] temoin_save(=uninitalized ) vector (same length as gamma_save for ex), with 0 or 1 depending whether the save for the corresponding value of gamma has been done (1) or not (0). Has to be saved, but not to be changed by the user. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wallDamping(=0.2 ) the vertical displacements done to to control (CNL or CNS cases) are in fact damped, through this wallDamping y0(=0.0 ) the height of the upper plate at the very rst time step : the engine nds its value [m]. Has to be saved, but not to be changed by the user. class yade.wrapper.KinemCNLEngine(inherits KinemSimpleShearBox BoundaryController GlobalEngine Engine Serializable) To apply a constant normal stress shear (i.e. Constant Normal Load : CNL) for a parallelogram box (simple shear box : SimpleShear Preprocessor or scripts/simpleShear.py) This engine allows one to translate horizontally the upper plate while the lateral ones rotate so that they always keep contact with the lower and upper walls. In fact the upper plate can move not only horizontally but also vertically, so that the normal stress acting on it remains constant (this constant value is not chosen by the user but is the one that exists at the beginning of the simulation) The right vertical displacements which will be allowed are computed from the rigidity Kn of the sample over the wall (so to cancel a deltaSigma, a normal dplt deltaSigma*S/(Kn) is set) The movement is moreover controlled by the user via a shearSpeed which will be the speed of the upper wall, and by a maximum value of horizontal displacement gammalim, after which the shear stops.
250
Note: Not only the positions of walls are updated but also their speeds, which is all but useless considering the fact that in the contact laws these velocities of bodies are used to compute values of tangential relative displacements. Warning: Because of this last point, if you want to use later saves of simulations executed with this Engine, but without that stopMovement was executed, your boxes will keep their speeds => you will have to cancel them by hand in the .xml. Key(= ) string to add at the names of the saved les LOG(=false) boolean controling the output of messages on the screen alpha(=Mathr::PI/2.0 ) the angle from the lower box to the left box (trigo wise). Measured by this Engine. Has to be saved, but not to be changed by the user. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f0(=0.0 ) the (vertical) force acting on the upper plate on the very rst time step (determined by the Engine). Controls of the loadings in case of KinemCNSEngine or KinemCNLEngine will be done according to this initial value [N]. Has to be saved, but not to be changed by the user. firstRun(=true) boolean set to false as soon as the engine has done its job one time : useful to know if initial height of, and normal force sustained by, the upper box are known or not (and thus if they have to be initialized). Has to be saved, but not to be changed by the user. gamma(=0.0 ) current value of tangential displacement [m] gamma_save(=uninitalized ) vector with the values of gamma at which a save of the simulation is performed [m] gammalim(=0.0 ) the value of tangential displacement (of upper plate) at wich the shearing is stopped [m] id_boxback(=4) the id of the wall at the back of the sample id_boxbas(=1) the id of the lower wall id_boxfront(=5 ) the id of the wall in front of the sample id_boxleft(=0 ) the id of the left wall id_boxright(=2 ) the id of the right wall
7.3. Global engines
251
id_topbox(=3 ) the id of the upper wall label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. max_vel(=1.0 ) to limit the speed of the vertical displacements done to control (CNL or CNS cases) [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. shearSpeed(=0.0 ) the speed at wich the shearing is performed : speed of the upper plate [m/s] temoin_save(=uninitalized ) vector (same length as gamma_save for ex), with 0 or 1 depending whether the save for the corresponding value of gamma has been done (1) or not (0). Has to be saved, but not to be changed by the user. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wallDamping(=0.2 ) the vertical displacements done to to control (CNL or CNS cases) are in fact damped, through this wallDamping y0(=0.0 ) the height of the upper plate at the very rst time step : the engine nds its value [m]. Has to be saved, but not to be changed by the user. class yade.wrapper.KinemCNSEngine(inherits KinemSimpleShearBox BoundaryController GlobalEngine Engine Serializable) To apply a Constant Normal Stifness (CNS) shear for a parallelogram box (simple shear) This engine, useable in simulations implying one deformable parallelepipedic box, allows one to translate horizontally the upper plate while the lateral ones rotate so that they always keep contact with the lower and upper walls. The upper plate can move not only horizontally but also vertically, so that the normal rigidity dened by DeltaF(upper plate)/DeltaU(upper plate) = constant (= KnC dened by the user). The movement is moreover controlled by the user via a shearSpeed which is the horizontal speed of the upper wall, and by a maximum value of horizontal displacement gammalim (of the upper plate), after which the shear stops. Note: not only the positions of walls are updated but also their speeds, which is all but useless considering the fact that in the contact laws these velocities of bodies are used to compute values of tangential relative displacements. Warning: But, because of this last point, if you want to use later saves of simulations executed with this Engine, but without that stopMovement was executed, your boxes will keep their speeds => you will have to cancel them by hand in the .xml
252
Key(= ) string to add at the names of the saved les KnC(=10.0e6 ) the normal rigidity chosen by the user [MPa/mm] - the conversion in Pa/m will be made LOG(=false) boolean controling the output of messages on the screen alpha(=Mathr::PI/2.0 ) the angle from the lower box to the left box (trigo wise). Measured by this Engine. Has to be saved, but not to be changed by the user. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f0(=0.0 ) the (vertical) force acting on the upper plate on the very rst time step (determined by the Engine). Controls of the loadings in case of KinemCNSEngine or KinemCNLEngine will be done according to this initial value [N]. Has to be saved, but not to be changed by the user. firstRun(=true) boolean set to false as soon as the engine has done its job one time : useful to know if initial height of, and normal force sustained by, the upper box are known or not (and thus if they have to be initialized). Has to be saved, but not to be changed by the user. gamma(=0.0 ) current value of tangential displacement [m] gammalim(=0.0 ) the value of tangential displacement (of upper plate) at wich the shearing is stopped [m] id_boxback(=4) the id of the wall at the back of the sample id_boxbas(=1) the id of the lower wall id_boxfront(=5 ) the id of the wall in front of the sample id_boxleft(=0 ) the id of the left wall id_boxright(=2 ) the id of the right wall id_topbox(=3 ) the id of the upper wall label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. max_vel(=1.0 ) to limit the speed of the vertical displacements done to control (CNL or CNS cases) [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can 7.3. Global engines 253
depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. shearSpeed(=0.0 ) the speed at wich the shearing is performed : speed of the upper plate [m/s] temoin_save(=uninitalized ) vector (same length as gamma_save for ex), with 0 or 1 depending whether the save for the corresponding value of gamma has been done (1) or not (0). Has to be saved, but not to be changed by the user. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wallDamping(=0.2 ) the vertical displacements done to to control (CNL or CNS cases) are in fact damped, through this wallDamping y0(=0.0 ) the height of the upper plate at the very rst time step : the engine nds its value [m]. Has to be saved, but not to be changed by the user. class yade.wrapper.KinemCTDEngine(inherits KinemSimpleShearBox BoundaryController GlobalEngine Engine Serializable) To compress a simple shear sample by moving the upper box in a vertical way only, so that the tangential displacement (dened by the horizontal gap between the upper and lower boxes) remains constant (thus, the CTD = Constant Tangential Displacement). The lateral boxes move also to keep always contact. All that until this box is submitted to a given stress (=*targetSigma*). Moreover saves are executed at each value of stresses stored in the vector sigma_save, and at targetSigma Key(= ) string to add at the names of the saved les LOG(=false) boolean controling the output of messages on the screen alpha(=Mathr::PI/2.0 ) the angle from the lower box to the left box (trigo wise). Measured by this Engine. Has to be saved, but not to be changed by the user. compSpeed(=0.0 ) (vertical) speed of the upper box : >0 for real compression, <0 for unloading [m/s] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f0(=0.0 ) the (vertical) force acting on the upper plate on the very rst time step (determined by the Engine). Controls of the loadings in case of KinemCNSEngine or KinemCNLEngine will be done according to this initial value [N]. Has to be saved, but not to be changed by the user.
254
firstRun(=true) boolean set to false as soon as the engine has done its job one time : useful to know if initial height of, and normal force sustained by, the upper box are known or not (and thus if they have to be initialized). Has to be saved, but not to be changed by the user. id_boxback(=4) the id of the wall at the back of the sample id_boxbas(=1) the id of the lower wall id_boxfront(=5 ) the id of the wall in front of the sample id_boxleft(=0 ) the id of the left wall id_boxright(=2 ) the id of the right wall id_topbox(=3 ) the id of the upper wall label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. max_vel(=1.0 ) to limit the speed of the vertical displacements done to control (CNL or CNS cases) [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. sigma_save(=uninitalized ) vector with the values of sigma at which a save of the simulation should be performed [kPa] targetSigma(=0.0 ) the value of sigma at which the compression should stop [kPa] temoin_save(=uninitalized ) vector (same length as gamma_save for ex), with 0 or 1 depending whether the save for the corresponding value of gamma has been done (1) or not (0). Has to be saved, but not to be changed by the user. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wallDamping(=0.2 ) the vertical displacements done to to control (CNL or CNS cases) are in fact damped, through this wallDamping y0(=0.0 ) the height of the upper plate at the very rst time step : the engine nds its value [m]. Has to be saved, but not to be changed by the user. class yade.wrapper.KinemSimpleShearBox(inherits BoundaryController GlobalEngine Engine Serializable) This class is supposed to be a mother class for all Engines performing loadings on the simple shear box of SimpleShear. It is not intended to be used by itself, but its declaration and implentation
7.3. Global engines
255
will thus contain all what is useful for all these Engines. The script simpleShear.py illustrates the use of the various corresponding Engines. Key(= ) string to add at the names of the saved les LOG(=false) boolean controling the output of messages on the screen alpha(=Mathr::PI/2.0 ) the angle from the lower box to the left box (trigo wise). Measured by this Engine. Has to be saved, but not to be changed by the user. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f0(=0.0 ) the (vertical) force acting on the upper plate on the very rst time step (determined by the Engine). Controls of the loadings in case of KinemCNSEngine or KinemCNLEngine will be done according to this initial value [N]. Has to be saved, but not to be changed by the user. firstRun(=true) boolean set to false as soon as the engine has done its job one time : useful to know if initial height of, and normal force sustained by, the upper box are known or not (and thus if they have to be initialized). Has to be saved, but not to be changed by the user. id_boxback(=4) the id of the wall at the back of the sample id_boxbas(=1) the id of the lower wall id_boxfront(=5 ) the id of the wall in front of the sample id_boxleft(=0 ) the id of the left wall id_boxright(=2 ) the id of the right wall id_topbox(=3 ) the id of the upper wall label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. max_vel(=1.0 ) to limit the speed of the vertical displacements done to control (CNL or CNS cases) [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level.
256
temoin_save(=uninitalized ) vector (same length as gamma_save for ex), with 0 or 1 depending whether the save for the corresponding value of gamma has been done (1) or not (0). Has to be saved, but not to be changed by the user. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wallDamping(=0.2 ) the vertical displacements done to to control (CNL or CNS cases) are in fact damped, through this wallDamping y0(=0.0 ) the height of the upper plate at the very rst time step : the engine nds its value [m]. Has to be saved, but not to be changed by the user. class yade.wrapper.Peri3dController(inherits BoundaryController GlobalEngine Engine Serializable) Class for controlling independently all 6 components of engineering stress and strain of periodic Cell. goal are the goal values, while stressMask determines which components prescribe stress and which prescribe strain. If the strain is prescribed, appropeiate strain rate is directly applied. If the stress is prescribed, the strain predictor is used: from stress values in two previous steps the value of strain rate is prescribed so as the value of stress in the next step is as close as possible to the ideal one. Current algorithm is extremly simple and probably will be changed in future, but is roboust enough and mostly works ne. Stress error (dierence between actual and ideal stress) is evaluated in current and previous steps (di , di1 . Linear extrapolation is used to estimate error in the next step di+1 = 2di di1
According to this error, the strain rate is modied by mod parameter { i )) mod i+1 = i max(abs( >0 di+1 i )) mod i+1 = i + max(abs( <0
According to this fact, the prescribed stress will (almost) never have exact prescribed value, but the dierence would be very small (and decreasing for increasing nSteps. This approach works good if one of the dominant strain rates is prescribed. If all stresses are prescribed or if all goal strains is prescribed as zero, a good estimation is needed for the rst step, therefore the compliance matrix is estimated (from user dened estimations of macroscopic material parameters youngEstimation and poissonEstimation) and respective strain rates is computed form prescribed stress rates and compliance matrix (the estimation of compliance matrix could be computed autamatically avoiding user inputs of this kind). The simulation on rotated periodic cell is also supported. Firstly, the polar decomposition is performed on cells transformation matrix trsf T = UP, where U is orthogonal (unitary) matrix representing rotation and P is a positive semi-denite Hermitian matrix representing strain. A logarithm of P should be used to obtain realistic values at higher strain values (not implemented is properly rotated to cells local yet). A prescribed strain increment in global coordinates dt coordinates and added to P Pi+1 = P + UT dt U
7.3. Global engines
257
The new value of trsf is computed at T i+1 = UPi+1 . From current and next trsf the cells velocity gradient velGrad is computed (according to its denition) as V = (T i+1 T 1 I)/dt
Current implementation allow user to dene independent loading path for each prescribed component. i.e. dene the prescribed value as a function of time (or progress or steps). See Paths. Examples scripts/test/peri3dController_example1 and scripts/test/peri3dController_triaxialCompression explain usage and inputs of Peri3dController, scripts/test/peri3dController_shear is an example of using shear components and also simulation on rotatd cell. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. doneHook(=uninitalized ) Python command (as string) to run when nSteps is achieved. If empty, the engine will be set dead. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). goal(=Vector6r::Zero()) Goal state; only the upper triangular matrix is considered; each component is either prescribed stress or strain, depending on stressMask. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. lenPe(=NaN ) Peri3dController internal variable lenPs(=NaN ) Peri3dController internal variable maxStrain(=1e6 ) Maximal asolute value of strain allowed in the simulation. If reached, the simulation is considered as nished maxStrainRate(=1e3 ) Maximal absolute value of strain rate (both normal and shear components of strain) mod(=.1) Predictor modicator, by trail-and-error analysis the value 0.1 was found as the best. nSteps(=1000 ) Number of steps of the simulation. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. pathSizes(=Vector6i::Zero()) Peri3dController internal variable
258
pathsCounter(=Vector6i::Zero()) Peri3dController internal variable pe(=Vector6i::Zero()) Peri3dController internal variable poissonEstimation(=.25 ) Estimation of macroscopic Poissons ratio, used used for the rst simulation step progress(=0.) Actual progress of the simulation with Controller. ps(=Vector6i::Zero()) Peri3dController internal variable strain(=Vector6r::Zero()) Current strain (deformation) vector (x ,y ,z ,yz ,zx ,xy ) (auto-updated). strainGoal(=Vector6r::Zero()) Peri3dController internal variable strainRate(=Vector6r::Zero()) Current strain rate vector. stress(=Vector6r::Zero()) Current stress vector (x ,y ,z ,yz ,zx ,xy )|yupdate|. stressGoal(=Vector6r::Zero()) Peri3dController internal variable stressIdeal(=Vector6r::Zero()) Ideal stress vector at current time step. stressMask(=0, all strains) mask determining whether components of goal are strain (0) or stress (1). The order is 00,11,22,12,02,01 from the least signicant bit. (e.g. 0b000011 is stress 00 and stress 11). stressRate(=Vector6r::Zero()) Current stress rate vector (that is prescribed, the actual one slightly dier). timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary xxPath Time function (piecewise linear) for xx direction. Sequence of couples of numbers. First number is time, second number desired value of respective quantity (stress or strain). The last couple is considered as nal state (equal to (nSteps, goal)), other values are relative to this state. Example: nSteps=1000, goal[0]=300, xxPath=((2,3),(4,1),(5,2)) at step 400 (=5*1000/2) the value is 450 (=3*300/2), at step 800 (=4*1000/5) the value is 150 (=1*300/2), at step 1000 (=5*1000/5=nSteps) the value is 300 (=2*300/2=goal[0]). See example scripts/test/peri3dController_example1 for illusration. xyPath(=vector<Vector2r>(1, Vector2r::Ones())) Time function for xy direction, see xxPath youngEstimation(=1e20 ) Estimation of macroscopic Youngs modulus, used for the rst simulation step yyPath(=vector<Vector2r>(1, Vector2r::Ones())) Time function for yy direction, see xxPath
7.3. Global engines
259
yzPath(=vector<Vector2r>(1, Vector2r::Ones())) Time function for yz direction, see xxPath zxPath(=vector<Vector2r>(1, Vector2r::Ones())) Time function for zx direction, see xxPath zzPath(=vector<Vector2r>(1, Vector2r::Ones())) Time function for zz direction, see xxPath class yade.wrapper.PeriIsoCompressor(inherits BoundaryController GlobalEngine Engine Serializable) Compress/decompress cloud of spheres by controlling periodic cell size until it reaches prescribed average stress, then moving to next stress value in given stress series. charLen(=-1.) Characteristic length, should be something like mean particle diameter (default -1=invalid value)) currUnbalanced Current value of unbalanced force dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. doneHook(= ) Python command to be run when reaching the last specied stress execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). globalUpdateInt(=20 ) how often to recompute average stress, stiness and unbalanced force keepProportions(=true) Exactly keep proportions of the cell (stress is controlled based on average, not its components label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxSpan(=-1.) Maximum body span in terms of bbox, to prevent periodic cell getting too small. (autocomputed) maxUnbalanced(=1e-4) if actual unbalanced force is smaller than this number, the packing is considered stable, ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. sigma Current stress value state(=0 ) Where are we at in the stress series
260
stresses(=uninitalized ) Stresses that should be reached, one after another timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.PeriTriaxController(inherits BoundaryController GlobalEngine Engine Serializable) Engine for independently controlling stress or strain in periodic simulations. strainStress contains absolute values for the controlled quantity, and stressMask determines meaning of those values (0 for strain, 1 for stress): e.g. ( 1<<0 | 1<<2 ) = 1 | 4 = 5 means that strainStress[0] and strainStress[2] are stress values, and strainStress[1] is strain. See scripts/test/periodic-triax.py for a simple example. absStressTol(=1e3 ) Absolute stress tolerance currUnbalanced(=NaN ) current unbalanced force (updated every globUpdate) (auto-updated) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. doneHook(=uninitalized ) python command to be run when the desired state is reached dynCell(=false) Imposed stress can be controlled using the packing stiness or by applying the laws of dynamic (dynCell=true). Dont forget to assign a mass to the cell. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). externalWork(=0 ) Work input from boundary controller. globUpdate(=5 ) How often to recompute average stress, stiness and unbalaced force. goal Desired stress or strain strain(i)=log(Fii). values (depending on stressMask), strains dened as
Warning: Strains are relative to the O.cell.refSize (reference cell size), not the current one (e.g. at the moment when the new strain value is set). growDamping(=.25 ) Damping of cell resizing (0=perfect control, 1=no control at all); see also wallDamping in TriaxialStressController. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python.
7.3. Global engines
261
mass(=NaN ) mass of the cell (user set); if not set and dynCell is used, it will be computed as sum of masses of all particles. maxBodySpan(=Vector3r::Zero()) maximum body dimension (auto-computed) maxStrainRate(=Vector3r(1, 1, 1)) Maximum strain rate of the periodic cell. maxUnbalanced(=1e-4) maximum unbalanced force. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. prevGrow(=Vector3r::Zero()) previous cell grow relStressTol(=3e-5 ) Relative stress tolerance stiff(=Vector3r::Zero()) average stiness (only every globUpdate steps recomputed from interactions) (auto-updated) strain(=Vector3r::Zero()) cell strain (auto-updated) strainRate(=Vector3r::Zero()) cell strain rate (auto-updated) stress(=Vector3r::Zero()) diagonal terms of the stress tensor stressMask(=0, all strains) mask determining strain/stress (0/1) meaning for goal components stressTensor(=Matrix3r::Zero()) average stresses, updated at every step (only every globUpdate steps recomputed from interactions if !dynCell) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary useDem3Dof(=false) For some constitutive laws (practicaly all laws based on Dem3Dof), normalForce and shearForce on interactions are in the reverse sense and this ag must be true. See this message. class yade.wrapper.ThreeDTriaxialEngine(inherits TriaxialStressController BoundaryController GlobalEngine Engine Serializable) The engine perform a triaxial compression with a control in direction i in stress (if stressControl_i) else in strain. For a stress control the imposed stress is specied by sigma_i with a max_veli depending on strainRatei. To obtain the same strain rate in stress control than in strain control you need to set wallDamping = 0.8. For a strain control the imposed strain is specied by strainRatei. With this engine you can also perform internal compaction by growing the size of particles by using
262
TriaxialStressController::controlInternalStress. For that, just switch on internalCompaction=1 and x sigma_iso=value of mean pressure that you want at the end of the internal compaction. Key(= ) A string appended at the end of all les, use it to name simulations. UnbalancedForce(=1) mean resultant forces divided by mean contact force boxVolume Total packing volume. computeStressStrainInterval(=10 ) currentStrainRate1(=0 ) current strain rate in direction 1 - converging to ThreeDTriaxialEngine::strainRate1 (./s) currentStrainRate2(=0 ) current strain rate in direction 2 - converging to ThreeDTriaxialEngine::strainRate2 (./s) currentStrainRate3(=0 ) current strain rate in direction 3 - converging to ThreeDTriaxialEngine::strainRate3 (./s) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. depth(=0 ) size of the box (2-axis) (auto-updated) depth0(=0 ) Reference size for strain denition. See TriaxialStressController::depth dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). externalWork(=0 ) Energy provided by boundaries.|yupdate| finalMaxMultiplier(=1.00001) max multiplier of diameters during internal compaction (secondary precise adjustment - TriaxialStressController::maxMultiplier is used in the initial stage) frictionAngleDegree(=-1) Value of friction used in the simulation if (updateFrictionAngle) goal1(=0 ) prescribed stress/strain rate on axis 1, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) goal2(=0 ) prescribed stress/strain rate on axis 2, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) goal3(=0 ) prescribed stress/strain rate on axis 3, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) height(=0 ) size of the box (1-axis) (auto-updated) height0(=0 ) Reference size for strain denition. See TriaxialStressController::height
7.3. Global engines
263
internalCompaction(=true) Switch between external (walls) and internal (growth of particles) compaction. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxMultiplier(=1.001) max multiplier of diameters during internal compaction (initial fast increase - TriaxialStressController::nalMaxMultiplier is used in a second stage) max_vel(=1) Maximum allowed walls velocity [m/s]. This value superseeds the one assigned by the stress controller if the later is higher. max_vel can be set to innity in many cases, but sometimes helps stabilizing packings. Based on this value, dierent maxima are computed for each axis based on the dimensions of the sample, so that if each boundary moves at its maximum velocity, the strain rate will be isotropic (see e.g. TriaxialStressController::max_vel1). max_vel1 see TriaxialStressController::max_vel (auto-computed) max_vel2 see TriaxialStressController::max_vel (auto-computed) max_vel3 see TriaxialStressController::max_vel (auto-computed) meanStress(=0 ) Mean stress in the packing. (auto-updated) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. porosity Porosity of the packing. previousMultiplier(=1) (auto-updated) previousStress(=0 ) (auto-updated) radiusControlInterval(=10 ) setContactProperties((oat)arg2 ) None Assign a new friction angle (degrees) to dynamic bodies and relative interactions spheresVolume Total volume pf spheres. stiffnessUpdateInterval(=10 ) target strain rate (./s) strain Current strain in a vector (exx,eyy,ezz). The values reect true (logarithmic) strain. strainDamping(=0.9997 ) factor used for smoothing changes in eective strain rate. If target rate is TR, then (1damping)*(TR-currentRate) will be added at each iteration. With damping=0, rate=target all the time. With damping=1, it doesnt change. strainRate Current strain rate in a vector d/dt(exx,eyy,ezz).
264
strainRate1(=0 ) target strain rate in direction 1 (./s) strainRate2(=0 ) target strain rate in direction 2 (./s) strainRate3(=0 ) target strain rate in direction 3 (./s) stress((int)id ) Vector3 Return the mean stress vector acting on boundary id, with id between 0 and 5. stressControl_1(=true) Switch to choose a stress or a strain control in directions 1 stressControl_2(=true) Switch to choose a stress or a strain control in directions 2 stressControl_3(=true) Switch to choose a stress or a strain control in directions 3 stressDamping(=0.25 ) wall damping coecient for the stress control - wallDamping=0 implies a (theoretical) perfect control, wallDamping=1 means no movement stressMask(=7 ) Bitmask determining if the components of TriaxialStressController::goal are stress (1) or strain (0). 0 for none, 7 for all, 1 for sigma1, etc. thickness(=-1) thickness of boxes (needed by some functions) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateFrictionAngle(=false) Switch to activate the update of the intergranular frictionto the value ThreeDTriaxialEngine::frictionAngleDegree. volumetricStrain(=0 ) Volumetric strain (see TriaxialStressController::strain).|yupdate| wall_back_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_back_id(=4) id of boundary ; coordinate 2- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_bottom_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_bottom_id(=2 ) id of boundary ; coordinate 1- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_front_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_front_id(=5 ) id of boundary ; coordinate 2+ (default value is ok if aabbWalls are appended BEFORE spheres.) wall_left_activated(=true) if true, this wall moves according to the target value (stress or strain rate).
7.3. Global engines
265
wall_left_id(=0 ) id of boundary ; coordinate 0- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_right_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_right_id(=1) id of boundary ; coordinate 0+ (default value is ok if aabbWalls are appended BEFORE spheres.) wall_top_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_top_id(=3 ) id of boundary ; coordinate 1+ (default value is ok if aabbWalls are appended BEFORE spheres.) width(=0 ) size of the box (0-axis) (auto-updated) width0(=0 ) Reference size for strain denition. See TriaxialStressController::width class yade.wrapper.TriaxialCompressionEngine(inherits TriaxialStressController BoundaryController GlobalEngine Engine Serializable) The engine is a state machine with the following states; transitions my be automatic, see below. 1.STATE_ISO_COMPACTION: isotropic compaction (compression) until the prescribed mean pressue sigmaIsoCompaction is reached and the packing is stable. The compaction happens either by straining the walls (!internalCompaction) or by growing size of grains (internalCompaction). 2.STATE_ISO_UNLOADING: isotropic unloading from the previously reached state, until the mean pressure sigmaLateralConnement is reached (and stabilizes). Note: this state will be skipped if sigmaLateralConnement == sigmaIsoCompaction. 3.STATE_TRIAX_LOADING: conned uniaxial compression: constant sigmaLateralConnement is kept at lateral walls (left, right, front, back), while top and bottom walls load the packing in their axis (by straining), until the value of epsilonMax (deformation along the loading axis) is reached. At this point, the simulation is stopped. 4.STATE_FIXED_POROSITY_COMPACTION: isotropic compaction (compression) until a chosen porosity value (parameter:xedPorosity). The six walls move with a chosen translation speed (parameter StrainRate). 5.STATE_TRIAX_LIMBO: currently unused, since simulation is hard-stopped in the previous state. Transition from COMPACTION to UNLOADING is done automatically if autoUnload==true; Transition from (UNLOADING to LOADING) or from (COMPACTION to LOADING: if UNLOADING is skipped) is done automatically if autoCompressionActivation=true; Both autoUnload and autoCompressionActivation are true by default. Note: Most of the algorithms used have been developed initialy for simulations reported in [Chareyre2002a] and [Chareyre2005]. They have been ported to Yade in a second step and used in e.g. [Kozicki2008],[Scholtes2009b]_,[Jerier2010b]. Key(= ) A string appended at the end of all les, use it to name simulations.
266
StabilityCriterion(=0.001) tolerance in terms of TriaxialCompressionEngine::UnbalancedForce to consider the packing is stable UnbalancedForce(=1) mean resultant forces divided by mean contact force autoCompressionActivation(=true) Auto-switch from isotropic compaction (or unloading state if sigmaLateralConnement<sigmaIsoCompaction) to deviatoric loading autoStopSimulation(=false) Stop the simulation when the sample reach STATE_LIMBO, or keep running autoUnload(=true) Auto-switch from isotropic compaction to unloading boxVolume Total packing volume. computeStressStrainInterval(=10 ) currentState(=1) There are 5 possible states in which TriaxialCompressionEngine can be. See above wrapper.TriaxialCompressionEngine currentStrainRate(=0 ) current strain rate - converging to TriaxialCompressionEngine::strainRate (./s) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. depth(=0 ) size of the box (2-axis) (auto-updated) depth0(=0 ) Reference size for strain denition. See TriaxialStressController::depth dict() dict Return dictionary of attributes. epsilonMax(=0.5 ) Value of axial deformation for which the loading must stop execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). externalWork(=0 ) Energy provided by boundaries.|yupdate| finalMaxMultiplier(=1.00001) max multiplier of diameters during internal compaction (secondary precise adjustment - TriaxialStressController::maxMultiplier is used in the initial stage) fixedPoroCompaction(=false) A special type of compaction with imposed nal porosity nEngine::xedPorosity (WARNING : can give unrealistic results!) fixedPorosity(=0 ) Value of porosity chosen by the user frictionAngleDegree(=-1) Value of friction assigned just before the deviatoric loading TriaxialCompressio-
7.3. Global engines
267
goal1(=0 ) prescribed stress/strain rate on axis 1, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) goal2(=0 ) prescribed stress/strain rate on axis 2, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) goal3(=0 ) prescribed stress/strain rate on axis 3, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) height(=0 ) size of the box (1-axis) (auto-updated) height0(=0 ) Reference size for strain denition. See TriaxialStressController::height internalCompaction(=true) Switch between external (walls) and internal (growth of particles) compaction. isAxisymetric(=false) if true, sigma_iso is assigned to sigma1, 2 and 3 (applies at each iteration and overrides user-set values of s1,2,3) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxMultiplier(=1.001) max multiplier of diameters during internal compaction (initial fast increase - TriaxialStressController::nalMaxMultiplier is used in a second stage) maxStress(=0 ) Max value of stress during the simulation (for post-processing) max_vel(=1) Maximum allowed walls velocity [m/s]. This value superseeds the one assigned by the stress controller if the later is higher. max_vel can be set to innity in many cases, but sometimes helps stabilizing packings. Based on this value, dierent maxima are computed for each axis based on the dimensions of the sample, so that if each boundary moves at its maximum velocity, the strain rate will be isotropic (see e.g. TriaxialStressController::max_vel1). max_vel1 see TriaxialStressController::max_vel (auto-computed) max_vel2 see TriaxialStressController::max_vel (auto-computed) max_vel3 see TriaxialStressController::max_vel (auto-computed) meanStress(=0 ) Mean stress in the packing. (auto-updated) noFiles(=false) If true, no les will be generated (*.xml, *.spheres,...) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. porosity Porosity of the packing.
268
previousMultiplier(=1) (auto-updated) previousSigmaIso(=1) Previous value of inherited sigma_iso (used to detect manual changes of the conning pressure) previousState(=1) Previous state (used to detect manual changes of the state in .xml) previousStress(=0 ) (auto-updated) radiusControlInterval(=10 ) setContactProperties((oat)arg2 ) None Assign a new friction angle (degrees) to dynamic bodies and relative interactions sigmaIsoCompaction(=1) Prescribed isotropic pressure during the compaction phase sigmaLateralConfinement(=1) Prescribed conning pressure in the deviatoric loading; might be dierent from TriaxialCompressionEngine::sigmaIsoCompaction sigma_iso(=0 ) prescribed conning stress (see :yref:TriaxialCompressionEngine::isAxisymetric) spheresVolume Total volume pf spheres. stiffnessUpdateInterval(=10 ) target strain rate (./s) strain Current strain in a vector (exx,eyy,ezz). The values reect true (logarithmic) strain. strainDamping(=0.99 ) coecient used for smoother transitions in the strain rate. The rate reaches the target value like dn reaches 0, where d is the damping coecient and n is the number of steps strainRate(=0 ) target strain rate (./s) stress((int)id ) Vector3 Return the mean stress vector acting on boundary id, with id between 0 and 5. stressDamping(=0.25 ) wall damping coecient for the stress control - wallDamping=0 implies a (theoretical) perfect control, wallDamping=1 means no movement stressMask(=7 ) Bitmask determining if the components of TriaxialStressController::goal are stress (1) or strain (0). 0 for none, 7 for all, 1 for sigma1, etc. testEquilibriumInterval(=20 ) interval of checks for transition between phases, higher than 1 saves computation time. thickness(=-1) thickness of boxes (needed by some functions) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. translationAxis(=TriaxialStressController::normal[wall_bottom] ) compression axis uniaxialEpsilonCurr(=1) Current value of axial deformation during conned loading (is reference to strain[1])
7.3. Global engines
269
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary volumetricStrain(=0 ) Volumetric strain (see TriaxialStressController::strain).|yupdate| wall_back_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_back_id(=4) id of boundary ; coordinate 2- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_bottom_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_bottom_id(=2 ) id of boundary ; coordinate 1- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_front_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_front_id(=5 ) id of boundary ; coordinate 2+ (default value is ok if aabbWalls are appended BEFORE spheres.) wall_left_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_left_id(=0 ) id of boundary ; coordinate 0- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_right_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_right_id(=1) id of boundary ; coordinate 0+ (default value is ok if aabbWalls are appended BEFORE spheres.) wall_top_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_top_id(=3 ) id of boundary ; coordinate 1+ (default value is ok if aabbWalls are appended BEFORE spheres.) warn(=0 ) counter used for sending a deprecation warning once width(=0 ) size of the box (0-axis) (auto-updated) width0(=0 ) Reference size for strain denition. See TriaxialStressController::width class yade.wrapper.TriaxialStressController(inherits BoundaryController GlobalEngine Engine Serializable) An engine maintaining constant stresses or constant strain rates on some boundaries of a parallepipedic packing. The stress/strain control is dened for each axis using TriaxialStressController::stressMask (a bitMask) and target values are dened by goal1,goal2, and goal3. sigmaIso has to be dened during growing phases. Note: The algorithms used have been developed initialy for simulations reported in [Chareyre2002a] and [Chareyre2005]. They have been ported to Yade in a second step and used in e.g. [Kozicki2008],[Scholtes2009b]_,[Jerier2010b].
270
boxVolume Total packing volume. computeStressStrainInterval(=10 ) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. depth(=0 ) size of the box (2-axis) (auto-updated) depth0(=0 ) Reference size for strain denition. See TriaxialStressController::depth dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). externalWork(=0 ) Energy provided by boundaries.|yupdate| finalMaxMultiplier(=1.00001) max multiplier of diameters during internal compaction (secondary precise adjustment - TriaxialStressController::maxMultiplier is used in the initial stage) goal1(=0 ) prescribed stress/strain rate on axis 1, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) goal2(=0 ) prescribed stress/strain rate on axis 2, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) goal3(=0 ) prescribed stress/strain rate on axis 3, as dened by TriaxialStressController::stressMask (see also TriaxialStressController::isAxisymetric) height(=0 ) size of the box (1-axis) (auto-updated) height0(=0 ) Reference size for strain denition. See TriaxialStressController::height internalCompaction(=true) Switch between external (walls) and internal (growth of particles) compaction. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxMultiplier(=1.001) max multiplier of diameters during internal compaction (initial fast increase - TriaxialStressController::nalMaxMultiplier is used in a second stage) max_vel(=1) Maximum allowed walls velocity [m/s]. This value superseeds the one assigned by the stress controller if the later is higher. max_vel can be set to innity in many cases, but sometimes helps stabilizing packings. Based on this value, dierent maxima are computed for each axis based on the dimensions of the sample, so that if each boundary moves at its maximum velocity, the strain rate will be isotropic (see e.g. TriaxialStressController::max_vel1). max_vel1 see TriaxialStressController::max_vel (auto-computed)
7.3. Global engines
271
max_vel2 see TriaxialStressController::max_vel (auto-computed) max_vel3 see TriaxialStressController::max_vel (auto-computed) meanStress(=0 ) Mean stress in the packing. (auto-updated) ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. porosity Porosity of the packing. previousMultiplier(=1) (auto-updated) previousStress(=0 ) (auto-updated) radiusControlInterval(=10 ) spheresVolume Total volume pf spheres. stiffnessUpdateInterval(=10 ) target strain rate (./s) strain Current strain in a vector (exx,eyy,ezz). The values reect true (logarithmic) strain. strainDamping(=0.99 ) coecient used for smoother transitions in the strain rate. The rate reaches the target value like dn reaches 0, where d is the damping coecient and n is the number of steps strainRate Current strain rate in a vector d/dt(exx,eyy,ezz). stress((int)id ) Vector3 Return the mean stress vector acting on boundary id, with id between 0 and 5. stressDamping(=0.25 ) wall damping coecient for the stress control - wallDamping=0 implies a (theoretical) perfect control, wallDamping=1 means no movement stressMask(=7 ) Bitmask determining if the components of TriaxialStressController::goal are stress (1) or strain (0). 0 for none, 7 for all, 1 for sigma1, etc. thickness(=-1) thickness of boxes (needed by some functions) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary volumetricStrain(=0 ) Volumetric strain (see TriaxialStressController::strain).|yupdate| wall_back_activated(=true) if true, this wall moves according to the target value (stress or strain rate). 272 Chapter 7. Class reference (yade.wrapper module)
wall_back_id(=4) id of boundary ; coordinate 2- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_bottom_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_bottom_id(=2 ) id of boundary ; coordinate 1- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_front_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_front_id(=5 ) id of boundary ; coordinate 2+ (default value is ok if aabbWalls are appended BEFORE spheres.) wall_left_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_left_id(=0 ) id of boundary ; coordinate 0- (default value is ok if aabbWalls are appended BEFORE spheres.) wall_right_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_right_id(=1) id of boundary ; coordinate 0+ (default value is ok if aabbWalls are appended BEFORE spheres.) wall_top_activated(=true) if true, this wall moves according to the target value (stress or strain rate). wall_top_id(=3 ) id of boundary ; coordinate 1+ (default value is ok if aabbWalls are appended BEFORE spheres.) width(=0 ) size of the box (0-axis) (auto-updated) width0(=0 ) Reference size for strain denition. See TriaxialStressController::width class yade.wrapper.UniaxialStrainer(inherits BoundaryController GlobalEngine Engine Serializable) Axial displacing two groups of bodies in the opposite direction with given strain rate. absSpeed(=NaN ) alternatively, absolute speed of boundary motion can be specied; this is eective only at the beginning and if strainRate is not set; changing absSpeed directly during simulation wil have no eect. [ms] active(=true) Whether this engine is activated asymmetry(=0, symmetric) If 0, straining is symmetric for negIds and posIds; for 1 (or -1), only posIds are strained and negIds dont move (or vice versa) avgStress(=0 ) Current average stress (auto-updated) [Pa] axis(=2 ) The axis which is strained (0,1,2 for x,y,z)
7.3. Global engines
273
blockDisplacements(=false) Whether displacement of boundary bodies perpendicular to the strained axis are blocked of are free blockRotations(=false) Whether rotations of boundary bodies are blocked. crossSectionArea(=NaN ) crossSection perpendicular to he strained axis; must be given explicitly [m] currentStrainRate(=NaN ) Current strain rate (update automatically). (auto-updated) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). idleIterations(=0 ) Number of iterations that will pass without straining activity after stopStrain has been reached initAccelTime(=-200 ) Time for strain reaching the requested value (linear interpolation). If negative, the time is dt*(-initAccelTime), where dt is the timestep at the rst iteration. [s] label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. limitStrain(=0, disabled ) Invert the sense of straining (sharply, without transition) one this value of strain is reached. Not eective if 0. negIds(=uninitalized ) Bodies on which strain will be applied (on the negative end along the axis) notYetReversed(=true) Flag whether the sense of straining has already been reversed (only used internally). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. originalLength(=NaN ) Distance of reference bodies in the direction of axis before straining started (computed automatically) [m] posIds(=uninitalized ) Bodies on which strain will be applied (on the positive end along the axis) setSpeeds(=false) should we set speeds at the beginning directly, instead of increasing strain rate progressively? stopStrain(=NaN ) Strain at which we will pause simulation; inactive (nan) by default; must be reached from below (in absolute value)
274
strain(=0 ) Current strain value, elongation/originalLength (auto-updated) [-] strainRate(=NaN ) Rate of strain, starting at 0, linearly raising to strainRate. [-] stressUpdateInterval(=10 ) How often to recompute stress on supports. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.3.3 Collider
S patialQuickS ortCollider
Pers is tentTriangulationCollider
Collider
FlatGridCollider
Ins ertionS ortCollider
Z ECollider
class yade.wrapper.Collider(inherits GlobalEngine Engine Serializable) Abstract class for nding spatial collisions between bodies. Special constructor Derived colliders (unless they override pyHandleCustomCtorArgs) can be given list of BoundFunctors which is used to initialize the internal boundDispatcher instance. avoidSelfInteractionMask This mask is used to avoid the interactions inside a group of particles. To do so, the particles must have the same mask and this mask have to be compatible with this one. boundDispatcher(=new BoundDispatcher ) BoundDispatcher object that is used for creating bounds on colliders request as necessary. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True).
7.3. Global engines
275
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.FlatGridCollider(inherits Collider GlobalEngine Engine Serializable) Non-optimized grid collider, storing grid as dense at array. Each body is assigned to (possibly multiple) cells, which are arranged in regular grid between aabbMin and aabbMax, with cell size step (same in all directions). Bodies outsize (aabbMin, aabbMax ) are handled gracefully, assigned to closest cells (this will create spurious potential interactions). verletDist determines how much is each body enlarged to avoid collision detection at every step. Note: This collider keeps all cells in linear memory array, therefore will be memory-inecient for sparse simulations. Warning: objects Body::bound are not used, BoundFunctors are not used either: assigning cells to bodies is hard-coded internally. Currently handles Shapes are: Sphere.
Note: Periodic boundary is not handled (yet). aabbMax(=Vector3r::Zero()) Upper corner of grid (approximate, might be rouded up to minStep. aabbMin(=Vector3r::Zero()) Lower corner of grid. avoidSelfInteractionMask This mask is used to avoid the interactions inside a group of particles. To do so, the particles must have the same mask and this mask have to be compatible with this one. boundDispatcher(=new BoundDispatcher ) BoundDispatcher object that is used for creating bounds on colliders request as necessary. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True).
276
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. step(=0 ) Step in the grid (cell size) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary verletDist(=0 ) Length by which enlarge space occupied by each particle; avoids running collision detection at every step. class yade.wrapper.InsertionSortCollider(inherits Collider GlobalEngine Engine Serializable) Collider with O(n log(n)) complexity, using Aabb for bounds. At the initial step, Bodies bounds (along sortAxis) are rst std::sorted along one axis (sortAxis), then collided. The initial sort has O(n2 ) complexity, see Colliders performance for some information (There are scripts in examples/collider-perf for measurements). Insertion sort is used for sorting the bound list that is already pre-sorted from last iteration, where each inversion calls checkOverlap which then handles either overlap (by creating interaction if necessary) or its absence (by deleting interaction if it is only potential). Bodies without bounding volume (such as clumps) are handled gracefully and never collide. Deleted bodies are handled gracefully as well. This collider handles periodic boundary conditions. There are some limitations, notably: 1.No body can have Aabb larger than cells half size in that respective dimension. You get exception if it does and gets in interaction. One way to explicitly by-pass this restriction is oered by allowBiggerThanPeriod, which can be turned on to insert a oor in the form of a very large box for instance (see examples/periodicSandPile.py). 2.No body can travel more than cells distance in one step; this would mean that the simulation is numerically exploding, and it is only detected in some cases. Stride can be used to avoid running collider at every step by enlarging the particles bounds, tracking their displacements and only re-run if they might have gone out of that bounds (see Verlet list for brief description and background) . This requires cooperation from NewtonIntegrator as well as BoundDispatcher, which will be found among engines automatically (exception is thrown if they are not found). If you wish to use strides, set verletDist (length by which bounds will be enlarged in all directions) to some value, e.g. 0.05 typical particle radius. This parameter expresses the tradeo between many potential interactions (running collider rarely, but with longer exact interaction resolution phase) and few potential interactions (running collider more frequently, but with less exact resolutions of interactions); it depends mainly on packing density and particle radius distribution. If targetInterv is >1, not all particles will have their bound enlarged by verletDist; instead, they will have bounds increased by a length in order to trigger a new colliding after targetInterv iteration, assuming they move at almost constant velocity. Ideally in this method, all particles would reach their bounds at the sime iteration. This is of course not the case as soon as velocities
7.3. Global engines
277
uctuate in time. Bound::sweepLength is tuned on the basis of the displacement recorded between the last two runs of the collider. In this situation, verletDist denes the maximum sweep length. allowBiggerThanPeriod If true, tests on bodies sizes will be disabled, and the simulation will run normaly even if bodies larger than period are found. It can be useful when the periodic problem include e.g. a oor modelized with wall/box/facet. Be sure you know what you are doing if you touch this ag. The result is undened if one large body moves out of the (0,0,0) period. avoidSelfInteractionMask This mask is used to avoid the interactions inside a group of particles. To do so, the particles must have the same mask and this mask have to be compatible with this one. boundDispatcher(=new BoundDispatcher ) BoundDispatcher object that is used for creating bounds on colliders request as necessary. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. doSort(=false) Do forced resorting of interactions. dumpBounds() tuple Return representation of the internal sort data. The format is ([...],[...],[...]) for 3 axes, where each ... is a list of entries (bounds). The entry is a tuple with the lowing items: coordinate (oat) body id (int), but negated for negative bounds period numer (int), if the collider is in the periodic regime. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). fastestBodyMaxDist(=-1) Normalized maximum displacement of the fastest body since last run; if >= 1, we could get out of bboxes and will trigger full run. (auto-updated) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. minSweepDistFactor(=0.1) Minimal distance by which enlarge all bounding boxes; superseeds computed value of verletDist when lower that (minSweepDistFactor x verletDist). numReinit(=0 ) Cummulative number of bound array re-initialization. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. periodic Whether the collider is in periodic mode (read-only; for debugging) (auto-updated)
278
sortAxis(=0 ) Axis for the initial contact detection. sortThenCollide(=false) Separate sorting and colliding phase; it is MUCH slower, but all interactions are processed at every step; this eectively makes the collider non-persistent, not remembering last state. (The default behavior relies on the fact that inversions during insertion sort are overlaps of bounding boxes that just started/ceased to exist, and only processes those; this makes the collider much more ecient.) strideActive Whether striding is active (read-only; for debugging). (auto-updated) targetInterv(=50 ) (experimental) Target number of iterations between bound update, used to dene a smaller sweep distance for slower grains if >0, else always use 1*verletDist. Useful in simulations with strong velocity contrasts between slow bodies and fast bodies. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updatingDispFactor(=-1) (experimental) Displacement factor used to trigger bound update: the bound is updated only if updatingDispFactor*disp>sweepDist when >0, else all bounds are updated. useless(=uninitalized ) for compatibility of scripts dening the old colliders attributes - see deprecated attributes verletDist(=-.5, Automatically initialized ) Length by which to enlarge particle bounds, to avoid running collider at every step. Stride disabled if zero. Negative value will trigger automatic computation, so that the real value will be verletDist minimum spherical particle radius; if there are no spherical particles, it will be disabled. The actual length added to one bound can be only a fraction of verletDist when InsertionSortCollider::targetInterv is > 0. class yade.wrapper.PersistentTriangulationCollider(inherits Collider GlobalEngine Engine Serializable) Collision detection engine based on regular triangulation. Handles spheres and at boundaries (considered as innite-sized bounding spheres). avoidSelfInteractionMask This mask is used to avoid the interactions inside a group of particles. To do so, the particles must have the same mask and this mask have to be compatible with this one. boundDispatcher(=new BoundDispatcher ) BoundDispatcher object that is used for creating bounds on colliders request as necessary. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). haveDistantTransient(=false) Keep distant interactions? If True, dont delete interactions once bodies dont overlap anymore; constitutive laws will be responsible for requesting deletion. If False, delete as soon as there is no object penetration. 7.3. Global engines 279
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.SpatialQuickSortCollider(inherits Collider GlobalEngine Engine Serializable) Collider using quicksort along axes at each step, using Aabb bounds. Its performance is lower than that of InsertionSortCollider (see Colliders performance), but the algorithm is simple enought to make it good for checking other colliders correctness. avoidSelfInteractionMask This mask is used to avoid the interactions inside a group of particles. To do so, the particles must have the same mask and this mask have to be compatible with this one. boundDispatcher(=new BoundDispatcher ) BoundDispatcher object that is used for creating bounds on colliders request as necessary. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ZECollider(inherits Collider GlobalEngine Engine Serializable) Collider with O(n log(n)) complexity, using a CGAL algorithm from Zomorodian and Edels-
280
brunner [Kettner2011] (http://www.cgal.org/Manual/beta/doc_html/cgal_manual/Box_intersection_d/Chapter_main.html) avoidSelfInteractionMask This mask is used to avoid the interactions inside a group of particles. To do so, the particles must have the same mask and this mask have to be compatible with this one. boundDispatcher(=new BoundDispatcher ) BoundDispatcher object that is used for creating bounds on colliders request as necessary. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). fastestBodyMaxDist(=-1) Maximum displacement of the fastest body since last run; if >= verletDist, we could get out of bboxes and will trigger full run. DEPRECATED, was only used without bins. (auto-updated) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. numReinit(=0 ) Cummulative number of bound array re-initialization. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. periodic Whether the collider is in periodic mode (read-only; for debugging) (auto-updated) sortAxis(=0 ) Axis for the initial contact detection. sortThenCollide(=false) Separate sorting and colliding phase; it is MUCH slower, but all interactions are processed at every step; this eectively makes the collider non-persistent, not remembering last state. (The default behavior relies on the fact that inversions during insertion sort are overlaps of bounding boxes that just started/ceased to exist, and only processes those; this makes the collider much more ecient.) strideActive Whether striding is active (read-only; for debugging). (auto-updated) targetInterv(=30 ) (experimental) Target number of iterations between bound update, used to dene a smaller sweep distance for slower grains if >0, else always use 1*verletDist. Useful in simulations with strong velocity contrasts between slow bodies and fast bodies. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True.
7.3. Global engines
281
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updatingDispFactor(=-1) (experimental) Displacement factor used to trigger bound update: the bound is updated only if updatingDispFactor*disp>sweepDist when >0, else all bounds are updated. verletDist(=-.15, Automatically initialized ) Length by which to enlarge particle bounds, to avoid running collider at every step. Stride disabled if zero. Negative value will trigger automatic computation, so that the real value will be verletDist minimum spherical particle radius; if there are no spherical particles, it will be disabled.
7.3.4 FieldApplier
CentralGravityEngine
FieldApplier
AxialGravityEngine
GravityEngine
Hdaps GravityEngine
class yade.wrapper.FieldApplier(inherits GlobalEngine Engine Serializable) Base for engines applying force les on particles. Not to be used directly. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.AxialGravityEngine(inherits FieldApplier GlobalEngine Engine Serializable) Apply acceleration (independent of distance) directed towards an axis. 282 Chapter 7. Class reference (yade.wrapper module)
acceleration(=0 ) Acceleration magnitude [kgms] axisDirection(=Vector3r::UnitX()) direction of the gravity axis (will be normalized automatically) axisPoint(=Vector3r::Zero()) Point through which the axis is passing. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=0 ) If mask dened, only bodies with corresponding groupMask will be aected by this engine. If 0, all bodies will be aected. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CentralGravityEngine(inherits FieldApplier GlobalEngine Engine Serializable) Engine applying acceleration to all bodies, towards a central body. accel(=0 ) Acceleration magnitude [kgms] centralBody(=Body::ID_NONE ) The body towards which all other bodies are attracted. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True).
7.3. Global engines
283
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=0 ) If mask dened, only bodies with corresponding groupMask will be aected by this engine. If 0, all bodies will be aected. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. reciprocal(=false) If true, acceleration will be applied on the central body as well. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GravityEngine(inherits FieldApplier GlobalEngine Engine Serializable) Engine applying constant acceleration to all bodies. DEPRECATED, use Newton::gravity unless you need energy tracking or selective gravity application using groupMask). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). gravity(=Vector3r::Zero()) Acceleration [kgms] label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=0 ) If mask dened, only bodies with corresponding groupMask will be aected by this engine. If 0, all bodies will be aected. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True.
284
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary warnOnce(=true) For deprecation warning once. class yade.wrapper.HdapsGravityEngine(inherits GravityEngine FieldApplier GlobalEngine Engine Serializable) Read accelerometer in Thinkpad laptops (HDAPS and accordingly set gravity within the simulation. This code draws from hdaps-gl . See scripts/test/hdaps.py for an example. accel(=Vector2i::Zero()) reading from the sysfs le calibrate(=Vector2i::Zero()) Zero position; if NaN, will be read from the hdapsDir / calibrate. calibrated(=false) Whether calibrate was already updated. Do not set to True by hand unless you also give a meaningful value for calibrate. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). gravity(=Vector3r::Zero()) Acceleration [kgms] hdapsDir(=/sys/devices/platform/hdaps ) Hdaps directory; contains position (with accelerometer readings) and calibration (zero acceleration). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mask(=0 ) If mask dened, only bodies with corresponding groupMask will be aected by this engine. If 0, all bodies will be aected. msecUpdate(=50 ) How often to update the reading. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateThreshold(=4) Minimum dierence of reading from the le before updating gravity, to avoid jitter. 7.3. Global engines 285
warnOnce(=true) For deprecation warning once. zeroGravity(=Vector3r(0, 0, -1)) Gravity if the accelerometer is in at (zero) position.
7.4 Partial engines

CombinedKinematicEngine
DragEngine
RadialForceEngine PeriodicFlowEngine FlowEngine Trans lationEngine KinematicEngine PartialEngine LawTes ter HarmonicMotionEngine LinearDragEngine HelixEngine InterpolatingHelixEngine RotationEngine HarmonicRotationEngine S ervoPIDController
ForceEngine
InterpolatingDirectedForceEngine
S tepDis placer
TorqueEngine
class yade.wrapper.PartialEngine(inherits Engine Serializable) Engine aecting only particular bodies in the simulation, dened by ids. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True.
286
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CombinedKinematicEngine(inherits PartialEngine Engine Serializable) Engine for applying combined displacements on pre-dened bodies. Constructed using + operator on regular KinematicEngines. The ids operated on are those of the rst engine in the combination (assigned automatically). comb(=uninitalized ) Kinematic engines that will be combined by this one, run in the order given. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.DragEngine(inherits PartialEngine Engine Serializable) Apply drag force on some particles at each step, decelerating them proportionally to their linear velocities. The applied force reads Fd = v 1 |v|2 Cd A |v| 2
where is the medium density (density), v is particles velocity, A is particle projected area (disc), Cd is the drag coecient (0.47 for Sphere), Note: Drag force is only applied to spherical particles, listed in ids. Cd(=0.47 ) Drag coecient <http://en.wikipedia.org/wiki/Drag_coecient>_. Rho(=1.225 ) Density of the medium (uid or air), by default - the density of the air.
7.4. Partial engines
287
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.FlowEngine(inherits PartialEngine Engine Serializable) An engine to solve ow problem in saturated granular media. Model description can be found in [Chareyre2012a] and [Catalano2013a]. See the example script FluidCouplingPFV/oedometer.py. More documentation to come. Note: Multi-threading seems to work ne for Cholesky decomposition, but it fails for the solve phase in which -j1 is the fastest, here we specify thread numbers independently using FlowEngine::numFactorizeThreads and FlowEngine::numSolveThreads. These multhreading settings are only impacting the behaviour of openblas library and are relatively independant of FlowEngine::multithread. However, the settings have to be globally consistent. For instance, :yref:FlowEngine::multithread=True with multithread :yref:FlowEngine::numSolveThreads=:yref:FlowEngine::numFactorizeThreads=4 means that openblas will mobilize 8 processors at some point, if the system doesnt have so many procs. it will hurt performance. Debug(=false) Activate debug messages P_zero(=0 ) The value used for initializing pore pressure. It is useless for incompressible uid, but important for compressible model. Relax(=1.9 ) Gauss-Seidel relaxation Tolerance(=1e-06 ) Gauss-Seidel Tolerance WaveAction(=false) Allow sinusoidal pressure condition to simulate ocean waves
288
avFlVelOnSph((int)Id_sph) object Compute a sphere-centered average uid velocity averagePressure() oat Measure averaged pore pressure in the entire volume averageSlicePressure((oat)posY ) oat Measure slice-averaged pore pressure at height posY bndCondIsPressure(=vector<bool>(6, false)) denes the type of boundary condition for each side. True if pressure is imposed, False for no-ux. Indexes can be retrieved with FlowEngine::xmin and friends. bndCondValue(=vector<double>(6, 0)) Imposed value of a boundary condition. Only applies if the boundary condition is imposed pressure, else the imposed ux is always zero presently (may be generalized to non-zero imposed uxes in the future). bodyNormalLubStress((int)Id_sph) Matrix3 Return the normal lubrication stress on sphere Id_sph. bodyShearLubStress((int)Id_sph) Matrix3 Return the shear lubrication stress on sphere Id_sph. boundaryUseMaxMin(=vector<bool>(6, true)) If true (default value) bounding sphere is added as function of max/min sphere coord, if false as function of yade wall position boundaryVelocity(=vector<Vector3r>(6, Vector3r::Zero())) velocity on top boundary, only change it using FlowEngine::setBoundaryVel clampKValues(=true) If true, clamp local permeabilities in [minKdivKmean,maxKdivKmean]*globalK. This clamping can avoid singular values in the permeability matrix and may reduce numerical errors in the solve phase. It will also hide junk values if they exist, or bias all values in very heterogeneous problems. So, use this with care. clearImposedFlux() None Clear the list of points with ux imposed. clearImposedPressure() None Clear the list of points with pressure imposed. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. defTolerance(=0.05 ) Cumulated deformation threshold for which retriangulation of pore space is performed. If negative, the triangulation update will occure with a xed frequency on the basis of FlowEngine::meshUpdateInterval dict() dict Return dictionary of attributes. dt(=0 ) timestep [s] emulateAction() None get scene and run action (may be used to manipulate an engine outside the timestepping loop). eps(=0.00001) roughness dened as a fraction of particles size, giving the minimum distance between particles in the lubrication model. eps_vol_max(=0 ) Maximal absolute volumetric strain computed at each iteration. (auto-updated)
289
execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). first(=true) Controls the initialization/update phases fluidBulkModulus(=0.) Bulk modulus of uid (inverse of compressibility) K=-dP*V/dV [Pa]. Flow is compressible if uidBulkModulus > 0, else incompressible. fluidForce((int)Id_sph) Vector3 Return the uid force on sphere Id_sph. getBoundaryFlux((int)boundary) oat Get total ux through boundary dened by its body id. The ux may be not zero even for no-ow condition. This artifact comes from cells which are incident to two or more boundaries (along the edges of the sample, typically). Such ux evaluation on impermeable boundary is just irrelevant, it does not imply that the boundary condition is not applied properly. getCell((Vector3)pos) int get id of the cell containing (X,Y,Z). getCellFlux((int)cond ) oat Get inux in cell associated to an imposed P (indexed using cond). getConstrictions([(bool)all=True ]) list Get the list of constriction radii (inscribed circle) for all nite facets (if all==True) or all facets not incident to a virtual bounding sphere (if all==False). When all facets are returned, negative radii denote facet incident to one or more ctious spheres. getConstrictionsFull([(bool)all=True ]) list Get the list of constrictions (inscribed circle) for all nite facets (if all==True), or all facets not incident to a ctious bounding sphere (if all==False). When all facets are returned, negative radii denote facet incident to one or more ctious spheres. The constrictions are returned in the format {{cell1,cell2}{rad,nx,ny,nz}} getPorePressure((Vector3)pos) oat Measure pore pressure in position pos[0],pos[1],pos[2] ids(=uninitalized ) Ids of bodies aected by this PartialEngine. ignoredBody(=-1) Id of a sphere to exclude from the triangulation.) imposeFlux((Vector3)pos, (oat)p) None Impose incoming ux in boundary cell of location pos. imposePressure((Vector3)pos, (oat)p) int Impose pressure in cell of location pos. The index of the condition is returned (for multiple imposed pressures at dierent points). isActivated(=true) Activates Flow Engine label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxKdivKmean(=100 ) dene the max K value (see FlowEngine::clampKValues) meanKStat(=false) report the local permeabilities correction 290 Chapter 7. Class reference (yade.wrapper module)
meshUpdateInterval(=1000 ) Maximum number of timesteps FlowEngine::defTolerance.
between
re-triangulation
events.
See
also
minKdivKmean(=0.0001) dene the min K value (see FlowEngine::clampKValues) multithread(=false) Build triangulation and factorize in the background (multi-thread mode) normalLubForce((int)Id_sph) Vector3 Return the normal lubrication force on sphere Id_sph. normalLubrication(=false) Compute normal lubrication force as developped by Brule ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. permeabilityFactor(=0.0 ) permability multiplier permeability_map(=false) Enable/disable stocking of average permeability scalar in cell infos. porosity(=0 ) Porosity computed at each retriangulation (auto-updated) pressureForce(=true) Compute the pressure eld and associated uid forces. WARNING: turning o means uid ow is not computed at all. pressureProfile((oat)wallUpY, (oat)wallDownY ) None Measure pore pressure in 6 equally-spaced points along the height of the sample saveVtk() None Save pressure eld in vtk format. setImposedPressure((int)cond, (oat)p) None Set pressure value at the point indexed cond. shearLubForce((int)Id_sph) Vector3 Return the shear lubrication force on sphere Id_sph. shearLubTorque((int)Id_sph) Vector3 Return the shear lubrication torque on sphere Id_sph. shearLubrication(=false) Compute shear lubrication force as developped by Brule (FIXME: ref.) sineAverage(=0 ) Pressure value (average) when sinusoidal pressure is applied sineMagnitude(=0 ) Pressure value (amplitude) when sinusoidal pressure is applied (p ) slip_boundary(=true) Controls friction condition on lateral walls stiffness(=10000 ) equivalent contact stiness used in the lubrication model timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. 7.4. Partial engines 291
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateBCs() None tells the engine to update its boundary conditions before running (especially useful when changing boundary pressure - should not be needed for point-wise imposed pressure) updateTriangulation(=0 ) If true the medium is retriangulated. Can be switched on to force retriangulation after some events (else it will be true periodicaly based on FlowEngine::defTolerance and FlowEngine::meshUpdateInterval. Of course, it costs CPU time. useSolver(=0 ) Solver to use 0=G-Seidel, 1=Taucs, 2-Pardiso, 3-CHOLMOD viscosity(=1.0 ) viscosity of the uid viscousNormalBodyStress(=false) Compute normal viscous stress applied on each body viscousShear(=false) Compute viscous shear terms as developped by Donia Marzougui (FIXME: ref.) viscousShearBodyStress(=false) Compute shear viscous stress applied on each body wallIds(=vector<int>(6)) body ids of the boundaries (default values are ok only if aabbWalls are appended before spheres, i.e. numbered 0,...,5) wall_thickness(=0.001) Walls thickness xmax(=1) See FlowEngine::xmin. xmin(=0 ) Index of the boundary xmin . This index is not equal the the id of the corresponding body in general, it may be used to access the corresponding attributes (e.g. ow.bndCondValue[ow.xmin], ow.wallId[ow.xmin],...). ymax(=3 ) See FlowEngine::xmin. ymin(=2 ) See FlowEngine::xmin. zmax(=5 ) See FlowEngine::xmin. zmin(=4) See FlowEngine::xmin. class yade.wrapper.ForceEngine(inherits PartialEngine Engine Serializable) Apply contact force on some particles at each step. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True).
292
force(=Vector3r::Zero()) Force to apply. ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.HarmonicMotionEngine(inherits KinematicEngine PartialEngine Engine Serializable) This engine implements the harmonic oscillation of bodies. http://en.wikipedia.org/wiki/Simple_harmonic_motion#Dynamics_of_simple_harmonic_motion A(=Vector3r::Zero()) Amplitude [m] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f(=Vector3r::Zero()) Frequency [hertz] fi(=Vector3r(Mathr::PI/2.0, Mathr::PI/2.0, Mathr::PI/2.0)) Initial phase [radians]. By default, the body oscillates around initial position. ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level.
293
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.HarmonicRotationEngine(inherits RotationEngine KinematicEngine PartialEngine Engine Serializable) This engine implements the harmonic-rotation oscillation of bodies. http://en.wikipedia.org/wiki/Simple_harmonic_motion#Dynamics_of_simple_harmonic_motion ; please, set dynamic=False for bodies, droven by this engine, otherwise amplitude will be 2x more, than awaited. A(=0 ) Amplitude [rad] angularVelocity(=0 ) Angular velocity. [rad/s] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). f(=0 ) Frequency [hertz] fi(=Mathr::PI/2.0 ) Initial phase [radians]. By default, the body oscillates around initial position. ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rotateAroundZero(=false) If True, bodies will not rotate around their centroids, but rather around zeroPoint. rotationAxis(=Vector3r::UnitX()) Axis of rotation (direction); will be normalized automatically. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary zeroPoint(=Vector3r::Zero()) Point around which bodies will rotate if rotateAroundZero is True 294 Chapter 7. Class reference (yade.wrapper module)
class yade.wrapper.HelixEngine(inherits RotationEngine KinematicEngine PartialEngine Engine Serializable) Engine applying both rotation and translation, along the same axis, whence the name HelixEngine angleTurned(=0 ) How much have we turned so far. (auto-updated) [rad] angularVelocity(=0 ) Angular velocity. [rad/s] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. linearVelocity(=0 ) Linear velocity [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rotateAroundZero(=false) If True, bodies will not rotate around their centroids, but rather around zeroPoint. rotationAxis(=Vector3r::UnitX()) Axis of rotation (direction); will be normalized automatically. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary zeroPoint(=Vector3r::Zero()) Point around which bodies will rotate if rotateAroundZero is True class yade.wrapper.InterpolatingDirectedForceEngine(inherits ForceEngine PartialEngine Engine Serializable) Engine for applying force of varying magnitude but constant direction on subscribed bodies. times and magnitudes must have the same length, direction (normalized automatically) gives the orientation. As usual with interpolating engines: the rst magnitude is used before the rst time point, last magnitude is used after the last time point. Wrap species whether time wraps around the last time point to the rst time point.
295
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. direction(=Vector3r::UnitX()) Contact force direction (normalized automatically) execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). force(=Vector3r::Zero()) Force to apply. ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. magnitudes(=uninitalized ) Force magnitudes readings [N] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. times(=uninitalized ) Time readings [s] timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wrap(=false) wrap to the beginning of the sequence if beyond the last time point class yade.wrapper.InterpolatingHelixEngine(inherits HelixEngine RotationEngine KinematicEngine PartialEngine Engine Serializable) Engine applying spiral motion, nding current angular velocity by linearly interpolating in times and velocities and translation by using slope parameter. The interpolation assumes the margin value before the rst time point and last value after the last time point. If wrap is specied, time will wrap around the last times value to the rst one (note that no interpolation between last and rst values is done). angleTurned(=0 ) How much have we turned so far. (auto-updated) [rad] angularVelocities(=uninitalized ) List of angular velocities; manadatorily of same length as times. [rad/s] angularVelocity(=0 ) Angular velocity. [rad/s]
296
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. linearVelocity(=0 ) Linear velocity [m/s] ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rotateAroundZero(=false) If True, bodies will not rotate around their centroids, but rather around zeroPoint. rotationAxis(=Vector3r::UnitX()) Axis of rotation (direction); will be normalized automatically. slope(=0 ) Axial translation per radian turn (can be negative) [m/rad] times(=uninitalized ) List of time points at which velocities are given; must be increasing [s] timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wrap(=false) Wrap t if t>times_n, i.e. t_wrapped=t-N*(times_n-times_0) zeroPoint(=Vector3r::Zero()) Point around which bodies will rotate if rotateAroundZero is True class yade.wrapper.KinematicEngine(inherits PartialEngine Engine Serializable) Abstract engine for applying prescribed displacement. Note: Derived classes should override the apply with given list of ids (not action with PartialEngine.ids), so that they work when combined together; velocity and angular velocity of all subscribed bodies is reset before the apply method is called, it should therefore only increment those quantities. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. 7.4. Partial engines 297
dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.LawTester(inherits PartialEngine Engine Serializable) Prescribe and apply deformations of an interaction in terms of local mutual displacements and rotations. The loading path is specied either using path (as sequence of 6-vectors containing generalized displacements ux , uy , uz , x , y , z ) or disPath (ux , uy , uz ) and rotPath (x , y , z ). Time function with time values (step numbers) corresponding to points on loading path is given by pathSteps. Loading values are linearly interpolated between given loading path points, and starting zero-value (the initial conguration) is assumed for both path and pathSteps. hooks can specify python code to run when respective point on the path is reached; when the path is nished, doneHook will be run. LawTester should be placed between InteractionLoop and NewtonIntegrator in the simulation loop, since it controls motion via setting linear/angular velocities on particles; those velocities are integrated by NewtonIntegrator to yield an actual position change, which in turn causes IGeom to be updated (and contact law applied) when InteractionLoop is executed. Constitutive law generating forces on particles will not aect prescribed particle motion, since both particles have all DoFs blocked when rst used with LawTester. LawTester uses, as much as possible, IGeom to provide useful data (such as local coordinate system), but is able to compute those independently if absent in the respective IGeom: IGeom L3Geom L6Geom ScGeom ScGeom6D Dem3DofGeom #DoFs 3 6 3 6 3 LawTester support level full full emulate local coordinate system emulate local coordinate system not supported
Depending on IGeom, 3 (ux , uy , uz ) or 6 (ux , uy , uz , x , y , z ) degrees of freedom (DoFs) are controlled with LawTester, by prescribing linear and angular velocities of both particles in contact. All DoFs controlled with LawTester are orthogonal (fully decoupled) and are controlled independently. When 3 DoFs are controlled, rotWeight controls whether local shear is applied by moving particle on arc around the other one, or by rotating without changing position; although such rotation
298
induces mutual rotation on the interaction, it is ignored with IGeom with only 3 DoFs. When 6 DoFs are controlled, only arc-displacement is applied for shear, since otherwise mutual rotation would occur. idWeight distributes prescribed motion between both particles (resulting local deformation is the same if id1 is moved towards id2 or id2 towards id1). This is true only for ux , uy , uz , x however ; bending rotations y , z are nevertheless always distributed regardless of idWeight to both spheres in inverse proportion to their radii, so that there is no shear induced. LawTester knows current contact deformation from 2 sources: from its own internal data (which are used for prescribing the displacement at every step), which can be accessed in uTest, and from IGeom itself (depending on which data it provides), which is stored in uGeom. These two values should be identical (disregarding numerical percision), and it is a way to test whether IGeom and related functors compute what they are supposed to compute. LawTester-operated interactions can be rendered with GlExtra_LawTester renderer. See scripts/test/law-test.py for an example. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. disPath(=uninitalized ) Loading path, where each Vector3 contains desired normal displacement and two components of the shear displacement (in local coordinate system, which is being tracked automatically. If shorter than rotPath, the last value is repeated. displIsRel(=true) Whether displacement values in disPath are normalized by reference contact length (r1+r2 for 2 spheres). doneHook(=uninitalized ) Python command (as string) to run when end of the path is achieved. If empty, the engine will be set dead. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). hooks(=uninitalized ) Python commands to be run when the corresponding point in path is reached, before doing other things in that particular step. See also doneHook. idWeight(=1) Float, usually 0,1, determining on how are displacements distributed between particles (0 for id1, 1 for id2); intermediate values will apply respective part to each of them. This parameter is ignored with 6-DoFs IGeom. ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes
299
openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. pathSteps(=vector<int>(1, 1), (constant step)) Step number for corresponding values in path; if shorter than path, distance between last 2 values is used for the rest. refLength(=0 ) Reference contact length, for rendering only. renderLength(=0 ) Characteristic length for the purposes of rendering, set equal to the smaller radius. rotPath(=uninitalized ) Rotational components of the loading path, where each item contains torsion and two bending rotations in local coordinates. If shorter than path, the last value is repeated. rotWeight(=1) Float 0,1 determining whether shear displacement is applied as rotation or displacement on arc (0 is displacement-only, 1 is rotation-only). Not eective when mutual rotation is specied. step(=1) Step number in which this engine is active; determines position in path, using pathSteps. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. trsf(=uninitalized ) Transformation matrix for the local coordinate system. (auto-updated) uGeom(=Vector6r::Zero()) Current generalized displacements (3 displacements, 3 rotations), as stored in the interation itself. They should corredpond to uTest, otherwise a bug is indicated. uTest(=Vector6r::Zero()) Current generalized displacements (3 displacements, 3 rotations), as they should be according to this LawTester. Should correspond to uGeom. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary uuPrev(=Vector6r::Zero()) Generalized displacement values reached in the previous step, for knowing which increment to apply in the current step. class yade.wrapper.LinearDragEngine(inherits PartialEngine Engine Serializable) Apply viscous resistance or linear drag on some particles at each step, decelerating them proportionally to their linear velocities. The applied force reads Fd = bv where b is the linear drag, v is particles velocity. b = 6r where is the medium viscosity, r is the Stokes radius of the particle (but in this case we accept it equal to sphere radius for simplication), Note: linear drag is only applied to spherical particles, listed in ids. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. 300 Chapter 7. Class reference (yade.wrapper module)
dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. nu(=0.001) Viscosity of the medium. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.PeriodicFlowEngine(inherits FlowEngine PartialEngine Engine Serializable) A variant of FlowEngine implementing periodic boundary conditions. The API is very similar. Debug(=false) Activate debug messages P_zero(=0 ) The value used for initializing pore pressure. It is useless for incompressible uid, but important for compressible model. Relax(=1.9 ) Gauss-Seidel relaxation Tolerance(=1e-06 ) Gauss-Seidel Tolerance WaveAction(=false) Allow sinusoidal pressure condition to simulate ocean waves avFlVelOnSph((int)Id_sph) object Compute a sphere-centered average uid velocity averagePressure() oat Measure averaged pore pressure in the entire volume averageSlicePressure((oat)posY ) oat Measure slice-averaged pore pressure at height posY bndCondIsPressure(=vector<bool>(6, false)) denes the type of boundary condition for each side. True if pressure is imposed, False for no-ux. Indexes can be retrieved with FlowEngine::xmin and friends. bndCondValue(=vector<double>(6, 0)) Imposed value of a boundary condition. Only applies if the boundary condition is imposed
301
pressure, else the imposed ux is always zero presently (may be generalized to non-zero imposed uxes in the future). bodyNormalLubStress((int)Id_sph) Matrix3 Return the normal lubrication stress on sphere Id_sph. bodyShearLubStress((int)Id_sph) Matrix3 Return the shear lubrication stress on sphere Id_sph. boundaryUseMaxMin(=vector<bool>(6, true)) If true (default value) bounding sphere is added as function of max/min sphere coord, if false as function of yade wall position boundaryVelocity(=vector<Vector3r>(6, Vector3r::Zero())) velocity on top boundary, only change it using FlowEngine::setBoundaryVel clampKValues(=true) If true, clamp local permeabilities in [minKdivKmean,maxKdivKmean]*globalK. This clamping can avoid singular values in the permeability matrix and may reduce numerical errors in the solve phase. It will also hide junk values if they exist, or bias all values in very heterogeneous problems. So, use this with care. clearImposedFlux() None Clear the list of points with ux imposed. clearImposedPressure() None Clear the list of points with pressure imposed. compTessVolumes() None Like TesselationWrapper::computeVolumes() dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. defTolerance(=0.05 ) Cumulated deformation threshold for which retriangulation of pore space is performed. If negative, the triangulation update will occure with a xed frequency on the basis of FlowEngine::meshUpdateInterval dict() dict Return dictionary of attributes. dt(=0 ) timestep [s] duplicateThreshold(=0.06 ) distance from cell borders that will triger periodic duplication in the triangulation (autoupdated) emulateAction() None get scene and run action (may be used to manipulate an engine outside the timestepping loop). eps(=0.00001) roughness dened as a fraction of particles size, giving the minimum distance between particles in the lubrication model. eps_vol_max(=0 ) Maximal absolute volumetric strain computed at each iteration. (auto-updated) execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). first(=true) Controls the initialization/update phases
302
fluidBulkModulus(=0.) Bulk modulus of uid (inverse of compressibility) K=-dP*V/dV [Pa]. Flow is compressible if uidBulkModulus > 0, else incompressible. fluidForce((int)Id_sph) Vector3 Return the uid force on sphere Id_sph. getBoundaryFlux((int)boundary) oat Get total ux through boundary dened by its body id. getCell((Vector3)pos) int get id of the cell containing pos. getCellFlux((int)cond ) oat Get inux in cell associated to an imposed P (indexed using cond). getConstrictions([(bool)all=True ]) list Get the list of constriction radii (inscribed circle) for all nite facets (if all==True) or all facets not incident to a virtual bounding sphere (if all==False). When all facets are returned, negative radii denote facet incident to one or more ctious spheres. getConstrictionsFull([(bool)all=True ]) list Get the list of constrictions (inscribed circle) for all nite facets (if all==True), or all facets not incident to a ctious bounding sphere (if all==False). When all facets are returned, negative radii denote facet incident to one or more ctious spheres. The constrictions are returned in the format {{cell1,cell2}{rad,nx,ny,nz}} getPorePressure((Vector3)pos) oat Measure pore pressure in position pos[0],pos[1],pos[2] gradP(=Vector3r::Zero()) Macroscopic pressure gradient ids(=uninitalized ) Ids of bodies aected by this PartialEngine. ignoredBody(=-1) Id of a sphere to exclude from the triangulation.) imposeFlux((Vector3)pos, (oat)p) None Impose incoming ux in boundary cell of location pos. imposePressure((Vector3)pos, (oat)p) int Impose pressure in cell of location pos. The index of the condition is returned (for multiple imposed pressures at dierent points). isActivated(=true) Activates Flow Engine label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxKdivKmean(=100 ) dene the max K value (see FlowEngine::clampKValues) meanKStat(=false) report the local permeabilities correction meanVelocity() Vector3 measure the mean velocity in the period meshUpdateInterval(=1000 ) Maximum number of timesteps FlowEngine::defTolerance. between re-triangulation events. See also
minKdivKmean(=0.0001) dene the min K value (see FlowEngine::clampKValues)
303
multithread(=false) Build triangulation and factorize in the background (multi-thread mode) normalLubForce((int)Id_sph) Vector3 Return the normal lubrication force on sphere Id_sph. normalLubrication(=false) Compute normal lubrication force as developped by Brule ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. permeabilityFactor(=0.0 ) permability multiplier permeability_map(=false) Enable/disable stocking of average permeability scalar in cell infos. porosity(=0 ) Porosity computed at each retriangulation (auto-updated) pressureForce(=true) Compute the pressure eld and associated uid forces. WARNING: turning o means uid ow is not computed at all. pressureProfile((oat)wallUpY, (oat)wallDownY ) None Measure pore pressure in 6 equally-spaced points along the height of the sample saveVtk() None Save pressure eld in vtk format. setImposedPressure((int)cond, (oat)p) None Set pressure value at the point indexed cond. shearLubForce((int)Id_sph) Vector3 Return the shear lubrication force on sphere Id_sph. shearLubTorque((int)Id_sph) Vector3 Return the shear lubrication torque on sphere Id_sph. shearLubrication(=false) Compute shear lubrication force as developped by Brule (FIXME: ref.) sineAverage(=0 ) Pressure value (average) when sinusoidal pressure is applied sineMagnitude(=0 ) Pressure value (amplitude) when sinusoidal pressure is applied (p ) slip_boundary(=true) Controls friction condition on lateral walls stiffness(=10000 ) equivalent contact stiness used in the lubrication model timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateBCs() None tells the engine to update its boundary conditions before running (especially useful when changing boundary pressure - should not be needed for point-wise imposed pressure) 304 Chapter 7. Class reference (yade.wrapper module)
updateTriangulation(=0 ) If true the medium is retriangulated. Can be switched on to force retriangulation after some events (else it will be true periodicaly based on FlowEngine::defTolerance and FlowEngine::meshUpdateInterval. Of course, it costs CPU time. useSolver(=0 ) Solver to use 0=G-Seidel, 1=Taucs, 2-Pardiso, 3-CHOLMOD viscosity(=1.0 ) viscosity of the uid viscousNormalBodyStress(=false) Compute normal viscous stress applied on each body viscousShear(=false) Compute viscous shear terms as developped by Donia Marzougui (FIXME: ref.) viscousShearBodyStress(=false) Compute shear viscous stress applied on each body volume([(int)id=0 ]) oat Returns the volume of Voronois cell of a sphere. wallIds(=vector<int>(6)) body ids of the boundaries (default values are ok only if aabbWalls are appended before spheres, i.e. numbered 0,...,5) wall_thickness(=0.001) Walls thickness xmax(=1) See FlowEngine::xmin. xmin(=0 ) Index of the boundary xmin . This index is not equal the the id of the corresponding body in general, it may be used to access the corresponding attributes (e.g. ow.bndCondValue[ow.xmin], ow.wallId[ow.xmin],...). ymax(=3 ) See FlowEngine::xmin. ymin(=2 ) See FlowEngine::xmin. zmax(=5 ) See FlowEngine::xmin. zmin(=4) See FlowEngine::xmin. class yade.wrapper.RadialForceEngine(inherits PartialEngine Engine Serializable) Apply force of given magnitude directed away from spatial axis. axisDir(=Vector3r::UnitX()) Axis direction (normalized automatically) axisPt(=Vector3r::Zero()) Point on axis dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True).
305
execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). fNorm(=0 ) Applied force magnitude ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.RotationEngine(inherits KinematicEngine PartialEngine Engine Serializable) Engine applying rotation (by setting angular velocity) to subscribed bodies. If rotateAroundZero is set, then each body is also displaced around zeroPoint. angularVelocity(=0 ) Angular velocity. [rad/s] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rotateAroundZero(=false) If True, bodies will not rotate around their centroids, but rather around zeroPoint.
306
rotationAxis(=Vector3r::UnitX()) Axis of rotation (direction); will be normalized automatically. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary zeroPoint(=Vector3r::Zero()) Point around which bodies will rotate if rotateAroundZero is True class yade.wrapper.ServoPIDController(inherits TranslationEngine KinematicEngine PartialEngine Engine Serializable) PIDController servo-engine for applying prescribed force on bodies. http://en.wikipedia.org/wiki/PID_controller axis(=Vector3r::Zero()) Unit vector along which apply the velocity [-] curVel(=0.0 ) Current applied velocity [m/s] dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. errorCur(=0.0 ) Current error [N] errorPrev(=0.0 ) Previous error [N] execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). iTerm(=0.0 ) Integral term [N] ids(=uninitalized ) Ids of bodies aected by this PartialEngine. iterPeriod(=100.0 ) Periodicity criterion of velocity correlation [-] iterPrevStart(=-1.0 ) Previous iteration of velocity correlation [-] kD(=0.0 ) Derivative gain/coecient for the PID-controller [-] kI(=0.0 ) Integral gain/coecient for the PID-controller [-] kP(=0.0 ) Proportional gain/coecient for the PID-controller [-] label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. maxVelocity(=0.0 ) Velocity [m/s]
307
ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. target(=0.0 ) Target value for the controller [N] timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. translationAxis(=uninitalized ) Direction [Vector3] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary velocity(=uninitalized ) Velocity [m/s] class yade.wrapper.StepDisplacer(inherits PartialEngine Engine Serializable) Apply generalized displacement (displacement or rotation) stepwise on subscribed bodies. Could be used for purposes of contact law tests (by moving one sphere compared to another), but in this case, see rather LawTester dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. mov(=Vector3r::Zero()) Linear displacement step to be applied per iteration, by addition to State.pos. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. rot(=Quaternionr::Identity()) Rotation step to be applied per iteration (via rotation composition with State.ori). setVelocities(=false) If false, positions and orientations are directly updated, without changing the speeds of concerned bodies. If true, only velocity and angularVelocity are modied. In this second case
308
integrator is supposed to be used, so that, thanks to this Engine, the bodies will have the prescribed jump over one iteration (dt). timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.TorqueEngine(inherits PartialEngine Engine Serializable) Apply given torque (momentum) value at every subscribed particle, at every step. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. moment(=Vector3r::Zero()) Torque value to be applied. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.TranslationEngine(inherits KinematicEngine PartialEngine Engine Serializable) This engine is the base class for dierent engines, which require any kind of motion. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True).
309
ids(=uninitalized ) Ids of bodies aected by this PartialEngine. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. translationAxis(=uninitalized ) Direction [Vector3] updateAttrs((dict)arg2 ) None Update object attributes from given dictionary velocity(=uninitalized ) Velocity [m/s]
7.5 Bounding volume creation

7.5.1 BoundFunctor
Bo1_ChainedCylinder_Aabb
Bo1_S phere_Aabb
Bo1_Box_Aabb
Bo1_Facet_Aabb BoundFunctor Bo1_GridConnection_Aabb
Bo1_Tetra_Aabb
Bo1_W all_Aabb
Bo1_Cylinder_Aabb
310
class yade.wrapper.BoundFunctor(inherits Functor Serializable) Functor for creating/updating Body::bound. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_Box_Aabb(inherits BoundFunctor Functor Serializable) Create/update an Aabb of a Box. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_ChainedCylinder_Aabb(inherits BoundFunctor Functor Serializable) Functor creating Aabb from ChainedCylinder. aabbEnlargeFactor Relative enlargement of the bounding box; deactivated if negative. Note: This attribute is used to create distant interaction, but is only meaningful with an IGeomFunctor which will not simply discard such interactions: Ig2_Cylinder_Cylinder_ScGeom::interactionDetectionFactor should have the same value as aabbEnlargeFactor. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.5. Bounding volume creation
311
class yade.wrapper.Bo1_Cylinder_Aabb(inherits BoundFunctor Functor Serializable) Functor creating Aabb from Cylinder. aabbEnlargeFactor Relative enlargement of the bounding box; deactivated if negative. Note: This attribute is used to create distant interaction, but is only meaningful with an IGeomFunctor which will not simply discard such interactions: Ig2_Cylinder_Cylinder_ScGeom::interactionDetectionFactor should have the same value as aabbEnlargeFactor. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_Facet_Aabb(inherits BoundFunctor Functor Serializable) Creates/updates an Aabb of a Facet. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_GridConnection_Aabb(inherits BoundFunctor Functor Serializable) Functor creating Aabb from a GridConnection. aabbEnlargeFactor(=-1, deactivated ) Relative enlargement of the bounding box; deactivated if negative. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True.
312
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_Sphere_Aabb(inherits BoundFunctor Functor Serializable) Functor creating Aabb from Sphere. aabbEnlargeFactor Relative enlargement of the bounding box; deactivated if negative. Note: This attribute is used to create distant interaction, but is only meaningful with an IGeomFunctor which will not simply discard such interactions: Ig2_Sphere_Sphere_ScGeom::interactionDetectionFactor should have the same value as aabbEnlargeFactor. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_Tetra_Aabb(inherits BoundFunctor Functor Serializable) Create/update Aabb of a Tetra bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Bo1_Wall_Aabb(inherits BoundFunctor Functor Serializable) Creates/updates an Aabb of a Wall bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True.
7.5. Bounding volume creation
313
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.5.2 BoundDispatcher
class yade.wrapper.BoundDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). activated(=true) Whether the engine is activated (only should be changed by the collider) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((Shape)arg2 ) BoundFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. minSweepDistFactor(=0.2 ) Minimal distance by which enlarge all bounding boxes; superseeds computed value of sweepDist when lower that (minSweepDistFactor x sweepDist). Updated by the collider. (autoupdated). ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. sweepDist(=0 ) Distance by which enlarge all bounding boxes, to prevent collider from being run at every step (only should be changed by the collider). targetInterv(=-1) see InsertionSortCollider::targetInterv (auto-updated) timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updatingDispFactor(=-1) see InsertionSortCollider::updatingDispFactor (auto-updated) 314 Chapter 7. Class reference (yade.wrapper module)
7.6 Interaction Geometry creation

7.6.1 IGeomFunctor
Ig2_S phere_S phere_Dem3DofGeom
Ig2_Box_S phere_S cGeom
Ig2_Box_S phere_S cGeom6D
Ig2_ChainedCylinder_ChainedCylinder_S cGeom6D
Ig2_S phere_S phere_L6Geom
Ig2_W all_S phere_S cGeom
Ig2_W all_S phere_L3Geom
Ig2_S phere_S phere_L3Geom
Ig2_Facet_S phere_L3Geom
Ig2_Facet_S phere_S cGeom
Ig2_Facet_S phere_S cGeom6D
IGeomFunctor
Ig2_S phere_S phere_S cGeom
Ig2_S phere_S phere_S cGeom6D
Ig2_S phere_GridConnection_S cGridCoGeom
Ig2_GridNode_GridNode_GridNodeGeom6D
Ig2_Facet_S phere_Dem3DofGeom
Ig2_S phere_ChainedCylinder_CylS cGeom
Ig2_S phere_ChainedCylinder_CylS cGeom6D
Ig2_Tetra_Tetra_TTetraGeom
Ig2_GridConnection_GridConnection_GridCoGridCoGeom
Ig2_W all_S phere_Dem3DofGeom
class yade.wrapper.IGeomFunctor(inherits Functor Serializable) Functor for creating/updating Interaction::geom objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Box_Sphere_ScGeom(inherits IGeomFunctor Functor Serializable) Create an interaction geometry ScGeom from Box and Sphere, representing the box with a projected virtual sphere of same radius. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
7.6. Interaction Geometry creation
315
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Box_Sphere_ScGeom6D(inherits Ig2_Box_Sphere_ScGeom IGeomFunctor Functor Serializable) Create an interaction geometry ScGeom6D from Box and Sphere, representing the box with a projected virtual sphere of same radius. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_ChainedCylinder_ChainedCylinder_ScGeom6D(inherits IGeomFunctor Functor Serializable) Create/update a ScGeom instance representing connexion between chained cylinders. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. halfLengthContacts(=true) If True, Cylinders nodes interact like spheres of radius 0.5*length, else one node has size length while the other has size 0. The dierence is mainly the locus of rotation denition. interactionDetectionFactor(=1) Enlarge both radii by this factor (if >1), to permit creation of distant interactions. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Facet_Sphere_Dem3DofGeom(inherits IGeomFunctor Functor Serializable) Compute geometry of facet-sphere contact with normal and shear DOFs. As in all other Dem3DofGeom-related classes, total formulation of both shear and normal deformations is used. See Dem3DofGeom_FacetSphere for more information.
316
bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Facet_Sphere_L3Geom(inherits Ig2_Sphere_Sphere_L3Geom IGeomFunctor Functor Serializable) Incrementally compute L3Geom for contact between Facet and Sphere. Uses attributes of Ig2_Sphere_Sphere_L3Geom. approxMask Selectively enable geometrical approximations (bitmask); add the values for approximations to be enabled. 1 use previous transformation to transform velocities (which are known at mid-steps), instead of mid-step transformation computed as quaternion slerp at t=0.5. 2 do not take average (mid-step) normal when computing relative shear displacement, use previous value instead 4 do not re-normalize average (mid-step) normal, if used. By default, the mask is zero, wherefore none of these approximations is used. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. distFactor(=1) Create interaction if spheres are not futher than distFactor *(r1+r2). If negative, zero normal deformation will be set to be the initial value (otherwise, the geometrical distance is the zero one). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noRatch(=true) See Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. trsfRenorm(=100 ) How often to renormalize trsf; if non-positive, never renormalized (simulation might be unstable) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Facet_Sphere_ScGeom(inherits IGeomFunctor Functor Serializable) Create/update a ScGeom instance representing intersection of Facet and Sphere.
317
bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). shrinkFactor(=0, no shrinking) The radius of the inscribed circle of the facet is decreased by the value of the spheres radius multipled by shrinkFactor. From the denition of contact point on the surface made of facets, the given surface is not continuous and becomes in eect surface covered with triangular tiles, with gap between the separate tiles equal to the spheres radius multiplied by 2*shrinkFactor*. If zero, no shrinking is done. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Facet_Sphere_ScGeom6D(inherits Ig2_Facet_Sphere_ScGeom IGeomFunctor Functor Serializable) Create an interaction geometry ScGeom6D from Facet and Sphere, representing the Facet with a projected virtual sphere of same radius. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). shrinkFactor(=0, no shrinking) The radius of the inscribed circle of the facet is decreased by the value of the spheres radius multipled by shrinkFactor. From the denition of contact point on the surface made of facets, the given surface is not continuous and becomes in eect surface covered with triangular tiles, with gap between the separate tiles equal to the spheres radius multiplied by 2*shrinkFactor*. If zero, no shrinking is done. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_GridConnection_GridConnection_GridCoGridCoGeom(inherits IGeomFunctor Functor Serializable) Create/update a GridCoGridCoGeom instance representing the geometry of a contact point between two GridConnection , including relative rotations. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
318
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_GridNode_GridNode_GridNodeGeom6D(inherits Ig2_Sphere_Sphere_ScGeom IGeomFunctor Functor Serializable) Create/update a GridNodeGeom6D instance representing the geometry of a contact point between two GridNode, including relative rotations. avoidGranularRatcheting Dene relative velocity so that ratcheting is avoided. It applies for sphere-sphere contacts. It eventualy also apply for sphere-emulating interactions (i.e. convertible into the ScGeom type), if the virtual spheres motion is dened correctly (see e.g. Ig2_Sphere_ChainedCylinder_CylScGeom. Short explanation of what we want to avoid : Numerical ratcheting is best understood considering a small elastic cycle at a contact between two grains : assuming b1 is xed, impose this displacement to b2 : 1.translation dx in the normal direction 2.rotation a 3.translation -dx (back to the initial position) 4.rotation -a (back to the initial orientation) If the branch vector used to dene the relative shear in rotationbranch is not constant (typically if it is dened from the vector centercontactPoint), then the shear displacement at the end of this cycle is not zero: rotations a and -a are multiplied by branches of dierent lengths. It results in a nite contact force at the end of the cycle even though the positions and orientations are unchanged, in total contradiction with the elastic nature of the problem. It could also be seen as an inconsistent energy creation or loss. Given that DEM simulations tend to generate oscillations around equilibrium (damped mass-spring), it can have a signicant impact on the evolution of the packings, resulting for instance in slow creep in iterations under constant load. The solution adopted here to avoid ratcheting is as proposed by McNamara and co-workers. They analyzed the ratcheting problem in detail - even though they comment on the basis of a cycle that diers from the one shown above. One will nd interesting discussions in e.g. [McNamara2008], even though solution it suggests is not fully applied here (equations of motion are not incorporating alpha, in contradiction with what is suggested by McNamara et al.). bases Ordered list of types (as strings) this functor accepts. creep(=false) Substract rotational creep from relative rotation. The rotational creep ScGeom6D::twistCreep is a quaternion and has to be updated inside a constitutive law, see for instance Law2_ScGeom6D_CohFrictPhys_CohesionMoment. dict() dict Return dictionary of attributes.
319
interactionDetectionFactor Enlarge both radii by this factor (if >1), to permit creation of distant interactions. InteractionGeometry will be computed when interactionDetectionFactor*(rad1+rad2) > distance. Note: This parameter is functionally coupled with Bo1_Sphere_Aabb::aabbEnlargeFactor, which will create larger bounding boxes and should be of the same value. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateRotations(=true) Precompute relative rotations. Turning this false can speed up simulations when rotations are not needed in constitutive laws (e.g. when spheres are compressed without cohesion and moment in early stage of a triaxial test), but is not foolproof. Change this value only if you know what you are doing. class yade.wrapper.Ig2_Sphere_ChainedCylinder_CylScGeom(inherits IGeomFunctor Functor Serializable) Create/update a ScGeom instance representing intersection of two Spheres. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. interactionDetectionFactor(=1) Enlarge both radii by this factor (if >1), to permit creation of distant interactions. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Sphere_ChainedCylinder_CylScGeom6D(inherits Ig2_Sphere_ChainedCylinder_CylScGeom IGeomFunctor Functor Serializable) Create/update a ScGeom6D instance representing the geometry of a contact point between two Spheres, including relative rotations. bases Ordered list of types (as strings) this functor accepts. creep(=false) Substract rotational creep from relative rotation. The rotational creep ScGeom6D::twistCreep is a quaternion and has to be updated inside a constitutive law, see for instance Law2_ScGeom6D_CohFrictPhys_CohesionMoment. dict() dict Return dictionary of attributes. 320 Chapter 7. Class reference (yade.wrapper module)
interactionDetectionFactor(=1) Enlarge both radii by this factor (if >1), to permit creation of distant interactions. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateRotations(=false) Precompute relative rotations. Turning this false can speed up simulations when rotations are not needed in constitutive laws (e.g. when spheres are compressed without cohesion and moment in early stage of a triaxial test), but is not foolproof. Change this value only if you know what you are doing. class yade.wrapper.Ig2_Sphere_GridConnection_ScGridCoGeom(inherits IGeomFunctor Functor Serializable) Create/update a ScGridCoGeom6D instance representing the geometry of a contact point between a GricConnection and a Sphere including relative rotations. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. interactionDetectionFactor(=1) Enlarge both radii by this factor (if >1), to permit creation of distant interactions. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Sphere_Sphere_Dem3DofGeom(inherits IGeomFunctor Functor Serializable) Functor handling contact of 2 spheres, producing Dem3DofGeom instance bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. distFactor(=-1) Factor of sphere radius such that sphere touch if their centers are not further than distFactor*(r1+r2); if negative, equilibrium distance is the sum of the spheres radii. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
321
class yade.wrapper.Ig2_Sphere_Sphere_L3Geom(inherits IGeomFunctor Functor Serializable) Functor for computing incrementally conguration of 2 Spheres stored in L3Geom; the conguration is positioned in global space by local origin c (contact point) and rotation matrix T (orthonormal transformation matrix), and its degrees of freedom are local displacement u (in one normal and two shear directions); with Ig2_Sphere_Sphere_L6Geom and L6Geom, there is additionally . The rst row of T , i.e. local x-axis, is the contact normal noted n for brevity. Additionally, quasiconstant values of u0 (and 0 ) are stored as shifted origins of u (and ); therefore, current value of displacement is always u u0 . Suppose two spheres with radii ri , positions xi , velocities vi , angular velocities i .
When there is not yet contact, it will be created if uN = |x 2 x1 | |fd |(r1 + r2) < 0, where fd is distFactor (sometimes also called interaction radius). If fd > 0, then u0x will be initalized to uN , otherwise to 0. In another words, contact will be created if spheres enlarged by |fd | touch, and the equilibrium distance (where ux u 0x is zero) will be set to the current distance if fd is positive, and to the geometrically-touching distance if negative.
Local axes (rows of T ) are initially dened as follows: local x-axis is n = xl = x2 x1 ; local y-axis positioned arbitrarily, but in a deterministic manner: aligned with the xz plane (if ny < nz ) or xy plane (otherwise); local z-axis zl = xl yl . If there has already been contact between the two spheres, it is updated to keep track of rigid motion of the contact (one that does not change mutual conguration of spheres) and mutual conguration changes. Rigid motion transforms local coordinate system and can be decomposed in rigid translation (aecting c), and rigid rotation (aecting T ), which can be split in rotation or perpendicular to the normal and rotation ot (twist) parallel with the normal:
o r =n n .
Since velocities are known at previous midstep (t t/2), we consider mid-step normal n = n + n . 2
For the sake of numerical stability, n is re-normalized after being computed, unless prohibited by approxMask. If approxMask has the appropriate bit set, the mid-normal is not compute, and we simply use n n . Rigid rotation parallel with the normal is ( ) 1 + 2 o = n n t. t 2
Branch vectors b1 , b2 (connecting x 1 , x2 with c { r1 n b1 = c x 1 { r2 n b2 = c x 2
are computed depending on noRatch (see here). with noRatch otherwise with noRatch otherwise
Relative velocity at c can be computed as v v r = ( 2 + 2 b2 ) (v1 + 1 b1 )
322
where v2 is v2 without mean-eld velocity gradient in periodic boundary conditions (see Cell.homoDeform). In the numerial implementation, the normal part of incident velocity is re moved (since it is computed directly) with v r2 = vr (n vr )n . Any vector a expressed in global coordinates transforms during one timestep as
a = a + v r t a or a tr
where the increments have the meaning of relative shear, rigid rotation normal to n and rigid rotation parallel with n. Local coordinate system orientation, rotation matrix T , is updated by rows, i.e. nx n n y z T = T 1, T 1, or T 1, ot T 2, T 2, or T , ot
This matrix is re-normalized (unless prevented by approxMask) and mid-step transformation is computed using quaternion spherical interpolation as ( ) T = Slerp T ; T ; t = 1/2 .
Depending on approxMask, this computation can be avoided by approximating T = T . Finally, current displacement is evaluated as u = u + T v r t.
For the normal component, non-incremental evaluation is preferred, giving

u x = |x2 x1 | (r1 + r2 )
If this functor is called for L6Geom, local rotation is updated as = + T t(2 1 )
approxMask Selectively enable geometrical approximations (bitmask); add the values for approximations to be enabled. 1 use previous transformation to transform velocities (which are known at mid-steps), instead of mid-step transformation computed as quaternion slerp at t=0.5. 2 do not take average (mid-step) normal when computing relative shear displacement, use previous value instead 4 do not re-normalize average (mid-step) normal, if used. By default, the mask is zero, wherefore none of these approximations is used. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
323
distFactor(=1) Create interaction if spheres are not futher than distFactor *(r1+r2). If negative, zero normal deformation will be set to be the initial value (otherwise, the geometrical distance is the zero one). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noRatch(=true) See Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. trsfRenorm(=100 ) How often to renormalize trsf; if non-positive, never renormalized (simulation might be unstable) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Sphere_Sphere_L6Geom(inherits Ig2_Sphere_Sphere_L3Geom IGeomFunctor Functor Serializable) Incrementally compute L6Geom for contact of 2 spheres. approxMask Selectively enable geometrical approximations (bitmask); add the values for approximations to be enabled. 1 use previous transformation to transform velocities (which are known at mid-steps), instead of mid-step transformation computed as quaternion slerp at t=0.5. 2 do not take average (mid-step) normal when computing relative shear displacement, use previous value instead 4 do not re-normalize average (mid-step) normal, if used. By default, the mask is zero, wherefore none of these approximations is used. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. distFactor(=1) Create interaction if spheres are not futher than distFactor *(r1+r2). If negative, zero normal deformation will be set to be the initial value (otherwise, the geometrical distance is the zero one). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noRatch(=true) See Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. trsfRenorm(=100 ) How often to renormalize trsf; if non-positive, never renormalized (simulation might be unstable) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
324
class yade.wrapper.Ig2_Sphere_Sphere_ScGeom(inherits IGeomFunctor Functor Serializable) Create/update a ScGeom instance representing the geometry of a contact point between two Spheres s. avoidGranularRatcheting Dene relative velocity so that ratcheting is avoided. It applies for sphere-sphere contacts. It eventualy also apply for sphere-emulating interactions (i.e. convertible into the ScGeom type), if the virtual spheres motion is dened correctly (see e.g. Ig2_Sphere_ChainedCylinder_CylScGeom. Short explanation of what we want to avoid : Numerical ratcheting is best understood considering a small elastic cycle at a contact between two grains : assuming b1 is xed, impose this displacement to b2 : 1.translation dx in the normal direction 2.rotation a 3.translation -dx (back to the initial position) 4.rotation -a (back to the initial orientation) If the branch vector used to dene the relative shear in rotationbranch is not constant (typically if it is dened from the vector centercontactPoint), then the shear displacement at the end of this cycle is not zero: rotations a and -a are multiplied by branches of dierent lengths. It results in a nite contact force at the end of the cycle even though the positions and orientations are unchanged, in total contradiction with the elastic nature of the problem. It could also be seen as an inconsistent energy creation or loss. Given that DEM simulations tend to generate oscillations around equilibrium (damped mass-spring), it can have a signicant impact on the evolution of the packings, resulting for instance in slow creep in iterations under constant load. The solution adopted here to avoid ratcheting is as proposed by McNamara and co-workers. They analyzed the ratcheting problem in detail - even though they comment on the basis of a cycle that diers from the one shown above. One will nd interesting discussions in e.g. [McNamara2008], even though solution it suggests is not fully applied here (equations of motion are not incorporating alpha, in contradiction with what is suggested by McNamara et al.). bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. interactionDetectionFactor Enlarge both radii by this factor (if >1), to permit creation of distant interactions. InteractionGeometry will be computed when interactionDetectionFactor*(rad1+rad2) > distance. Note: This parameter is functionally coupled with Bo1_Sphere_Aabb::aabbEnlargeFactor, which will create larger bounding boxes and should be of the same value. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True.
325
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Sphere_Sphere_ScGeom6D(inherits Ig2_Sphere_Sphere_ScGeom IGeomFunctor Functor Serializable) Create/update a ScGeom6D instance representing the geometry of a contact point between two Spheres, including relative rotations. avoidGranularRatcheting Dene relative velocity so that ratcheting is avoided. It applies for sphere-sphere contacts. It eventualy also apply for sphere-emulating interactions (i.e. convertible into the ScGeom type), if the virtual spheres motion is dened correctly (see e.g. Ig2_Sphere_ChainedCylinder_CylScGeom. Short explanation of what we want to avoid : Numerical ratcheting is best understood considering a small elastic cycle at a contact between two grains : assuming b1 is xed, impose this displacement to b2 : 1.translation dx in the normal direction 2.rotation a 3.translation -dx (back to the initial position) 4.rotation -a (back to the initial orientation) If the branch vector used to dene the relative shear in rotationbranch is not constant (typically if it is dened from the vector centercontactPoint), then the shear displacement at the end of this cycle is not zero: rotations a and -a are multiplied by branches of dierent lengths. It results in a nite contact force at the end of the cycle even though the positions and orientations are unchanged, in total contradiction with the elastic nature of the problem. It could also be seen as an inconsistent energy creation or loss. Given that DEM simulations tend to generate oscillations around equilibrium (damped mass-spring), it can have a signicant impact on the evolution of the packings, resulting for instance in slow creep in iterations under constant load. The solution adopted here to avoid ratcheting is as proposed by McNamara and co-workers. They analyzed the ratcheting problem in detail - even though they comment on the basis of a cycle that diers from the one shown above. One will nd interesting discussions in e.g. [McNamara2008], even though solution it suggests is not fully applied here (equations of motion are not incorporating alpha, in contradiction with what is suggested by McNamara et al.). bases Ordered list of types (as strings) this functor accepts. creep(=false) Substract rotational creep from relative rotation. The rotational creep ScGeom6D::twistCreep is a quaternion and has to be updated inside a constitutive law, see for instance Law2_ScGeom6D_CohFrictPhys_CohesionMoment. dict() dict Return dictionary of attributes. interactionDetectionFactor Enlarge both radii by this factor (if >1), to permit creation of distant interactions. InteractionGeometry will be computed when interactionDetectionFactor*(rad1+rad2) > distance. Note: This parameter is functionally coupled with Bo1_Sphere_Aabb::aabbEnlargeFactor, which will create larger bounding boxes and should be of the same value.
326
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary updateRotations(=true) Precompute relative rotations. Turning this false can speed up simulations when rotations are not needed in constitutive laws (e.g. when spheres are compressed without cohesion and moment in early stage of a triaxial test), but is not foolproof. Change this value only if you know what you are doing. class yade.wrapper.Ig2_Tetra_Tetra_TTetraGeom(inherits IGeomFunctor Functor Serializable) Create/update geometry of collision between 2 tetrahedra (TTetraGeom instance) bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Wall_Sphere_Dem3DofGeom(inherits IGeomFunctor Functor Serializable) Create/update contact of Wall and Sphere (Dem3DofGeom_WallSphere instance) bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Wall_Sphere_L3Geom(inherits Ig2_Sphere_Sphere_L3Geom IGeomFunctor Functor Serializable) Incrementally compute L3Geom for contact between Wall and Sphere. Uses attributes of Ig2_Sphere_Sphere_L3Geom. approxMask Selectively enable geometrical approximations (bitmask); add the values for approximations to be enabled.
327
1 2 4
use previous transformation to transform velocities (which are known at mid-steps), instead of mid-step transformation computed as quaternion slerp at t=0.5. do not take average (mid-step) normal when computing relative shear displacement, use previous value instead do not re-normalize average (mid-step) normal, if used.
By default, the mask is zero, wherefore none of these approximations is used. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. distFactor(=1) Create interaction if spheres are not futher than distFactor *(r1+r2). If negative, zero normal deformation will be set to be the initial value (otherwise, the geometrical distance is the zero one). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noRatch(=true) See Ig2_Sphere_Sphere_ScGeom.avoidGranularRatcheting. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. trsfRenorm(=100 ) How often to renormalize trsf; if non-positive, never renormalized (simulation might be unstable) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ig2_Wall_Sphere_ScGeom(inherits IGeomFunctor Functor Serializable) Create/update a ScGeom instance representing intersection of Wall and Sphere. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noRatch(=true) Avoid granular ratcheting timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.6.2 IGeomDispatcher
class yade.wrapper.IGeomDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s).
328
dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((Shape)arg2, (Shape)arg3 ) IGeomFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
329
7.7 Interaction Physics creation

7.7.1 IPhysFunctor
Ip2_FrictMat_FrictMat_CapillaryPhys
Ip2_LudingMat_LudingMat_LudingPhys
Ip2_FrictMat_FrictMat_MindlinCapillaryPhys
Ip2_FrictMat_CpmMat_FrictPhys
Ip2_JCFpmMat_JCFpmMat_JCFpmPhys
Ip2_RpmMat_RpmMat_RpmPhys
Ip2_2xNormalInelas ticMat_NormalInelas ticityPhys
Ip2_FrictMat_FrictMat_FrictPhys IPhys Functor Ip2_2xInelas tCohFrictMat_Inelas tCohFrictPhys
Ip2_FrictMat_FrictMat_Vis coFrictPhys
Ip2_FrictMat_FrictMat_MindlinPhys
Ip2_WireMat_WireMat_WirePhys
Ip2_CpmMat_CpmMat_CpmPhys
Ip2_CFpmMat_CFpmMat_CFpmPhys
Ip2_CohFrictMat_CohFrictMat_CohFrictPhys
Ip2_BubbleMat_BubbleMat_BubblePhys
Ip2_Vis cElMat_Vis cElMat_Vis cElPhys
class yade.wrapper.IPhysFunctor(inherits Functor Serializable) Functor for creating/updating Interaction::phys objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True.
330
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_2xInelastCohFrictMat_InelastCohFrictPhys(inherits IPhysFunctor Functor Serializable) Generates cohesive-frictional interactions with moments. Used in the contact law Law2_ScGeom6D_InelastCohFrictPhys_CohesionMoment. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_2xNormalInelasticMat_NormalInelasticityPhys(inherits IPhysFunctor Functor Serializable) The RelationShips for using Law2_ScGeom6D_NormalInelasticityPhys_NormalInelasticity In these RelationShips all the attributes of the interactions (which are of NormalInelasticityPhys type) are computed. Warning: as in the others Ip2 functors, most of the attributes are computed only once, when the interaction is new. bases Ordered list of types (as strings) this functor accepts. betaR(=0.12 ) Parameter for computing the torque-stifness : T-stifness = betaR * Rmoy^2 dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_BubbleMat_BubbleMat_BubblePhys(inherits IPhysFunctor Functor Serializable) Generates bubble interactions.Used in the contact law Law2_ScGeom_BubblePhys_Bubble. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
7.7. Interaction Physics creation
331
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_CFpmMat_CFpmMat_CFpmPhys(inherits IPhysFunctor Functor Serializable) Converts 2 CFpmmat instances to CFpmPhys with corresponding parameters. Alpha(=0 ) Denes the ratio ks/kn. Beta(=0 ) Denes the ratio kr/(ks*meanRadius^2) to compute the resistive moment in rotation. [-] bases Ordered list of types (as strings) this functor accepts. cohesion(=0 ) Denes the maximum admissible tangential force in shear FsMax=cohesion*crossSection. [Pa] cohesiveTresholdIteration(=1) Should new contacts be cohesive? They will before this iter, they wont afterward. dict() dict Return dictionary of attributes. eta(=0 ) Denes the maximum admissible resistive moment in rotation MtMax=eta*meanRadius*Fn. [-] label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). strengthSoftening(=0 ) Denes the softening when Dtensile is reached to avoid explosion of the contact. Typically, when D > Dtensile, Fn=FnMax - (kn/strengthSoftening)*(Dtensile-D). [-] tensileStrength(=0 ) Denes the maximum admissible Max=tensileStrength*crossSection. [Pa] normal force in traction Fn-
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary useAlphaBeta(=false) If true, stinesses are computed based on Alpha and Beta. class yade.wrapper.Ip2_CohFrictMat_CohFrictMat_CohFrictPhys(inherits IPhysFunctor Functor Serializable) Generates cohesive-frictional interactions with moments. Used in the contact law Law2_ScGeom6D_CohFrictPhys_CohesionMoment. The normal/shear stiness and friction denitions are the same as in Ip2_FrictMat_FrictMat_FrictPhys, check the documentation there for details. bases Ordered list of types (as strings) this functor accepts.
332
dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). setCohesionNow(=false) If true, assign cohesion to all existing contacts in current time-step. The ag is turned false automatically, so that assignment is done in the current timestep only. setCohesionOnNewContacts(=false) If true, assign cohesion at all new contacts. If false, only existing contacts can be cohesive (also see Ip2_CohFrictMat_CohFrictMat_CohFrictPhys::setCohesionNow), and new contacts are only frictional. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_CpmMat_CpmMat_CpmPhys(inherits IPhysFunctor Functor Serializable) Convert 2 CpmMat instances to CpmPhys with corresponding parameters. Uses simple (arithmetic) averages if material are dierent. Simple copy of parameters is performed if the material is shared between both particles. See cpm-model for detals. bases Ordered list of types (as strings) this functor accepts. cohesiveThresholdIter(=10 ) Should new contacts be cohesive? They will before this iter#, they will not be afterwards. If 0, they will never be. If negative, they will always be created as cohesive (10 by default). dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_FrictMat_CpmMat_FrictPhys(inherits IPhysFunctor Functor Serializable) Convert CpmMat instance and FrictMat instance to FrictPhys with corresponding parameters (young, poisson, frictionAngle). Uses simple (arithmetic) averages if material parameters are different. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier).
333
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_FrictMat_FrictMat_CapillaryPhys(inherits IPhysFunctor Functor Serializable) RelationShips to use with Law2_ScGeom_CapillaryPhys_Capillarity In these RelationShips all the interaction attributes are computed. Warning: as in the others Ip2 functors, most of the attributes are computed only once, when the interaction is new. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_FrictMat_FrictMat_FrictPhys(inherits IPhysFunctor Functor Serializable) Create a FrictPhys from two FrictMats. The compliance of one sphere under point load is dened here as 1/(E.D), with E the stiness of the sphere and D its diameter. The compliance of the contact itself will be the sum of compliances from each sphere, i.e. 1/(E1 .D1 ) + 1/(E2 .D2 ) in the general case, or 2/(E.D) in the special case of equal sizes and equal stiness. Note that summing compliances corresponds to an harmonic average of stinesss (as in e.g. [Scholtes2009a]), which is how kn is actually computed in the Ip2_FrictMat_FrictMat_FrictPhys functor: kn =
E1 D1 E2 D2 E1 D1 +E2 D2
k1 k2 k1 +k2 ,
with ki = Ei Di .
The shear stiness ks of one sphere is dened via the material parameter ElastMat::poisson, as ks=poisson*kn, and the resulting shear stiness of the interaction will be also an harmonic average. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. frictAngle(=uninitalized ) Instance of MatchMaker determining how to compute interactions friction angle. If None, minimum value is used. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
334
class yade.wrapper.Ip2_FrictMat_FrictMat_MindlinCapillaryPhys(inherits IPhysFunctor Functor Serializable) RelationShips to use with Law2_ScGeom_CapillaryPhys_Capillarity In these RelationShips all the interaction attributes are computed. Warning: as in the others Ip2 functors, most of the attributes are computed only once, when the interaction is new. bases Ordered list of types (as strings) this functor accepts. betan(=uninitalized ) Normal viscous damping coecient n . betas(=uninitalized ) Shear viscous damping coecient s . dict() dict Return dictionary of attributes. en(=uninitalized ) Normal coecient of restitution en . es(=uninitalized ) Shear coecient of restitution es . eta(=0.0 ) Coecient to determine the plastic bending moment gamma(=0.0 ) Surface energy parameter [J/m^2] per each unit contact surface, to derive DMT formulation from HM krot(=0.0 ) Rotational stiness for moment contact law ktwist(=0.0 ) Torsional stiness for moment contact law label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_FrictMat_FrictMat_MindlinPhys(inherits IPhysFunctor Functor Serializable) Calculate some physical parameters needed to obtain the normal and shear stinesses according to the Hertz-Mindlins formulation (as implemented in PFC). Viscous parameters can be specied either using coecients of restitution (en , es ) or viscous damping coecient (n , s ). The following rules apply: #. If the n (s ) coecient is given, it is directly. #. If en is given, MindlinPhys.betan assigned to MindlinPhys.betan (MindlinPhys.betas) is computed using n = (log en )/ 2 + (log en )2 . The same applies to es , MindlinPhys.betas. #. It is an error (exception) to specify both en and n (es and s ). #. If neither en nor n is given, zero value for MindlinPhys.betan is used; there will be no viscous eects. #.If neither es nor s is given, the value of MindlinPhys.betan is used for MindlinPhys.betas as well. The en , n , es , s are MatchMaker objects; they can be constructed from oat values to always return constant value. See scripts/test/shots.py for an example of specifying en based on combination of parameters.
335
bases Ordered list of types (as strings) this functor accepts. betan(=uninitalized ) Normal viscous damping coecient n . betas(=uninitalized ) Shear viscous damping coecient s . dict() dict Return dictionary of attributes. en(=uninitalized ) Normal coecient of restitution en . es(=uninitalized ) Shear coecient of restitution es . eta(=0.0 ) Coecient to determine the plastic bending moment gamma(=0.0 ) Surface energy parameter [J/m^2] per each unit contact surface, to derive DMT formulation from HM krot(=0.0 ) Rotational stiness for moment contact law ktwist(=0.0 ) Torsional stiness for moment contact law label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_FrictMat_FrictMat_ViscoFrictPhys(inherits Ip2_FrictMat_FrictMat_FrictPhys IPhysFunctor Functor Serializable) Create a FrictPhys from two FrictMats. The compliance of one sphere under symetric point loads is dened here as 1/(E.r), with E the stiness of the sphere and r its radius, and corresponds to a compliance 1/(2.E.r)=1/(E.D) from each contact point. The compliance of the contact itself will be the sum of compliances from each sphere, i.e. 1/(E.D1)+1/(E.D2) in the general case, or 1/(E.r) in the special case of equal sizes. Note that summing compliances corresponds to an harmonic average of stinesss, which is how kn is actually computed in the Ip2_FrictMat_FrictMat_FrictPhys functor. The shear stiness ks of one sphere is dened via the material parameter ElastMat::poisson, as ks=poisson*kn, and the resulting shear stiness of the interaction will be also an harmonic average. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. frictAngle(=uninitalized ) Instance of MatchMaker determining how to compute interactions friction angle. If None, minimum value is used.
336
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_JCFpmMat_JCFpmMat_JCFpmPhys(inherits IPhysFunctor Functor Serializable) converts 2 JCFpmMat instances to JCFpmPhys with corresponding parameters. alpha(=0.) denes the shear to normal stiness ratio ks/kn. bases Ordered list of types (as strings) this functor accepts. cohesion(=0.) denes the maximum admissible tangential force in shear FsMax=cohesion*crossSection in the matrix. [Pa] cohesiveTresholdIteration(=1) should new contacts be cohesive? They will before this iter, they wont afterward. dict() dict Return dictionary of attributes. jointCohesion(=0.) denes the maximum admissible tangential force in shear FsMax=cohesion*crossSection on the joint surface. [Pa] jointDilationAngle(=0 ) denes the dilatancy of the joint surface (only valid for smooth contact logic). [rad] jointFrictionAngle(=-1) denes Coulomb friction on the joint surface. [rad] jointNormalStiffness(=0.) denes the normal stiness on the joint surface. jointShearStiffness(=0.) denes the shear stiness on the joint surface. jointTensileStrength(=0.) denes the maximum admissible normal force in traction FnMax=tensileStrength*crossSection on the joint surface. [Pa] label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). tensileStrength(=0.) denes the maximum admissible normal force in traction FnMax=tensileStrength*crossSection in the matrix. [Pa] timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_LudingMat_LudingMat_LudingPhys(inherits IPhysFunctor Functor Serializable) Convert 2 instances of LudingMat to LudingPhys using the rule of consecutive connection.
337
bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_RpmMat_RpmMat_RpmPhys(inherits IPhysFunctor Functor Serializable) Convert 2 RpmMat instances to RpmPhys with corresponding parameters. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_ViscElMat_ViscElMat_ViscElPhys(inherits IPhysFunctor Functor Serializable) Convert 2 instances of ViscElMat to ViscElPhys using the rule of consecutive connection. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Ip2_WireMat_WireMat_WirePhys(inherits IPhysFunctor Functor Serializable) Converts 2 WireMat instances to WirePhys with corresponding parameters. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
338
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). linkThresholdIteration(=1) Iteration to create the link. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.7.2 IPhysDispatcher
class yade.wrapper.IPhysDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((Material)arg2, (Material)arg3 ) IPhysFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
339
7.8 Constitutive laws

7.8.1 LawFunctor
Law2_S cGeom_BubblePhys _Bubble
Law2_S cGeom_CpmPhys _Cpm
Law2_Dem3DofGeom_RockPMPhys _Rpm
Law2_ChCylGeom6D_CohFrictPhys _Cohes ionMoment
Law2_S cGeom_CFpmPhys _Cohes iveFrictionalPM
Law2_CylS cGeom_FrictPhys _CundallS track
Law2_S cGeom_MindlinPhys _HertzWithLinearS hear
Law2_S cGeom6D_NormalInelas ticityPhys _NormalInelas ticity
Law2_CylS cGeom6D_CohFrictPhys _Cohes ionMoment
Law2_L3Geom_FrictPhys _ElPerfPl
Law2_L6Geom_FrictPhys _Linear
LawFunctor
Law2_S cGeom_FrictPhys _CundallS track
Law2_S cGeom_Vis coFrictPhys _CundallS track
Law2_S cGeom_WirePhys _WirePM
Law2_GridCoGridCoGeom_FrictPhys _CundallS track
Law2_S cGeom6D_Inelas tCohFrictPhys _Cohes ionMoment
Law2_S cGeom6D_CohFrictPhys _Cohes ionMoment
Law2_S cGeom_Vis cElPhys _Bas ic
Law2_S cGridCoGeom_CohFrictPhys _CundallS track
Law2_S cGeom_MindlinPhys _Mindlin
Law2_S cGeom_LudingPhys _Bas ic
Law2_S cGeom_MindlinPhys _MindlinDeres iewitz
Law2_S cGridCoGeom_FrictPhys _CundallS track
Law2_S cGeom_JCFpmPhys _JointedCohes iveFrictionalPM
class yade.wrapper.LawFunctor(inherits Functor Serializable) Functor for applying constitutive laws on interactions. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron
340
python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ChCylGeom6D_CohFrictPhys_CohesionMoment(inherits LawFunctor Functor Serializable) Law for linear compression, and Mohr-Coulomb plasticity surface without cohesion. This law implements the classical linear elastic-plastic law from [CundallStrack1979] (see also [Pfc3dManual30]). The normal force is (with the convention of positive tensile forces) Fn = min(kn un , 0). The shear force is Fs = ks us , the plasticity condition denes the maximum value of the shear force : Fmax = Fn tan(), with the friction angle. s Note: This law uses ScGeom; there is also functionally equivalent Law2_Dem3DofGeom_FrictPhys_CundallStrack, which uses Dem3DofGeom (sphere-box interactions are not implemented for the latest).
Note: This law is well tested in the context of triaxial simulation, and has been used for a number of published results (see e.g. [Scholtes2009b] and other papers from the same authors). It is generalised by Law2_ScGeom6D_CohFrictPhys_CohesionMoment, which adds cohesion and moments at contact. always_use_moment_law(=false) If true, use bending/twisting moments at all contacts. If false, compute moments only for cohesive contacts. bases Ordered list of types (as strings) this functor accepts. creep_viscosity(=1) creep viscosity [Pa.s/m]. probably should be moved to Ip2_CohFrictMat_CohFrictMat_CohFrictPhys... dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) shear_creep(=false) activate creep on the shear force, using CohesiveFrictionalContactLaw::creep_viscosity. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. twist_creep(=false) activate creep on the twisting moment, using CohesiveFrictionalContactLaw::creep_viscosity. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.8. Constitutive laws
341
useIncrementalForm(=false) use the incremental formulation to compute bending and twisting moments. Creep on the twisting moment is not included in such a case. class yade.wrapper.Law2_CylScGeom6D_CohFrictPhys_CohesionMoment(inherits LawFunctor Functor Serializable) This law generalises Law2_CylScGeom_FrictPhys_CundallStrack by adding cohesion and moments at contact. always_use_moment_law(=false) If true, use bending/twisting moments at all contacts. If false, compute moments only for cohesive contacts. bases Ordered list of types (as strings) this functor accepts. creep_viscosity(=1) creep viscosity [Pa.s/m]. probably should be moved to Ip2_CohFrictMat_CohFrictMat_CohFrictPhys... dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) shear_creep(=false) activate creep on the shear force, using CohesiveFrictionalContactLaw::creep_viscosity. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. twist_creep(=false) activate creep on the twisting moment, using CohesiveFrictionalContactLaw::creep_viscosity. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary useIncrementalForm(=false) use the incremental formulation to compute bending and twisting moments. Creep on the twisting moment is not included in such a case. class yade.wrapper.Law2_CylScGeom_FrictPhys_CundallStrack(inherits LawFunctor Functor Serializable) Law for linear compression, and Mohr-Coulomb plasticity surface without cohesion. This law implements the classical linear elastic-plastic law from [CundallStrack1979] (see also [Pfc3dManual30]). The normal force is (with the convention of positive tensile forces) Fn = min(kn un , 0). The shear force is Fs = ks us , the plasticity condition denes the maximum value of the shear force : Fmax = Fn tan(), with the friction angle. s Note: This law uses ScGeom.
Note: This law is well tested in the context of triaxial simulation, and has been used for a number of published results (see e.g. [Scholtes2009b] and other papers from the same authors). It is generalised by Law2_ScGeom6D_CohFrictPhys_CohesionMoment, which adds cohesion and moments at contact.
342
bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_Dem3DofGeom_RockPMPhys_Rpm(inherits LawFunctor Functor Serializable) Constitutive law for the Rpm model bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_GridCoGridCoGeom_FrictPhys_CundallStrack(inherits Law2_ScGeom_FrictPhys_CundallStrack LawFunctor Functor Serializable) Frictional elastic contact law between two gridConnection . See Law2_ScGeom_FrictPhys_CundallStrack for more details. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. elasticEnergy() oat Compute and return the total elastic energy in all FrictPhys contacts initPlasticDissipation((oat)arg2 ) None Initialize cummulated plastic dissipation to a value (0 by default). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier).
343
neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) plasticDissipation() oat Total energy dissipated in plastic slips at all FrictPhys contacts. Computed only if Law2_ScGeom_FrictPhys_CundallStrack::traceEnergy is true. sphericalBodies(=true) If true, compute branch vectors from radii (faster), else use contactPoint-position. Turning this ag true is safe for sphere-sphere contacts and a few other specic cases. It will give wrong values of torques on facets or boxes. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. traceEnergy(=false) Dene the total energy dissipated in plastic slips at all contacts. This will trace only plastic energy in this law, see O.trackEnergy for a more complete energies tracing updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_L3Geom_FrictPhys_ElPerfPl(inherits LawFunctor Functor Serializable) Basic law for testing L3Geom; it bears no cohesion (unless noBreak is True), and plastic slip obeys the Mohr-Coulomb criterion (unless noSlip is True). bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noBreak(=false) Do not break contacts when particles separate. noSlip(=false) No plastic slipping. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_L6Geom_FrictPhys_Linear(inherits Law2_L3Geom_FrictPhys_ElPerfPl LawFunctor Functor Serializable) Basic law for testing L6Geom linear in both normal and shear sense, without slip or breakage. bases Ordered list of types (as strings) this functor accepts. charLen(=1) Characteristic length with the meaning of the stiness ratios bending/shear and torsion/normal. dict() dict Return dictionary of attributes.
344
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). noBreak(=false) Do not break contacts when particles separate. noSlip(=false) No plastic slipping. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom6D_CohFrictPhys_CohesionMoment(inherits LawFunctor Functor Serializable) Law for linear traction-compression-bending-twisting, with cohesion+friction and Mohr-Coulomb plasticity surface. This law adds adhesion and moments to Law2_ScGeom_FrictPhys_CundallStrack. The normal force is (with the convention of positive tensile forces) Fn = min(kn un , an ), with an the normal adhesion. The shear force is Fs = ks us , the plasticity condition denes the maximum value of the shear force, by default Fmax = Fn tan() + as , with the friction angle and an s the shear adhesion. If CohFrictPhys::cohesionDisableFriction is True, friction is ignored as long as adhesion is active, and the maximum shear force is only Fmax = as . s If the maximum tensile or maximum shear force is reached and CohFrictPhys::fragile =True (default), the cohesive link is broken, and an , as are set back to zero. If a tensile force is present, the contact is lost, else the shear strength is Fmax = Fn tan(). If CohFrictPhys::fragile =False s (in course of implementation), the behaviour is perfectly plastic, and the shear strength is kept constant. If Law2_ScGeom6D_CohFrictPhys_CohesionMoment::momentRotationLaw =True, bending and twisting moments are computed using a linear law with moduli respectively kt and kr (the two values are the same currently), so that the moments are : Mb = kb b and Mt = kt t , with b,t the relative rotations between interacting bodies. The maximum value of moments can be dened and takes the form of rolling friction. Cohesive -type moment may also be included in the future. Creep at contact is implemented in this law, as dened in [Hassan2010]. If activated, there is a viscous behaviour of the shear and twisting components, and the evolution of the elastic parts of shear displacement and relative twist is given by dus,e /dt = Fs /s and dt,e /dt = Mt /t . always_use_moment_law(=false) If true, use bending/twisting moments at all contacts. If false, compute moments only for cohesive contacts. bases Ordered list of types (as strings) this functor accepts. creep_viscosity(=1) creep viscosity [Pa.s/m]. probably should be moved to Ip2_CohFrictMat_CohFrictMat_CohFrictPhys. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity)
345
normElastEnergy() oat Compute normal elastic energy. shearElastEnergy() oat Compute shear elastic energy. shear_creep(=false) activate creep on the shear force, using CohesiveFrictionalContactLaw::creep_viscosity. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. twist_creep(=false) activate creep on the twisting moment, using CohesiveFrictionalContactLaw::creep_viscosity. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary useIncrementalForm(=false) use the incremental formulation to compute bending and twisting moments. Creep on the twisting moment is not included in such a case. class yade.wrapper.Law2_ScGeom6D_InelastCohFrictPhys_CohesionMoment(inherits LawFunctor Functor Serializable) Law for linear traction-compression-bending-twisting, with cohesion+friction and Mohr-Coulomb plasticity surface. This law adds adhesion and moments to Law2_ScGeom_FrictPhys_CundallStrack. The normal force is (with the convention of positive tensile forces) Fn = min(kn un , an ), with an the normal adhesion. The shear force is Fs = ks us , the plasticity condition denes the maximum = Fn tan() + as , with the friction angle and an value of the shear force, by default Fmax s the shear adhesion. If CohFrictPhys::cohesionDisableFriction is True, friction is ignored as long as = as . adhesion is active, and the maximum shear force is only Fmax s If the maximum tensile or maximum shear force is reached and CohFrictPhys::fragile =True (default), the cohesive link is broken, and an , as are set back to zero. If a tensile force is present, the contact is lost, else the shear strength is Fmax = Fn tan(). If CohFrictPhys::fragile =False s (in course of implementation), the behaviour is perfectly plastic, and the shear strength is kept constant. If Law2_ScGeom6D_CohFrictPhys_CohesionMoment::momentRotationLaw =True, bending and twisting moments are computed using a linear law with moduli respectively kt and kr (the two values are the same currently), so that the moments are : Mb = kb b and Mt = kt t , with b,t the relative rotations between interacting bodies. The maximum value of moments can be dened and takes the form of rolling friction. Cohesive -type moment may also be included in the future. Creep at contact is implemented in this law, as dened in [Hassan2010]. If activated, there is a viscous behaviour of the shear and twisting components, and the evolution of the elastic parts of shear displacement and relative twist is given by dus,e /dt = Fs /s and dt,e /dt = Mt /t . bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). normElastEnergy() oat Compute normal elastic energy.
346
shearElastEnergy() oat Compute shear elastic energy. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary useIncrementalForm(=false) use the incremental formulation to compute bending and twisting moments. Creep on the twisting moment is not included in such a case. class yade.wrapper.Law2_ScGeom6D_NormalInelasticityPhys_NormalInelasticity(inherits LawFunctor Functor Serializable) Contact law used to simulate granular ller in rock joints [Duriez2009a], [Duriez2011]. It includes possibility of cohesion, moment transfer and inelastic compression behaviour (to reproduce the normal inelasticity observed for rock joints, for the latter). The moment transfer relation corresponds to the adaptation of the work of Plassiard & Belheine (see in [DeghmReport2006] for example), which was realized by J. Kozicki, and is now coded in ScGeom6D. As others LawFunctor, it uses pre-computed data of the interactions (rigidities, friction angles -with their tan()-, orientations of the interactions); this work is done here in Ip2_2xNormalInelasticMat_NormalInelasticityPhys. To use this you should also use NormalInelasticMat as material type of the bodies. The eects of this law are illustrated in examples/normalInelasticityTest.py bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). momentAlwaysElastic(=false) boolean, true=> the part of the contact torque (caused by relative rotations, which is computed only if momentRotationLaw..) is not limited by a plastic threshold momentRotationLaw(=true) boolean, true=> computation of a torque (against relative rotation) exchanged between particles timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_BubblePhys_Bubble(inherits LawFunctor Functor Serializable) Constitutive law for Bubble model. bases Ordered list of types (as strings) this functor accepts.
347
dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_CFpmPhys_CohesiveFrictionalPM(inherits LawFunctor Functor Serializable) Constitutive law for the CFpm model. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). preventGranularRatcheting(=true) If true rotations are computed such as granular ratcheting is prevented. [Alonso2004], pg. 3-10 and a lot more papers from the same authors). See article
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_CpmPhys_Cpm(inherits LawFunctor Functor Serializable) Constitutive law for the cpm-model. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. elasticEnergy() oat Compute and return the total elastic energy in all CpmPhys contacts epsSoft(=-3e-3, approximates connement -20MPa precisely, -100MPa a little over, -200 and -400 are OK (secant)) Strain at which softening in compression starts (non-negative to deactivate) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). omegaThreshold(=1., >=1. to deactivate, i.e. never delete any contacts) damage after which the contact disappears (<1), since omega reaches 1 only for strain + relKnSoft(=.3 ) Relative rigidity of the softening branch in compression (0=perfect elastic-plastic, <0 softening, >0 hardening) timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True.
348
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary yieldEllipseShift(=NaN ) horizontal scaling of the ellipse (shifts on the +x axis as interactions with +y are given) yieldLogSpeed(=.1) scaling in the logarithmic yield surface (should be <1 for realistic results; >=0 for meaningful results) yieldSigmaTMagnitude((oat)sigmaN, (oat)omega, (oat)undamagedCohesion, (oat)tanFrictionAngle) oat Return radius of yield surface for given material and state parameters; uses attributes of the current instance (yieldSurfType etc), change them before calling if you need that. yieldSurfType(=2 ) yield function: 0: mohr-coulomb (original); 1: parabolic; 2: logarithmic, 3: log+lin_tension, 4: elliptic, 5: elliptic+log class yade.wrapper.Law2_ScGeom_FrictPhys_CundallStrack(inherits LawFunctor Functor Serializable) Law for linear compression, and Mohr-Coulomb plasticity surface without cohesion. This law implements the classical linear elastic-plastic law from [CundallStrack1979] (see also [Pfc3dManual30]). The normal force is (with the convention of positive tensile forces) Fn = min(kn un , 0). The shear force is Fs = ks us , the plasticity condition denes the maximum value of the shear force : Fmax = Fn tan(), with the friction angle. s This law is well tested in the context of triaxial simulation, and has been used for a number of published results (see e.g. [Scholtes2009b] and other papers from the same authors). It is generalised by Law2_ScGeom6D_CohFrictPhys_CohesionMoment, which adds cohesion and moments at contact. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. elasticEnergy() oat Compute and return the total elastic energy in all FrictPhys contacts initPlasticDissipation((oat)arg2 ) None Initialize cummulated plastic dissipation to a value (0 by default). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) plasticDissipation() oat Total energy dissipated in plastic slips at all FrictPhys contacts. Computed only if Law2_ScGeom_FrictPhys_CundallStrack::traceEnergy is true. sphericalBodies(=true) If true, compute branch vectors from radii (faster), else use contactPoint-position. Turning this ag true is safe for sphere-sphere contacts and a few other specic cases. It will give wrong values of torques on facets or boxes. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. traceEnergy(=false) Dene the total energy dissipated in plastic slips at all contacts. This will trace only plastic energy in this law, see O.trackEnergy for a more complete energies tracing 7.8. Constitutive laws 349
updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_JCFpmPhys_JointedCohesiveFrictionalPM(inherits LawFunctor Functor Serializable) interaction law for jointed frictional material. Basically, this law adds the possibility to dene joint surfaces into a cohesive frictional material as dened by :yref:Law2_ScGeom_CFpmPhys_CohesiveFrictionalPM. Joint surfaces can be dened in a preprocessing phase through .stl meshes (see ref for details of the procedure). bases Ordered list of types (as strings) this functor accepts. cracksFileExist(=false) If true, text le already exists and its content wont be reset. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). recordCracks(=false) if true a text le cracks.txt will be created (iteration, position, type (tensile or shear), cross section and contact normal). smoothJoint(=false) if true, particles belonging to joint surface have a smooth contact logic. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_LudingPhys_Basic(inherits LawFunctor Functor Serializable) Linear viscoelastic model operating on ScGeom and LudingPhys. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_MindlinPhys_HertzWithLinearShear(inherits LawFunctor Functor Serializable) Constitutive law for the Hertz formulation (using MindlinPhys.kno) and linear beahvior in shear (using MindlinPhys.kso for stiness and FrictPhys.tangensOfFrictionAngle). Note: No viscosity or damping. If you need those, look at Law2_ScGeom_MindlinPhys_Mindlin, which also includes non-linear Mindlin shear. 350 Chapter 7. Class reference (yade.wrapper module)
bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) nonLin(=0 ) Shear force nonlinearity (the value determines how many features of the non-linearity are taken in account). 1: ks as in HM 2: shearElastic increment computed as in HM 3. granular ratcheting disabled. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_MindlinPhys_Mindlin(inherits LawFunctor Functor Serializable) Constitutive law for the Hertz-Mindlin formulation. It includes non linear elasticity in the normal direction as predicted by Hertz for two non-conforming elastic contact bodies. In the shear direction, instead, it reseambles the simplied case without slip discussed in Mindlins paper, where a linear relationship between shear force and tangential displacement is provided. Finally, the MohrCoulomb criterion is employed to established the maximum friction force which can be developed at the contact. Moreover, it is also possible to include the eect of linear viscous damping through the denition of the parameters n and s . bases Ordered list of types (as strings) this functor accepts. calcEnergy(=false) bool to calculate energy terms (shear potential energy, dissipation of energy due to friction and dissipation of energy due to normal and tangential damping) contactsAdhesive() oat Compute total number of adhesive contacts. dict() dict Return dictionary of attributes. frictionDissipation(=uninitalized ) Energy dissipation due to sliding includeAdhesion(=false) bool to include the adhesion force following the DMT formulation. If true, also the normal elastic energy takes into account the adhesion eect. includeMoment(=false) bool to consider rolling resistance (if Ip2_FrictMat_FrictMat_MindlinPhys::eta is 0.0, no plastic condition is applied.) label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier).
351
neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) normDampDissip(=uninitalized ) Energy dissipated by normal damping normElastEnergy() oat Compute normal elastic potential energy. It handles the DMT formulation if Law2_ScGeom_MindlinPhys_Mindlin::includeAdhesion is set to true. preventGranularRatcheting(=true) bool to avoid granular ratcheting ratioSlidingContacts() oat Return the ratio between the number of contacts sliding to the total number at a given time. shearDampDissip(=uninitalized ) Energy dissipated by tangential damping shearEnergy(=uninitalized ) Shear elastic potential energy timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_MindlinPhys_MindlinDeresiewitz(inherits LawFunctor Functor Serializable) Hertz-Mindlin contact law with partial slip solution, as described in [Thornton1991]. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_ViscElPhys_Basic(inherits LawFunctor Functor Serializable) Linear viscoelastic model operating on ScGeom and ViscElPhys. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier).
352
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGeom_ViscoFrictPhys_CundallStrack(inherits Law2_ScGeom_FrictPhys_CundallStrack LawFunctor Functor Serializable) Law similar to Law2_ScGeom_FrictPhys_CundallStrack with the addition of shear creep at contacts. bases Ordered list of types (as strings) this functor accepts. creepStiffness(=1) dict() dict Return dictionary of attributes. elasticEnergy() oat Compute and return the total elastic energy in all FrictPhys contacts initPlasticDissipation((oat)arg2 ) None Initialize cummulated plastic dissipation to a value (0 by default). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) plasticDissipation() oat Total energy dissipated in plastic slips at all FrictPhys contacts. Computed only if Law2_ScGeom_FrictPhys_CundallStrack::traceEnergy is true. shearCreep(=false) sphericalBodies(=true) If true, compute branch vectors from radii (faster), else use contactPoint-position. Turning this ag true is safe for sphere-sphere contacts and a few other specic cases. It will give wrong values of torques on facets or boxes. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. traceEnergy(=false) Dene the total energy dissipated in plastic slips at all contacts. This will trace only plastic energy in this law, see O.trackEnergy for a more complete energies tracing updateAttrs((dict)arg2 ) None Update object attributes from given dictionary viscosity(=1) class yade.wrapper.Law2_ScGeom_WirePhys_WirePM(inherits LawFunctor Functor Serializable) Constitutive law for the wire model. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
353
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). linkThresholdIteration(=1) Iteration to create the link. timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGridCoGeom_CohFrictPhys_CundallStrack(inherits LawFunctor Functor Serializable) Law between a cohesive frictional GridConnection and a cohesive frictional Sphere. Almost the same than Law2_ScGeom6D_CohFrictPhys_CohesionMoment, but THE ROTATIONAL MOMENTS ARE NOT COMPUTED. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Law2_ScGridCoGeom_FrictPhys_CundallStrack(inherits Functor Law between a frictional GridConnection and a frictional Sphere. Almost ScGeom_FrictPhys_CundallStrack, but the force is divided and applied only. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). neverErase(=false) Keep interactions even if particles go away from each other (only in case another constitutive law is in the scene, e.g. Law2_ScGeom_CapillaryPhys_Capillarity) timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary LawFunctor Serializable) the same than Law2_on the two GridNodes
354
7.8.2 LawDispatcher
class yade.wrapper.LawDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((IGeom)arg2, (IPhys)arg3 ) LawFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.9 Callbacks
IntrCallback
S umIntrForces Cb
class yade.wrapper.IntrCallback(inherits Serializable) Abstract callback object which will be called for every (real) Interaction after the interaction has been processed by InteractionLoop. At the beginning of the interaction loop, stepInit is called, initializing the object; it returns either NULL (to deactivate the callback during this time step) or pointer to function, which will then be passed (1) pointer to the callback object itself and (2) pointer to Interaction.
7.9. Callbacks
355
Note: (NOT YET DONE) This functionality is accessible from python by passing 4th argument to InteractionLoop constructor, or by appending the callback object to InteractionLoop::callbacks. dict() dict Return dictionary of attributes. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.SumIntrForcesCb(inherits IntrCallback Serializable) Callback summing magnitudes of forces over all interactions. IPhys of interactions must derive from NormShearPhys (responsability fo the user). dict() dict Return dictionary of attributes. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.10 Preprocessors
CapillaryTriaxialTes t
Cohes iveTriaxialTes t FileGenerator S impleS hear
TriaxialTes t
class yade.wrapper.FileGenerator(inherits Serializable) Base class for scene generators, preprocessors. dict() dict Return dictionary of attributes. generate((str)out ) None Generate scene, save to given le load() None Generate scene, save to temporary le and load immediately updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.CapillaryTriaxialTest(inherits FileGenerator Serializable) This preprocessor is a variant of TriaxialTest, including the model of capillary forces developed as part of the PhD of Luc Scholts. See the documentation of Law2_ScGeom_CapillaryPhys_Capillarity or the main page https://yade-dem.org/wiki/CapillaryTriaxialTest, for more details. Results obtained with this preprocessor were reported for instance in Scholtes et al. Micromechanics of granular materials with capillary eects. International Journal of Engineering Science 2009,(47)1, 64-75. Key(= ) A code that is added to output lenames.
356
Rdispersion(=0.3 ) Normalized standard deviation of generated sizes. StabilityCriterion(=0.01) Value of unbalanced force for which the system is considered stable. Used in conditionals to switch between loading stages. WallStressRecordFile(=./WallStressesWater+Key) autoCompressionActivation(=true) Do we just want to generate a stable packing under isotropic pressure (false) or do we want the triaxial loading to start automatically right after compaction stage (true)? autoStopSimulation(=false) freeze the simulation when conditions are reached (dont activate this if you want to be able to run/stop from Qt GUI) autoUnload(=true) auto adjust the isotropic stress state from TriaxialTest::sigmaIsoCompaction to TriaxialTest::sigmaLateralConnement if they have dierent values. See docs for TriaxialCompressionEngine::autoUnload biaxial2dTest(=false) FIXME : what is that? binaryFusion(=true) Denes how overlapping bridges aect the capillary forces (see CapillaryTriaxialTest::fusionDetection). If binary=true, the force is null as soon as there is an overlap detected, if not, the force is divided by the number of overlaps. boxFrictionDeg(=0.0 ) Friction angle [] of boundaries contacts. boxKsDivKn(=0.5 ) Ratio of shear vs. normal contact stiness for boxes. boxWalls(=true) Use boxes for boundaries (recommended). boxYoungModulus(=15000000.0 ) Stiness of boxes. capillaryPressure(=0 ) Dene succion in the packing [Pa]. This is the value used in the capillary model. capillaryStressRecordFile(=./capStresses+Key) compactionFrictionDeg(=sphereFrictionDeg) Friction angle [] of spheres during compaction (dierent values result in dierent porosities)]. This value is overridden by TriaxialTest::sphereFrictionDeg before triaxial testing. contactStressRecordFile(=./contStresses+Key) dampingForce(=0.2 ) Coecient of Cundal-Non-Viscous damping (applied on on the 3 components of forces) dampingMomentum(=0.2 ) Coecient of Cundal-Non-Viscous damping (applied on on the 3 components of torques) defaultDt(=0.0001) Max time-step. Used as initial value if dened. Latter adjusted by the time stepper. density(=2600 ) density of spheres dict() dict Return dictionary of attributes. facetWalls(=false) Use facets for boundaries (not tested)
7.10. Preprocessors
357
finalMaxMultiplier(=1.001) max multiplier of diameters during internal compaction (secondary precise adjustment) fixedBoxDims(= ) string that contains some subset (max. 2) of {x,y,z} ; contains axes will have box dimension hardcoded, even if box is scaled as mean_radius is prescribed: scaling will be applied on the rest. fixedPoroCompaction(=false) ag to choose an isotropic compaction until a xed porosity choosing a same translation speed for the six walls fixedPorosity(=1) FIXME : what is that? fusionDetection(=false) test overlaps between liquid bridges on modify forces if overlaps exist generate((str)out ) None Generate scene, save to given le importFilename(= ) File with positions and sizes of spheres. internalCompaction(=false) ag for choosing between moving boundaries or increasing particles sizes during the compaction stage. load() None Generate scene, save to temporary le and load immediately lowerCorner(=Vector3r(0, 0, 0)) Lower corner of the box. maxMultiplier(=1.01) max multiplier of diameters during internal compaction (initial fast increase) maxWallVelocity(=10 ) max velocity of boundaries. Usually useless, but can help stabilizing the system in some cases. noFiles(=false) Do not create any les during run (.xml, .spheres, wall stress records) numberOfGrains(=400 ) Number of generated spheres. radiusControlInterval(=10 ) interval between size changes when growing spheres. radiusMean(=-1) Mean radius. If negative (default), autocomputed to as a function of box size and TriaxialTest::numberOfGrains recordIntervalIter(=20 ) interval between le outputs sigmaIsoCompaction(=50000 ) Conning stress during isotropic compaction. sigmaLateralConfinement(=50000 ) Lateral stress during triaxial loading. An isotropic unloading is performed if the value is not equal to CapillaryTriaxialTest::SigmaIsoCompaction. sphereFrictionDeg(=18.0 ) Friction angle [] of spheres assigned just before triaxial testing. sphereKsDivKn(=0.5 ) Ratio of shear vs. normal contact stiness for spheres.
358
sphereYoungModulus(=15000000.0 ) Stiness of spheres. strainRate(=1) Strain rate in triaxial loading. thickness(=0.001) thickness of boundaries. It is arbitrary and should have no eect timeStepOutputInterval(=50 ) interval for outputing general information on the simulation (stress,unbalanced force,...) timeStepUpdateInterval(=50 ) interval for GlobalStinessTimeStepper updateAttrs((dict)arg2 ) None Update object attributes from given dictionary upperCorner(=Vector3r(1, 1, 1)) Upper corner of the box. wallOversizeFactor(=1.3 ) Make boundaries larger than the packing to make sure spheres dont go out during deformation. wallStiffnessUpdateInterval(=10 ) interval for updating the stiness of sample/boundaries contacts wallWalls(=false) Use walls for boundaries (not tested) water(=true) activate capillary model class yade.wrapper.CohesiveTriaxialTest(inherits FileGenerator Serializable) This preprocessor is a variant of TriaxialTest using the cohesive-frictional contact law with moments. It sets up a scene for cohesive triaxial tests. See full documentation at http://yadedem.org/wiki/TriaxialTest. Cohesion is initially 0 by default. The suggested usage is to dene cohesion values in a second step, after isotropic compaction : dene shear and normal cohesions in Ip2_CohFrictMat_CohFrictMat_CohFrictPhys, then turn Ip2_CohFrictMat_CohFrictMat_CohFrictPhys::setCohesionNow true to assign them at each contact at next iteration. Key(= ) A code that is added to output lenames. StabilityCriterion(=0.01) Value of unbalanced force for which the system is considered stable. Used in conditionals to switch between loading stages. WallStressRecordFile(=./CohesiveWallStresses+Key) autoCompressionActivation(=true) Do we just want to generate a stable packing under isotropic pressure (false) or do we want the triaxial loading to start automatically right after compaction stage (true)? autoStopSimulation(=false) freeze the simulation when conditions are reached (dont activate this if you want to be able to run/stop from Qt GUI) autoUnload(=true) auto adjust the isotropic stress state from TriaxialTest::sigmaIsoCompaction to TriaxialTest::sigmaLateralConnement if they have dierent values. See docs for TriaxialCompressionEngine::autoUnload biaxial2dTest(=false) FIXME : what is that?
7.10. Preprocessors
359
boxFrictionDeg(=0.0 ) Friction angle [] of boundaries contacts. boxKsDivKn(=0.5 ) Ratio of shear vs. normal contact stiness for boxes. boxWalls(=true) Use boxes for boundaries (recommended). boxYoungModulus(=15000000.0 ) Stiness of boxes. compactionFrictionDeg(=sphereFrictionDeg) Friction angle [] of spheres during compaction (dierent values result in dierent porosities)]. This value is overridden by TriaxialTest::sphereFrictionDeg before triaxial testing. dampingForce(=0.2 ) Coecient of Cundal-Non-Viscous damping (applied on on the 3 components of forces) dampingMomentum(=0.2 ) Coecient of Cundal-Non-Viscous damping (applied on on the 3 components of torques) defaultDt(=0.001) Max time-step. Used as initial value if dened. Latter adjusted by the time stepper. density(=2600 ) density of spheres dict() dict Return dictionary of attributes. facetWalls(=false) Use facets for boundaries (not tested) finalMaxMultiplier(=1.001) max multiplier of diameters during internal compaction (secondary precise adjustment) fixedBoxDims(= ) string that contains some subset (max. 2) of {x,y,z} ; contains axes will have box dimension hardcoded, even if box is scaled as mean_radius is prescribed: scaling will be applied on the rest. fixedPoroCompaction(=false) ag to choose an isotropic compaction until a xed porosity choosing a same translation speed for the six walls fixedPorosity(=1) FIXME : what is that? generate((str)out ) None Generate scene, save to given le importFilename(= ) File with positions and sizes of spheres. internalCompaction(=false) ag for choosing between moving boundaries or increasing particles sizes during the compaction stage. load() None Generate scene, save to temporary le and load immediately lowerCorner(=Vector3r(0, 0, 0)) Lower corner of the box. maxMultiplier(=1.01) max multiplier of diameters during internal compaction (initial fast increase) maxWallVelocity(=10 ) max velocity of boundaries. Usually useless, but can help stabilizing the system in some cases.
360
noFiles(=false) Do not create any les during run (.xml, .spheres, wall stress records) normalCohesion(=0 ) Material parameter used to dene contact strength in tension. numberOfGrains(=400 ) Number of generated spheres. radiusControlInterval(=10 ) interval between size changes when growing spheres. radiusDeviation(=0.3 ) Normalized standard deviation of generated sizes. radiusMean(=-1) Mean radius. If negative (default), autocomputed to as a function of box size and TriaxialTest::numberOfGrains recordIntervalIter(=20 ) interval between le outputs setCohesionOnNewContacts(=false) create cohesionless (False) or cohesive (True) interactions for new contacts. shearCohesion(=0 ) Material parameter used to dene shear strength of contacts. sigmaIsoCompaction(=50000 ) Conning stress during isotropic compaction. sigmaLateralConfinement(=50000 ) Lateral stress during triaxial loading. An isotropic unloading is performed if the value is not equal to TriaxialTest::sigmaIsoCompaction. sphereFrictionDeg(=18.0 ) Friction angle [] of spheres assigned just before triaxial testing. sphereKsDivKn(=0.5 ) Ratio of shear vs. normal contact stiness for spheres. sphereYoungModulus(=15000000.0 ) Stiness of spheres. strainRate(=0.1) Strain rate in triaxial loading. thickness(=0.001) thickness of boundaries. It is arbitrary and should have no eect timeStepUpdateInterval(=50 ) interval for GlobalStinessTimeStepper updateAttrs((dict)arg2 ) None Update object attributes from given dictionary upperCorner(=Vector3r(1, 1, 1)) Upper corner of the box. wallOversizeFactor(=1.3 ) Make boundaries larger than the packing to make sure spheres dont go out during deformation. wallStiffnessUpdateInterval(=10 ) interval for updating the stiness of sample/boundaries contacts wallWalls(=false) Use walls for boundaries (not tested) class yade.wrapper.SimpleShear(inherits FileGenerator Serializable) Preprocessor for creating a numerical model of a simple shear box. 7.10. Preprocessors 361
Boxes (6) constitute the dierent sides of the box itself Spheres are contained in the box. The sample is generated by default via the same method used in TriaxialTest Preprocesor (=> see in source function GenerateCloud). But import of a list of spheres from a text le can be also performed after few changes in the source code. Launching this preprocessor will carry out an oedometric compression, until a value of normal stress equal to 2 MPa (and stable). But with others Engines KinemCNDEngine, KinemCNSEngine and KinemCNLEngine, respectively constant normal displacement, constant normal rigidity and constant normal stress paths can be carried out for such simple shear boxes. NB about micro-parameters : their default values correspond to those used in [Duriez2009a] and [Duriez2011] to simulate inlled rock joints. boxPoissonRatio(=0.04) value of ElastMat::poisson for the spheres [-] boxYoungModulus(=4.0e9 ) value of ElastMat::young for the boxes [Pa] density(=2600 ) density of the spheres [kg/m3 ] dict() dict Return dictionary of attributes. generate((str)out ) None Generate scene, save to given le gravApplied(=false) depending on this, GravityEngine is added or not to the scene to take into account the weight of particles gravity(=Vector3r(0, -9.81, 0)) vector corresponding to used gravity (if gravApplied) [m/s2 ] height(=0.02 ) initial height (along y-axis) of the shear box [m] length(=0.1) initial length (along x-axis) of the shear box [m] load() None Generate scene, save to temporary le and load immediately sphereFrictionDeg(=37 ) value of ElastMat::poisson for the spheres [ ] (the necessary conversion in rad is done automatically) spherePoissonRatio(=0.04) value of ElastMat::poisson for the spheres [-] sphereYoungModulus(=4.0e9 ) value of ElastMat::young for the spheres [Pa] thickness(=0.001) thickness of the boxes constituting the shear box [m] timeStepUpdateInterval(=50 ) value of TimeStepper::timeStepUpdateInterval for the TimeStepper used here updateAttrs((dict)arg2 ) None Update object attributes from given dictionary width(=0.04) initial width (along z-axis) of the shear box [m]
362
class yade.wrapper.TriaxialTest(inherits FileGenerator Serializable) Create a scene for triaxal test. Introduction Yade includes tools to simulate triaxial tests on particles assemblies. This preprocessor (and variants like e.g. CapillaryTriaxialTest) illustrate how to use them. It generates a scene which will - by default - go through the following steps : generate random loose packings in a parallelepiped. compress the packing isotropicaly, either squeezing the packing between moving rigid boxes or expanding the particles while boxes are xed (depending on ag internalCompaction). The conning pressure in this stage is dened via sigmaIsoCompaction. when the packing is dense and stable, simulate a loading path and get the mechanical response as a result. The default loading path corresponds to a constant lateral stress (sigmaLateralConnement) in 2 directions and constant strain rate on the third direction. This default loading path is performed when the ag autoCompressionActivation it True, otherwise the simulation stops after isotropic compression. Dierent loading paths might be performed. In order to dene them, the user can modify the ags found in engine TriaxialStressController at any point in the simulation (in c++). If TriaxialStressController.wall_X_activated is true boundary X is moved automatically to maintain the dened stress level sigmaN (see axis conventions below). If false the boundary is not controlled by the engine at all. In that case the user is free to prescribe xed position, constant velocity, or more complex conditions. Note: Axis conventions. Boundaries perpendicular to the x axis are called left and right, y corresponds to top and bottom, and axis z to front and back. In the default loading path, strain rate is assigned along y, and constant stresses are assigned on x and z. Essential engines 1. The TrixaialCompressionEngine is used for controlling the state of the sample and simulating loading paths. TriaxialCompressionEngine inherits from TriaxialStressController, which computes stress- and strain-like quantities in the packing and maintain a constant level of stress at each boundary. TriaxialCompressionEngine has few more members in order to impose constant strain rate and control the transition between isotropic compression and triaxial test. Transitions are dened by changing some ags of the TriaxialStressController, switching from/to imposed strain rate to/from imposed stress. 2. The class TriaxialStateRecorder is used to write to a le the history of stresses and strains. 3. TriaxialTest is using GlobalStinessTimeStepper to compute an appropriate t for the numerical scheme. Note: TriaxialStressController::ComputeUnbalancedForce returns a value that can be useful for evaluating the stability of the packing. It is dened as (mean force on particles)/(mean contact force), so that it tends to 0 in a stable packing. This parameter is checked by TriaxialCompressionEngine to switch from one stage of the simulation to the next one (e.g. stop isotropic connment and start axial loading)
Frequently Asked Questions
1.How is generated the packing? How to change particles sizes distribution? Why do I have a m The initial positioning of spheres is done by generating random (x,y,z) in a box and checking if a sphere of radius R (R also randomly generated with respect to a uniform distribution between mean*(1-std_dev) and mean*(1+std_dev) can be inserted at this location without overlaping with others.
7.10. Preprocessors
363
If the sphere overlaps, new (x,y,z)s are generated until a free position for the new sphere is found. This explains the message you have: after 3000 trial-and-error, the sphere couldnt be placed, and the algorithm stops. You get the message above if you try to generate an initialy dense packing, which is not possible with this algorithm. It can only generate clouds. You should keep the default value of porosity (n~0.7), or even increase if it is still to low in some cases. The dense state will be obtained in the second step (compaction, see below).
2.How is the compaction done, what are the parameters maxWallVelocity and nalMaxMultiplie Compaction is done (a) by moving rigid boxes or (b) by increasing the sizes of the particles (decided using the option internalCompaction size increase). Both algorithm needs numerical parameters to prevent instabilities. For instance, with the method (1) maxWallVelocity is the maximum wall velocity, with method (2) nalMaxMultiplier is the max value of the multiplier applied on sizes at each iteration (always something like 1.00001).
3.During the simulation of triaxial compression test, the wall in one direction moves with an incr The control of stress on a boundary is based on the total stiness K of all contacts between the packing and this boundary. In short, at each step, displacement=stress_error/K. This algorithm is implemented in TriaxialStressController, and the control itself is in TriaxialStressController::ControlExternalStress. The control can be turned o independently for each boundary, using the ags wall_XXX_activated, with XXX {top, bottom, left, right, back, front }. The imposed sress is a unique value (sigma_iso) for all directions if TriaxialStressController.isAxisymetric, or 3 independent values sigma1, sigma2, sigma3. 4.Which value of friction angle do you use during the compaction phase of the Triaxial Test? The friction during the compaction (whether you are using the expansion method or the compression one for the specimen generation) can be anything between 0 and the nal value used during the Triaxial phase. Note that higher friction than the nal one would result in volumetric collapse at the beginning of the test. The purpose of using a dierent value of friction during this phase is related to the fact that the nal porosity you get at the end of the sample generation essentially depends on it as well as on the assumed Particle Size Distribution. Changing the initial value of friction will get to a dierent value of the nal porosity. 5.Which is the aim of the bool isRadiusControlIteration? This internal variable (updated automatically) is true each N timesteps (with N =radiusControlInterval). For other timesteps, there is no expansion. Cycling without expanding is just a way to speed up the simulation, based on the idea that 1% increase each 10 iterations needs less operations than 0.1% at each iteration, but will give similar results.
6.How comes the unbalanced force reaches a low value only after many timesteps in the compact The value of unbalanced force (dimensionless) is expected to reach low value (i.e. identifying a static-equilibrium condition for the specimen) only at the end of the compaction phase. The code is not aiming at simulating a quasistatic isotropic compaction process, it is only giving a stable packing at the end of it. Key(= ) A code that is added to output lenames. StabilityCriterion(=0.01) Value of unbalanced force for which the system is considered stable. Used in conditionals to switch between loading stages. WallStressRecordFile(=./WallStresses+Key)
364
autoCompressionActivation(=true) Do we just want to generate a stable packing under isotropic pressure (false) or do we want the triaxial loading to start automatically right after compaction stage (true)? autoStopSimulation(=false) freeze the simulation when conditions are reached (dont activate this if you want to be able to run/stop from Qt GUI) autoUnload(=true) auto adjust the isotropic stress state from TriaxialTest::sigmaIsoCompaction to TriaxialTest::sigmaLateralConnement if they have dierent values. See docs for TriaxialCompressionEngine::autoUnload biaxial2dTest(=false) FIXME : what is that? boxFrictionDeg(=0.0 ) Friction angle [] of boundaries contacts. boxKsDivKn(=0.5 ) Ratio of shear vs. normal contact stiness for boxes. boxYoungModulus(=15000000.0 ) Stiness of boxes. compactionFrictionDeg(=sphereFrictionDeg) Friction angle [] of spheres during compaction (dierent values result in dierent porosities)]. This value is overridden by TriaxialTest::sphereFrictionDeg before triaxial testing. dampingForce(=0.2 ) Coecient of Cundal-Non-Viscous damping (applied on on the 3 components of forces) dampingMomentum(=0.2 ) Coecient of Cundal-Non-Viscous damping (applied on on the 3 components of torques) defaultDt(=-1) Max time-step. Used as initial value if dened. Latter adjusted by the time stepper. density(=2600 ) density of spheres dict() dict Return dictionary of attributes. facetWalls(=false) Use facets for boundaries (not tested) finalMaxMultiplier(=1.001) max multiplier of diameters during internal compaction (secondary precise adjustment) fixedBoxDims(= ) string that contains some subset (max. 2) of {x,y,z} ; contains axes will have box dimension hardcoded, even if box is scaled as mean_radius is prescribed: scaling will be applied on the rest. generate((str)out ) None Generate scene, save to given le importFilename(= ) File with positions and sizes of spheres. internalCompaction(=false) ag for choosing between moving boundaries or increasing particles sizes during the compaction stage. load() None Generate scene, save to temporary le and load immediately lowerCorner(=Vector3r(0, 0, 0)) Lower corner of the box. 7.10. Preprocessors 365
maxMultiplier(=1.01) max multiplier of diameters during internal compaction (initial fast increase) maxWallVelocity(=10 ) max velocity of boundaries. Usually useless, but can help stabilizing the system in some cases. noFiles(=false) Do not create any les during run (.xml, .spheres, wall stress records) numberOfGrains(=400 ) Number of generated spheres. radiusControlInterval(=10 ) interval between size changes when growing spheres. radiusMean(=-1) Mean radius. If negative (default), autocomputed to as a function of box size and TriaxialTest::numberOfGrains radiusStdDev(=0.3 ) Normalized standard deviation of generated sizes. recordIntervalIter(=20 ) interval between le outputs sigmaIsoCompaction(=50000 ) Conning stress during isotropic compaction. sigmaLateralConfinement(=50000 ) Lateral stress during triaxial loading. An isotropic unloading is performed if the value is not equal to TriaxialTest::sigmaIsoCompaction. sphereFrictionDeg(=18.0 ) Friction angle [] of spheres assigned just before triaxial testing. sphereKsDivKn(=0.5 ) Ratio of shear vs. normal contact stiness for spheres. sphereYoungModulus(=15000000.0 ) Stiness of spheres. strainRate(=0.1) Strain rate in triaxial loading. thickness(=0.001) thickness of boundaries. It is arbitrary and should have no eect timeStepUpdateInterval(=50 ) interval for GlobalStinessTimeStepper updateAttrs((dict)arg2 ) None Update object attributes from given dictionary upperCorner(=Vector3r(1, 1, 1)) Upper corner of the box. wallOversizeFactor(=1.3 ) Make boundaries larger than the packing to make sure spheres dont go out during deformation. wallStiffnessUpdateInterval(=10 ) interval for updating the stiness of sample/boundaries contacts wallWalls(=false) Use walls for boundaries (not tested)
366
7.11 Rendering
7.11.1 OpenGLRenderer
class yade.wrapper.OpenGLRenderer(inherits Serializable) Class responsible for rendering scene on OpenGL devices. bgColor(=Vector3r(.2, .2, .2)) Color of the background canvas (RGB) bound(=false) Render body Bound cellColor(=Vector3r(1, 1, 0)) Color of the periodic cell (RGB). clipPlaneActive(=vector<bool>(numClipPlanes, false)) Activate/deactivate respective clipping planes clipPlaneSe3(=vector<Se3r>(numClipPlanes, nionr::Identity()))) Position and orientation of clipping planes dict() dict Return dictionary of attributes. dispScale(=Vector3r::Ones(), disable scaling) Articially enlarge (scale) dispalcements from bodies reference positions by this relative amount, so that they become better visible (independently in 3 dimensions). Disbled if (1,1,1). dof(=false) Show which degrees of freedom are blocked for each body extraDrawers(=uninitalized ) Additional rendering components (GlExtraDrawer). ghosts(=true) Render objects crossing periodic cell edges by cloning them in multiple places (periodic simulations only). id(=false) Show body ids intrAllWire(=false) Draw wire for all interactions, blue for potential and green for real ones (mostly for debugging) intrGeom(=false) Render Interaction::geom objects. intrPhys(=false) Render Interaction::phys objects intrWire(=false) If rendering interactions, use only wires to represent them. light1(=true) Turn light 1 on. light2(=true) Turn light 2 on. light2Color(=Vector3r(0.5, 0.5, 0.1)) Per-color intensity of secondary light (RGB). light2Pos(=Vector3r(-130, 75, 30)) Position of secondary OpenGL light source in the scene. lightColor(=Vector3r(0.6, 0.6, 0.6)) Per-color intensity of primary light (RGB). Se3r(Vector3r::Zero(), Quater-
7.11. Rendering
367
lightPos(=Vector3r(75, 130, 0)) Position of OpenGL light source in the scene. mask(=~0, draw everything) Bitmask for showing only bodies where ((mask & Body::mask)!=0) render() None Render the scene in the current OpenGL context. rotScale(=1., disable scaling) Articially enlarge (scale) rotations of bodies relative to their reference orientation, so the they are better visible. selId(=Body::ID_NONE ) Id of particle that was selected by the user. setRefSe3() None Make current positions and orientation reference for scaleDisplacements and scaleRotations. shape(=true) Render body Shape updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire(=false) Render all bodies with wire only (faster)
7.11.2 GlShapeFunctor
Gl1_S phere
Gl1_Tetra
Gl1_W all
GlS hapeFunctor
Gl1_GridConnection
Gl1_Box
Gl1_Cylinder
Gl1_ChainedCylinder
Gl1_Facet
class yade.wrapper.GlShapeFunctor(inherits Functor Serializable) Abstract functor for rendering Shape objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes.
368
label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_Box(inherits GlShapeFunctor Functor Serializable) Renders Box object bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_ChainedCylinder(inherits Gl1_Cylinder GlShapeFunctor Functor Serializable) Renders ChainedCylinder object including a shift for compensating exion. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. glutNormalize = True glutSlices = 8 glutStacks = 4 label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire = False class yade.wrapper.Gl1_Cylinder(inherits GlShapeFunctor Functor Serializable) Renders Cylinder object wire(=false) [static] Only show wireframe (controlled by glutSlices and glutStacks. glutNormalize(=true) [static] Fix normals for non-wire rendering glutSlices(=8 ) [static] Number of sphere slices.
7.11. Rendering
369
glutStacks(=4) [static] Number of sphere stacks. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. glutNormalize = True glutSlices = 8 glutStacks = 4 label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire = False class yade.wrapper.Gl1_Facet(inherits GlShapeFunctor Functor Serializable) Renders Facet object normals(=false) [static] In wire mode, render normals of facets and edges; facets colors are disregarded in that case. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). normals = False timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_GridConnection(inherits GlShapeFunctor Functor Serializable) Renders Cylinder object wire(=false) [static] Only show wireframe (controlled by glutSlices and glutStacks. glutNormalize(=true) [static] Fix normals for non-wire rendering glutSlices(=8 ) [static] Number of cylinder slices. glutStacks(=4) [static] Number of cylinder stacks. bases Ordered list of types (as strings) this functor accepts.
370
dict() dict Return dictionary of attributes. glutNormalize = True glutSlices = 8 glutStacks = 4 label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire = False class yade.wrapper.Gl1_Sphere(inherits GlShapeFunctor Functor Serializable) Renders Sphere object quality(=1.0 ) [static] Change discretization level of spheres. quality>1 for better image quality, at the price of more cpu/gpu usage, 0<quality<1 for faster rendering. If mono-color spheres are displayed (Gl1_Sphere::stripes = False), quality mutiplies Gl1_Sphere::glutSlices and Gl1_Sphere::glutStacks. If striped spheres are displayed (Gl1_Sphere::stripes = True), only integer increments are meaningfull : quality=1 and quality=1.9 will give the same result, quality=2 will give ner result. wire(=false) [static] Only show wireframe (controlled by glutSlices and glutStacks. stripes(=false) [static] In non-wire rendering, show stripes clearly showing particle rotation. localSpecView(=true) [static] Compute specular light in local eye coordinate system. glutSlices(=12 ) [static] Base number of sphere slices, multiplied by Gl1_Sphere::quality before use); not used with stripes (see glut{Solid,Wire}Sphere reference) glutStacks(=6 ) [static] Base number of sphere stacks, multiplied by Gl1_Sphere::quality before use; not used with stripes (see glut{Solid,Wire}Sphere reference) bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. glutSlices = 12 glutStacks = 6 label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). localSpecView = True quality = 1.0 stripes = False
7.11. Rendering
371
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary wire = False class yade.wrapper.Gl1_Tetra(inherits GlShapeFunctor Functor Serializable) Renders Tetra object bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_Wall(inherits GlShapeFunctor Functor Serializable) Renders Wall object div(=20 ) [static] Number of divisions of the wall inside visible scene part. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. div = 20 label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.11.3 GlStateFunctor
class yade.wrapper.GlStateFunctor(inherits Functor Serializable) Abstract functor for rendering State objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier).
372
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.11.4 GlBoundFunctor
GlBoundFunctor
Gl1_Aabb
class yade.wrapper.GlBoundFunctor(inherits Functor Serializable) Abstract functor for rendering Bound objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_Aabb(inherits GlBoundFunctor Functor Serializable) Render Axis-aligned bounding box (Aabb). bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.11. Rendering
373
7.11.5 GlIGeomFunctor
Gl1_L3Geom
Gl1_L6Geom
Gl1_Dem3DofGeom_S phereS phere GlIGeomFunctor Gl1_Dem3DofGeom_FacetS phere
Gl1_Dem3DofGeom_W allS phere
class yade.wrapper.GlIGeomFunctor(inherits Functor Serializable) Abstract functor for rendering IGeom objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_Dem3DofGeom_FacetSphere(inherits GlIGeomFunctor Functor Serializable) Render interaction of facet and sphere (represented by Dem3DofGeom_FacetSphere) normal(=false) [static] Render interaction normal rolledPoints(=false) [static] Render points rolled on the sphere & facet (original contact point) unrolledPoints(=false) [static] Render original contact points unrolled to the contact plane shear(=false) [static] Render shear line in the contact plane shearLabel(=false) [static] Render shear magnitude as number bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). normal = False
374
rolledPoints = False shear = False shearLabel = False timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. unrolledPoints = False updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_Dem3DofGeom_SphereSphere(inherits GlIGeomFunctor Functor Serializable) Render interaction of 2 spheres (represented by Dem3DofGeom_SphereSphere) normal(=false) [static] Render interaction normal rolledPoints(=false) [static] Render points rolled on the spheres (tracks the original contact point) unrolledPoints(=false) [static] Render original contact points unrolled to the contact plane shear(=false) [static] Render shear line in the contact plane shearLabel(=false) [static] Render shear magnitude as number bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). normal = False rolledPoints = False shear = False shearLabel = False timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. unrolledPoints = False updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_Dem3DofGeom_WallSphere(inherits GlIGeomFunctor Functor Serializable) Render interaction of wall and sphere (represented by Dem3DofGeom_WallSphere) normal(=false) [static] Render interaction normal rolledPoints(=false) [static] Render points rolled on the spheres (tracks the original contact point) unrolledPoints(=false) [static] Render original contact points unrolled to the contact plane
7.11. Rendering
375
shear(=false) [static] Render shear line in the contact plane shearLabel(=false) [static] Render shear magnitude as number bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). normal = False rolledPoints = False shear = False shearLabel = False timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. unrolledPoints = False updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_L3Geom(inherits GlIGeomFunctor Functor Serializable) Render L3Geom geometry. axesLabels(=false) [static] Whether to display labels for local axes (x,y,z) axesScale(=1.) [static] Scale local axes, their reference length being half of the minimum radius. axesWd(=1.) [static] Width of axes lines, in pixels; not drawn if non-positive uPhiWd(=2.) [static] Width of lines for drawing displacements (and rotations for L6Geom); not drawn if nonpositive. uScale(=1.) [static] Scale local displacements (u - u0); 1 means the true scale, 0 disables drawing local displacements; negative values are permissible. axesLabels = False axesScale = 1.0 axesWd = 1.0 bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True.
376
uPhiWd = 2.0 uScale = 1.0 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_L6Geom(inherits Gl1_L3Geom GlIGeomFunctor Functor Serializable) Render L6Geom geometry. phiScale(=1.) [static] Scale local rotations (phi - phi0). The default scale is to draw rotation with length equal to minimum radius. axesLabels = False axesScale = 1.0 axesWd = 1.0 bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). phiScale = 1.0 timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. uPhiWd = 2.0 uScale = 1.0 updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.11.6 GlIPhysFunctor
Gl1_CpmPhys GlIPhys Functor Gl1_NormPhys
class yade.wrapper.GlIPhysFunctor(inherits Functor Serializable) Abstract functor for rendering IPhys objects. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier).
7.11. Rendering
377
timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_CpmPhys(inherits GlIPhysFunctor Functor Serializable) Render CpmPhys objects of interactions. contactLine(=true) [static] Show contact line dmgLabel(=true) [static] Numerically show contact damage parameter dmgPlane(=false) [static] [what is this?] epsT(=false) [static] Show shear strain epsTAxes(=false) [static] Show axes of shear plane normal(=false) [static] Show contact normal colorStrainRatio(=-1) [static] If positive, set the interaction (wire) color based on N normalized by 0 colorStrainRatio (0 = CpmPhys.epsCrackOnset ). Otherwise, color based on the residual strength. epsNLabel(=false) [static] Numerically show normal strain bases Ordered list of types (as strings) this functor accepts. colorStrainRatio = -1.0 contactLine = True dict() dict Return dictionary of attributes. dmgLabel = True dmgPlane = False epsNLabel = False epsT = False epsTAxes = False label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). normal = False timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Gl1_NormPhys(inherits GlIPhysFunctor Functor Serializable) Renders NormPhys objects as cylinders of which diameter and color depends on NormPhys::normForce magnitude.
378
maxFn(=0 ) [static] Value of NormPhys.normalForce corresponding to maxDiameter. This value will be increased (but not decreased ) automatically. signFilter(=0 ) [static] If non-zero, only display contacts with negative (-1) or positive (+1) normal forces; if zero, all contacts will be displayed. refRadius(=std::numeric_limits<Real>::innity()) [static] Reference (minimum) particle radius; used only if maxRadius is negative. This value will be decreased (but not increased ) automatically. (auto-updated) maxRadius(=-1) [static] Cylinder radius corresponding to the maximum normal force. If negative, auto-updated refRadius will be used instead. slices(=6 ) [static] Number of sphere slices; (see glutCylinder reference) stacks(=1) [static] Number of sphere stacks; (see glutCylinder reference) maxWeakFn(=NaN ) [static] Value that divides contacts by their normal force into the weak fabric and strong fabric. This value is set as side-eect by utils.fabricTensor. weakFilter(=0 ) [static] If non-zero, only display contacts belonging to the weak (-1) or strong (+1) fabric. weakScale(=1.) [static] If maxWeakFn is set, scale radius of the weak fabric by this amount (usually smaller than 1). If zero, 1 pixel line is displayed. Colors are not aected by this value. bases Ordered list of types (as strings) this functor accepts. dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). maxFn = 0.0 maxRadius = -1.0 maxWeakFn = nan refRadius = inf signFilter = 0 slices = 6 stacks = 1 timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary weakFilter = 0 weakScale = 1.0
7.11. Rendering
379
7.12 Simulation data

7.12.1 Omega
class yade.wrapper.Omega addScene() int Add new scene to Omega, returns its number bodies Bodies in the current simulation (container supporting index access by id and iteration) cell Periodic cell of the current scene (None if the scene is aperiodic). childClassesNonrecursive((str)arg2 ) list Return list of all classes deriving from given class, as registered in the class factory disableGdb() None Revert SEGV and ABRT handlers to system defaults. dt Current timestep (t) value. dynDt Whether a TimeStepper is used for dynamic t control. See dt on how to enable/disable TimeStepper. dynDtAvailable Whether a TimeStepper is amongst O.engines, activated or not. energy EnergyTracker of the current simulation. (meaningful only with O.trackEnergy) engines List of engines in the simulation (Scene::engines). exitNoBacktrace([(int)status=0 ]) None Disable SEGV handler and exit, optionally with given status number. filename Filename under which the current simulation was saved (None if never saved). forceSyncCount Counter for number of syncs in ForceContainer, for proling purposes. forces ForceContainer (forces, torques, displacements) in the current simulation. interactions Interactions in the current simulation (container supporting index acces by either (id1,id2) or interactionNumber and iteration) isChildClassOf((str)arg2, (str)arg3 ) bool Tells whether the rst class derives from the second one (both given as strings). iter Get current step number labeledEngine((str)arg2 ) object Return instance of engine/functor with the given label. This function shouldnt be called by the user directly; every ehange in O.engines will assign respective global python variables according to labels. For example: O.engines=[InsertionSortCollider(label=collider)]
380
collider.nBins=5 # collider has become a variable after assignment to O.engines automatically load((str)le[, (bool)quiet=False ]) None Load simulation from le. The le should be saved in the same version of Yade, otherwise compatibility is not guaranteed. loadTmp([(str)mark= [, (bool)quiet=False ]]) None Load simulation previously stored in memory by saveTmp. mark optionally distinguishes multiple saved simulations lsTmp() list Return list of all memory-saved simulations. materials Shared materials; they can be accessed by id or by label miscParams MiscParams in the simulation (Scene::mistParams), usually used to save serializables that dont t anywhere else, like GL functors numThreads Get maximum number of threads openMP can use. pause() None Stop simulation execution. (May be called from within the loop, and it will stop after the current step). periodic Get/set whether the scene is periodic or not (True/False). plugins() list Return list of all plugins registered in the class factory. realtime Return clock (human world) time the simulation has been running. reload([(bool)quiet=False ]) None Reload current simulation reset() None Reset simulations completely (including another scenes!). resetAllScenes() None Reset all scenes. resetCurrentScene() None Reset current scene. resetThisScene() None Reset current scene. resetTime() None Reset simulation time: step number, virtual and real time. (Doesnt touch anything else, including timings). run([(int)nSteps=-1[, (bool)wait=False ]]) None Run the simulation. nSteps how many steps to run, then stop (if positive); wait will cause not returning to python until simulation will have stopped. runEngine((Engine)arg2 ) None Run given engine exactly once; simulation time, step number etc. will not be incremented (use only if you know what you do). running Whether background thread is currently running a simulation. save((str)le[, (bool)quiet=False ]) None Save current simulation to le (should be .xml or .xml.bz2). The le should be loaded in the same version of Yade, otherwise compatibility is not guaranteed. 7.12. Simulation data 381
saveTmp([(str)mark= [, (bool)quiet=False ]]) None Save simulation to memory (disappears at shutdown), can be loaded later with loadTmp. mark optionally distinguishes dierent memory-saved simulations. sceneToString() str Return the entire scene as a string. Equivalent to using O.save(...) except that the scene goes to a string instead of a le. (see also stringToScene()) speed Return current calculation speed [iter/sec]. step() None Advance the simulation by one step. Returns after the step will have nished. stopAtIter Get/set number of iteration after which the simulation will stop. stringToScene((str)arg2 [, (str)mark= ]) None Load simulation from a string passed as argument (see also sceneToString). subStep Get the current subStep number (only meaningful if O.subStepping==True); -1 when outside the loop, otherwise either 0 (O.subStepping==False) or number of engine to be run (O.subStepping==True) subStepping Get/set whether subStepping is active. switchScene() None Switch to alternative simulation (while keeping the old one). Calling the function again switches back to the rst one. Note that most variables from the rst simulation will still refer to the rst simulation even after the switch (e.g. b=O.bodies[4]; O.switchScene(); [b still refers to the body in the rst simulation here]) switchToScene((int)arg2 ) None Switch to dened scene. Default scene has number 0, other scenes have to be created by addScene method. tags Tags (string=string dictionary) of the current simulation (container supporting string-index access/assignment) time Return virtual (model world) time of the simulation. timingEnabled Globally enable/disable timing services (see documentation of the timing module). tmpFilename() str Return unique name of le in temporary directory which will be deleted when yade exits. tmpToFile((str)leName[, (str)mark= ]) None Save XML of saveTmpd simulation into leName. tmpToString([(str)mark= ]) str Return XML of saveTmpd simulation as string. trackEnergy When energy tracking is enabled or disabled in this simulation. wait() None Dont return until the simulation will have been paused. (Returns immediately if not running).
7.12.2 BodyContainer
class yade.wrapper.BodyContainer
382
__init__((BodyContainer)arg2 ) None addToClump((object)arg2, (int)arg3 ) None Add body b (or a list of bodies) to an existing clump c. c must be clump and b may not be a clump member of c. See /examples/clumps/addToClump-example.py for an example script. Note: If b is a clump itself, then all members will be added to c and b will be deleted. If b is a clump member of clump d, then all members from d will be added to c and d will be deleted. If you need to add just clump member b, release this member from d rst. append((Body)arg2 ) int Append one Body instance, return its id. append( (BodyContainer)arg1, (object)arg2) object : Append list of Body instance, return list of ids appendClumped((object)arg2 ) tuple Append given list of bodies as a clump (rigid aggregate); returns a tuple of (clumpId,[memberId1,memberId2,...]). Clump masses and inertia are adapted automatically. If clump members are overlapping this is done by integration/summation over mass points using a regular grid of cells. For non-overlapping members inertia of the clump is the sum of inertias from members. clear() None Remove all bodies (interactions not checked) clump((object)arg2 ) int Clump given bodies together (creating a rigid aggregate); returns clumpId. Clump masses and inertia are adapted automatically (see appendClumped()). erase((int)arg2 ) bool Erase body with the given id; all interaction will be deleted by InteractionLoop in the next step. getRoundness((list)arg2 ) oat Returns roundness coecient RC = R2/R1. R1 is the theoretical radius of a sphere, with same volume as clump. R2 is the minimum radius of a sphere, that imbeds clump. If just spheres are present RC = 1. If clumps are present 0 < RC < 1. Bodies can be excluded from the calculation by giving a list of ids: O.bodies.getRoundness([ids]). See /examples/clumps/replaceByClumps-example.py for an example script. releaseFromClump((int)arg2, (int)arg3 ) None Release body b from clump c. b must be a clump member of c. See /examples/clumps/releaseFromClump-example.py for an example script. Note: If c contains only 2 members b will not be released and a warning will appear. In this case clump c should be erased. replace((object)arg2 ) object replaceByClumps((list)arg2, (object)arg3 ) list Replace spheres by clumps using a list of clump templates and a list of amounts; returns a list of tuples: [(clumpId1,[memberId1,memberId2,...]),(clumpId2,[memberId1,memberId2,...]),...]. A new clump will have the same volume as the sphere, that was replaced (clump volume/mass/inertia is accounting for overlaps assuming that there are only pair overlaps, to adapt masses of clumps with multiple overlaps use adaptClumpMasses()).
7.12. Simulation data
383
O.bodies.replaceByClumps( [utils.clumpTemplate([1,1],[.5,.5])] , [.9] ) #will replace 90 % of all standalone spheres by dyads See /examples/clumps/replaceByClumps-example.py for an example script.
7.12.3 InteractionContainer
class yade.wrapper.InteractionContainer Access to interactions of simulation, by using 1.ids of both Bodies of the interactions, e.g. O.interactions[23,65] 2.iteraction over the whole container:
for i in O.interactions: print i.id1,i.id2
Note: Iteration silently skips interactions that are not real. __init__((InteractionContainer)arg2 ) None clear() None Remove all interactions, and invalidate persistent collider data (if the collider supports it). countReal() int Return number of interactions that are real, i.e. they have phys and geom. erase((int)arg2, (int)arg3 ) None Erase one interaction, given by id1, id2 (internally, requestErase is called the interaction might still exist as potential, if the Collider decides so). eraseNonReal() None Erase all interactions that are not real . nth((int)arg2 ) Interaction Return n-th interaction from the container (usable for picking random interaction). serializeSorted withBody((int)arg2 ) list Return list of real interactions of given body. withBodyAll((int)arg2 ) list Return list of all (real as well as non-real) interactions of given body.
7.12.4 ForceContainer
class yade.wrapper.ForceContainer __init__((ForceContainer)arg2 ) None addF((int)id, (Vector3)f [, (bool)permanent=False ]) None Apply force on body (accumulates). # If permanent=false (default), the force applies for one iteration, then it is reset by ForceResetter. # If permanent=true, it persists over iterations, until it is overwritten by another call to addF(id,f,True) or removed by reset(resetAll=True). The permanent force on a body can be checked with permF(id). addMove((int)id, (Vector3)m) None Apply displacement on body (accumulates). addRot((int)id, (Vector3)r ) None Apply rotation on body (accumulates).
384
addT((int)id, (Vector3)t [, (bool)permanent=False ]) None Apply torque on body (accumulates). # If permanent=false (default), the torque applies for one iteration, then it is reset by ForceResetter. # If permanent=true, it persists over iterations, until it is overwritten by another call to addT(id,f,True) or removed by reset(resetAll=True). The permanent torque on a body can be checked with permT(id). f((int)id [, (bool)sync=False ]) Vector3 Force applied on body. For clumps in openMP, synchronize the force container with sync=True, else the value will be wrong. getPermForceUsed() bool Check wether permanent forces are present. m((int)id [, (bool)sync=False ]) Vector3 Deprecated alias for t (torque). move((int)id ) Vector3 Displacement applied on body. permF((int)id ) Vector3 read the value of permanent force on body (set with setPermF()). permT((int)id ) Vector3 read the value of permanent torque on body (set with setPermT()). reset([(bool)resetAll=True ]) None Reset the force container, including user dened permanent forces/torques. resetAll=False will keep permanent forces/torques unchanged. rot((int)id ) Vector3 Rotation applied on body. syncCount Number of synchronizations of ForceContainer (cummulative); if signicantly higher than number of steps, there might be unnecessary syncs hurting performance. t((int)id [, (bool)sync=False ]) Vector3 Torque applied on body. For clumps in openMP, synchronize the force container with sync=True, else the value will be wrong.
7.12.5 MaterialContainer
class yade.wrapper.MaterialContainer Container for Materials. A material can be accessed using 1.numerical index in range(0,len(cont)), like cont[2]; 2.textual label that was given to the material, like cont[steel]. This etails traversing all materials and should not be used frequently. __init__((MaterialContainer)arg2 ) None append((Material)arg2 ) int Add new shared Material; changes its id and return it. append( (MaterialContainer)arg1, (object)arg2) object : Append list of Material instances, return list of ids. index((str)arg2 ) int Return id of material, given its label.
385
7.12.6 Scene
class yade.wrapper.Scene(inherits Serializable) Object comprising the whole simulation. compressionNegative Whether the convention is that compression has negative sign (set by Ig2Functor. dict() dict Return dictionary of attributes. doSort(=false) Used, when new body is added to the scene. dt(=1e-8 ) Current timestep for integration. flags(=0 ) Various ags of the scene; 1 (Scene::LOCAL_COORDS): use local coordinate system rather than global one for per-interaction quantities (set automatically from the functor). isPeriodic(=false) Whether periodic boundary conditions are active. iter(=0 ) Current iteration (computational step) number localCoords Whether local coordianate system is used on interactions (set by Ig2Functor. selectedBody(=-1) Id of body that is selected by the user speed(=0 ) Current calculation speed [iter/s] stopAtIter(=0 ) Iteration after which to stop the simulation. stopAtTime(=0 ) Time after which to stop the simulation subStep(=-1) Number of sub-step; not to be changed directly. -1 means to run loop prologue (cell integration), 0n-1 runs respective engines (n is number of engines), n runs epilogue (increment step number and time. subStepping(=false) Whether we currently advance by one engine in every step (rather than by single run through all engines). tags(=uninitalized ) Arbitrary key=value associations (tags like mp3 tags: author, date, version, description etc.) time(=0 ) Simulation time (virtual time) [s] trackEnergy(=false) Whether energies are being traced. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
7.12.7 Cell
class yade.wrapper.Cell(inherits Serializable) Parameters of periodic boundary conditions. Only applies if O.isPeriodic==True.
386
dict() dict Return dictionary of attributes. getDefGrad() Matrix3 Returns deformation gradient tensor F (http://en.wikipedia.org/wiki/Finite_strain_theory) of the cell deformation
getEulerianAlmansiStrain() Matrix3 1 Returns Eulerian-Almansi strain tensor e = 1 ) = 2 (I b (http://en.wikipedia.org/wiki/Finite_strain_theory) getLCauchyGreenDef() Matrix3 Returns left Cauchy-Green deformation tensor (http://en.wikipedia.org/wiki/Finite_strain_theory) getLagrangianStrain() Matrix3 Returns Lagrangian strain tensor E = 1 2 ( C I) = (http://en.wikipedia.org/wiki/Finite_strain_theory) b
1 2 (I
(FFT )1 ) of the cell
FFT
of
the
cell
1 T 2 (F F
I) =
2 1 2 (U
I) of the cell
getLeftStretch() Matrix3 Returns left (spatial) stretch tensor of the cell (matrix U from polar decomposition F = RU ) getPolarDecOfDefGrad() tuple Returns orthogonal matrix R and symmetric positive semi-denite matrix U as polar decomposition of deformation gradient F of the cell ( F = RU ) getRCauchyGreenDef() Matrix3 Returns right Cauchy-Green deformation tensor (http://en.wikipedia.org/wiki/Finite_strain_theory) C = FT F of the cell
getRightStretch() Matrix3 Returns right (material) stretch tensor of the cell (matrix V from polar decomposition F = RU = VR V = FRT ) getRotation() Matrix3 Returns rotation of the cell (orthogonal matrix R from polar decomposition F = RU ) getSmallStrain() Matrix3 T Returns small strain tensor = 1 2 (F + F ) I of the cell (http://en.wikipedia.org/wiki/Finite_strain_theory) hSize Base cell vectors (columns of the matrix), updated at every step from velGrad (trsf accumulates applied velGrad transformations). Setting hSize during a simulation is not supported by most contact laws, it is only meant to be used at iteration 0 before any interactions have been created. hSize0 Value of untransformed hSize, with respect to current trsf (computed as trsf hSize. homoDeform(=true) Deform (velGrad) the cell homothetically, by adjusting positions and velocities of bodies. The velocity change is obtained by deriving the expression v=v.x, where v is the macroscopic velocity gradient, giving in an incremental form: v= v x + v x. As a result, velocities are modied as soon as velGrad changes, according to the rst term: v(t)= v x(t), while the 2nd term reects a convective term: v= v v(t-dt/2). nextVelGrad(=Matrix3r::Zero()) see Cell.velGrad. prevHSize(=Matrix3r::Identity()) hSize from the previous step, used in the denition of relative velocity across periods. prevVelGrad(=Matrix3r::Zero()) Velocity gradient in the previous step.
387
refHSize(=Matrix3r::Identity()) Reference cell conguration, only used with OpenGLRenderer.dispScale. Updated automatically when hSize or trsf is assigned directly; also modied by utils.setRefSe3 (called e.g. by the :gui:Reference button in the UI). refSize Reference size of the cell (lengths of initial cell vectors, i.e. column norms of hSize). Note: Modifying this value is deprecated, use setBox instead. setBox((Vector3)arg2 ) None Set Cell shape to be rectangular, with dimensions along axes specied by given argument. Shorthand for assigning diagonal matrix with respective entries to hSize. setBox( (Cell)arg1, (oat)arg2, (oat)arg3, (oat)arg4) None : Set Cell shape to be rectangular, with dimensions along x, y, z specied by arguments. Shorthand for assigning diagonal matrix with the respective entries to hSize. shearPt((Vector3)arg2 ) Vector3 Apply shear (cell skew+rot) on the point shearTrsf Current skew+rot transformation (no resize) size Current size of the cell, i.e. lengths of the 3 cell lateral vectors contained in Cell.hSize columns. Updated automatically at every step. trsf Current transformation matrix of the cell, obtained from time integration of Cell.velGrad. unshearPt((Vector3)arg2 ) Vector3 Apply inverse shear on the point (removes skew+rot of the cell) unshearTrsf Inverse of the current skew+rot transformation (no resize) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary velGrad Velocity gradient of the transformation; used in NewtonIntegrator. Values of velGrad accumulate in trsf at every step. NOTE: changing velGrad at the beginning of a simulation loop would lead to inacurate integration for one step, as it should normaly be changed after the contact laws (but before Newton). To avoid this problem, assignment is deferred automatically. The target value typed in terminal is actually stored in Cell.nextVelGrad and will be applied right in time by Newton integrator. velGradChanged(=false) true when velGrad has been changed manually (see also Cell.nextVelGrad) volume Current volume of the cell. wrap((Vector3)arg2 ) Vector3 Transform an arbitrary point into a point in the reference cell wrapPt((Vector3)arg2 ) Vector3 Wrap point inside the reference cell, assuming the cell has no skew+rot.
388
7.13 Other classes

class yade.wrapper.Engine(inherits Serializable) Basic execution unit of simulation, called from the simulation loop (O.engines) dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Cell(inherits Serializable) Parameters of periodic boundary conditions. Only applies if O.isPeriodic==True. dict() dict Return dictionary of attributes. getDefGrad() Matrix3 Returns deformation gradient tensor F (http://en.wikipedia.org/wiki/Finite_strain_theory) of the cell deformation
getEulerianAlmansiStrain() Matrix3 1 Returns Eulerian-Almansi strain tensor e = 1 ) = 2 (I b (http://en.wikipedia.org/wiki/Finite_strain_theory) getLCauchyGreenDef() Matrix3 Returns left Cauchy-Green deformation tensor (http://en.wikipedia.org/wiki/Finite_strain_theory) getLagrangianStrain() Matrix3 Returns Lagrangian strain tensor E = 1 2 ( C I) = (http://en.wikipedia.org/wiki/Finite_strain_theory) b
1 2 (I
(FFT )1 ) of the cell
FFT
of
the
cell
1 T 2 (F F
I) =
2 1 2 (U
I) of the cell
getLeftStretch() Matrix3 Returns left (spatial) stretch tensor of the cell (matrix U from polar decomposition F = RU ) getPolarDecOfDefGrad() tuple Returns orthogonal matrix R and symmetric positive semi-denite matrix U as polar decomposition of deformation gradient F of the cell ( F = RU )
7.13. Other classes
389
getRCauchyGreenDef() Matrix3 Returns right Cauchy-Green deformation tensor (http://en.wikipedia.org/wiki/Finite_strain_theory)
FT F
of
the
cell
getRightStretch() Matrix3 Returns right (material) stretch tensor of the cell (matrix V from polar decomposition F = RU = VR V = FRT ) getRotation() Matrix3 Returns rotation of the cell (orthogonal matrix R from polar decomposition F = RU ) getSmallStrain() Matrix3 T Returns small strain tensor = 1 2 (F + F ) I of the cell (http://en.wikipedia.org/wiki/Finite_strain_theory) hSize Base cell vectors (columns of the matrix), updated at every step from velGrad (trsf accumulates applied velGrad transformations). Setting hSize during a simulation is not supported by most contact laws, it is only meant to be used at iteration 0 before any interactions have been created. hSize0 Value of untransformed hSize, with respect to current trsf (computed as trsf hSize. homoDeform(=true) Deform (velGrad) the cell homothetically, by adjusting positions and velocities of bodies. The velocity change is obtained by deriving the expression v=v.x, where v is the macroscopic velocity gradient, giving in an incremental form: v= v x + v x. As a result, velocities are modied as soon as velGrad changes, according to the rst term: v(t)= v x(t), while the 2nd term reects a convective term: v= v v(t-dt/2). nextVelGrad(=Matrix3r::Zero()) see Cell.velGrad. prevHSize(=Matrix3r::Identity()) hSize from the previous step, used in the denition of relative velocity across periods. prevVelGrad(=Matrix3r::Zero()) Velocity gradient in the previous step. refHSize(=Matrix3r::Identity()) Reference cell conguration, only used with OpenGLRenderer.dispScale. Updated automatically when hSize or trsf is assigned directly; also modied by utils.setRefSe3 (called e.g. by the :gui:Reference button in the UI). refSize Reference size of the cell (lengths of initial cell vectors, i.e. column norms of hSize). Note: Modifying this value is deprecated, use setBox instead. setBox((Vector3)arg2 ) None Set Cell shape to be rectangular, with dimensions along axes specied by given argument. Shorthand for assigning diagonal matrix with respective entries to hSize. setBox( (Cell)arg1, (oat)arg2, (oat)arg3, (oat)arg4) None : Set Cell shape to be rectangular, with dimensions along x, y, z specied by arguments. Shorthand for assigning diagonal matrix with the respective entries to hSize. shearPt((Vector3)arg2 ) Vector3 Apply shear (cell skew+rot) on the point shearTrsf Current skew+rot transformation (no resize)
390
size Current size of the cell, i.e. lengths of the 3 cell lateral vectors contained in Cell.hSize columns. Updated automatically at every step. trsf Current transformation matrix of the cell, obtained from time integration of Cell.velGrad. unshearPt((Vector3)arg2 ) Vector3 Apply inverse shear on the point (removes skew+rot of the cell) unshearTrsf Inverse of the current skew+rot transformation (no resize) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary velGrad Velocity gradient of the transformation; used in NewtonIntegrator. Values of velGrad accumulate in trsf at every step. NOTE: changing velGrad at the beginning of a simulation loop would lead to inacurate integration for one step, as it should normaly be changed after the contact laws (but before Newton). To avoid this problem, assignment is deferred automatically. The target value typed in terminal is actually stored in Cell.nextVelGrad and will be applied right in time by Newton integrator. velGradChanged(=false) true when velGrad has been changed manually (see also Cell.nextVelGrad) volume Current volume of the cell. wrap((Vector3)arg2 ) Vector3 Transform an arbitrary point into a point in the reference cell wrapPt((Vector3)arg2 ) Vector3 Wrap point inside the reference cell, assuming the cell has no skew+rot. class yade.wrapper.TimingDeltas data Get timing data as list of tuples (label, execTime[nsec], execCount) (one tuple per checkpoint) reset() None Reset timing information class yade.wrapper.GlExtraDrawer(inherits Serializable) Performing arbitrary OpenGL drawing commands; called from OpenGLRenderer (see OpenGLRenderer.extraDrawers) once regular rendering routines will have nished. This class itself does not render anything, derived classes should override the render method. dead(=false) Deactivate the object (on error/exception). dict() dict Return dictionary of attributes. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlIGeomDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point.
7.13. Other classes
391
dict() dict Return dictionary of attributes. dispFunctor((IGeom)arg2 ) GlIGeomFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.ParallelEngine(inherits Engine Serializable) Engine for running other Engine in parallel. __init__() None object __init__(tuple args, dict kwds) __init__((list)arg2) object : Construct from (possibly nested) list of slaves. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes
392
openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. slaves List of lists of Engines; each top-level group will be run in parallel with other groups, while Engines inside each group will be run sequentially, in given order. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlShapeDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((Shape)arg2 ) GlShapeFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Functor(inherits Serializable) Function-like object that is called by Dispatcher, if types of arguments match those the Functor declares to accept. bases Ordered list of types (as strings) this functor accepts.
7.13. Other classes
393
dict() dict Return dictionary of attributes. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly fron python (must be a valid python identier). timingDeltas Detailed information about timing inside the Dispatcher itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Serializable dict() dict Return dictionary of attributes. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlExtra_LawTester(inherits GlExtraDrawer Serializable) Find an instance of LawTester and show visually its data. dead(=false) Deactivate the object (on error/exception). dict() dict Return dictionary of attributes. tester(=uninitalized ) Associated LawTester object. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlStateDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((State)arg2 ) GlStateFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be
394
typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.MatchMaker(inherits Serializable) Class matching pair of ids to return pre-dened (for a pair of ids dened in matches) or derived value (computed using algo) of a scalar parameter. It can be called (id1, id2, val1=NaN, val2=NaN) in both python and c++. Note: There is a converter from python number dened for this class, which creates a new MatchMaker returning the value of that number; instead of giving the object instance therefore, you can only pass the number value and it will be converted automatically. algo Alogorithm used to compute value when no match for ids is found. Possible values are avg (arithmetic average) min (minimum value) max (maximum value) harmAvg (harmonic average) The following algo algorithms do not require meaningful input values in order to work: val (return value specied by val) zero (always return 0.)
computeFallback((oat)val1, (oat)val2 ) oat Compute algo value for val1 and val2, using algorithm specied by algo. dict() dict Return dictionary of attributes. matches(=uninitalized ) Array of (id1,id2,value) items; queries matching id1 + id2 or id2 + id1 will return value updateAttrs((dict)arg2 ) None Update object attributes from given dictionary val(=NaN ) Constant value returned if there is no match and algo is val class yade.wrapper.GlBoundDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes.
7.13. Other classes
395
dispFunctor((Bound)arg2 ) GlBoundFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlIPhysDispatcher(inherits Dispatcher Engine Serializable) Dispatcher calling functors based on received argument type(s). dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. dispFunctor((IPhys)arg2 ) GlIPhysFunctor Return functor that would be dispatched for given argument(s); None if no dispatch; ambiguous dispatch throws. dispMatrix([(bool)names=True ]) dict Return dictionary with contents of the dispatch matrix. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). functors Functors associated with this dispatcher. label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can
396
depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level. timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.GlExtra_OctreeCubes(inherits GlExtraDrawer Serializable) Render boxed read from le boxesFile(=uninitalized ) File to read boxes from; ascii les with x0 y0 z0 x1 y1 z1 c records, where c is an integer specifying ll (0 for wire, 1 for lled). dead(=false) Deactivate the object (on error/exception). dict() dict Return dictionary of attributes. fillRangeDraw(=Vector2i(-2, 2)) Range of ll indices that will be rendered. fillRangeFill(=Vector2i(2, 2)) Range of ll indices that will be lled. levelRangeDraw(=Vector2i(-2, 2)) Range of levels that will be rendered. noFillZero(=true) Do not ll 0-ll boxed (those that are further subdivided) updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.Dispatcher(inherits Engine Serializable) Engine dispatching control to its associated functors, based on types of argument it receives. This abstract base class provides no functionality in itself. dead(=false) If true, this engine will not run at all; can be used for making an engine temporarily deactivated and only resurrect it at a later point. dict() dict Return dictionary of attributes. execCount Cummulative count this engine was run (only used if O.timingEnabled==True). execTime Cummulative time this Engine took to run (only used if O.timingEnabled==True). label(=uninitalized ) Textual label for this object; must be valid python identier, you can refer to it directly from python. ompThreads(=-1) Number of threads to be used in the engine. If ompThreads<0 (default), the number will be typically OMP_NUM_THREADS or the number N dened by yade -jN (this behavior can depend on the engine though). This attribute will only aect engines whose code includes openMP parallel regions (e.g. InteractionLoop). This attribute is mostly useful for experiments or when combining ParallelEngine with engines that run parallel regions, resulting in nested OMP loops with dierent number of threads at each level.
7.13. Other classes
397
timingDeltas Detailed information about timing inside the Engine itself. Empty unless enabled in the source code and O.timingEnabled==True. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary class yade.wrapper.EnergyTracker(inherits Serializable) Storage for tracing energies. Only to be used if O.trackEnergy is True. clear() None Clear all stored values. dict() dict Return dictionary of attributes. energies(=uninitalized ) Energy values, in linear array items() list Return contents as list of (name,value) tuples. keys() list Return dened energies. total() oat Return sum of all energies. updateAttrs((dict)arg2 ) None Update object attributes from given dictionary
398
Chapter 8
Yade modules
8.1 yade.bodiesHandling module
Miscellaneous functions, which are useful for handling bodies. yade.bodiesHandling.facetsDimensions(idFacets=[], mask=-1) The function accepts the list of facet ids or list of facets and calculates max and min dimensions, geometrical center. Parameters idFacets (list ) list of spheres mask (int ) Body.mask for the checked bodies Returns dictionary with keys min (minimal dimension, Vector3), max (maximal dimension, Vector3), minId (minimal dimension facet Id, Vector3), maxId (maximal dimension facet Id, Vector3), center (central point of bounding box, Vector3), extends (sizes of bounding box, Vector3), number (number of facets, int), yade.bodiesHandling.sphereDuplicate(idSphere) The functions makes a copy of sphere yade.bodiesHandling.spheresModify(idSpheres=[], mask=-1, shift=Vector3(0, 0, 0), scale=1.0, orientation=Quaternion((1, 0, 0), 0), copy=False) The function accepts the list of spheres ids or list of bodies and modies them: rotating, scaling, shifting. if copy=True copies bodies and modies them. Also the mask can be given. If idSpheres not empty, the function aects only bodies, where the mask passes. If idSpheres is empty, the function search for bodies, where the mask passes. Parameters shift (Vector3 ) Vector3(X,Y,Z) parameter moves spheres. scale (oat ) factor scales given spheres. orientation (Quaternion) orientation of spheres mask (int ) Body.mask for the checked bodies Returns list of bodies if copy=True, and Boolean value if copy=False yade.bodiesHandling.spheresPackDimensions(idSpheres=[], mask=-1) The function accepts the list of spheres ids or list of bodies and calculates max and min dimensions, geometrical center. Parameters idSpheres (list ) list of spheres mask (int ) Body.mask for the checked bodies 399
Returns dictionary with keys min (minimal dimension, Vector3), max (maximal dimension, Vector3), minId (minimal dimension sphere Id, Vector3), maxId (maximal dimension sphere Id, Vector3), center (central point of bounding box, Vector3), extends (sizes of bounding box, Vector3), volume (volume of spheres, Real), mass (mass of spheres, Real), number (number of spheres, int),
8.2 yade.eudoxos module

Miscillaneous functions that are not believed to be generally usable, therefore kept in my private module here. They comprise notably oofem export and various CPM-related functions. class yade.eudoxos.IntrSmooth3d Return spatially weigted gaussian average of arbitrary quantity dened on interactions. At construction time, all real interactions are put inside spatial grid, permitting fast search for points in neighbourhood dened by distance. Parameters for the distribution are standard deviation and relative cuto distance relThreshold (3 by default) which will discard points farther than relThreshold . Given central point p0 , points are weighted by gaussian function ( ) 1 ||p0 p||2 (p0 , p) = exp 22 2 To get the averaged value, simply call the instance, passing central point and callable object which received interaction object and returns the desired quantity:
>>> O.reset() >>> from yade import utils >>> O.bodies.append([utils.sphere((0,0,0),1),utils.sphere((0,0,1.9),1)]) [0, 1] >>> O.engines=[InteractionLoop([Ig2_Sphere_Sphere_ScGeom(),],[Ip2_FrictMat_FrictMat_FrictPhys()],[])] >>> utils.createInteraction(0,1) <Interaction instance at 0x...>
>> is3d=IntrSmooth3d(0.003) >> is3d((0,0,0),lambda i: tor3(0,0,0) bounds() count() yade.eudoxos.displacementsInteractionsExport(fName)
i.phys.normalForce) Vec-
yade.eudoxos.eliminateJumps(eps, sigma, numSteep=10, gapWidth=5, movWd=40 ) yade.eudoxos.estimatePoissonYoung(principalAxis, stress=0, plot=False, cuto=0.0 ) Estimate Poissons ration given the principal axis of straining. For every base direction, homogenized strain is computed (slope in linear regression on discrete function particle coordinate particle displacement in the same direction as returned by utils.coordsAndDisplacements) and, (if axis 0 is the strained axis) the poissons ratio is given as -(1+2)/. Youngs modulus is computed as /; if stress is not given (default 0), the result is 0. cuto, if > 0., will take only smaller part (centered) or the specimen into account yade.eudoxos.estimateStress(strain, cuto=0.0 ) Use summed stored energy in contacts to compute macroscopic stress over the same volume, provided known strain. yade.eudoxos.oofemDirectExport(leBase, title=None, negIds=[], posIds=[]) yade.eudoxos.oofemPrescribedDisplacementsExport(leName)
400
Chapter 8. Yade modules
yade.eudoxos.oofemTextExport(fName) Export simulation data in text format The format is line-oriented as follows:
E G # elastic material parameters epsCrackOnset relDuctility xiShear transStrainCoeff # tensile parameters; epsFr=epsCrackOnset*relDuctility cohesionT tanPhi # shear parameters number_of_spheres number_of_links id x y z r boundary # spheres; boundary: -1 negative, 0 none, 1 positive id1 id2 cp_x cp_y cp_z A # interactions; cp = contact point; A = cross-section
class yade._eudoxos.HelixInteractionLocator2d Locate all real interactions in 2d plane (reduced by spiral projection from 3d, using Shop::spiralProject, which is the same as utils.spiralProject) using their contact points. Note: Do not run simulation while using this object. __init__((oat)dH_dTheta[, Parameters dH_dTheta (oat ) Spiral inclination, i.e. height increase per 1 radian turn; axis (int ) axis of rotation (0=x,1=y,2=z) theta (oat ) spiral angle at zero height (theta intercept) thetaMin (oat ) only interactions with thetaMin will be considered (NaN to deactivate) thetaMax (oat ) only interactions with thetaMax will be considered (NaN to deactivate) See utils.spiralProject. hi Return upper corner of the rectangle containing all interactions. intrsAroundPt((Vector2)pt2d, (oat)radius) list Return list of interaction objects that are not further from pt2d than radius in the projection plane lo Return lower corner of the rectangle containing all interactions. macroAroundPt((Vector2)pt2d, (oat)radius) tuple Compute macroscopic stress around given point; the interaction (n and T are rotated to the projection plane by (as given by utils.spiralProject) rst, but no skew is applied). The formula used is [ )] 1 IJ IJ N,IJ IJ IJ 1 ( T,IJ IJ T,IJ IJ d A ni nj + nj + j ni ij = V 2 i
IJ
(int)axis=0 [,
(oat)periodStart=nan[,
(oat)theta0=0 [,
(oat)thetaMin=nan[, (oat)thetaMax=nan ]]]]]) None
where the sum is taken over volume V containing interactions IJ between spheres I and J; i, j indices denote Cartesian components of vectors and tensors, dIJ is current distance between spheres I and J, AIJ is area of contact IJ, n is (-rotated) interaction normal (unit vector pointing from center of I to the center of J) N,IJ is normal stress (as scalar) in contact IJ,
8.2. yade.eudoxos module
401
T,IJ is shear stress in contact IJ in global coordinates and -rotated. ) and CpmPhys.kappaD ( D ). N is Additionally, computes average of CpmPhys.omega ( the number of interactions in the volume given. , D ). Returns tuple of (N, , class yade._eudoxos.InteractionLocator Locate all (real) interactions in space by their contact point. When constructed, all real interactions are spatially indexed (uses vtkPointLocator internally). Use instance methods to use those data. Note: Data might become inconsistent with real simulation state if simulation is being run between creation of this object and spatial queries. bounds Return coordinates of lower and uppoer corner of axis-aligned abounding box of all interactions count Number of interactions held intrsAroundPt((Vector3)point, (oat)maxDist ) list Return list of real interactions that are not further than maxDist from point. macroAroundPt((Vector3)point, (oat)maxDist [, (oat)forceVolume=-1 ]) tuple Return tuple of averaged stress tensor (as Matrix3), average omega and average kappa values. forceVolume can be used (if positive) rather than the sphere (with maxDist radius) volume for the computation. (This is useful if point and maxDist encompass empty space that you want to avoid.) yade._eudoxos.particleConfinement() None yade._eudoxos.velocityTowardsAxis((Vector3)axisPoint, (oat)timeToAxis[, (oat)perturbation ]]) None (Vector3)axisDirection, (oat)subtractDist [,
8.3 yade.export module

Export (not only) geometry to various formats. class yade.export.VTKExporter Class for exporting data to VTK Simple Legacy File (for example if, for some reason, you are not able to use VTKRecorder). Export of spheres, facets and interactions is supported. USAGE: create object vtkExporter = VTKExporter(baseFileName), add to engines PyRunner with command=vtkExporter.exportSomething(params) Example: examples/test/vtk-exporter/vtkExporter.py, read/unvReadVTKExport.py. Parameters baseName (string) name of the exported les. The les would be named baseName-spheres-snapNb.vtk or baseName-facets-snapNb.vtk startSnap (int=0 ) the numbering of les will start form startSnap exportFacets() exports facets (positions) and dened properties. Facets are exported with multiplicated nodes :param [int]|all ids: if all, then export all facets, otherwise only facets from integer list :param [tuple(2)] what: see exportSpheres :param string comment: comment to add to vtk le :param int numLabel: number of le (e.g. time step), if unspecied, the last used value + 1 will be used examples/test/unv-
402
exportFacetsAsMesh() exports facets (positions) and dened properties. Facets are exported as mesh (not with multiplicated nodes). Therefore additional parameters connectivityTable is needed :param [int]|all ids: if all, then export all facets, otherwise only facets from integer list :param [tuple(2)] what: see exportSpheres :param string comment: comment to add to vtk le :param int numLabel: number of le (e.g. time step), if unspecied, the last used value + 1 will be used :param [(oat,oat,oat)|Vector3] nodes: list of coordinates of nodes :param [(int,int,int)] connectivityTable: list of node ids of individual elements (facets) exportInteractions() exports interactions and dened properties. :param ids: if all, then export all spheres, otherwise only spheres from integer list :type ids: [int] | all :param what: what to export. parameter is list of couple (name,command). Name is string under which it is save to vtk, command is string to evaluate. Node that the interactions are labeled as i in this function. Scalar, vector and tensor variables are supported. For example, to export stiness dierence from certain value (1e9) (named as dSti) you should write: ... what=[(dSti,i.phys.kn-1e9), ... :type what: [tuple(2)] :param string comment: comment to add to vtk le :param int numLabel: number of le (e.g. time step), if unspecied, the last used value + 1 will be used exportSpheres() exports spheres (positions and radius) and dened properties. :param ids: if all, then export all spheres, otherwise only spheres from integer list :type ids: [int] | all :param what: what other than then position and radius export. parameter is list of couple (name,command). Name is string under which it is save to vtk, command is string to evaluate. Node that the bodies are labeled as b in this function. Scalar, vector and tensor variables are supported. For example, to export velocity (with name particleVelocity) and the distance form point (0,0,0) (named as dist) you should write: ... what=[(particleVelocity,b.state.vel),(dist,b.state.pos.norm(), ... :type what: [tuple(2)] :param string comment: comment to add to vtk le :param int numLabel: number of le (e.g. time step), if unspecied, the last used value + 1 will be used class yade.export.VTKWriter USAGE: create object vtk_writer = VTKWriter(base_le_name), add to engines PyRunner with command=vtk_writer.snapshot() snapshot() yade.export.gmshGeo(lename, comment=, mask=-1, accuracy=-1) Save spheres in geo-le for the following using in GMSH (http://www.geuz.org/gmsh/doc/texinfo/) program. The spheres can be there meshed. Parameters lename (string) the name of the le, where sphere coordinates will be exported. mask (int ) export only spheres with the corresponding mask export only spheres with the corresponding mask accuracy (oat ) the accuracy parameter, which will be set for the poinst in geo-le. By default: 1./10. of the minimal sphere diameter. Returns number of spheres which were exported. Return type int yade.export.text(lename, mask=-1) Save sphere coordinates into a text le; the format of the line is: x y z r. Non-spherical bodies are silently skipped. Example added to examples/regular-sphere-pack/regular-sphere-pack.py Parameters lename (string) the name of the le, where sphere coordinates will be exported.
8.3. yade.export module
403
mask (int ) export only spheres with the corresponding mask Returns number of spheres which were written. Return type int yade.export.textExt(lename, format=x_y_z_r, comment=, mask=-1) Save sphere coordinates and other parameters into a text le in specic format. Non-spherical bodies are silently skipped. Users can add here their own specic format, giving meaningful names. The rst le row will contain the format name. Be sure to add the same format specication in ymport.textExt. Parameters lename (string) the name of the le, where sphere coordinates will be exported. format (string) the name of output format. Supported x_y_z_r(default), x_y_z_r_matId comment (string) the text, which will be added as a comment at the top of le. If you want to create several lines of text, please use \n# for next lines. mask (int ) export only spheres with the corresponding mask export only spheres with the corresponding mask Returns number of spheres which were written. Return type int
8.4 yade.geom module

Creates geometry objects from facets. yade.geom.facetBox(center, extents, orientation=Quaternion((1, 0, 0), 0), wallMask=63, **kw ) Create arbitrarily-aligned box composed of facets, with given center, extents and orientation. If any of the box dimensions is zero, corresponding facets will not be created. The facets are oriented outwards from the box. Parameters center (Vector3 ) center of the box extents (Vector3 ) lengths of the box sides orientation (Quaternion) orientation of the box wallMask (bitmask ) determines which walls will be created, in the order -x (1), +x (2), -y (4), +y (8), -z (16), +z (32). The numbers are ANDed; the default 63 means to create all walls **kw (unused keyword arguments) passed to utils.facet Returns list of facets forming the box yade.geom.facetBunker(center, dBunker, dOutput, hBunker, hOutput, hPipe=0.0, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, wallMask=4, angleRange=None, closeGap=False, **kw ) Create arbitrarily-aligned bunker, composed of facets, with given center, radii, heights and orientation. Return List of facets forming the bunker;
dBunker ______________ | | | | | | hBunker | | | | | |
404
|____________| \ / \ / \ / hOutput \ / \____/ | | |____| hPipe dOutput
Parameters center (Vector3 ) center of the created bunker dBunker (oat ) bunker diameter, top dOutput (oat ) bunker output diameter hBunker (oat ) bunker height hOutput (oat ) bunker output height hPipe (oat ) bunker pipe height orientation (Quaternion) orientation of the bunker; the reference orientation has axis along the +x axis. segmentsNumber (int ) number of edges on the bunker surface (>=5) wallMask (bitmask ) determines which walls will be created, in the order up (1), down (2), side (4). The numbers are ANDed; the default 7 means to create all walls angleRange ((min,max)) allows one to create only part of bunker by specifying range of angles; if None, (0,2*pi) is assumed. closeGap (bool ) close range skipped in angleRange with triangular facets at cylinder bases. **kw (unused keyword arguments) passed to utils.facet; yade.geom.facetCone(center, radiusTop, radiusBottom, height, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, wallMask=7, angleRange=None, closeGap=False, **kw ) Create arbitrarily-aligned cone composed of facets, with given center, radius, height and orientation. Return List of facets forming the cone; Parameters center (Vector3 ) center of the created cylinder radiusTop (oat ) cone top radius radiusBottom (oat ) cone bottom radius height (oat ) cone height orientation (Quaternion) orientation of the cone; the reference orientation has axis along the +x axis. segmentsNumber (int ) number of edges on the cone surface (>=5) wallMask (bitmask ) determines which walls will be created, in the order up (1), down (2), side (4). The numbers are ANDed; the default 7 means to create all walls angleRange ((min,max)) allows one to create only part of cone by specifying range of angles; if None, (0,2*pi) is assumed. closeGap (bool ) close range skipped in angleRange with triangular facets at cylinder bases.
8.4. yade.geom module
405
**kw (unused keyword arguments) passed to utils.facet; yade.geom.facetCylinder(center, radius, height, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, wallMask=7, angleRange=None, closeGap=False, **kw ) Create arbitrarily-aligned cylinder composed of facets, with given center, radius, height and orientation. Return List of facets forming the cylinder; Parameters center (Vector3 ) center of the created cylinder radius (oat ) cylinder radius height (oat ) cylinder height orientation (Quaternion) orientation of the cylinder; the reference orientation has axis along the +x axis. segmentsNumber (int ) number of edges on the cylinder surface (>=5) wallMask (bitmask ) determines which walls will be created, in the order up (1), down (2), side (4). The numbers are ANDed; the default 7 means to create all walls angleRange ((min,max)) allows one to create only part of bunker by specifying range of angles; if None, (0,2*pi) is assumed. closeGap (bool ) close range skipped in angleRange with triangular facets at cylinder bases. **kw (unused keyword arguments) passed to utils.facet; yade.geom.facetCylinderConeGenerator(center, radiusTop, height, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, wallMask=7, angleRange=None, closeGap=False, radiusBottom=-1, **kw ) Please, do not use this function directly! Use geom.facetCylinder and geom.facetCone instead. This is the base function for generating cylinders and cones from facets. :param oat radiusTop: top radius :param oat radiusBottom: bottom radius :param **kw: (unused keyword arguments) passed to utils.facet; yade.geom.facetHelix(center, radiusOuter, pitch, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, angleRange=None, radiusInner=0, **kw ) Create arbitrarily-aligned helix composed of facets, with given center, radius (outer and inner), pitch and orientation. Return List of facets forming the helix; Parameters center (Vector3 ) center of the created cylinder radiusOuter (oat ) outer radius radiusInner (oat ) inner height (can be 0) orientation (Quaternion) orientation of the helix; the reference orientation has axis along the +x axis. segmentsNumber (int ) number of edges on the helix surface (>=3) angleRange ((min,max)) range of angles; if None, (0,2*pi) is assumed. **kw (unused keyword arguments) passed to utils.facet; yade.geom.facetParallelepiped(center, extents, height, orientation=Quaternion((1, 0, 0), 0), wallMask=63, **kw ) Create arbitrarily-aligned Parallelepiped composed of facets, with given center, extents, height and orientation. If any of the parallelepiped dimensions is zero, corresponding facets will not be created. The facets are oriented outwards from the parallelepiped. Parameters center (Vector3 ) center of the parallelepiped 406 Chapter 8. Yade modules
extents (Vector3 ) lengths of the parallelepiped sides height (Real ) height of the parallelepiped (along axis z) orientation (Quaternion) orientation of the parallelepiped wallMask (bitmask ) determines which walls will be created, in the order -x (1), +x (2), -y (4), +y (8), -z (16), +z (32). The numbers are ANDed; the default 63 means to create all walls **kw (unused keyword arguments) passed to utils.facet Returns list of facets forming the parallelepiped yade.geom.facetPolygon(center, radiusOuter, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, angleRange=None, radiusInner=0, **kw ) Create arbitrarily-aligned polygon composed of facets, with given center, radius (outer and inner) and orientation. Return List of facets forming the polygon; Parameters center (Vector3 ) center of the created cylinder radiusOuter (oat ) outer radius radiusInner (oat ) inner height (can be 0) orientation (Quaternion) orientation of the polygon; the reference orientation has axis along the +x axis. segmentsNumber (int ) number of edges on the polygon surface (>=3) angleRange ((min,max)) allows one to create only part of polygon by specifying range of angles; if None, (0,2*pi) is assumed. **kw (unused keyword arguments) passed to utils.facet; yade.geom.facetPolygonHelixGenerator(center, radiusOuter, pitch=0, orientation=Quaternion((1, 0, 0), 0), segmentsNumber=10, angleRange=None, radiusInner=0, **kw ) Please, do not use this function directly! Use geom.facetPloygon and geom.facetHelix instead. This is the base function for generating polygons and helixes from facets. yade.geom.facetSphere(center, radius, thetaResolution=8, phiResolution=8, returnElementMap=False, **kw ) Create arbitrarily-aligned sphere composed of facets, with given center, radius and orientation. Return List of facets forming the sphere. Parameters inspired by ParaView sphere glyph Parameters center (Vector3 ) center of the created sphere radius (oat ) sphere radius thetaResolution (int ) number of facets around equator phiResolution (int ) number of facets between poles + 1 returnElementMap (bool ) returns also tuple of nodes ((x1,y1,z1),(x2,y2,z2),...) and elements ((id01,id02,id03),(id11,id12,id13),...) if true, only facets otherwise **kw (unused keyword arguments) passed to utils.facet;
8.5 yade.linterpolation module

Module for rudimentary support of manipulation with piecewise-linear functions (which are usually interpolations of higher-order functions, whence the module name). Interpolation is always given as two lists of the same length, where the x-list must be increasing.
8.5. yade.linterpolation module
407
Periodicity is supported by supposing that the interpolation can wrap from the last x-value to the rst x-value (which should be 0 for meaningful results). Non-periodic interpolation can be converted to periodic one by padding the interpolation with constant head and tail using the sanitizeInterpolation function. There is a c++ template function for interpolating on such sequences in pkg/common/Engine/PartialEngine/LinearInterpolate.hpp (stateful, therefore fast for sequential reads). TODO: Interpolating from within python is not (yet) supported. yade.linterpolation.integral(x, y) Return integral of piecewise-linear function given by points x0,x1, and y0,y1, yade.linterpolation.revIntegrateLinear(I, x0, y0, x1, y1) Helper function, returns value of integral variable x for linear function f passing through (x0,y0),(x1,y1) such that 1. x[x0,x1] 2. _x0^x f dx=I and raise exception if such number doesnt exist or the solution is not unique (possible?) yade.linterpolation.sanitizeInterpolation(x, y, x0, x1) Extends piecewise-linear function in such way that it spans at least the x0x1 interval, by adding constant padding at the beginning (using y0) and/or at the end (using y1) or not at all. yade.linterpolation.xFractionalFromIntegral(integral, x, y) Return x within range x0xn such that _x0^x f dx==integral. Raises error if the integral value is not reached within the x-range. yade.linterpolation.xFromIntegral(integralValue, x, y) Return x such that _x0^x f dx==integral. x wraps around at xn. For meaningful results, therefore, x0 should == 0
8.6 yade.pack module

Creating packings and lling volumes dened by boundary representation or constructive solid geometry. For examples, see scripts/test/gts-operators.py scripts/test/gts-random-pack-obb.py scripts/test/gts-random-pack.py scripts/test/pack-cloud.py scripts/test/pack-predicates.py examples/packs/packs.py examples/gts-horse/gts-horse.py examples/WireMatPM/wirepackings.py yade.pack.SpherePack_toSimulation(self, rot=Matrix3(1, 0, 0, 0, 1, 0, 0, 0, 1), **kw ) Append spheres directly to the simulation. In addition calling O.bodies.append, this method also appropriately sets periodic cell information of the simulation.
>>> from yade import pack; from math import * >>> sp=pack.SpherePack()
Create random periodic packing with 20 spheres:

>>> sp.makeCloud((0,0,0),(5,5,5),rMean=.5,rRelFuzz=.5,periodic=True,num=20) 20
Virgin simulation is aperiodic:
408
>>> O.reset() >>> O.periodic False
Add generated packing to the simulation, rotated by 45 along +z

>>> sp.toSimulation(rot=Quaternion((0,0,1),pi/4),color=(0,0,1)) [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
Periodic properties are transferred to the simulation correctly, including rotation (this could be avoided by explicitly passing hSize=O.cell.hSize as an argument):
>>> O.periodic True >>> O.cell.refSize Vector3(5,5,5)
yade.pack.filterSpherePack(predicate, spherePack, returnSpherePack=None, **kw ) Using given SpherePack instance, return spheres the satisfy predicate. The packing will be recentered to match the predicate and warning is given if the predicate is larger than the packing. yade.pack.gtsSurface2Facets(surf, **kw ) Construct facets from given GTS surface. **kw is passed to utils.facet. yade.pack.gtsSurfaceBestFitOBB(surf ) Return (Vector3 center, Vector3 halfSize, Quaternion orientation) describing best-t oriented bounding box (OBB) for the given surface. See cloudBestFitOBB for details. yade.pack.hexaNet(radius, cornerCoord=[0, 0, 0], xLength=1.0, yLength=0.5, mos=0.08, a=0.04, b=0.04, startAtCorner=True, isSymmetric=False, **kw ) Denition of the particles for a hexagonal wire net in the x-y-plane for the WireMatPM. Parameters radius radius of the particle cornerCoord coordinates of the lower left corner of the net xLenght net length in x-direction yLenght net length in y-direction mos mesh opening size (horizontal distance between the double twists) a length of double-twist b height of single wire section startAtCorner if true the generation starts with a double-twist at the lower left corner isSymmetric denes if the net is symmetric with respect to the y-axis Returns set of spheres which denes the net (net) and exact dimensions of the net (lx,ly). note:: This packing works for the WireMatPM only. The particles at the corner are always generated rst. For examples on how to use this packing see examples/WireMatPM. In order to create the proper interactions for the net the interaction radius has to be adapted in the simulation. class yade.pack.inGtsSurface_py(inherits Predicate) This class was re-implemented in c++, but should stay here to serve as reference for implementing Predicates in pure python code. C++ allows us to play dirty tricks in GTS which are not accessible through pygts itself; the performance penalty of pygts comes from fact that if constructs and destructs bb tree for the surface at every invocation of gts.Point().is_inside(). That is cached in the c++ code, provided that the surface is not manipulated with during lifetime of the object (users responsibility). Predicate for GTS surfaces. Constructed using an already existing surfaces, which must be closed. 8.6. yade.pack module 409
import gts surf=gts.read(open(horse.gts)) inGtsSurface(surf) Note: Padding is optionally supported by testing 6 points along the axes in the pad distance. This must be enabled in the ctor by saying doSlowPad=True. If it is not enabled and pad is not zero, warning is issued. aabb() center() Vector3 dim() Vector3 class yade.pack.inSpace(inherits Predicate) Predicate returning True for any points, with innite bounding box. aabb() center() dim() yade.pack.randomDensePack(predicate, radius, material=-1, dim=None, cropLayers=0, rRelFuzz=0.0, spheresInCell=0, memoizeDb=None, useOBB=False, memoDbg=False, color=None, returnSpherePack=None) Generator of random dense packing with given geometry properties, using TriaxialTest (aperiodic) or PeriIsoCompressor (periodic). The periodicity depens on whether the spheresInCell parameter is given. O.switchScene() magic is used to have clean simulation for TriaxialTest without deleting the original simulation. This function therefore should never run in parallel with some code accessing your simulation. Parameters predicate solid-dening predicate for which we generate packing spheresInCell if given, the packing will be periodic, with given number of spheres in the periodic cell. radius mean radius of spheres rRelFuzz relative fuzz of the radius e.g. radius=10, rRelFuzz=.2, then spheres will have radii 10 (10*.2)). 0 by default, meaning all spheres will have exactly the same radius. cropLayers (aperiodic only) how many layers of spheres will be added to the computed dimension of the box so that there no (or not so much, at least) boundary eects at the boundaries of the predicate. dim dimension of the packing, to override dimensions of the predicate (if it is innite, for instance) memoizeDb name of sqlite database (existent or nonexistent) to nd an already generated packing or to store the packing that will be generated, if not found (the technique of caching results of expensive computations is known as memoization). Fuzzy matching is used to select suitable candidate packing will be scaled, rRelFuzz and dimensions compared. Packing that are too small are dictarded. From the remaining candidate, the one with the least number spheres will be loaded and returned. useOBB eective only if a inGtsSurface predicate is given. If true (not default), oriented bounding box will be computed rst; it can reduce substantially number of spheres for the triaxial compression (like 10 depending on how much asymmetric the body is), see scripts/test/gts-triax-pack-obb.py. memoDbg show packigns that are considered and reasons why they are rejected/accepted returnSpherePack see lterSpherePack 410 Chapter 8. Yade modules
Returns SpherePack object with spheres, ltered by the predicate. yade.pack.randomPeriPack(radius, initSize, rRelFuzz=0.0, memoizeDb=None, noPrint=False) Generate periodic dense packing. A cell of initSize is stued with as many spheres as possible, then we run periodic compression with PeriIsoCompressor, just like with randomDensePack. Parameters radius mean sphere radius rRelFuzz relative fuzz of sphere radius (equal distribution); see the same param for randomDensePack. initSize initial size of the periodic cell. Returns SpherePack object, which also contains periodicity information. yade.pack.regularHexa(predicate, radius, gap, **kw ) Return set of spheres in regular hexagonal grid, clipped inside solid given by predicate. Created spheres will have given radius and will be separated by gap space. yade.pack.regularOrtho(predicate, radius, gap, **kw ) Return set of spheres in regular orthogonal grid, clipped inside solid given by predicate. Created spheres will have given radius and will be separated by gap space. yade.pack.revolutionSurfaceMeridians(sects, angles, origin=Vector3(0, 0, 0), orientation=Quaternion((1, 0, 0), 0)) Revolution surface given sequences of 2d points and sequence of corresponding angles, returning sequences of 3d points representing meridian sections of the revolution surface. The 2d sections are turned around z-axis, but they can be transformed using the origin and orientation arguments to give arbitrary orientation. yade.pack.sweptPolylines2gtsSurface(pts, threshold=0, capStart=False, capEnd=False) Create swept suface (as GTS triangulation) given same-length sequences of points (as 3-tuples). If threshold is given (>0), then degenerate faces (with edges shorter than threshold) will not be created gts.Surface().cleanup(threshold) will be called before returning, which merges vertices mutually closer than threshold. In case your pts are closed (last point concident with the rst one) this will the surface strip of triangles. If you additionally have capStart==True and capEnd==True, the surface will be closed. Note: capStart and capEnd make the most naive polygon triangulation (diagonals) and will perhaps fail for non-convex sections. Warning: the algorithm connects points sequentially; if two polylines are mutually rotated or have inverse sense, the algorithm will not detect it and connect them regardless in their given order. Creation, manipulation, IO for generic sphere packings. class yade._packSpheres.SpherePack Set of spheres represented as centers and radii. This class is returned by pack.randomDensePack, pack.randomPeriPack and others. The object supports iteration over spheres, as in
>>> sp=SpherePack() >>> for center,radius in sp: print center,radius >>> for sphere in sp: print sphere[0],sphere[1] ## same, but without unpacking the tuple automatically ## same, but accessing spheres by index
>>> for i in range(0,len(sp)): print sp[i][0], sp[i][1]
8.6. yade.pack module
411
Special constructors Construct from list of [(c1,r1),(c2,r2),]. To convert two same-length lists of centers and radii, construct with zip(centers,radii). __init__([(list)list ]) None Empty constructor, optionally taking list [ ((cx,cy,cz),r), ] for initial data. aabb() tuple Get axis-aligned bounding box coordinates, as 2 3-tuples. add((Vector3)arg2, (oat)arg3 ) None Add single sphere to packing, given center as 3-tuple and radius appliedPsdScaling A factor between 0 and 1, uniformly applied on all sizes of of the PSD. cellFill((Vector3)arg2 ) None Repeat the packing (if periodic) so that the results has dim() >= given size. The packing retains periodicity, but changes cellSize. Raises exception for non-periodic packing. cellRepeat((Vector3i)arg2 ) None Repeat the packing given number of times in each dimension. Periodicity is retained, cellSize changes. Raises exception for non-periodic packing. cellSize Size of periodic cell; is Vector3(0,0,0) if not periodic. (Change this property only if you know what youre doing). center() Vector3 Return coordinates of the bounding box center. dim() Vector3 Return dimensions of the packing in terms of aabb(), as a 3-tuple. fromList((list)arg2 ) None Make packing from given list, same format as for constructor. Discards current data. fromList( (SpherePack)arg1, (object)centers, (object)radii) None : Make packing from given list, same format as for constructor. Discards current data. fromSimulation() None Make packing corresponding to the current simulation. Discards current data. getClumps() tuple Return lists of sphere ids sorted by clumps they belong to. The return value is (standalones,[clump1,clump2,]), where each item is list of ids of spheres. hasClumps() bool Whether this object contains clumps. isPeriodic was the packing generated in periodic boundaries? load((str)leName) None Load packing from external text le (current data will be discarded). makeCloud([(Vector3)minCorner=Vector3(0, 0, 0)[, (Vector3)maxCorner=Vector3(0, 0, 0)[, (oat)rMean=-1[, (oat)rRelFuzz=0 [, (int)num=-1[, (bool)periodic=False[, (object)psdCumm=[] [, (oat)porosity=0.65 [, (object)psdSizes=[] [,
(bool)distributeMass=False[, (int)seed=0 [, (Matrix3)hSize=Matrix3(0, 0, 0, 0, 0, 0, 0, 0, 0) ]]]]]]]]]]]]) int Create random loose packing enclosed in a parallelepiped (also works in 2D if min-
412
Corner[k]=maxCorner[k] for one coordinate). Sphere radius distribution can be specied using one of the following ways: 1.rMean, rRelFuzz and num gives uniform radius distribution in rMean(1 rRelFuzz). Less than num spheres can be generated if it is too high. 2.rRelFuzz, num and (optional) porosity, which estimates mean radius so that porosity is attained at the end. rMean must be less than 0 (default). porosity is only an initial guess for the generation algorithm, which will retry with higher porosity until the prescibed num is obtained. 3.psdSizes and psdCumm, two arrays specifying points of the particle size distribution function. As many spheres as possible are generated. 4.psdSizes, psdCumm, num, and (optional) porosity, like above but if num is not obtained, psdSizes will be scaled down uniformly, until num is obtained (see appliedPsdScaling). By default (with distributeMass==False), the distribution is applied to particle radii. The usual sense of particle size distribution is the distribution of mass fraction (rather than particle count); this can be achieved with distributeMass=True. If num is dened, then sizes generation is deterministic, giving the best t of target distribution. It enables spheres placement in descending size order, thus giving lower porosity than the random generation. Parameters minCorner (Vector3 ) lower corner of an axis-aligned box maxCorner (Vector3 ) upper corner of an axis-aligned box hSize (Matrix3 ) base vectors of a generalized box (arbitrary parallelepiped, typically Cell::hSize), superseeds minCorner and maxCorner if dened. For periodic boundaries only. rMean (oat ) mean radius or spheres rRelFuzz (oat ) dispersion of radius relative to rMean num (int ) number of spheres to be generated. If negavite (default), generate as many as possible with stochastic sizes, ending after a xed number of tries to place the sphere in space, else generate exactly num spheres with deterministic size distribution. periodic (bool ) whether the packing to be generated should be periodic porosity (oat ) initial guess for the iterative generation procedure (if num>1). The algorithm will be retrying until the number of generated spheres is num. The rst iteration tries with the provided porosity, but next iterations increase it if necessary (hence an initialy high porosity can speed-up the algorithm). If psdSizes is not dened, rRelFuzz (z) and num (N) are used so that the porosity given () is approximately achieved at the end of generation,
V (1) rm = 3 4 . The default is =0.5. The optimal value depends on 2 3 (1+z )N rRelFuzz or psdSizes.
psdSizes sieve sizes (particle diameters) when particle size distribution (PSD) is specied psdCumm cummulative fractions of particle sizes given by psdSizes; must be the same length as psdSizes and should be non-decreasing distributeMass (bool ) if True, given distribution will be used to distribute spheres mass rather than radius of them. seed number used to initialize the random number generator. Returns number of created spheres, which can be lower than num depending on the method used.
413
makeClumpCloud((Vector3)minCorner,
(Vector3)maxCorner,
(object)clumps[,
(bool)periodic=False[, (int)num=-1[, (int)seed=0 ]]]) int Create random loose packing of clumps within box given by minCorner and maxCorner. Clumps are selected with equal probability. At most num clumps will be positioned if num is positive; otherwise, as many clumps as possible will be put in space, until maximum number of attempts to place a new clump randomly is attained. :param seed: number used to initialize the random number generator. particleSD((Vector3)minCorner, (Vector3)maxCorner, (oat)rMean, (bool)periodic=False, (str)name, (int)numSph[, (object)radii=[] [, (object)passing=[] [, (bool)passingIsNotPercentageButCount=False[, (int)seed=0 ]]]]) int Create random packing enclosed in box given by minCorner and maxCorner, containing numSph spheres. Returns number of created spheres, which can be < num if the packing is too tight. The computation is done according to the given psd. particleSD2((object)radii, (object)passing, (int)numSph[, (bool)periodic=False[, (oat)cloudPorosity=0.8 [, (int)seed=0 ]]]) int Create random packing following the given particle size distribution (radii and volume/mass passing for each fraction) and total number of particles numSph. The cloud size (periodic or aperiodic) is computed from the PSD and is always cubic. particleSD_2d((Vector2)minCorner, (Vector2)maxCorner, (oat)rMean, (bool)periodic=False, (str)name, (int)numSph[, (object)radii=[] [, (object)passing=[] [, (bool)passingIsNotPercentageButCount=False[, (int)seed=0
]]]]) int Create random packing on XY plane, dened by minCorner and maxCorner, containing numSph spheres. Returns number of created spheres, which can be < num if the packing is too tight. The computation is done according to the given psd.
psd([(int)bins=50 [, (bool)mass=True ]]) tuple Return particle size distribution of the packing. :param int bins: number of bins between minimum and maximum diameter :param mass: Compute relative mass rather than relative particle count for each bin. Corresponds to distributeMass parameter for makeCloud. :returns: tuple of (cumm,edges), where cumm are cummulative fractions for respective diameters and edges are those diameter values. Dimension of both arrays is equal to bins+1. relDensity() oat Relative packing density, measured as sum of spheres volumes / aabb volume. (Sphere overlaps are ignored.) rotate((Vector3)axis, (oat)angle) None Rotate all spheres around packing center (in terms of aabb()), given axis and angle of the rotation. save((str)leName) None Save packing to external text le (will be overwritten). scale((oat)arg2 ) None Scale the packing around its center (in terms of aabb()) by given factor (may be negative). toList() list Return packing data as python list. toSimulation() Append spheres directly to the simulation. In addition calling O.bodies.append, this method also appropriately sets periodic cell information of the simulation.
>>> from yade import pack; from math import * >>> sp=pack.SpherePack()
Create random periodic packing with 20 spheres:
414
>>> sp.makeCloud((0,0,0),(5,5,5),rMean=.5,rRelFuzz=.5,periodic=True,num=20) 20
Virgin simulation is aperiodic:

>>> O.reset() >>> O.periodic False
Add generated packing to the simulation, rotated by 45 along +z

>>> sp.toSimulation(rot=Quaternion((0,0,1),pi/4),color=(0,0,1)) [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
Periodic properties are transferred to the simulation correctly, including rotation (this could be avoided by explicitly passing hSize=O.cell.hSize as an argument):
>>> O.periodic True >>> O.cell.refSize Vector3(5,5,5)
translate((Vector3)arg2 ) None Translate all spheres by given vector. class yade._packSpheres.SpherePackIterator __init__((SpherePackIterator)arg2 ) None next() tuple Spatial predicates for volumes (dened analytically or by triangulation). class yade._packPredicates.Predicate aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.PredicateBoolean(inherits Predicate) Boolean operation on 2 predicates (abstract class) A B __init__() Raises an exception This class cannot be instantiated from Python aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.PredicateDifference(inherits PredicateBoolean Predicate) Dierence (conjunction with negative predicate) of 2 predicates. A point has to be inside the rst and outside the second predicate. Can be constructed using the - operator on predicates: pred1 - pred2. A B
415
__init__((object)arg2, (object)arg3 ) None aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.PredicateIntersection(inherits PredicateBoolean Predicate) Intersection (conjunction) of 2 predicates. A point has to be inside both predicates. Can be constructed using the & operator on predicates: pred1 & pred2. A B __init__((object)arg2, (object)arg3 ) None aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.PredicateSymmetricDifference(inherits PredicateBoolean Predicate) SymmetricDierence (exclusive disjunction) of 2 predicates. A point has to be in exactly one predicate of the two. Can be constructed using the ^ operator on predicates: pred1 ^ pred2. A B __init__((object)arg2, (object)arg3 ) None aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.PredicateUnion(inherits PredicateBoolean Predicate) Union (non-exclusive disjunction) of 2 predicates. A point has to be inside any of the two predicates to be inside. Can be constructed using the | operator on predicates: pred1 | pred2. A B __init__((object)arg2, (object)arg3 ) None aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.inAlignedBox(inherits Predicate) Axis-aligned box predicate __init__((Vector3)minAABB, (Vector3)maxAABB) None Ctor taking minumum and maximum points of the box (as 3-tuples). aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.inCylinder(inherits Predicate) Cylinder predicate 416 Chapter 8. Yade modules
__init__((Vector3)centerBottom, (Vector3)centerTop, (oat)radius) None Ctor taking centers of the lateral walls (as 3-tuples) and radius. aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.inEllipsoid(inherits Predicate) Ellipsoid predicate __init__((Vector3)centerPoint, (Vector3)abc) None Ctor taking center of the ellipsoid (3-tuple) and its 3 radii (3-tuple). aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.inGtsSurface(inherits Predicate) GTS surface predicate __init__((object)surface[, (bool)noPad ]) None Ctor taking a gts.Surface() instance, which must not be modied during instance lifetime. The optional noPad can disable padding (if set to True), which speeds up calls several times. Note: padding checks inclusion of 6 points along +- cardinal directions in the pad distance from given point, which is not exact. aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 surf The associated gts.Surface object. class yade._packPredicates.inHyperboloid(inherits Predicate) Hyperboloid predicate __init__((Vector3)centerBottom, (Vector3)centerTop, (oat)radius, (oat)skirt ) None Ctor taking centers of the lateral walls (as 3-tuples), radius at bases and skirt (middle radius). aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.inParallelepiped(inherits Predicate) Parallelepiped predicate __init__((Vector3)o, (Vector3)a, (Vector3)b, (Vector3)c) None Ctor taking four points: o (for origin) and then a, b, c which dene endpoints of 3 respective edges from o. aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.inSphere(inherits Predicate) Sphere predicate.
417
__init__((Vector3)center, (oat)radius) None Ctor taking center (as a 3-tuple) and radius aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 class yade._packPredicates.notInNotch(inherits Predicate) Outside of innite, rectangle-shaped notch predicate __init__((Vector3)centerPoint, (Vector3)edge, (Vector3)normal, (oat)aperture) None Ctor taking point in the symmetry plane, vector pointing along the edge, plane normal and aperture size. The side inside the notch is edgenormal. Normal is made perpendicular to the edge. All vectors are normalized at construction time. aabb() tuple aabb( (Predicate)arg1) None center() Vector3 dim() Vector3 Computation of oriented bounding box for cloud of points. yade._packObb.cloudBestFitOBB((tuple)arg1) tuple Return (Vector3 center, Vector3 halfSize, Quaternion orientation) of best-t oriented bounding-box for given tuple of points (uses brute-force velome minimization, do not use for very large clouds).
8.7 yade.plot module

Module containing utility functions for plotting inside yade. See examples/simple-scene/simple-sceneplot.py or examples/concrete/uniax.py for example of usage.
yade.plot.data = {force: [nan, nan, 1000.0, nan, nan, 1000.0], sigma: [12, nan, nan, 12, nan, nan], eps Global dictionary containing all data values, common for all plots, in the form {name:[value,...],...}. Data should be added using plot.addData function. All [value,...] columns have the same length, they are padded with NaN if unspecied. yade.plot.plots = {i: (t,), i : (z1, v1)} dictionary x-name -> (yspec,...), where yspec is either y-name or (y-name,line-specication). If (yspec,...) is None, then the plot has meaning of image, which will be taken from respective eld of plot.imgData. yade.plot.labels = {} Dictionary converting names in data to human-readable names (TeX names, for instance); if a variable is not specied, it is left untranslated. yade.plot.live = True Enable/disable live plot updating. Disabled by default for now, since it has a few rough edges. yade.plot.liveInterval = 1 Interval for the live plot updating, in seconds. yade.plot.autozoom = True Enable/disable automatic plot rezooming after data update. yade.plot.plot(noShow=False, subPlots=True) Do the actual plot, which is either shown on screen (and nothing is returned: if noShow is False - note that your yade compilation should present qt4 feature so that gures can be displayed) or, if noShow is True, returned as matplotlibs Figure object or list of them. You can use
418
>>> from yade import plot >>> plot.resetData() >>> plot.plots={'foo':('bar',)} >>> plot.plot(noShow=True).savefig('someFile.pdf') >>> import os >>> os.path.exists('someFile.pdf') True
to save the gure to le automatically. Note: For backwards compatibility reasons, noShow option will return list of gures for multiple gures but a single gure (rather than list with 1 element) if there is only 1 gure. yade.plot.reset() Reset all plot-related variables (data, plots, labels) yade.plot.resetData() Reset all plot data; keep plots and labels intact. yade.plot.splitData() Make all plots discontinuous at this point (adds nans to all data elds) yade.plot.reverseData() Reverse yade.plot.data order. Useful for tension-compression test, where the initial (zero) state is loaded and, to make data continuous, last part must end in the zero state. yade.plot.addData(*d_in, **kw ) Add data from arguments name1=value1,name2=value2 to yade.plot.data. {name1:value1,name2:value2} is deprecated, but still supported) (the old
New data will be padded with nans, unspecied data will be nan (nans dont appear in graphs). This way, equal length of all data is assured so that they can be plotted one against any other.
>>> from yade import plot >>> from pprint import pprint >>> plot.resetData() >>> plot.addData(a=1) >>> plot.addData(b=2) >>> plot.addData(a=3,b=4) >>> pprint(plot.data) {'a': [1, nan, 3], 'b': [nan, 2, 4]}
Some sequence types can be given to addData; they will be saved in synthesized columns for individual components.
>>> plot.resetData() >>> plot.addData(c=Vector3(5,6,7),d=Matrix3(8,9,10, 11,12,13, 14,15,16)) >>> pprint(plot.data) {'c_x': [5.0], 'c_y': [6.0], 'c_z': [7.0], 'd_xx': [8.0], 'd_xy': [9.0], 'd_xz': [10.0], 'd_yy': [12.0], 'd_yz': [11.0], 'd_zx': [14.0], 'd_zy': [15.0], 'd_zz': [16.0]}
yade.plot.addAutoData() Add data by evaluating contents of plot.plots. Expressions rasing exceptions will be handled gracefully, but warning is printed for each. 8.7. yade.plot module 419
>>> from yade import plot >>> from pprint import pprint >>> O.reset() >>> plot.resetData() >>> plot.plots={'O.iter':('O.time',None,'numParticles=len(O.bodies)')} >>> plot.addAutoData() >>> pprint(plot.data) {'O.iter': [0], 'O.time': [0.0], 'numParticles': [0]}
Note that each item in plot.plots can be an expression to be evaluated (using the eval builtin); name=expression string, where name will appear as label in plots, and expression will be evaluated each time; a dictionary-like object current keys are labels of plots and current values are added to plot.data. The contents of the dictionary can change over time, in which case new lines will be created as necessary. A simple simulation with plot can be written in the following way; note how the energy plot is specied.
>>> from yade import plot, utils >>> plot.plots={'i=O.iter':(O.energy,None,'total energy=O.energy.total()')} >>> # we create a simple simulation with one ball falling down >>> plot.resetData() >>> O.bodies.append(utils.sphere((0,0,0),1)) 0 >>> O.dt=utils.PWaveTimeStep() >>> O.engines=[ ... ForceResetter(), ... GravityEngine(gravity=(0,0,-10),warnOnce=False), ... NewtonIntegrator(damping=.4,kinSplit=True), ... # get data required by plots at every step ... PyRunner(command='yade.plot.addAutoData()',iterPeriod=1,initRun=True) ... ] >>> O.trackEnergy=True >>> O.run(2,True) >>> pprint(plot.data) {'gravWork': [0.0, -25.13274...], 'i': [0, 1], 'kinRot': [0.0, 0.0], 'kinTrans': [0.0, 7.5398...], 'nonviscDamp': [0.0, 10.0530...], 'total energy': [0.0, -7.5398...]}
yade.plot.saveGnuplot(baseName, term=wxt, extension=None, timestamp=False, comment=None, title=None, varData=False) Save data added with plot.addData into (compressed) le and create .gnuplot le that attempts to mimick plots specied with plot.plots. Parameters baseName used for creating baseName.gnuplot (command le for gnuplot), associated baseName.data.bz2 (data) and output les (if applicable) in the form baseName.[plot number].extension term specify the gnuplot terminal; defaults to x11, in which case gnuplot will draw persistent windows to screen and terminate; other useful terminals are png, cairopdf and so on extension extension for baseName defaults to terminal name; ne for png for example; if you use cairopdf, you should also say extension='pdf' however timestamp (bool ) append numeric time to the basename
420
varData (bool ) whether le to plot will be declared as variable or be in-place in the plot expression comment a user comment (may be multiline) that will be embedded in the control le Returns name of the gnuplot le created. yade.plot.saveDataTxt(leName, vars=None) Save plot data into a (optionally compressed) text le. The rst line contains a comment (starting with #) giving variable name for each of the columns. This format is suitable for being loaded for further processing (outside yade) with numpy.genfromtxt function, which recognizes those variable names (creating numpy array with named entries) and handles decompression transparently.
>>> from yade import plot >>> from pprint import pprint >>> plot.reset() >>> plot.addData(a=1,b=11,c=21,d=31) # add some data here >>> plot.addData(a=2,b=12,c=22,d=32) >>> pprint(plot.data) {'a': [1, 2], 'b': [11, 12], 'c': [21, 22], 'd': [31, 32]} >>> plot.saveDataTxt('/tmp/dataFile.txt.bz2',vars=('a','b','c')) >>> import numpy >>> d=numpy.genfromtxt('/tmp/dataFile.txt.bz2',dtype=None,names=True) >>> d['a'] array([1, 2]) >>> d['b'] array([11, 12])
Parameters leName le to save data to; if it ends with .bz2 / .gz, the le will be compressed using bzip2 / gzip. vars Sequence (tuple/list/set) of variable names to be saved. If None (default), all variables in plot.plot are saved. yade.plot.savePlotSequence(leBase, stride=1, imgRatio=(5, 7), title=None, titleFrames=20, lastFrames=30 ) Save sequence of plots, each plot corresponding to one line in history. It is especially meant to be used for utils.makeVideo. Parameters stride only consider every stride-th line of history (default creates one frame per each line) title Create title frame, where lines of title are separated with newlines (\n) and optional subtitle is separated from title by double newline. titleFrames (int ) Create this number of frames with title (by repeating its lename), determines how long the title will stand in the movie. lastFrames (int ) Repeat the last frame this number of times, so that the movie does not end abruptly. Returns List of lenames with consecutive frames.
8.8 yade.post2d module

Module for 2d postprocessing, containing classes to project points from 3d to 2d in various ways, providing basic but exible framework for extracting arbitrary scalar values from bodies/interactions and plotting the results. There are 2 basic components: atteners and extractors.
8.8. yade.post2d module
421
The algorithms operate on bodies (default) or interactions, depending on the intr parameter of post2d.data.
8.8.1 Flatteners
Instance of classes that convert 3d (model) coordinates to 2d (plot) coordinates. Their interface is dened by the post2d.Flatten class (__call__, planar, normal).
8.8.2 Extractors
Callable objects returning scalar or vector value, given a body/interaction object. If a 3d vector is returned, Flattener.planar is called, which should return only in-plane components of the vector.
8.8.3 Example
This example can be found in examples/concrete/uniax-post.py
from yade import post2d import pylab # the matlab-like interface of matplotlib O.load('/tmp/uniax-tension.xml.bz2') # flattener that project to the xz plane flattener=post2d.AxisFlatten(useRef=False,axis=1) # return scalar given a Body instance extractDmg=lambda b: b.state.normDmg # will call flattener.planar implicitly # the same as: extractVelocity=lambda b: flattener.planar(b,b.state.vel) extractVelocity=lambda b: b.state.vel # create new figure pylab.figure() # plot raw damage post2d.plot(post2d.data(extractDmg,flattener)) # plot smooth damage into new figure pylab.figure(); ax,map=post2d.plot(post2d.data(extractDmg,flattener,stDev=2e-3)) # show color scale pylab.colorbar(map,orientation='horizontal') # raw velocity (vector field) plot pylab.figure(); post2d.plot(post2d.data(extractVelocity,flattener)) # smooth velocity plot; data are sampled at regular grid pylab.figure(); ax,map=post2d.plot(post2d.data(extractVelocity,flattener,stDev=1e-3)) # save last (current) figure to file pylab.gcf().savefig('/tmp/foo.png') # show the figures pylab.show()
class yade.post2d.AxisFlatten(inherits Flatten) __init__() :param bool useRef: use reference positions rather than actual positions (only meaningful when operating on Bodies) :param {0,1,2} axis: axis normal to the plane; the return value will be simply position with this component dropped. normal()
422
planar() class yade.post2d.CylinderFlatten(inherits Flatten) Class for converting 3d point to 2d based on projection onto plane from circle. The y-axis in the projection corresponds to the rotation axis; the x-axis is distance form the axis. __init__() :param useRef: (bool) use reference positions rather than actual positions :param axis: axis of the cylinder, {0,1,2} normal() planar() class yade.post2d.Flatten Abstract class for converting 3d point into 2d. Used by post2d.data2d. normal() Given position and vector value, return lenght of the vector normal to the at plane. planar() Given position and vector value, project the vector value to the at plane and return its 2 in-plane components. class yade.post2d.HelixFlatten(inherits Flatten) Class converting 3d point to 2d based on projection from helix. The y-axis in the projection corresponds to the rotation axis __init__() :param bool useRef: use reference positions rather than actual positions :param (min,max) thetaRange: bodies outside this range will be discarded :param oat dH_dTheta: inclination of the spiral (per radian) :param {0,1,2} axis: axis of rotation of the spiral :param oat periodStart: height of the spiral for zero angle normal() planar() yade.post2d.data(extractor, attener, intr=False, onlyDynamic=True, stDev=None, relThreshold=3.0, perArea=0, div=(50, 50), margin=(0, 0), radius=1) Filter all bodies/interactions, project them to 2d and extract required scalar value; return either discrete array of positions and values, or smoothed data, depending on whether the stDev value is specied. The intr parameter determines whether we operate on bodies or interactions; the extractor provided should expect to receive body/interaction. Parameters extractor (callable) receives Body (or Interaction, if intr is True) instance, should return scalar, a 2-tuple (vector elds) or None (to skip that body/interaction) attener (callable) post2d.Flatten instance, receiving body/interaction, returns its 2d coordinates or None (to skip that body/interaction) intr (bool ) operate on interactions rather than bodies onlyDynamic (bool ) skip all non-dynamic bodies stDev (oat/None) standard deviation for averaging, enables smoothing; None (default) means raw mode, where discrete points are returned relThreshold (oat ) threshold for the gaussian weight function relative to stDev (smooth mode only) perArea (int ) if 1, compute weightedSum/weightedArea rather than weighted average (weightedSum/sumWeights); the rst is useful to compute average stress; if 2, compute averages on subdivision elements, not using weight function div ((int,int)) number of cells for the gaussian grid (smooth mode only)
8.8. yade.post2d module
423
margin ((oat,oat)) x,y margins around bounding box for data (smooth mode only) radius (oat/callable) Fallback value for radius (for raw plotting) for nonspherical bodies or interactions; if a callable, receives body/interaction and returns radius Returns dictionary Returned dictionary always containing keys type (one of rawScalar,rawVector,smoothScalar,smoothVector, depending on value of smooth and on return value from extractor), x, y, bbox. Raw data further contains radii. Scalar elds contain val (value from extractor ), vector elds have valX and valY (2 components returned by the extractor ). yade.post2d.plot(data, axes=None, alpha=0.5, clabel=True, cbar=False, aspect=equal, **kw ) Given output from post2d.data, plot the scalar as discrete or smooth plot. For raw discrete data, plot lled circles with radii of particles, colored by the scalar value. For smooth discrete data, plot image with optional contours and contour labels. For vector data (raw or smooth), plot quiver (vector eld), with arrows colored by the magnitude. Parameters axes matplotlib.axesinstance where the gure will be plotted; if None, will be created from scratch. data value returned by post2d.data clabel (bool ) show contour labels (smooth mode only), or annotate cells with numbers inside (with perArea==2) cbar (bool ) show colorbar (equivalent to calling pylab.colorbar(mappable) on the returned mappable) Returns tuple of (axes,mappable); mappable can be used in further calls to pylab.colorbar.
8.9 yade.qt module

Common initialization core for yade. This le is executed when anything is imported from yade for the rst time. It loads yade plugins and injects c++ class constructors to the __builtins__ (that might change in the future, though) namespace, making them available everywhere. yade.qt.bin(QTextStream) QTextStream yade.qt.hex(QTextStream) QTextStream yade.qt.oct(QTextStream) QTextStream class yade.qt._GLViewer.GLViewer __init__() Raises an exception This class cannot be instantiated from Python axes Show arrows for axes. center([(bool)median=True ]) None Center view. View is centered either so that all bodies t inside (median = False), or so that 75% of bodies t inside (median = True). close() None 424 Chapter 8. Yade modules
eyePosition Camera position. fitAABB((Vector3)mn, (Vector3)mx ) None Adjust scene bounds so that Axis-aligned bounding box given by its lower and upper corners mn, mx ts in. fitSphere((Vector3)center, (oat)radius) None Adjust scene bounds so that sphere given by center and radius ts in. fps Show frames per second indicator. grid Display square grid in zero planes, as 3-tuple of bools for yz, xz, xy planes. loadState((int)slot ) None Load display parameters from slot saved previously into, identied by its number. lookAt Point at which camera is directed. ortho Whether orthographic projection is used; if false, use perspective projection. saveState((int)slot ) None Save display parameters into numbered memory slot. Saves state for both GLViewer and associated OpenGLRenderer. scale Scale of the view (?) sceneRadius Visible scene radius. screenSize Size of the viewers window, in scree pixels selection showEntireScene() None timeDisp Time displayed on in the vindow; is a string composed of characters r, v, i standing respectively for real time, virtual time, iteration number. upVector Vector that will be shown oriented up on the screen. viewDir Camera orientation (as vector). yade.qt._GLViewer.Renderer() OpenGLRenderer Return the active OpenGLRenderer object. yade.qt._GLViewer.View() GLViewer Create a new 3d view. yade.qt._GLViewer.center() None Center all views. yade.qt._GLViewer.views() list Return list of all open qt.GLViewer objects
8.10 yade.timing module

Functions for accessing timing information stored in engines and functors. See Timing section of the programmers manual, wiki page for some examples. 8.10. yade.timing module 425
yade.timing.reset() Zero all timing data. yade.timing.stats() Print summary table of timing information from engines and functors. Absolute times as well as percentages are given. Sample output:
Name Count Time Rel. time ------------------------------------------------------------------------------------------------------ForceResetter 102 2150us 0.02% "collider" 5 64200us 0.60% InteractionLoop 102 10571887us 98.49% "combEngine" 102 8362us 0.08% "newton" 102 73166us 0.68% "cpmStateUpdater" 1 9605us 0.09% PyRunner 1 136us 0.00% "plotDataCollector" 1 291us 0.00% TOTAL 10733564us 100.00%
sample output (compiled with -DCMAKE_CXX_FLAGS=-DUSE_TIMING_DELTAS option):

Name Count Time Rel. time ------------------------------------------------------------------------------------------------------ForceResetter 102 2150us 0.02% "collider" 5 64200us 0.60% InteractionLoop 102 10571887us 98.49% Ig2_Sphere_Sphere_ScGeom 1222186 1723168us 16.30% Ig2_Sphere_Sphere_ScGeom 1222186 1723168us 100.00% Ig2_Facet_Sphere_ScGeom 753 1157us 0.01% Ig2_Facet_Sphere_ScGeom 753 1157us 100.00% Ip2_CpmMat_CpmMat_CpmPhys 11712 26015us 0.25% end of Ip2_CpmPhys 11712 26015us 100.00% Ip2_FrictMat_CpmMat_FrictPhys 0 0us 0.00% Law2_ScGeom_CpmPhys_Cpm 3583872 4819289us 45.59% GO A 1194624 1423738us 29.54% GO B 1194624 1801250us 37.38% rest 1194624 1594300us 33.08% TOTAL 3583872 4819289us 100.00% Law2_ScGeom_FrictPhys_CundallStrack 0 0us 0.00% "combEngine" 102 8362us 0.08% "newton" 102 73166us 0.68% "cpmStateUpdater" 1 9605us 0.09% PyRunner 1 136us 0.00% "plotDataCollector" 1 291us 0.00% TOTAL 10733564us 100.00%
8.11 yade.utils module

Heap of functions that dont (yet) t anywhere else. Devs: please DO NOT ADD more functions here, it is getting too crowded! yade.utils.NormalRestitution2DampingRate(en) Compute the normal damping rate as a function of the normal coecient of restitution en . For en 0, 1 damping rate equals log en 2 e2 n+
426
yade.utils.SpherePWaveTimeStep(radius, density, young) Compute P-wave critical timestep for a single (presumably representative) sphere, using formula for P-Wave propagation speed tc = r . If you want to compute minimum critical timestep
E/
for all spheres in the simulation, use utils.PWaveTimeStep instead.

>>> SpherePWaveTimeStep(1e-3,2400,30e9) 2.8284271247461903e-07
class yade.utils.TableParamReader Class for reading simulation parameters from text le. Each parameter is represented by one column, each parameter set by one line. Colums are separated by blanks (no quoting). First non-empty line contains column titles (without quotes). You may use special column named description to describe this parameter set; if such colum is absent, description will be built by concatenating column names and corresponding values (param1=34,param2=12.22,param4=foo) from columns ending in ! (the ! is not included in the column name) from all columns, if no columns end in !. Empty lines within the le are ignored (although counted); # starts comment till the end of line. Number of blank-separated columns must be the same for all non-empty lines. A special value = can be used instead of parameter value; value from the previous non-empty line will be used instead (works recursively). This class is used by utils.readParamsFromTable. __init__() Setup the reader class, read data into memory. paramDict() Return dictionary containing data from le given to constructor. Keys are line numbers (which might be non-contiguous and refer to real line numbers that one can see in text editors), values are dictionaries mapping parameter names to their values given in the le. The special value = has already been interpreted, ! (bangs) (if any) were already removed from column titles, description column has already been added (if absent). yade.utils.aabbDim(cuto=0.0, centers=False) Return dimensions of the axis-aligned bounding box, optionally with relative part cuto cut away. yade.utils.aabbExtrema2d(pts) Return 2d bounding box for a sequence of 2-tuples. yade.utils.aabbWalls(extrema=None, thickness=0, oversizeFactor=1.5, **kw ) Return 6 boxes that will wrap existing packing as walls from all sides; extrema are extremal points of the Aabb of the packing (will be calculated if not specied) thickness is wall thickness (will be 1/10 of the X-dimension if not specied) Walls will be enlarged in their plane by oversizeFactor. returns list of 6 wall Bodies enclosing the packing, in the order minX,maxX,minY,maxY,minZ,maxZ. yade.utils.avgNumInteractions(cuto=0.0, skipFree=False, considerClumps=False) Return average numer of interactions per particle, also known as coordination number Z. This number is dened as Z = 2C/N
where C is number of contacts and N is number of particles. When clumps are present, number of particles is the sum of standalone spheres plus the sum of clumps. Clumps are considered in the calculation if cuto != 0 or skipFree = True. If cuto=0 (default) and skipFree=False (default) one needs to set considerClumps=True to consider clumps in the calculation.
8.11. yade.utils module
427
With skipFree, particles not contributing to stable state of the packing are skipped, following equation (8) given in [Thornton2000]: Zm = 2C N1 N N0 N1
Parameters cuto cut some relative part of the samples bounding box away. skipFree see above. considerClumps also consider clumps if cuto=0 and skipFree=False; for further explanation see above. yade.utils.box(center, extents, orientation=Quaternion((1, 0, 0), 0), dynamic=None, xed=False, wire=False, color=None, highlight=False, material=-1, mask=1) Create box (cuboid) with given parameters. Parameters extents (Vector3 ) half-sizes along x,y,z axes See utils.spheres documentation for meaning of other parameters. yade.utils.chainedCylinder(begin=Vector3(0, 0, 0), end=Vector3(1, 0, 0), radius=0.2, dynamic=None, xed=False, wire=False, color=None, highlight=False, material=-1, mask=1) Create and connect a chainedCylinder with given parameters. The shape generated by repeted calls of this function is the Minkowski sum of polyline and sphere. Parameters radius (Real ) radius of sphere in the Minkowski sum. begin (Vector3 ) rst point positioning the line in the Minkowski sum last (Vector3 ) last point positioning the line in the Minkowski sum In order to build a correct chain, last point of element of rank N must correspond to rst point of element of rank N+1 in the same chain (with some tolerance, since bounding boxes will be used to create connections. Returns Body object with the ChainedCylinder shape. class yade.utils.clumpTemplate Create a clump template by a list of relative radii and a list of relative positions. Both lists must have the same length. Parameters relRadii ([oat,oat,...] ) list of relative radii (minimum length = 2) relPositions ([Vector3,Vector3,...] ) list of relative positions (minimum length = 2) yade.utils.defaultMaterial() Return default material, when creating bodies with utils.sphere and friends, material is unspecied and there is no shared material dened yet. By default, this function returns:
.. code-block:: python
FrictMat(density=1e3,young=1e7,poisson=.3,frictionAngle=.5,label=defaultMat) yade.utils.facet(vertices, dynamic=None, xed=True, wire=True, color=None, light=False, noBound=False, material=-1, mask=1, chain=-1) Create facet with given parameters. Parameters vertices ([Vector3,Vector3,Vector3] ) coordinates of vertices in the global coordinate system. 428 Chapter 8. Yade modules high-
wire (bool ) if True, facets are shown as skeleton; otherwise facets are lled noBound (bool ) set Body.bounded color (Vector3-or-None) color of the facet; random color will be assigned if None. See utils.spheres documentation for meaning of other parameters. yade.utils.fractionalBox(fraction=1.0, minMax=None) retrurn (min,max) that is the original minMax box (or aabb of the whole simulation if not specied) linearly scaled around its center to the fraction factor yade.utils.gridConnection(id1, id2, radius, wire=False, color=None, highlight=False, material=-1, mask=none, cellDist=None) yade.utils.gridNode(center, radius, dynamic=None, xed=False, wire=False, color=None, highlight=False, material=-1) yade.utils.loadVars(mark=None) Load variables from utils.saveVars, which are saved inside the simulation. If mark==None, all save variables are loaded. Otherwise only those with the mark passed. yade.utils.makeVideo(frameSpec, out, renameNotOverwrite=True, fps=24, kbps=6000, bps=None) Create a video from external image les using mencoder. Two-pass encoding using the default mencoder codec (mpeg4) is performed, running multi-threaded with number of threads equal to number of OpenMP threads allocated for Yade. Parameters frameSpec wildcard | sequence of lenames. If list or tuple, lenames to be encoded in given order; otherwise wildcard understood by mencoders mf:// URI option (shell wildcards such as /tmp/snap-*.png or and printf-style pattern like /tmp/snap-%05d.png) out (str ) le to save video into renameNotOverwrite (bool ) if True, existing same-named video le will have -number appended; will be overwritten otherwise. fps (int ) Frames per second (-mf fps=) kbps (int ) Bitrate (-lavcopts vbitrate=) in kb/s yade.utils.perpendicularArea(axis) Return area perpendicular to given axis (0=x,1=y,2=z) generated by bodies for which the function consider returns True (defaults to returning True always) and which is of the type Sphere. yade.utils.plotDirections(aabb=(), mask=0, bins=20, numHist=True, noShow=False) Plot 3 histograms for distribution of interaction directions, in yz,xz and xy planes and (optional but default) histogram of number of interactions per body. Returns If noShow is False, displays the gure and returns nothing. If noShow, the gure object is returned without being displayed (works the same way as plot.plot). yade.utils.plotNumInteractionsHistogram(cuto=0.0 ) Plot histogram with number of interactions per body, optionally cutting away cuto relative axisaligned box from specimen margin. yade.utils.psd(bins=5, mass=True, mask=-1) Calculates particle size distribution. Parameters bins (int ) number of bins mass (bool ) if true, the mass-PSD will be calculated mask (int ) Body.mask for the body Returns
429
binsSizes: list of bins sizes binsProc: how much material (in percents) are in the bin, cumulative binsSumCum: how much material (in units) are in the bin, cumulative binsSizes, binsProc, binsSumCum yade.utils.randomColor() Return random Vector3 with each component in interval 01 (uniform distribution) yade.utils.randomizeColors(onlyDynamic=False) Assign random colors to Shape::color. If onlyDynamic is true, only dynamic bodies will have the color changed. yade.utils.readParamsFromTable(tableFileLine=None, noTableOk=True, unknownOk=False, **kw ) Read parameters from a le and assign them to __builtin__ variables. The format of the le is as follows (commens starting with # and empty lines allowed):
# commented lines allowed anywhere name1 name2 # first non-blank line are column headings # empty line is OK, with or without comment val1 val2 # 1st parameter set val2 val2 # 2nd
Assigned tags (the description column is synthesized if absent,see utils.TableParamReader); O.tags[description]= # assigns the description column; might be synthesized O.tags[params]=name1=val1,name2=val2, # all explicitly assigned parameters O.tags[defaultParams]=unassignedName1=defaultValue1, # parameters that were left at their defaults O.tags[d.id]=O.tags[id]+.+O.tags[description] O.tags[id.d]=O.tags[description]+.+O.tags[id] All parameters (default as well as settable) are saved using utils.saveVars('table'). Parameters tableFile text le (with one value per blank-separated columns) tableLine (int ) number of line where to get the values from noTableOk (bool ) if False, raise exception if the le cannot be open; use default values otherwise unknownOk (bool ) do not raise exception if unknown column name is found in the le, and assign it as well Returns number of assigned parameters yade.utils.replaceCollider(colliderEngine) Replaces collider (Collider) engine with the engine supplied. Raises error if no collider is in engines. yade.utils.runningInBatch() Tell whether we are running inside the batch or separately. yade.utils.saveVars(mark=, loadNow=True, **kw ) Save passed variables into the simulation so that it can be recovered when the simulation is loaded again. For example, variables a, b and c are dened. To save them, use:
>>> >>> >>> (1, saveVars('something',a=1,b=2,c=3) from yade.params.something import * a,b,c 2, 3)
those variables will be save in the .xml le, when the simulation itself is saved. To recover those variables once the .xml is loaded again, use
430
>>> loadVars('something')
and they will be dened in the yade.params.mark module. utils.loadVars after saving automatically. If something already exists, given variables will be inserted.
The loadNow parameter calls
yade.utils.sphere(center, radius, dynamic=None, xed=False, wire=False, color=None, highlight=False, material=-1, mask=1) Create sphere with given parameters; mass and inertia computed automatically. Last assigned material is used by default (material = -1), and utils.defaultMaterial() will be used if no material is dened at all. Parameters center (Vector3 ) center radius (oat ) radius dynamic (oat ) deprecated, see xed xed (oat ) generate the body with all DOFs blocked? material specify Body.material; dierent types are accepted: int: O.materials[material] will be used; as a special case, if material==1 and there is no shared materials dened, utils.defaultMaterial() will be assigned to O.materials[0] string: label of an existing material that will be used Material instance: this instance will be used callable: will be called without arguments; returned Material value will be used (Material factory object, if you like) mask (int ) Body.mask for the body wire display as wire sphere? highlight highlight this body in the viewer? Vector3-or-None bodys color, as normalized RGB; random color will be assigned if None. Returns A Body instance with desired characteristics. Creating default shared material if none exists neither is given:
>>> >>> >>> 0 >>> >>> 1 O.reset() from yade import utils len(O.materials) s0=utils.sphere([2,0,0],1) len(O.materials)
Instance of material can be given:

>>> s1=utils.sphere([0,0,0],1,wire=False,color=(0,1,0),material=ElastMat(young=30e9,density=2e3)) >>> s1.shape.wire False >>> s1.shape.color Vector3(0,1,0) >>> s1.mat.density 2000.0
Material can be given by label:
431
>>> O.materials.append(FrictMat(young=10e9,poisson=.11,label='myMaterial')) 1 >>> s2=utils.sphere([0,0,2],1,material='myMaterial') >>> s2.mat.label 'myMaterial' >>> s2.mat.poisson 0.11
Finally, material can be a callable object (taking no arguments), which returns a Material instance. Use this if you dont call this function directly (for instance, through yade.pack.randomDensePack), passing only 1 material parameter, but you dont want material to be shared. For instance, randomized material properties can be created like this:
>>> >>> ... >>> >>> import random def matFactory(): return ElastMat(young=1e10*random.random(),density=1e3+1e3*random.random()) s3=utils.sphere([0,2,0],1,material=matFactory) s4=utils.sphere([1,2,0],1,material=matFactory)
yade.utils.trackPerfomance(updateTime=5 ) Track perfomance of a simulation. (Experimental) Will create new thread to produce some plots. Useful for track perfomance of long run simulations (in bath mode for example). yade.utils.typedEngine(name) Return rst engine from current O.engines, identied by its type (as string). For example:
>>> from yade import utils >>> O.engines=[InsertionSortCollider(),NewtonIntegrator(),GravityEngine()] >>> utils.typedEngine("NewtonIntegrator") == O.engines[1] True
yade.utils.uniaxialTestFeatures(lename=None, areaSections=10, axis=-1, distFactor=2.2, **kw ) Get some data about the current packing useful for uniaxial test: 1.Find the dimensions that is the longest (uniaxial loading axis) 2.Find the minimum cross-section area of the specimen by examining several (areaSections) sections perpendicular to axis, computing area of the convex hull for each one. This will work also for non-prismatic specimen. 3.Find the bodies that are on the negative/positive boundary, to which the straining condition should be applied. Parameters lename if given, spheres will be loaded from this le (ASCII format); if not, current simulation will be used. areaSection (oat ) number of section that will be used to estimate crosssection axis ({0,1,2}) if given, force strained axis, rather than computing it from predominant length Returns dictionary with keys negIds, posIds, axis, area. Warning: The function utils.approxSectionArea uses convex hull algorithm to nd the area, but the implementation is reported to be buggy (bot works in some cases). Always check this number, or x the convex hull algorithm (it is documented in the source, see py/_utils.cpp). yade.utils.vmData() Return memory usage data from Linuxs /proc/[pid]/status, line VmData.
432
yade.utils.voxelPorosityTriaxial(triax, resolution=200, oset=0 ) Calculate the porosity of a sample, given the TriaxialCompressionEngine. A function utils.voxelPorosity is invoked, with the volume of a box enclosed by TriaxialCompressionEngine walls. The additional parameter oset allows using a smaller volume inside the box, where each side of the volume is at oset distance from the walls. By this way it is possible to nd a more precise porosity of the sample, since at walls contact the porosity is usually reduced. A recommended value of oset is bigger or equal to the average radius of spheres inside. The value of resolution depends on size of spheres used. It can be calibrated by invoking voxelPorosityTriaxial with oset=0 and comparing the result with TriaxialCompressionEngine.porosity. After calibration, the oset can be set to radius, or a bigger value, to get the result. Parameters triax the TriaxialCompressionEngine handle resolution voxel grid resolution oset oset distance Returns the porosity of the sample inside given volume Example invocation:
from yade import utils rAvg=0.03 TriaxialTest(numberOfGrains=200,radiusMean=rAvg).load() O.dt=-1 O.run(1000) O.engines[4].porosity 0.44007807740143889 utils.voxelPorosityTriaxial(O.engines[4],200,0) 0.44055412500000002 utils.voxelPorosityTriaxial(O.engines[4],200,rAvg) 0.36798199999999998
yade.utils.waitIfBatch() Block the simulation if running inside a batch. Typically used at the end of script so that it does not nish prematurely in batch mode (the execution would be ended in such a case). yade.utils.wall(position, axis, sense=0, color=None, material=-1, mask=1) Return ready-made wall body. Parameters position (oat-or-Vector3 ) center of the wall. If oat, it is the position along given axis, the other 2 components being zero axis ({0,1,2}) orientation of the wall normal (0,1,2) for x,y,z (sc. planes yz, xz, xy) sense ({-1,0,1}) sense in which to interact (0: both, -1: negative, +1: positive; see Wall) See utils.spheres documentation for meaning of other parameters. yade.utils.xMirror(half ) Mirror a sequence of 2d points around the x axis (changing sign on the y coord). The sequence should start up and then it will wrap from y downwards (or vice versa). If the last points x coord is zero, it will not be duplicated. yade._utils.PWaveTimeStep() oat Get timestep accoring to the velocity of P-Wave propagation; computed from sphere radii, rigidities and masses. yade._utils.RayleighWaveTimeStep() oat Determination of time step according to Rayleigh wave speed of force propagation.
433
yade._utils.aabbExtrema([(oat)cuto=0.0 [, (bool)centers=False ]]) tuple Return coordinates of box enclosing all bodies Parameters centers (bool ) do not take sphere radii in account, only their centroids cuto (oat01) relative dimension by which the box will be cut away at its boundaries. Returns (lower corner, upper corner) as (Vector3,Vector3) yade._utils.approxSectionArea((oat)arg1, (int)arg2 ) oat Compute area of convex hull when when taking (swept) spheres crossing the plane at coord, perpendicular to axis. yade._utils.bodyNumInteractionsHistogram([(tuple)aabb ]) tuple yade._utils.bodyStressTensors() list Compute and return a table with per-particle stress tensors. Each tensor represents the average stress in one particle, obtained from the contour integral of applied load as detailed below. This denition is considering each sphere as a continuum. It can be considered exact in the context of spheres at static equilibrium, interacting at contact points with negligible volume changes of the solid phase (this last assumption is not restricting possible deformations and volume changes at the packing scale). Proof: First, we remark the identity: ij = ik kj = xi,k kj = (xi kj ),k xi kj,k . Atequilibrium, the . Consequently, after divergence theorem: divergence of stress is null: kj,k = 0 1 1 1 1 b b dV = ( x ) dV = x n dS = ij i kj ,k i kj k b xi fj . V V V V V V V The last equality is implicitely based on the representation of external loads as Dirac distributions whose zeros are the so-called contact points: 0-sized surfaces on which the contact forces are applied, located at xi in the deformed conguration. A weighted average of per-body stresses will give the average stress inside the solid phase. There is a simple relation between the stress inside the solid phase and the stress in an equivalent continuum in the absence of uid pressure. For porosity n, the relation reads: equ. = (1 n)solid . ij ij yade._utils.calm([(int)mask=-1 ]) None Set translational and rotational velocities of all bodies to zero. yade._utils.coordsAndDisplacements((int)axis[, (tuple)Aabb=() ]) tuple Return tuple of 2 same-length lists for coordinates and displacements (coordinate minus reference coordinate) along given axis (1st arg); if the Aabb=((x_min,y_min,z_min),(x_max,y_max,z_max)) box is given, only bodies within this box will be considered. yade._utils.createInteraction((int)id1, (int)id2 ) Interaction Create interaction between given bodies by hand. Current engines are searched for IGeomDispatcher and IPhysDispatcher (might be both hidden in InteractionLoop). Geometry is created using force parameter of the geometry dispatcher, wherefore the interaction will exist even if bodies do not spatially overlap and the functor would return false under normal circumstances. Warning: This function will very likely behave incorrectly for periodic simulations (though it could be extended it to handle it farily easily). yade._utils.elasticEnergy((tuple)arg1) oat yade._utils.fabricTensor([(bool)splitTensor=False[, (bool)revertSign=False[, (oat)thresholdForce=nan ]]]) tuple Compute the fabric tensor of the periodic cell. The original paper can be found in [Satake1982]. Parameters
434
splitTensor (bool ) split the fabric tensor into two parts related to the strong and weak contact forces respectively. revertSign (bool ) it must be set to true if the contact laws convention takes compressive forces as positive. thresholdForce (Real ) if the fabric tensor is split into two parts, a threshold value can be specied otherwise the mean contact force is considered by default. It is worth to note that this value has a sign and the user needs to set it according to the convention adopted for the contact law. To note that this value could be set to zero if one wanted to make distinction between compressive and tensile forces. yade._utils.flipCell([(Matrix3)ip=Matrix3(0, 0, 0, 0, 0, 0, 0, 0, 0) ]) Matrix3 Flip periodic cell so that angles between R3 axes and transformed axes are as small as possible. This function relies on the fact that periodic cell denes by repetition or its corners regular grid of points in R3 ; however, all cells generating identical grid are equivalent and can be ipped one over another. This necessiatates adjustment of Interaction.cellDist for interactions that cross boundary and didnt before (or vice versa), and re-initialization of collider. The ip argument can be used to specify desired ip: integers, each column for one axis; if zero matrix, best t (minimizing the angles) is computed automatically. In c++, this function is accessible as Shop::flipCell. Warning: This function is currently broken and should not be used. yade._utils.forcesOnCoordPlane((oat)arg1, (int)arg2 ) Vector3 yade._utils.forcesOnPlane((Vector3)planePt, (Vector3)normal ) Vector3 Find all interactions deriving from NormShearPhys that cross given plane and sum forces (both normal and shear) on them. Parameters planePt (Vector3 ) a point on the plane normal (Vector3 ) plane normal (will be normalized). yade._utils.getBodyIdsContacts([(int)bodyID=0 ]) list Get a list of body-ids, which contacts the given body. yade._utils.getCapillaryStress([(oat)volume=0 ]) Matrix3 Compute and return Love-Weber capillary stress tensor: b cap,b 1 cap = V , where the sum is over all interactions, with l the branch vector ij b li fj (joining centers of the bodies) and fcap is the capillary force. V can be passed to the function. If it is not, it will be equal to one in non-periodic cases, or equal to the volume of the cell in periodic cases. Only the CapillaryPhys interaction type is supported presently. yade._utils.getSpheresMass([(int)mask=-1 ]) oat Compute the total mass of spheres in the simulation (might crash for now if dynamic bodies are not spheres), mask parameter is considered yade._utils.getSpheresVolume([(int)mask=-1 ]) oat Compute the total volume of spheres in the simulation (might crash for now if dynamic bodies are not spheres), mask parameter is considered yade._utils.getStress([(oat)volume=0 ]) Matrix3 Compute and return Love-Weber stress tensor: b b 1 ij = V b li fj , where the sum is over all interactions, with l the branch vector (joining centers of the bodies) and f is the contact force. V can be passed to the function. If it is not, it will be equal to one in non-periodic cases, or equal to the volume of the cell in periodic cases.
435
yade._utils.getViscoelasticFromSpheresInteraction((oat)tc, (oat)en, (oat)es) dict Compute viscoelastic interaction parameters from analytical solution of a pair spheres collision problem: ) m( 2 + (ln en )2 2 tc 2m cn = tc ( ) 2m 2 kt = + (ln et )2 2 7 tc 2m ct = ln et 7 tc kn = where kn , cn are normal elastic and viscous coecients and kt , ct shear elastic and viscous coecients. For details see [Pournin2001]. Parameters m (oat ) sphere mass m tc (oat ) collision time tc en (oat ) normal restitution coecient en es (oat ) tangential restitution coecient es Returns dictionary with keys kn (the value of kn ), cn (cn ), kt (kt ), ct (ct ). yade._utils.growParticles((oat)multiplier [, (bool)updateMass=True[, (bool)dynamicOnly=True ]]) None Change the size of spheres and sphere clumps by the multiplier. If updateMass=True, then the mass is updated. dynamicOnly=True is mandatory in many cases since in current implementation the function would crash on the non-spherical and non-dynamic bodies (e.g. facets, boxes, etc.) yade._utils.highlightNone() None Reset highlight on all bodies. yade._utils.inscribedCircleCenter((Vector3)v1, (Vector3)v2, (Vector3)v3 ) Vector3 Return center of inscribed circle for triangle given by its vertices v1, v2, v3. yade._utils.interactionAnglesHistogram((int)axis[, (int)mask [, (int)bins[, (tuple)aabb ]]]) tuple yade._utils.intrsOfEachBody() list returns list of lists of interactions of each body yade._utils.kineticEnergy([(bool)ndMaxId=False ]) object Compute overall kinetic energy of the simulation as 1( 2 ) T mi v2 i + (I ) .
For aspherical bodies, the inertia tensor I is transformed to global frame, before multiplied by , therefore the value should be accurate. yade._utils.maxOverlapRatio() oat Return maximum overlap ration in interactions (with ScGeom) of two spheres. The ratio is comuN puted as 2(r1 r2 )/r1 +r2 , where uN is the current overlap distance and r1 , r2 are radii of the two spheres in contact. yade._utils.negPosExtremeIds((int)axis[, (oat)distFactor ]) tuple Return list of ids for spheres (only) that are on extremal ends of the specimen along given axis; distFactor multiplies their radius so that sphere that do not touch the boundary coordinate can also be returned. 436 Chapter 8. Yade modules
yade._utils.normalShearStressTensors([(bool)compressionPositive=False[, (bool)splitNormalTensor=False[, (oat)thresholdForce=nan ]]]) tuple Compute overall stress tensor of the periodic cell decomposed in 2 parts, one contributed by normal forces, the other by shear forces. The formulation can be found in [Thornton2000], eq. (3): ij = 2 2 RNni nj + RT ni tj V V
where V is the cell volume, R is contact radius (in our implementation, current distance between particle centroids), n is the normal vector, t is a vector perpendicular to n, N and T are norms of normal and shear forces. Parameters splitNormalTensor (bool ) if true the function returns normal stress tensor split into two parts according to the two subnetworks of strong an weak forces. thresholdForce (Real ) threshold value according to which the normal stress tensor can be split (e.g. a zero value would make distinction between tensile and compressive forces). yade._utils.numIntrsOfEachBody() list returns list of number of interactions of each body yade._utils.pointInsidePolygon((tuple)arg1, (object)arg2 ) bool yade._utils.porosity([(oat)volume=-1 ]) oat Vs Compute packing poro sity V where V is overall volume and Vs is volume of spheres. V Parameters volume (oat ) overall volume which must be specied for aperiodic simulations. For periodic simulations, current volume of the Cell is used. yade._utils.ptInAABB((Vector3)arg1, (Vector3)arg2, (Vector3)arg3 ) bool Return True/False whether the point p is within box given by its min and max corners yade._utils.scalarOnColorScale((oat)arg1, (oat)arg2, (oat)arg3 ) Vector3 yade._utils.setContactFriction((oat)angleRad ) None Modify the friction angle (in radians) inside the material classes and existing contacts. The friction for non-dynamic bodies is not modied. yade._utils.setNewVerticesOfFacet((Body)b, (Vector3)v1, (Vector3)v2, (Vector3)v3 ) None Sets new vertices (in global coordinates) to given facet. yade._utils.setRefSe3() None Set reference positions and orientations of all bodies equal to their current positions and orientations. yade._utils.shiftBodies((list)ids, (Vector3)shift ) oat Shifts bodies listed in ids without updating their velocities. yade._utils.spiralProject((Vector3)pt, (oat)dH_dTheta[, (int)axis=2 [, (oat)periodStart=nan[, (oat)theta0=0 ]]]) tuple yade._utils.sumFacetNormalForces((object)ids[, (int)axis=-1 ]) oat Sum force magnitudes on given bodies (must have shape of the Facet type), considering only part of forces perpendicular to each facets face; if axis has positive value, then the specied axis (0=x, 1=y, 2=z) will be used instead of facets normals. yade._utils.sumForces((list)ids, (Vector3)direction) oat Return summary force on bodies with given ids, projected on the direction vector. yade._utils.sumTorques((list)ids, (Vector3)axis, (Vector3)axisPt ) oat Sum forces and torques on bodies given in ids with respect to axis specied by a point axisPt and its direction axis. 8.11. yade.utils module 437
yade._utils.totalForceInVolume() tuple Return summed forces on all interactions and average isotropic stiness, as tuple (Vector3,oat) yade._utils.unbalancedForce([(bool)useMaxForce=False ]) oat Compute the ratio of mean (or maximum, if useMaxForce) summary force on bodies and mean force magnitude on interactions. For perfectly static equilibrium, summary force on all bodies is zero (since forces from interactions cancel out and induce no acceleration of particles); this ratio will tend to zero as simulation stabilizes, though zero is never reached because of nite precision computation. Suciently small value can be e.g. 1e-2 or smaller, depending on how much equilibrium it should be. yade._utils.voxelPorosity([(int)resolution=200 [, (Vector3)start=Vector3(0, 0, 0)[, (Vector3)end=Vector3(0, 0, 0) ]]]) oat Vv Compute packing porosity V where V is a specied volume (from start to end) and Vv is volume V of voxels that fall inside any sphere. The calculation method is to divide whole volume into a dense grid of voxels (at given resolution), and count the voxels that fall inside any of the spheres. This method allows one to calculate porosity in any given sub-volume of a whole sample. It is properly excluding part of a sphere that does not fall inside a specied volume. Parameters resolution (int ) voxel grid resolution, values bigger than resolution=1600 require a 64 bit operating system, because more than 4GB of RAM is used, a resolution=800 will use 500MB of RAM. start (Vector3 ) start corner of the volume. end (Vector3 ) end corner of the volume. yade._utils.wireAll() None Set Shape::wire on all bodies to True, rendering them with wireframe only. yade._utils.wireNoSpheres() None Set Shape::wire to True on non-spherical bodies (Facets, Walls). yade._utils.wireNone() None Set Shape::wire on all bodies to False, rendering them as solids.
8.12 yade.ymport module

Import geometry from various formats (import is python keyword, hence the name ymport). yade.ymport.gengeo(mntable, shift=Vector3(0, 0, 0), scale=1.0, **kw ) Imports geometry from LSMGenGeo library and creates spheres. Since 2012 the package is available in Debian/Ubuntu and known as python-demgengeo http://packages.qa.debian.org/p/pythondemgengeo.html Parameters mntable: mntable object, which creates by LSMGenGeo library, see example shift : [oat,oat,oat] [X,Y,Z] parameter moves the specimen. scale: oat factor scales the given data. **kw: (unused keyword arguments) is passed to utils.sphere LSMGenGeo library allows one to create pack of spheres with given [Rmin:Rmax] with null stress inside the specimen. Can be useful for Mining Rock simulation. Example: examples/packs/packs.py, ples/test/genCylLSM.py. usage of LSMGenGeo library in exam-
https://answers.launchpad.net/esys-particle/+faq/877 http://www.access.edu.au/lsmgengeo_python_doc/current/pythonapi/html/GenGeomodule.html
438
https://svn.esscc.uq.edu.au/svn/esys3/lsm/contrib/LSMGenGeo/ yade.ymport.gengeoFile(leName=le.geo, shift=Vector3(0, 0, 0), scale=1.0, orientation=Quaternion((1, 0, 0), 0), **kw ) Imports geometry from LSMGenGeo .geo le and creates spheres. Since 2012 the package is available in Debian/Ubuntu and known as python-demgengeo http://packages.qa.debian.org/p/pythondemgengeo.html Parameters lename: string le which has 4 colums [x, y, z, radius]. shift : Vector3 Vector3(X,Y,Z) parameter moves the specimen. scale: oat factor scales the given data. orientation : quaternion orientation of the imported geometry **kw: (unused keyword arguments) is passed to utils.sphere Returns list of spheres. LSMGenGeo library allows one to create pack of spheres with given [Rmin:Rmax] with null stress inside the specimen. Can be useful for Mining Rock simulation. Example: examples/packs/packs.py, ples/test/genCylLSM.py. usage of LSMGenGeo library in exam-
https://answers.launchpad.net/esys-particle/+faq/877 http://www.access.edu.au/lsmgengeo_python_doc/current/pythonapi/html/GenGeomodule.html https://svn.esscc.uq.edu.au/svn/esys3/lsm/contrib/LSMGenGeo/ yade.ymport.gmsh(meshle=le.mesh, shift=Vector3(0, tion=Quaternion((1, 0, 0), 0), **kw ) Imports geometry from mesh le and creates facets. Parameters shift : [oat,oat,oat] [X,Y,Z] parameter moves the specimen. scale: oat factor scales the given data. orientation : quaternion orientation of the imported mesh **kw: (unused keyword arguments) is passed to utils.facet Returns list of facets forming the specimen. mesh les can be easily created with GMSH. Example added to examples/regular-spherepack/regular-sphere-pack.py Additional examples of mesh-les roc.inria.fr/gamma/download/download.php can be downloaded from http://www0, 0), scale=1.0, orienta-
yade.ymport.gts(meshle, shift=(0, 0, 0), scale=1.0, **kw ) Read given meshle in gts format. Parameters meshle: string name of the input le. shift : [oat,oat,oat] [X,Y,Z] parameter moves the specimen. scale: oat factor scales the given data. **kw: (unused keyword arguments) is passed to utils.facet Returns list of facets. yade.ymport.iges(leName, shift=(0, 0, 0), scale=1.0, returnConnectivityTable=False, **kw ) Import triangular mesh from .igs le, return list of created facets. Parameters
8.12. yade.ymport module
439
leName (string) name of iges le shift ((oat,oat,oat)|Vector3 ) (X,Y,Z) parameter moves the specimen. scale (oat ) factor scales the given data. **kw (unused keyword arguments) is passed to utils.facet returnConnectivityTable (bool ) if True, apart from facets returns also nodes (list of (x,y,z) nodes coordinates) and elements (list of (id1,id2,id3) element nodes ids). If False (default), returns only facets yade.ymport.stl(le, dynamic=None, xed=True, wire=True, color=None, highlight=False, noBound=False, material=-1) Import geometry from stl le, return list of created facets. yade.ymport.text(leName, shift=Vector3(0, 0, 0), scale=1.0, **kw ) Load sphere coordinates from le, create spheres, insert them to the simulation. Parameters lename (string) le which has 4 colums [x, y, z, radius]. shift ([oat,oat,oat] ) [X,Y,Z] parameter moves the specimen. scale (oat ) factor scales the given data. **kw (unused keyword arguments) is passed to utils.sphere Returns list of spheres. Lines starting with # are skipped yade.ymport.textExt(leName, format=x_y_z_r, shift=Vector3(0, 0, 0), scale=1.0, **kw ) Load sphere coordinates from le in specic format, create spheres, insert them to the simulation. Parameters lename (str ) le name format (str ) the name of output format. Supported x_y_z_r(default), x_y_z_r_matId shift ([oat,oat,oat] ) [X,Y,Z] parameter moves the specimen. scale (oat ) factor scales the given data. **kw (unused keyword arguments) is passed to utils.sphere Returns list of spheres. Lines starting with # are skipped yade.ymport.unv(leName, shift=(0, 0, 0), scale=1.0, returnConnectivityTable=False, **kw ) Import geometry from unv le, return list of created facets. param string leName name of unv le param (oat,oat,oat)|Vector3 shift (X,Y,Z) parameter moves the specimen. param oat scale factor scales the given data. param **kw (unused keyword arguments) is passed to utils.facet param bool returnConnectivityTable if True, apart from facets returns also nodes (list of (x,y,z) nodes coordinates) and elements (list of (id1,id2,id3) element nodes ids). If False (default), returns only facets unv les are mainly used for FEM analyses (are used by OOFEM and Abaqus), but triangular elements can be imported as facets. These les cen be created e.g. with open-source free software Salome. Example: examples/test/unv-read/unvRead.py.
440
Chapter 9
Acknowledging Yade
In order to let users cite Yade consistently in publications, we provide a list of bibliographic references for the dierent parts of the documentation, as in the citation model pushed by CGAL. This way of acknowledging Yade is also a way to make developments and documentation of Yade more attractive for researchers, who are evaluated on the basis of citations of their work by others. We therefore kindly ask users to cite Yade as accurately as possible in their papers. A more detailed discussion of the citation model and its application to Yade can be found here. If new developments are presented and explained in self-contained papers (at the end of a PhD, typically), we will be glad to include them in the documentation and to reference them in the list below. Any other substantial improvement is welcome and can be discussed in the yade-dev mailing list.
9.1 Citing the Yade Project as a whole (the lazy citation method)
If it is not possible to choose the right chapter (but please try), you may cite the documentation [yade:doc] as a whole: 22. milauer, E. Catalano, B. Chareyre, S. Dorofeenko, J. Duriez, A. Gladky, J. Kozicki, C. Modenese, L. Scholts, L. Sibille, J. Strnsk, and K. Thoeni, Yade Documentation (V. milauer, ed.), The Yade Project, 1st ed., 2010. http://yade-dem.org/doc/.
9.2 Citing chapters of Yade Documentation

The rst edition of Yade documentation is seen as a collection with the three volumes (or chapters) listed below, also provided as bibtex entries. Please cite the chapter that is the most relevant in your case. For instance, a paper using one of the documented contact laws will cite the reference documentation [yade:reference]; if programing concepts are discussed, Yades manual [yade:manual] will be cited; the theoretical background [yade:background] can be used as the refence for contact detection, time-step determination, or periodic boundary conditions.
The reference documentation includes details on equations and algorithms found at the highest leve 22. milauer, E. Catalano, B. Chareyre, S. Dorofeenko, J. Duriez, A. Gladky, J. Kozicki, C. Modenese, L. Scholts, L. Sibille, J. Strnsk, and K. Thoeni, Yade Reference Documentation, in Yade Documentation (V. milauer, ed.), The Yade Project, 1st ed., 2010. http://yade-dem.org/doc/. Software design, users and programmers manuals are in (pdf version): 22. milauer, A. Gladky, J. Kozicki, C. Modenese, and J. Strnsk, Yade Using and Programming, in Yade Documentation (V. milauer, ed.), The Yade Project, 1st ed., 2010. http://yade-dem.org/doc/.
441
Fundamentals of the DEM as implemented in Yade are explained in (pdf version): 22. milauer and B. Chareyre, Yade Dem Formulation, in Yade Documentation (V. milauer, ed.), The Yade Project, 1st ed., 2010. http://yade-dem.org/doc/formulation.html.
442
Chapter 9. Acknowledging Yade
Chapter 10
Publications on Yade
This page should be a relatively complete list of publications on Yade itself or done with Yade. If you publish something, do not hesitate to add it on the list. If PDF is freely available, add url for direct fulltext downlad. If not, consider uploading fulltext in PDF, either to Yade wiki or to other website, if legally permitted. The last section gives the references that we recommend to use for citing Yade in publications, as explained in the Acknowledging Yade section. Note: This le is generated from doc/yade-articles.bib, doc/yade-conferences.bib, doc/yade-theses.bib and doc/yade-docref.bib.
10.1 Citing Yade

Corresponding bibtex entries here. Pdf versions are available for each of them. See also Acknowledging Yade.
10.2 Journal articles 10.3 Conference materials 10.4 Master and PhD theses
443
444
Chapter 10. Publications on Yade
Chapter 11
References
All external articles referenced in Yade documentation. Note: This le is generated from doc/references.bib.
445
446
Chapter 11. References
Chapter 12
Indices and tables

genindex modindex search
447
448
Chapter 12. Indices and tables
Bibliography
[yade:background] V. milauer, B. Chareyre (2010), Yade dem formulation. In Yade Documentation ( V. milauer, ed.), The Yade Project , 1st ed. (fulltext) (http://yade-dem.org/doc/formulation.html) [yade:doc] V. milauer, E. Catalano, B. Chareyre, S. Dorofeenko, J. Duriez, A. Gladky, J. Kozicki, C. Modenese, L. Scholts, L. Sibille, J. Strnsk, K. Thoeni (2010), Yade Documentation. The Yade Project. (http://yade-dem.org/doc/) [yade:manual] V. milauer, A. Gladky, J. Kozicki, C. Modenese, J. Strnsk (2010), Yade, using and programming. In Yade Documentation ( V. milauer, ed.), The Yade Project , 1st ed. (fulltext) (http://yade-dem.org/doc/) [yade:reference] V. milauer, E. Catalano, B. Chareyre, S. Dorofeenko, J. Duriez, A. Gladky, J. Kozicki, C. Modenese, L. Scholts, L. Sibille, J. Strnsk, K. Thoeni (2010), Yade Reference Documentation. In Yade Documentation ( V. milauer, ed.), The Yade Project , 1st ed. (fulltext) (http://yade-dem.org/doc/) [Boon2012a] Boon,C.W., Houlsby, G.T., Utili, S. (2012), A new algorithm for contact detection between convex polygonal and polyhedral particles in the discrete element method. Computers and Geotechnics (44), pages 73 - 82. DOI 10.1016/j.compgeo.2012.03.012 (fulltext) [Boon2012b] Boon,C.W., Houlsby, G.T., Utili, S. (2013), A new contact detection algorithm for three-dimensional non-spherical particles. Powder Technology. DOI 10.1016/j.powtec.2012.12.040 (fulltext) [Bourrier2013] Bourrier, F., Kneib, F., Chareyre, B., Fourcaud, T. (2013), Discrete modeling of granular soils reinforcement by plant roots. Ecological Engineering. DOI 10.1016/j.ecoleng.2013.05.002 (fulltext) [Catalano2013a] Catalano, E., Chareyre, B., Barthlmy, E. (2013), Pore-scale modeling of uidparticles interaction and emerging poromechanical eects. International Journal for Numerical and Analytical Methods in Geomechanics. DOI 10.1002/nag.2198 (fulltext) (in press, arxiv version available) [Chareyre2012a] Chareyre, B., Cortis, A., Catalano, E., Barthlemy, E. (2012), Pore-scale modeling of viscous ow and induced forces in dense sphere packings. Transport in Porous Media (92), pages 473-493. DOI 10.1007/s11242-011-9915-6 (fulltext) [Chen2007] Chen, F., Drumm, E. C., Guiochon, G. (2007), Prediction/verication of particle motion in one dimension with the discrete-element method. International Journal of Geomechanics, ASCE (7), pages 344352. DOI 10.1061/(ASCE)1532-3641(2007)7:5(344) [Chen2011a] Chen, F., Drumm, E., Guiochon G. (2011), Coupled discrete element and nite volume solution of two classical soil mechanics problems. Computers and Geotechnics. DOI 10.1016/j.compgeo.2011.03.009 (fulltext) [Chen2012] Chen, Jingsong, Huang, Baoshan, Chen, Feng, Shu, Xiang (2012), Application of discrete element method to superpave gyratory compaction. Road Materials and Pavement Design (13), pages 480-500. DOI 10.1080/14680629.2012.694160 (fulltext) [Dang2010a] Dang, H. K., Meguid, M. A. (2010), Algorithm to generate a discrete element specimen with predened properties. International Journal of Geomechanics (10), pages 85-91. DOI 10.1061/(ASCE)GM.1943-5622.0000028
449
[Dang2010b] Dang, H. K., Meguid, M. A. (2010), Evaluating the performance of an explicit dynamic relaxation technique in analyzing non-linear geotechnical engineering problems. Computers and Geotechnics (37), pages 125 - 131. DOI 10.1016/j.compgeo.2009.08.004 [Donze2008] Donz, F.V. (2008), Impacts on cohesive frictional geomaterials. European Journal of Environmental and Civil Engineering (12), pages 967985. [Duriez2011] Duriez,J., Darve, F., Donz, F.V. (2011), A discrete modeling-based constitutive relation for inlled rock joints. International Journal of Rock Mechanics & Mining Sciences (48), pages 458468. DOI 10.1016/j.ijrmms.2010.09.008 [Duriez2013] Duriez, J., Darve, F., Donz, F.V. (2013), Incrementally non-linear plasticity applied to rock joint modelling. International Journal for Numerical and Analytical Methods in Geomechanics (37), pages 453477. DOI 10.1002/nag.1105 (fulltext) [Epifancev2013] Epifancev, K., Nikulin, A., Kovshov, S., Mozer, S., Brigadnov, I. (2013), Modeling of peat mass process formation based on 3d analysis of the screw machine by the code yade. American Journal of Mechanical Engineering (1), pages 7375. DOI 10.12691/ajme-1-3-3 (fulltext) [Epifantsev2012] Epifantsev, K., Mikhailov, A., Gladky, A. (2012), Proizvodstvo kuskovogo torfa, ekstrudirovanie, forma zakhodnoi i kalibriruyushchei chasti lery matritsy, metod diskretnykh elementov [rus]. Mining informational and analytical bulletin (scientic and technical journal), pages 212-219. [Favier2009a] Favier, L., Daudon, D., Donz, F.V., Mazars, J. (2009), Predicting the drag coecient of a granular ow using the discrete element method. Journal of Statistical Mechanics: Theory and Experiment (2009), pages P06012. [Favier2012] Favier, L., Daudon, D., Donz, F.V. (2012), Rigid obstacle impacted by a supercritical cohesive granular ow using a 3d discrete element model. Cold Regions Science and Technology (85), pages 232241. (fulltext) [Grujicic2013] Grujicic, M, Snipes, JS, Ramaswami, S, Yavari, R (2013), Discrete element modeling and analysis of structural collapse/survivability of a building subjected to improvised explosive device (ied) attack. Advances in Materials Science and Applications (2), pages 924. [Gusenbauer2012] Gusenbauer, M., Kovacs, A., Reichel, F., Exl, L., Bance, S., zelt, H., Schre, T. (2012), Self-organizing magnetic beads for biomedical applications. Journal of Magnetism and Magnetic Materials (324), pages 977982. [Hadda2013] Hadda, Nejib, Nicot, Franois, Bourrier, Franck, Sibille, Luc, Radjai, Farhang, Darve, Flix (2013), Micromechanical analysis of second order work in granular media. Granular Matter (15), pages 221235. DOI 10.1007/s10035-013-0402-3 (fulltext) [Harthong2009] Harthong, B., Jerier, J.F., Doremus, P., Imbault, D., Donz, F.V. (2009), Modeling of high-density compaction of granular materials by the discrete element method. International Journal of Solids and Structures (46), pages 33573364. DOI 10.1016/j.ijsolstr.2009.05.008 [Harthong2012b] Harthong, B., Jerier, J.-F., Richefeu, V., Chareyre, B., Doremus, P., Imbault, D., Donz, F.V. (2012), Contact impingement in packings of elasticplastic spheres, application to powder compaction. International Journal of Mechanical Sciences (61), pages 3243. [Hartong2012a] Harthong, B., Scholts, L., Donz, F.-V. (2012), Strength characterization of rock masses, using a coupled demdfn model. Geophysical Journal International (191), pages 467 480. DOI 10.1111/j.1365-246X.2012.05642.x (fulltext) [Hassan2010] Hassan, A., Chareyre, B., Darve, F., Meyssonier, J., Flin, F. (2010 (submitted)), Microtomography-based discrete element modelling of creep in snow. Granular Matter. [Jerier2009] Jerier, J.-F., Imbault, D.and Donz, F.V., Doremus, P. (2009), A geometric algorithm based on tetrahedral meshes to generate a dense polydisperse sphere packing. Granular Matter (11). DOI 10.1007/s10035-008-0116-0 [Jerier2010a] Jerier, J.-F., Richefeu, V., Imbault, D., Donz, F.V. (2010), Packing spherical discrete elements for large scale simulations. Computer Methods in Applied Mechanics and Engineering. DOI 10.1016/j.cma.2010.01.016
450
Bibliography
[Jerier2010b] Jerier, J.-F., Hathong, B., Richefeu, V., Chareyre, B., Imbault, D., Donz, F.-V., Doremus, P. (2010), Study of cold powder compaction by using the discrete element method. Powder Technology (In Press). DOI 10.1016/j.powtec.2010.08.056 [Kozicki2005a] Kozicki, J. (2005), Discrete lattice model used to describe the fracture process of concrete. Discrete Element Group for Risk Mitigation Annual Report 1, Grenoble University of Joseph Fourier, France, pages 95101. (fulltext) [Kozicki2006a] Kozicki, J., Tejchman, J. (2006), 2D lattice model for fracture in brittle materials. Archives of Hydro-Engineering and Environmental Mechanics (53), pages 7188. (fulltext) [Kozicki2007a] Kozicki, J., Tejchman, J. (2007), Eect of aggregate structure on fracture process in concrete using 2d lattice model. Archives of Mechanics (59), pages 365384. (fulltext) [Kozicki2008] Kozicki, J., Donz, F.V. (2008), A new open-source software developed for numerical simulations using discrete modeling methods. Computer Methods in Applied Mechanics and Engineering (197), pages 44294443. DOI 10.1016/j.cma.2008.05.023 (fulltext) [Kozicki2009] Kozicki, J., Donz, F.V. (2009), Yade-open dem: an open-source software using a discrete element method to simulate granular material. Engineering Computations (26), pages 786805. DOI 10.1108/02644400910985170 (fulltext) [Lomine2013] Lomin, F., Scholts, L., Sibille, L., Poullain, P. (2013), Modelling of uid-solid interaction in granular media with coupled lb/de methods: application to piping erosion. International Journal for Numerical and Analytical Methods in Geomechanics (37), pages 577-596. DOI 10.1002/nag.1109 [Nicot2011] Nicot, F., Hadda, N., Bourrier, F., Sibille, L., Darve, F. (2011), Failure mechanisms in granular media: a discrete element analysis. Granular Matter (13), pages 255-260. DOI 10.1007/s10035-010-0242-3 [Nicot2012] Nicot, F., Sibille, L., Darve, F. (2012), Failure in rate-independent granular materials as a bifurcation toward a dynamic regime. International Journal of Plasticity (29), pages 136154. DOI 10.1016/j.ijplas.2011.08.002 [Nicot2013a] Nicot, F., Hadda, N., Darve, F. (2013), Second-order work analysis for granular materials using a multiscale approach. International Journal for Numerical and Analytical Methods in Geomechanics. DOI 10.1002/nag.2175 [Nicot2013b] Nicot, F., Hadda, N., Guessasma, M., Fortin, J., Millet, O. (2013), On the denition of the stress tensor in granular media. International Journal of Solids and Structures. DOI 10.1016/j.ijsolstr.2013.04.001 (fulltext) [Puckett2011] Puckett, J.G., Lechenault, F., Daniels, K.E. (2011), Local origins of volume fraction uctuations in dense granular materials. *Physical Review E*4 (83), pages 041301. DOI 10.1103/PhysRevE.83.041301 (fulltext) [Sayeed2011] Sayeed, M.A., Suzuki, K., Rahman, M.M., Mohamad, W.H.W., Razlan, M.A., Ahmad, Z., Thumrongvut, J., Seangatith, S., Sobhan, MA, Moz, SA, others (2011), Strength and deformation characteristics of granular materials under extremely low to high conning pressures in triaxial compression. International Journal of Civil & Environmental Engineering IJCEE-IJENS (11). [Scholtes2009a] Scholts, L., Chareyre, B., Nicot, F., Darve, F. (2009), Micromechanics of granular materials with capillary eects. International Journal of Engineering Science (47), pages 6475. DOI 10.1016/j.ijengsci.2008.07.002 (fulltext) [Scholtes2009b] Scholts, L., Hicher, P.-Y., Chareyre, B., Nicot, F., Darve, F. (2009), On the capillary stress tensor in wet granular materials. International Journal for Numerical and Analytical Methods in Geomechanics (33), pages 12891313. DOI 10.1002/nag.767 (fulltext) [Scholtes2009c] Scholts, L., Chareyre, B., Nicot, F., Darve, F. (2009), Discrete modelling of capillary mechanisms in multi-phase granular media. Computer Modeling in Engineering and Sciences (52), pages 297318. (fulltext)
Bibliography
451
[Scholtes2010] Scholts, L., Hicher, P.-Y., Sibille, L. (2010), Multiscale approaches to describe mechanical responses induced by particle removal in granular materials. Comptes Rendus Mcanique (338), pages 627638. DOI 10.1016/j.crme.2010.10.003 (fulltext) [Scholtes2011] Scholts, L., Donz, F.V., Khanal, M. (2011), Scale eects on strength of geomaterials, case study: coal. Journal of the Mechanics and Physics of Solids (59), pages 11311146. DOI 10.1016/j.jmps.2011.01.009 (fulltext) [Scholtes2012] Scholts, L., Donz, F.V. (2012), Modelling progressive failure in fractured rock masses using a 3d discrete element method. International Journal of Rock Mechanics and Mining Sciences (52), pages 1830. DOI 10.1016/j.ijrmms.2012.02.009 (fulltext) [Scholtes2013] Scholts, L., Donz, F.V. (2013), A DEM model for soft and hard rocks: role of grain interlocking on strength. Journal of the Mechanics and Physics of Solids (61), pages 352369. DOI 10.1016/j.jmps.2012.10.005 (fulltext) [Shiu2008] Shiu, W., Donz, F.V., Daudeville, L. (2008), Compaction process in concrete during missile impact: a dem analysis. Computers and Concrete (5), pages 329342. [Shiu2009] Shiu, W., Donz, F.V., Daudeville, L. (2009), Discrete element modelling of missile impacts on a reinforced concrete target. International Journal of Computer Applications in Technology (34), pages 3341. [Smilauer2006] Vclav milauer (2006), The splendors and miseries of yade design. Annual Report of Discrete Element Group for Hazard Mitigation. (fulltext) [Tejchman2011] Tejchman, J. (2011), Comparative modelling of shear zone patterns in granular bodies with nite and discrete element model. Advances in Bifurcation and Degradation in Geomaterials, pages 255260. [Thoeni2013] K. Thoeni, C. Lambert, A. Giacomini, S.W. Sloan (2013), Discrete modelling of hexagonal wire meshes with a stochastically distorted contact model. Computers and Geotechnics (49), pages 158169. DOI 10.1016/j.compgeo.2012.10.014 (fulltext) [Tong2012] Tong, A.-T., Catalano, E., Chareyre, B. (2012), Pore-scale ow simulations: model predictions compared with experiments on bi-dispersed granular assemblies. Oil & Gas Science and Technology - Rev. IFP Energies nouvelles. DOI 10.2516/ogst/2012032 (fulltext) [Tran2011] Tran, V.T., Donz, F.V., Marin, P. (2011), A discrete element model of concrete under high triaxial loading. Cement and Concrete Composites. [Tran2012] Tran, V.D.H., Meguid, M.A., Chouinard, L.E. (2012), An algorithm for the propagation of uncertainty in soils using the discrete element method. The Electronic Journal of Geotechnical Engineering. (fulltext) [Tran2012c] Tran, V.D.H., Meguid, M.A., Chouinard, L.E. (2012), Discrete element and experimental investigations of the earth pressure distribution on cylindrical shafts. International Journal of Geomechanics. DOI 10.1061/(ASCE)GM.1943-5622.0000277 [Tran2013] Tran, V.D.H., Meguid, M.A., Chouinard, L.E. (2013), A nitediscrete element framework for the 3d modeling of geogridsoil interaction under pullout loading conditions. Geotextiles and Geomembranes (37), pages 1-9. DOI 10.1016/j.geotexmem.2013.01.003 [Catalano2009] E. Catalano, B. Chareyre, E. Barthlmy (2009), Fluid-solid coupling in discrete models. In Alert Geomaterials Workshop 2009. [Catalano2010a] E. Catalano, B. Chareyre, E. Barthlmy (2010), A coupled model for uid-solid interaction analysis in geomaterials. In Alert Geomaterials Workshop 2010. [Catalano2010b] E. Catalano, B. Chareyre, E. Barthlmy (2010), Pore scale modelling of stokes ow. In GdR MeGe. [Catalano2011a] E. Catalano, B. Chareyre, A. Cortis, E. Barthlmy (2011), A pore-scale hydromechanical coupled model for geomaterials. In II International Conference on Particle-based Methods - Fundamentals and Applications. (fulltext) [Catalano2013b] Catalano E., Chareyre B., Barthlmy E. (2013), Dem-pfv analysis of solid-uid transition in granular sediments under the action of waves. In Powders and Grains 2013:
452
Bibliography
Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI 10.1063/1.4812118 (fulltext) [Chareyre2009] Chareyre B., Scholts L. (2009), Micro-statics and micro-kinematics of capillary phenomena in dense granular materials. In POWDERS AND GRAINS 2009: Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI 10.1063/1.3180083 [Chareyre2011] B. Chareyre, E. Catalano, E. Barthlmy (2011), Numerical simulation of hydromechanical couplings by combined discrete element method and nite-volumes. In International Conference on Flows and Mechanics in Natural Porous Media from Pore to Field Scale Pore2Field. (fulltext) [Chareyre2012b] B. Chareyre, L. Scholts, F. Darve (2012), The properties of some stress tensors investigated by dem simulations. In Euromech Colloquium 539; Mechanics of Unsaturated Porous Media: Eective Stress principle; from micromechanics to thermodynamics (fulltext) [Chareyre2012c] B. Chareyre (2012), Micro-poromechanics: recent advances in numerical models and perspectives. In ICACM symposium 2012, The role of structure on emerging material properties [Chen2008a] Chen, F., Drumm, E.C., Guiochon, G., Suzuki, K. (2008), Discrete element simulation of 1d upward seepage ow with particle-uid interaction using coupled open source software. In Proceedings of The 12th International Conference of the International Association for Computer Methods and Advances in Geomechanics (IACMAG) Goa, India. [Chen2009b] Chen, F., Drumm, E.C., Guiochon, G. (2009), 3d dem analysis of graded rock ll sinkhole repair: particle size eects on the probability of stability. In Transportation Research Board Conference (Washington DC). [Dang2008a] Dang, H.K., Mohamed, M.A. (2008), An algorithm to generate a specimen for discrete element simulations with a predened grain size distribution.. In 61th Canadian Geotechnical Conference, Edmonton, Alberta. [Dang2008b] Dang, H.K., Mohamed, M.A. (2008), 3d simulation of the trap door problem using the discrete element method.. In 61th Canadian Geotechnical Conference, Edmonton, Alberta. [Favier2009b] L. Favier, D. Daudon, F. Donz, J. Mazars (2009), Validation of a dem granular ow model aimed at forecasting snow avalanche pressure. In AIP Conference Proceedings. DOI 10.1063/1.3180002 (fulltext) [Gillibert2009] Gillibert L., Flin F., Rolland du Roscoat S., Chareyre B., Philip A., Lesare B., Meyssonier J. (2009), Curvature-driven grain segmentation: application to snow images from x-ray microtomography. In IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Miami, USA). [Guo2013] Ning Guo, Jidong Zhao (2013), A hierarchical model for cross-scale simulation of granular media. In Powders and Grains 2013: Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI 10.1063/1.4812158 (fulltext) [Hadda2013] Nejib Hadda, Franois Nicot, Luc Sibille, Farhang Radjai, Antoinette Tordesillas, Flix Darve (2013), A multiscale description of failure in granular materials. In Powders and Grains 2013: Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI 10.1063/1.4811999 (fulltext) [Harthong2013] Barthlmy Harthong, Richard G Wan (2009), Directional plastic ow and fabric dependencies in granular materials. In Powders and Grains 2013: Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI http://dx.doi.org/10.1063/1.4811900 (fulltext) [Hasan2010b] A. Hasan, B. Chareyre, J. Kozicki, F. Flin, F. Darve, J. Meyssonnier (2010), Microtomography-based discrete element modeling to simulate snow microstructure deformation. In AGU Fall Meeting Abstracts [Hicher2009] Hicher P.-Y., Scholts L., Chareyre B., Nicot F., Darve F. (2008), On the capillary stress tensor in wet granular materials. In Inaugural International Conference of the Engineering Mechanics Institute (EM08) - (Minneapolis, USA). Bibliography 453
[Hicher2011] Hicher, P.Y, Scholts, L., Sibille, L. (2011), Multiscale modeling of particle removal impact on granular material behavior. In Engineering Mechanics Institute, EMI 2011. [Hilton2013] J. E. Hilton, P. W. Cleary, A. Tordesillas (2013), Unitary stick-slip motion in granular beds. In Powders and Grains 2013: Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI 10.1063/1.4812063 (fulltext) [Kozicki2003a] J. Kozicki, J. Tejchman (2003), Discrete methods to describe the behaviour of quasi-brittle and granular materials. In 16th Engineering Mechanics Conference, University of Washington, Seattle, CDROM. [Kozicki2003c] J. Kozicki, J. Tejchman (2003), Lattice method to describe the behaviour of quasibrittle materials. In CURE Workshop, Eective use of building materials, Sopot. [Kozicki2004a] J. Kozicki, J. Tejchman (2004), Study of fracture process in concrete using a discrete lattice model. In CURE Workshop, Simulations in Urban Engineering, Gdansk. [Kozicki2005b] J. Kozicki, J. Tejchman (2005), Simulations of fracture in concrete elements using a discrete lattice model. In Proc. Conf. Computer Methods in Mechanics (CMM 2005), Czestochowa, Poland. [Kozicki2005c] J. Kozicki, J. Tejchman (2005), Simulation of the crack propagation in concrete with a discrete lattice model. In Proc. Conf. Analytical Models and New Concepts in Concrete and Masonry Structures (AMCM 2005), Gliwice, Poland. [Kozicki2006b] J. Kozicki, J. Tejchman (2006), Modelling of fracture process in brittle materials using a lattice model. In Computational Modelling of Concrete Structures, EURO-C (eds.: G. Meschke, R. de Borst, H. Mang and N. Bicanic), Taylor anf Francis. [Kozicki2006c] J. Kozicki, J. Tejchman (2006), Lattice type fracture model for brittle materials. In 35th Solid Mechanics Conference (SOLMECH 2006), Krakow. [Kozicki2007c] J. Kozicki, J. Tejchman (2007), Simulations of fracture processes in concrete using a 3d lattice model. In Int. Conf. on Computational Fracture and Failure of Materials and Structures (CFRAC 2007), Nantes. (fulltext) [Kozicki2007d] J. Kozicki, J. Tejchman (2007), Eect of aggregate density on fracture process in concrete using 2d discrete lattice model. In Proc. Conf. Computer Methods in Mechanics (CMM 2007), Lodz-Spala. [Kozicki2007e] J. Kozicki, J. Tejchman (2007), Modelling of a direct shear test in granular bodies with a continuum and a discrete approach. In Proc. Conf. Computer Methods in Mechanics (CMM 2007), Lodz-Spala. [Kozicki2007f] J. Kozicki, J. Tejchman (2007), Investigations of size eect in tensile fracture of concrete using a lattice model. In Proc. Conf. Modelling of Heterogeneous Materials with Applications in Construction and Biomedical Engineering (MHM 2007), Prague. [Kozicki2011] J. Kozicki, J. Tejchman (2011), Numerical simulation of sand behaviour using dem with two dierent descriptions of grain roughness. In II International Conference on Particlebased Methods - Fundamentals and Applications. (fulltext) [Kozicki2013] Jan Kozicki, Jacek Tejchman, Danuta Lesniewska (2013), Study of some microstructural phenomena in granular shear zones. In Powders and Grains 2013: Proceedings of the 6th International Conference on Micromechanics of Granular Media. AIP Conference Proceedings. DOI 10.1063/1.4811976 (fulltext) [Lomine2010a] Lomin, F., Scholts, L., Poullain, P., Sibille, L. (2010), Soil microstructure changes induced by internal uid ow: investigations with coupled de/lb methods. In Proc. of 3rd Euromediterranean Symposium on Advances in Geomaterials and Structures, AGS10. [Lomine2010b] Lomin, F., Poullain, P., Sibille, L. (2010), Modelling of uid-solid interaction in granular media with coupled lb/de methods: application to solid particle detachment under hydraulic loading. In 19th Discrete Simulation of Fluid Dynamics, DSFD 2010. [Lomine2011] Lomin, F., Sibille, L., Marot, D. (2011), A coupled discrete element - lattice botzmann method to investigate internal erosion in soil. In Proc. 2nd Int. Symposium on Computational Geomechanics (ComGeo II).
454
Bibliography
[Michallet2012] H. Michallet, E. Catalano, C. Berni, V. Rameliarison, E. Barthlmy (2012), Physical and numerical modelling of sand liquefaction in waves interacting with a vertical wall. In ICSE6 - 6th International conference on Scour and Erosion [Modenese2012] C. Modenese, S. Utili, G.T. Houlsby (2012), Dem modelling of elastic adhesive particles with application to lunar soil. In Earth and Space 2012: Engineering, Science, Construction, and Operations in Challenging Environments 2012 ASCE. DOI 10.1061/9780784412190.006 (fulltext) [Modenese2012a] Modenese, C, Utili, S, Houlsby, G T (2012), A study of the inuence of surface energy on the mechanical properties of lunar soil using DEM. In Discrete Element Modelling of Particulate Media ( Wu, Chuan-Yu, ed.), Royal Society of Chemistry , DOI 10.1039/978184973503200069 (fulltext) [Modenese2012b] Modenese, C, Utili, S, Houlsby, G T (2012), A numerical investigation of quasistatic conditions for granular media. In Discrete Element Modelling of Particulate Media ( Wu, Chuan-Yu, ed.), Royal Society of Chemistry , DOI 10.1039/9781849735032-00187 (fulltext) [Nicot2010] Nicot, F., Sibille, L., Daouadji, A., Hicher, P.Y., Darve, F. (2010), Multiscale modeling of instability in granular materials. In Engineering Mechanics Institute, EMI 2010. [Nicot2011b] Nicot, F., Hadda, N., Bourrier, F., Sibille, L., Darve, F. (2011), A discrete element analysis of collapse mechanisms in granular materials. In Proc. 2nd Int. Symposium on Computational Geomechanics (ComGeo II). [Sari2011] H. Sari, B. Chareyre, E. Catalano, P. Philippe, E. Vincens (2011), Investigation of internal erosion processes using a coupled dem-uid method. In II International Conference on Particle-based Methods - Fundamentals and Applications. (fulltext) [Scholtes2007a] L. Scholts, B. Chareyre, F. Nicot, F. Darve (2007), Micromechanical modelling of unsaturated granular media. In Proceedings ECCOMAS-MHM07, Prague. [Scholtes2008a] L. Scholts, B. Chareyre, F. Nicot, F. Darve (2008), Capillary eects modelling in unsaturated granular materials. In 8th World Congress on Computational Mechanics - 5th European Congress on Computational Methods in Applied Sciences and Engineering, Venice. [Scholtes2008b] L. Scholts, P.-Y. Hicher, F.Nicot, B. Chareyre, F. Darve (2008), On the capillary stress tensor in unsaturated granular materials. In EM08: Inaugural International Conference of the Engineering Mechanics Institute, Minneapolis. [Scholtes2009e] Scholtes L, Chareyre B, Darve F (2009), Micromechanics of partialy saturated granular material. In Int. Conf. on Particle Based Methods, ECCOMAS-Particles. [Scholtes2011b] L. Scholts, F. Donz (2011), Progressive failure mechanisms in jointed rock: insight from 3d dem modelling. In II International Conference on Particle-based Methods Fundamentals and Applications. (fulltext) [Scholtes2011c] Scholts, L., Hicher, P.Y., Sibille, L. (2011), A micromechanical approach to describe internal erosion eects in soils. In Proc. of Geomechanics and Geotechnics: from micro to macro, IS-Shanghai 2011. [Shiu2007a] W. Shiu, F.V. Donze, L. Daudeville (2007), Discrete element modelling of missile impacts on a reinforced concrete target. In Int. Conf. on Computational Fracture and Failure of Materials and Structures (CFRAC 2007), Nantes. [Sibille2009] Sibille, L., Scholts, L. (2009), Eects of internal erosion on mechanical behaviour of soils: a dem approach. In Proc. of International Conference on Particle-Based Methods, Particles 2009. [Smilauer2007a] V. milauer (2007), Discrete and hybrid models: applications to concrete damage. In Unpublished. (fulltext) [Smilauer2008] Vclav milauer (2008), Commanding c++ with python. In ALERT Doctoral school talk. (fulltext) [Smilauer2010a] Vclav milauer (2010), Yade: past, present, future. In Internal seminary in Laboratoire 3S-R, Grenoble. (fulltext) (LaTeX sources)
Bibliography
455
[Stransky2010] Jan Strnsk, Milan Jirsek, Vclav milauer (2010), Macroscopic elastic properties of particle models. In Proceedings of the International Conference on Modelling and Simulation 2010, Prague. (fulltext) [Stransky2011] J. Stransky, M. Jirasek (2011), Calibration of particle-based models using cells with periodic boundary conditions. In II International Conference on Particle-based Methods Fundamentals and Applications. (fulltext) [Thoeni2011] K. Thoeni, C. Lambert, A. Giacomini, S.W. Sloan (2011), Discrete modelling of a rockfall protective system. In Particles 2011: Fundamentals and Applications. (fulltext) [Tran2012b] Tran, V.D.H, Meguid, M.A, Chouinard, L.E (2012), A discrete element study of the earth pressure distribution on cylindrical shafts. In Tunnelling Association of Canada (TAC) Conference 2012, Montreal [Catalano2008a] E. Catalano (2008), Inltration eects on a partially saturated slope - an application of the discrete element method and its implementation in the open-source software yade. Master thesis at UJF-Grenoble. (fulltext) [Charlas2013] Benoit Charlas (2013), Etude du comportement mcanique dun hydrure intermtallique utilis pour le stockage dhydrogne. PhD thesis at Universit de Grenoble. [Chen2009a] Chen, F. (2009), Coupled ow discrete element method application in granular porous media using open source codes. PhD thesis at University of Tennessee, Knoxville. (fulltext) [Chen2011b] Chen, J. (2011), Discrete element method (dem) analyses for hot-mix asphalt (hma) mixture compaction. PhD thesis at University of Tennessee, Knoxville. (fulltext) [Duriez2009a] J. Duriez (2009), Stabilit des massifs rocheux : une approche mcanique. PhD thesis at Institut polytechnique de Grenoble. (fulltext) [Favier2009c] Favier, L. (2009), Approche numrique par lments discrets 3d de la sollicitation dun coulement granulaire sur un obstacle. PhD thesis at Universit Grenoble I Joseph Fourier. [Jerier2009b] Jerier, J.F. (2009), Modlisation de la compression haute densit des poudres mtalliques ductiles par la mthode des lments discrets (in french). PhD thesis at Universit Grenoble I Joseph Fourier. (fulltext) [Kozicki2007b] J. Kozicki (2007), Application of discrete models to describe the fracture process in brittle materials. PhD thesis at Gdansk University of Technology. (fulltext) [Marzougui2011] Marzougui, D. (2011), Hydromechanical modeling of the transport and deformation in bed load sediment with discrete elements and nite volume. Master thesis at Ecole Nationale dIngnieur de Tunis. (fulltext) [Scholtes2009d] Luc Scholts (2009), modlisation micromcanique des milieux granulaires partiellement saturs. PhD thesis at Institut National Polytechnique de Grenoble. (fulltext) [Smilauer2010b] Vclav milauer (2010), Cohesive particle model using the discrete element method on the yade platform. PhD thesis at Czech Technical University in Prague, Faculty of Civil Engineering & Universit Grenoble I Joseph Fourier, cole doctorale I-MEP2. (fulltext) (LaTeX sources) [Smilauer2010c] Vclav milauer (2010), Doctoral thesis statement. (PhD thesis summary). (fulltext) (LaTeX sources) [Tran2011b] Van Tieng TRAN (2011), Structures en bton soumises des chargements mcaniques extrmes: modlisation de la rponse locale par la mthode des lments discrets (in french). PhD thesis at Universit Grenoble I Joseph Fourier. (fulltext) [Addetta2001] G.A. DAddetta, F. Kun, E. Ramm, H.J. Herrmann (2001), From solids to granulates - Discrete element simulations of fracture and fragmentation processes in geomaterials.. In Continuous and Discontinuous Modelling of Cohesive-Frictional Materials. (fulltext) [Allen1989] M. P. Allen, D. J. Tildesley (1989), Computer simulation of liquids. Clarendon Press.
456
Bibliography
[Alonso2004] F. Alonso-Marroquin, R. Garcia-Rojo, H.J. Herrmann (2004), Micro-mechanical investigation of the granular ratcheting. In Cyclic Behaviour of Soils and Liquefaction Phenomena. (fulltext) [Bagi2006] Katalin Bagi (2006), Analysis of microstructural strain tensors for granular assemblies. International Journal of Solids and Structures (43), pages 3166 - 3184. DOI 10.1016/j.ijsolstr.2005.07.016 [Bertrand2005] D. Bertrand, F. Nicot, P. Gotteland, S. Lambert (2005), Modelling a geo-composite cell using discrete analysis. Computers and Geotechnics (32), pages 564577. [Bertrand2008] D. Bertrand, F. Nicot, P. Gotteland, S. Lambert (2008), Discrete element method (dem) numerical modeling of double-twisted hexagonal mesh. Canadian Geotechnical Journal (45), pages 11041117. [Calvetti1997] Calvetti, F., Combe, G., Lanier, J. (1997), Experimental micromechanical analysis of a 2d granular material: relation between structure evolution and loading path. Mechanics of Cohesive-frictional Materials (2), pages 121163. [Camborde2000a] F. Camborde, C. Mariotti, F.V. Donz (2000), Numerical study of rock and concrete behaviour by discrete element modelling. Computers and Geotechnics (27), pages 225 247. [Chareyre2002a] B. Chareyre, L. Briancon, P. Villard (2002), Theoretical versus experimental modeling of the anchorage capacity of geotextiles in trenches.. Geosynthet. Int. (9), pages 97123. [Chareyre2002b] B. Chareyre, P. Villard (2002), Discrete element modeling of curved geosynthetic anchorages with known macro-properties.. In Proc., First Int. PFC Symposium, Gelsenkirchen, Germany. [Chareyre2003] Bruno Chareyre (2003), Modlisation du comportement douvrages composites sol-gosynthtique par lments discrets - application aux tranches dancrage en tte de talus.. PhD thesis at Grenoble University. (fulltext) [Chareyre2005] Bruno Chareyre, Pascal Villard (2005), Dynamic spar elements and discrete element methods in two dimensions for the modeling of soil-inclusion problems. Journal of Engineering Mechanics (131), pages 689698. DOI 10.1061/(ASCE)0733-9399(2005)131:7(689) (fulltext) [CundallStrack1979] P.A. Cundall, O.D.L. Strack (1979), A discrete numerical model for granular assemblies. Geotechnique (), pages 4765. DOI 10.1680/geot.1979.29.1.47 [DeghmReport2006] F. V. Donz (ed.), Annual report 2006 (2006). Discrete Element Group for Hazard Mitigation. Universit Joseph Fourier, Grenoble (fulltext) [Donze1994a] F.V. Donz, P. Mora, S.A. Magnier (1994), Numerical simulation of faults and shear zones. Geophys. J. Int. (116), pages 4652. [Donze1995a] F.V. Donz, S.A. Magnier (1995), Formulation of a three-dimensional numerical model of brittle behavior. Geophys. J. Int. (122), pages 790802. [Donze1999a] F.V. Donz, S.A. Magnier, L. Daudeville, C. Mariotti, L. Davenne (1999), Study of the behavior of concrete at high strain rate compressions by a discrete element method. ASCE J. of Eng. Mech (125), pages 11541163. DOI 10.1016/S0266-352X(00)00013-6 [Donze2004a] F.V. Donz, P. Bernasconi (2004), Simulation of the blasting patterns in shaft sinking using a discrete element method. Electronic Journal of Geotechnical Engineering (9), pages 144. [GarciaRojo2004] R. Garca-Rojo, S. McNamara, H. J. Herrmann (2004), Discrete element methods for the micro-mechanical investigation of granular ratcheting. In Proceedings ECCOMAS 2004. (fulltext) [Hentz2003] Sebastien Hentz (2003), Modlisation dune structure en bton arm soumise un choc par la mthode des elments discrets. PhD thesis at Universit Grenoble 1 Joseph Fourier.
Bibliography
457
[Hentz2004a] S. Hentz, F.V. Donz, L.Daudeville (2004), Discrete element modelling of concrete submitted to dynamic loading at high strain rates. Computers and Structures (82), pages 25092524. DOI 10.1016/j.compstruc.2004.05.016 [Hentz2004b] S. Hentz, L. Daudeville, F.V. Donz (2004), Identication and validation of a discrete element model for concrete. ASCE Journal of Engineering Mechanics (130), pages 709719. DOI 10.1061/(ASCE)0733-9399(2004)130:6(709) [Hentz2005a] S. Hentz, F.V. Donz, L.Daudeville (2005), Discrete elements modeling of a reinforced concrete structure submitted to a rock impact. Italian Geotechnical Journal (XXXIX), pages 8394. [Herminghaus2005] Herminghaus, S. (2005), Dynamics of wet granular matter. Advances in Physics (54), pages 221-261. DOI 10.1080/00018730500167855 (fulltext) [Hubbard1996] Philip M. Hubbard (1996), Approximating polyhedra with spheres for timecritical collision detection. ACM Trans. Graph. (15), pages 179210. DOI 10.1145/231731.231732 [Johnson2008] Scott M. Johnson, John R. Williams, Benjamin K. Cook (2008), Quaternion-based rigid body rotation integration algorithms for use in particle methods. International Journal for Numerical Methods in Engineering (74), pages 13031313. DOI 10.1002/nme.2210 [Jung1997] Derek Jung, Kamal K. Gupta (1997), Octree-based hierarchical distance maps for collision detection. Journal of Robotic Systems (14), pages 789806. DOI 10.1002/(SICI)10974563(199711)14:11<789::AID-ROB3>3.0.CO;2-Q [Kettner2011] Lutz Kettner, Andreas Meyer, Afra Zomorodian (2011), Intersecting sequences of dD iso-oriented boxes. In CGAL User and Reference Manual. (fulltext) [Klosowski1998] James T. Klosowski, Martin Held, Joseph S. B. Mitchell, Henry Sowizral, Karel Zikan (1998), Ecient collision detection using bounding volume hierarchies of k-dops. IEEE Transactions on Visualization and Computer Graphics (4), pages 2136. (fulltext) [Kuhl2001] E. Kuhl, G. A. DAddetta, M. Leukart, E. Ramm (2001), Microplane modelling and particle modelling of cohesive-frictional materials. In Continuous and Discontinuous Modelling of Cohesive-Frictional Materials. DOI 10.1007/3-540-44424-6_3 (fulltext) [Lu1998] Ya Yan Lu (1998), Computing the logarithm of a symmetric positive denite matrix. Appl. Numer. Math (26), pages 483496. DOI 10.1016/S0168-9274(97)00103-7 (fulltext) [Luding2008] Luding, Stefan (2008), Cohesive, frictional powders: contact models for tension. Granular Matter (10), pages 235-246. DOI 10.1007/s10035-008-0099-x (fulltext) [Magnier1998a] S.A. Magnier, F.V. Donz (1998), Numerical simulation of impacts using a discrete element method. Mech. Cohes.-frict. Mater. (3), pages 257276. DOI 10.1002/(SICI)10991484(199807)3:3<257::AID-CFM50>3.0.CO;2-Z [McNamara2008] S. McNamara, R. Garca-Rojo, H. J. Herrmann (2008), Microscopic origin of granular ratcheting. Physical Review E (77). DOI 11.1103/PhysRevE.77.031304 [Munjiza1998] A. Munjiza, K. R. F. Andrews (1998), Nbs contact detection algorithm for bodies of similar size. International Journal for Numerical Methods in Engineering (43), pages 131149. DOI 10.1002/(SICI)1097-0207(19980915)43:1<131::AID-NME447>3.0.CO;2-S [Munjiza2006] A. Munjiza, E. Rougier, N. W. M. John (2006), Mr linear contact detection algorithm. International Journal for Numerical Methods in Engineering (66), pages 4671. DOI 10.1002/nme.1538 [Neto2006] Natale Neto, Luca Bellucci (2006), A new algorithm for rigid body molecular dynamics. Chemical Physics (328), pages 259268. DOI 10.1016/j.chemphys.2006.07.009 [Omelyan1999] Igor P. Omelyan (1999), A new leapfrog integrator of rotational motion. the revised angular-momentum approach. Molecular Simulation (22). DOI 10.1080/08927029908022097 (fulltext) [Pfc3dManual30] ICG (2003), Pfc3d (particle ow code in 3d) theory and background manual, version 3.0. Itasca Consulting Group.
458
Bibliography
[Pion2011] Sylvain Pion, Monique Teillaud (2011), 3D triangulations. In CGAL User and Reference Manual. (fulltext) [Pournin2001] L. Pournin, Th. M. Liebling, A. Mocellin (2001), Molecular-dynamics force models for better control of energy dissipation in numerical simulations of dense granular media. Phys. Rev. E (65), pages 011302. DOI 10.1103/PhysRevE.65.011302 [Price2007] Mathew Price, Vasile Murariu, Garry Morrison (2007), Sphere clump generation and trajectory comparison for real particles. In Proceedings of Discrete Element Modelling 2007. (fulltext) [Rabinov2005] RABINOVICH Yakov I., ESAYANUR Madhavan S., MOUDGIL Brij M. (2005), Capillary forces between two spheres with a xed volume liquid bridge : theory and experiment. Langmuir (21), pages 1099210997. (fulltext) (eng) [Satake1982] M. Satake (1982), Fabric tensor in granular materials.. In Proc., IUTAM Symp. on Deformation and Failure of Granular materials, Delft, The Netherlands. [Thoeni2013] K. Thoeni, C. Lambert, A. Giacomini, S.W. Sloan (2013), Discrete modelling of hexagonal wire meshes with a stochastically distorted contact model. Computers and Geotechnics (49), pages 158169. DOI 10.1016/j.compgeo.2012.10.014 (fulltext) [Thornton1991] Colin Thornton, K. K. Yin (1991), Impact of elastic spheres with and without adhesion. Powder technology (65), pages 153166. DOI 10.1016/0032-5910(91)80178-L [Thornton2000] Colin Thornton (2000), Numerical simulations of deviatoric shear deformation of granular media. Gotechnique (50), pages 4353. DOI 10.1680/geot.2000.50.1.43 [Verlet1967] Loup Verlet (1967), Computer experiments on classical uids. i. thermodynamical properties of lennard-jones molecules. Phys. Rev. (159), pages 98. DOI 10.1103/PhysRev.159.98 [Villard2004a] P. Villard, B. Chareyre (2004), Design methods for geosynthetic anchor trenches on the basis of true scale experiments and discrete element modelling. Canadian Geotechnical Journal (41), pages 11931205. [Wang2009] Yucang Wang (2009), A new algorithm to model the dynamics of 3-d bonded rigid bodies with rotations. Acta Geotechnica (4), pages 117127. DOI 10.1007/s11440-008-00721 (fulltext) [Weigert1999] Weigert, Tom, Ripperger, Siegfried (1999), Calculation of the liquid bridge volume and bulk saturation from the half-lling angle. Particle & Particle Systems Characterization (16), pages 238242. DOI 10.1002/(SICI)1521-4117(199910)16:5<238::AID-PPSC238>3.0.CO;2E (fulltext) [Willett2000] Willett, Christopher D., Adams, Michael J., Johnson, Simon A., Seville, Jonathan P. K. (2000), Capillary bridges between two spherical bodies. Langmuir (16), pages 9396-9405. DOI 10.1021/la000657y (fulltext) [Zhou1999536] Y.C. Zhou, B.D. Wright, R.Y. Yang, B.H. Xu, A.B. Yu (1999), Rolling friction in the dynamic simulation of sandpile formation. Physica A: Statistical Mechanics and its Applications (269), pages 536553. DOI 10.1016/S0378-4371(99)00183-1 (fulltext) [cgal] Jean-Daniel Boissonnat, Olivier Devillers, Sylvain Pion, Monique Teillaud, Mariette Yvinec (2002), Triangulations in cgal. Computational Geometry: Theory and Applications (22), pages 519.
Bibliography
459
460
Bibliography
Python Module Index

_
yade._eudoxos, 401 yade._packObb, 418 yade._packPredicates, 415 yade._packSpheres, 411 yade._utils, 433
b
yade.bodiesHandling, 399
e
yade.eudoxos, 400 yade.export, 402
g
yade.geom, 404
l
yade.linterpolation, 407
p
yade.pack, 408 yade.plot, 418 yade.post2d, 421
q
yade.qt, 424 yade.qt._GLViewer, 424
t
yade.timing, 425
u
yade.utils, 426
y
yade.ymport, 438
461

Yade DEM Documentation

Uploaded by

Copyright:

Available Formats

Yade DEM Documentation

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Yade DEM Documentation

Uploaded by

Copyright:

Available Formats

Yade Documentation

Release 2013-08-23.git-3aeba29, August 23, 2013

Yade Documentation, Release 2013-08-23.git-3aeba29

Yade Documentation, Release 2013-08-23.git-3aeba29

1.1.1 Starting yade

(In case you did not install

Yade Documentation, Release 2013-08-23.git-3aeba29

Saving such script as hello.py, it can be given as argument to yade:

If you want Yade to quit immediately

Yade Documentation, Release 2013-08-23.git-3aeba29

1.1.2 Creating simulation

1.1.3 Running simulation

1.1. Getting started

Yade Documentation, Release 2013-08-23.git-3aeba29

Yade [21]: O.iter Out[21]: 100533

1.1.4 Saving and loading

1.1.5 Graphical interface

Yade Documentation, Release 2013-08-23.git-3aeba29

1.2 Architecture overview

1.2.1 Data and functions

1.2. Architecture overview

Yade Documentation, Release 2013-08-23.git-3aeba29

# functional components, empty at the moment

Yade Documentation, Release 2013-08-23.git-3aeba29

Bodies can be iterated over using standard python iteration syntax:

1.2. Architecture overview

Yade Documentation, Release 2013-08-23.git-3aeba29

Yade Documentation, Release 2013-08-23.git-3aeba29

collision detection pass 1

other forces (gravity, BC, ...)

constitutive law compute forces from strains

1.2. Architecture overview

Yade Documentation, Release 2013-08-23.git-3aeba29

Dispatchers and functors

Yade Documentation, Release 2013-08-23.git-3aeba29

Interaction Geometry Interaction Physics

1.2. Architecture overview

Yade Documentation, Release 2013-08-23.git-3aeba29

Yade Documentation, Release 2013-08-23.git-3aeba29

| --lib | --local | --doc --home --user --user1

Useful keys on the command-line are: <tab> ^C (=Ctrl+C) ^D ^C ^\ Shift-PgUp Shift-PgDown

Yade Documentation, Release 2013-08-23.git-3aeba29

For a complete help, see man yade

2.2.2 Python basics

# gain access to "module" of functions

Yade Documentation, Release 2013-08-23.git-3aeba29

# use the module under a different name

# import everything so that it can be used without module name

# negative counts from the end

# from second element to the end

# extend the list

Lists can be created in various ways:

List of squares of even number smaller than 20, i.e.

{ 2 } a a {0, , 19} 2a (note the similarity):

Yade Documentation, Release 2013-08-23.git-3aeba29

Tuples are constant sequences:

Mapping from keys to values:

Yade [5]: c=0 if a<1 else 1

# conditional expression # define a new function

# and call that function

a=range(5) b=[(aa2 if aa%2==0 else -aa2) for aa in a]