Chapter 1: Using User Programmable Features (UPFs)

Go to the Next Chapter Go to the Table of Contents for This Manual. Chapter 1 * Chapter 2 * Chapter 3 * Appendix A

1.1 Introduction to UPFs

Because the ANSYS program has an open architecture, you can write your own routines or subroutines in C or FORTRAN and either link them to ANSYS or use them as external commands. In fact, some of the ANSYS features you see today as "standard" offerings originated as user programmable features (UPFs). You can take advantage of UPFs if you're licensed for any of the following products:

ANSYS/Multiphysics ANSYS/Mechanical ANSYS/Structural ANSYS/PrepPost ANSYS/ED (Research Faculty/Student version)

Other versions of the ANSYS program don't support UPFs.

1.1.1 What Are UPFs?

User programmable features are ANSYS capabilities you can use to write your own routines. Using UPFs, you can tailor the ANSYS program to your organization's needs. For instance, you may need to define a new material behavior, a special element, or a modified failure criterion for composites. You can even write your own design optimization algorithm that calls the entire ANSYS program as a subroutine. UPFs provide the following capabilities:

To read information into or fetch information from the ANSYS database, you can create subroutines and either link them into the ANSYS program or use them in the external command feature (see Appendix A for more information

about external commands). If you link these subroutines into ANSYS, you are lilmited to 10 database access commands. Such commands, created through either method, operate at all levels of ANSYS operation, including the begin, preprocessor, general postprocessor, time-history postprocessor, and solution levels. For more information about accessing the ANSYS database, see Chapter 2. ANSYS provides a set of routines you can use to specify various types of loads, including BF or BFE loads, pressures, convections, heat fluxes, and charge densities. These routines are described under section 1.7. Another set of UPF routines enables you to define the following material properties: plasticity, creep, swelling law, viscoplasticity, hyperelasticity, and layered element failure criteria. To see inputs and outputs for these routines, see section 1.6. Several sets of UPFs enable you to define new elements and to adjust the nodal orientation matrix. See section 1.3 for more information. Another group of UPFs enables you to modify and monitor existing elements. For details, see section 1.5. You can customize UPF userop to create a custom design optimization routine. For more information, see section 1.12. You can call the ANSYS program as a subroutine in a program you've written. To learn how, see section 1.8.

1.1.2 What You Should Know Before Using UPFs

Before you do anything with linked UPFs, contact your on-site ANSYS system support person to get the permissions needed to access the appropriate ANSYS files. The UPF subroutines are written in FORTRAN 77; some extensions are used. They contain comments intended to give you enough detail to develop your own versions of the subroutines. To use UPFs successfully, you need strong working knowledge of the following:

The ANSYS program. The UPF subroutines themselves. Study the UPF subroutines before customizing them, and make sure that you fully understand the subroutines, as well as any applicable functions. Unless you review them carefully, a few UPF subroutines may seem like a maze with many logic paths to consider. You may have to set special variables correctly in order to run your customized ANSYS program without errors. Even if you have in-depth knowledge of the ANSYS input and your desired outputs, you still need to ensure that everything that

needs to be done in the UPF subroutines is done properly in your custom version. FORTRAN 77. Besides knowing how to write FORTRAN 77 subroutines, you must be sure that the level of the FORTRAN 77 compiler is as least as high as the level mentioned in your ANSYS installation manual. You also need to know what to do should the computer abort the program due to an arithmetic error, a file read error, a memory access error, and so on. The mathematics of the phenomenon you're planning to include.

Important

UPFs aren't available in certain data center environments or on some hardware configurations, such as machines using parallel processing for element formulation. For additional information, consult your ANSYS installation manual or your on-site ANSYS system support person Carefully consider whether you wish to use UPFs, especially if you are linking them into ANSYS (rather than into a shared library for use as external commands). When you add your own routines to ANSYS by either method, you're creating a customized, site-dependent version of the program. ANSYS, Inc. considers the use of UPFs a non-standard use of the program, one that the ANSYS Quality Assurance verification testing program doesn't cover. Therefore, you're responsible for verifying that the results produced are accurate and that your customizations don't adversely affect other, standard areas of the ANSYS program. Although the flexibility that UPFs offer can be highly attractive, UPF usage is a complicated process that can introduce errors. Consider what you want your customizations to accomplish. You may be able to customize ANSYS more easily and safely with macros than with UPFs.

For other guidelines for non-standard uses of the ANSYS program, see Chapter 6 of the ANSYS Advanced Analysis Techniques Guide.

1.1.3 Planning Your UPFs

UPFs can range from a simple element output routine for customized output to a complex user optimization. Before you start programming, ask yourself these questions:

Does the capability you want already exist in the ANSYS program? Remember, a capability may not be obvious at first, especially to a novice ANSYS user.

Does your proposed subroutine fit into the ANSYS program architecture and specifications? For example, you can't program a user element that has more than 32 degrees of freedom per node or more than 20 nodes per element.

Use your experience and judgment to answer these questions. If you need help to do so, consult your ANSYS Support Distributor. If you can respond "no" to both questions, then the user routine you're planning will be both useful and feasible.

1.1.4 Studying the ANSYS User Routines

Your ANSYS distribution medium contains the source codes for all user routines:

If you have a UNIX version of ANSYS, the source code for the UPF routines resides in directory /ansys55/customize/user. If you're running the ANSYS program under Windows NT or Windows 95, the UPF source code resides in directory \ansys55\custom\user\intel.

Most of the user routines have at least simple functionality, so print out the routines of interest before you start programming. All source routines are concatenated onto file user.f or user.for. Delete the routines you don't want and make appropriate changes to the others.

1.1.5 Programming in Languages Other than FORTRAN

If you access UPFs by compiling and linking a custom version of ANSYS, the preferred method is to design and program your custom routine in FORTRAN 77. Although you can use languages other than FORTRAN 77, in each case FORTRAN 77 must provide the interface to the rest of the ANSYS program. If you do use a language other than FORTRAN 77, such as the C programming language, your code may require a FORTRAN shell. You need to take care when calling FORTRAN subroutines from C subroutines. You must use the symbol associated with the FORTRAN subroutine when invoking the subroutine from a C function. This symbol typically differs slightly from the FORTRAN subroutine name, and is extremely system dependent. On many UNIX systems, you build this symbol name by taking the FORTRAN subroutine name, converting it to lower case, and appending an underscore. For example, the symbol name for the FORTRAN subroutine HeapInquire would be heapinquire_. You would have to use the symbol heapinquire_ in the invoking C function to avoid an unsatisfied external reference when the program is linked.

Keep in mind that the instance described above is just an example. Compilers from different vendors may construct the symbols differently. Please consult the manuals for your specific compiler for information on how to call FORTRAN subroutines from C functions.

1.1.6 Developing UPFs: a Suggested Strategy

When developing UPFs by compiling and linking a custom version of ANSYS, you can avoid problems and reduce debugging time by following a gradual, orderly process. Start with a trivial test. Then, add a few changes at a time so that if something goes wrong, the error that caused the problem should be isolated and relatively easy to locate. The example procedure below illustrates this type of gradual process. The example assumes that you're creating a new element for the ANSYS program. You develop and test it by performing these steps: 1. Get the applicable element subroutines for uel101 from the ANSYS distribution medium. Add a small change (such as a misspelling in an output heading), then compile and link the subroutines. 2. Using a production version of the ANSYS program, run several analysis problems using LINK8 (and maybe other elements) to form a base for comparison. 3. Replacing LINK8 with USER101, run the same problem on your custom version of ANSYS. 4. Compare the results from Steps 2 and 3. If they show discrepancies other than the misspelled output heading, resolve them before you go on to Step 5. 5. Choose the standard ANSYS element that most closely resembles your new custom element, and run some problems on a production version of ANSYS using that element. 6. Modify the element subroutines to match the element in chose in Step 5. Then, compile and link those subroutines into a custom version of ANSYS. 7. Again, compare the results from Steps 5 and 6. If they don't match, resolve the discrepancies before moving on to Step 8. 8. Modify your element subroutines to include the features you want. Then, compile and link the subroutines into a custom version of ANSYS.

9. Test the changes with a series of increasingly complex problems for which you already know the answers.

1.1.7 Include Decks

In addition to the subroutines and functions described in this chapter, most of the include decks (files with the extension .inc) used by ANSYS are on your ANSYS distribution medium. These include decks, also called commons, contain important but relatively small amounts of data. The ANSYS program also handles large amounts of data using various access routines (GET and PUT), as described elsewhere in this manual. To insert include decks in a subroutine or function, use the INCLUDE (or an analogous) statement. Do not modify an include deck under any circumstances. The following table lists some of the more commonly used ANSYS include files and the definitions they contain:
Include Description File acelcm.inc Contains accelerations and angular velocities Defines general ANSYS parameters. You must include this common to retrieve the ansysdef.inc parameter values of HEAP_INTEGER, HEAP_DOUBLE, HEAP_COMPLEX, or HEAP_REAL. cmopt.inc Contains optimization variables echprm.inc Defines parameters for element characteristics elccmt.inc Defines element characteristics (comments only) elecom.inc Contains element-specific information elparm.inc Defines pointers for the element data array elucom.inc Defines the element degree-of-freedom pointers etycom.inc Element type data impcom.inc Used by all routines and functions in the ANSYS program outpcm.inc Defines output control information soptcm.inc Contains solution options and keys Defines stack storage. You must include this common in any routines that access stack.inc stack space. stepcm.inc Contains load step information usvrcm.inc Defines storage of user-defined variables

1.1.8 Linking User Routines

After you make your changes to the user routines supplied on your ANSYS distribution medium, you can either

Link your routines into shared libraries (as discussed starting in Appendix A). Compile and link your custom routines into the ANSYS program itself. This is discussed for UNIX systems in 1.1.9 and for Windows systems in 1.1.10. You may need super-user or root privileges to run the procedure that does the linking.

1.1.9 Compiling and Linking UPFs on UNIX Systems

As mentioned previously, the source files for the user routines reside in subdirectory /ansys55/customize/user. If you modify any of these subroutines, run the procedure/ansys55/customize/user/ANSCUSTOM to link these changes into the ANSYS program. When you run a user-linked version of the ANSYS program, the output will include the following:
NOTE: This ANSYS version was linked by Licensee

The ANSCUSTOM procedure compiles all FORTRAN files (files ending with .F) and all C files (files ending with .c) in the current working directory. The procedure then loads all object files (files ending with .o) along with the default ANSYS objects and libraries in /ansys55/customize/user/platform (where platform is a directory that uniquely identifies the hardware platform version). The new executable file created will be named ansys.e55t and will reside in the current directory. FORTRAN files are assumed to be FORTRAN 77 (some extensions are allowed), and C files are assumed to be ANSI C. Chapter 1 of the ANSYS Installation and Configuration Guide for UNIX lists the compilers you'll need to use UPFs.

1.1.10 Compiling and Linking UPFs on Windows Systems

As mentioned previously, the source files for the user routines reside in subdirectory \ansys55\custom\user\platform. If you modify any of these subroutines, perform the following tasks to link your changes into the ANSYS program: 1. Create and go to a working directory for building ANSYS.

2. Copy the following files from \ansys55\custom\user\platform to your working directory: anscust.bat, makefile, and any user routines you plan to modify. If you're running Windows 95, also copy filefcomp.bat. 3. Run the procedure ANSCUST. This procedure compiles all FORTRAN files (files ending with .F) and all C files (files ending with .c) in the current working directory. The procedure then loads all object files (files ending with .obj), along with the default ANSYS objects and libraries (in \ansys55\custom\user\platform ). The executable file created will be named ansys.exe and will reside in the current working directory. The executable will be named ansys.exe and will reside in the current working directory. Windows 95 Users Only: Set the environment variable ANSYS55_DIR to the root ANSYS directory, as shown below:
SET ANSYS55_DIR=C:\ANSYS55

If you installed the program in a directory other than c:\ansys55, you'll also need to modify the ANSYS55_DIR variable in the makefile. Important-Be sure to answer "no" when the linking procedure asks, "Do you want to overwrite it (current executable)?" If your answer isn't "no," all other ANSYS program users will also get your changes.

1.1.11 Activating UPFs

The ANSYS program activates many UPFs through a specific user action. This can be through a command option or a user selection. For example, to activate a user element, all you need to do is select it as one of the element types in a model (using either the ET command or menu path Main Menu>Preprocessor>Element Type>Add/Edit/Delete). You then set the element attribute pointer (TYPEcommand or menu path Main Menu>Preprocessor>-Meshing->Attributes>Default Attribs), and define elements using the solid modeling or direct generation method. UPFs that are not activated by the means described above must be activated by either of the following methods:

Issuing the USRCAL command Choosing menu path Main Menu>Preprocessor>Loads>-Load Step Opts>Other>User Routines or Main Menu>Solution>-Load Step Opts>Other>User Routines.

To activate or deactivate the routines, issue the command USRCAL,Rnam1,...Rnam9, where Rnam1 and Rnam9 are the names of specific routines. You can specify up to

nine routines with one USRCALcommand, or you can issue multiple USRCAL commands. Issue the command USRCAL,NONE to deactivate all valid user subroutines. To list the status of the routines, issue the command USRCAL,STAT. For a list of the user routines that the USRCAL command (or its equivalent menu paths) affects, see the USRCAL command description in the ANSYS Commands Reference. If you don't activate the UPFs in this manner, standard ANSYS logic will be used by default. For instance, when you apply a convection load, standard ANSYS logic is the default even if you have a user convection routine linked in. The user convection routine must be activated by the USRCAL command or its menu equivalent.

1.1.12 Running Your Custom Executable

You can run your custom executable using a supplied procedure called ansys55cust in UNIX (ansys55cust.exe in Windows). This procedure/executable resides in the ansys55\bin\platform directory. When run, the procedure searches the current working directory for the custom ANSYS executable (either ansys.e55t or ansys.exe, depending on your platform). If the custom ANSYS executable resides in a separate directory, you can specify the path on the ansys55cust command line, as follows:
ansys55cust -custom /pathname/ansys.e55t

or, for Windows

ansys55cust -custom /pathname/ansys.exe

1.1.13 Verifying Your Routines

After compiling and linking your new user routine, test and verify it using whatever procedures you think are adequate. Remember, verifying that your customized version of the ANSYS program works properly is your responsibility. Make certain that your custom version of the ANSYS program performs correctly for the combinations of elements, analysis types, materials, boundary conditions, etc. that you plan to use. Confirm that the logic you introduced is correct and doesn't produce any unwanted side effects. In testing your custom user routines, you also should verify that the changes you've made don't affect standard, non-customized ANSYS features. To do so, you can compare the results of a set of problems from the ANSYS Verification Manual run on

the standard version and on the customized version. Input for these problems is also available on your ANSYS distribution medium. Always remember: your last step, a series of steps, or even your concept may be wrong. Proceed in clear steps, and verify your work as often as possible. Keep intermediate versions of your modified source code on backup media. Note-If you contact your site's ANSYS system support person or any ANSYS, Inc. representative about the performance of a custom version of ANSYS, always tell him or her explicitly that you're using a user programmable feature. If you feel that an error exists in an unrelated feature of the ANSYS program, demonstrate the suspected error in an non-customized, production version of the program before you report the error to an ANSYS. Inc. representative.

1.1.14 Debugging Commands

To debug errors in your user routines, you can use commands and other features not documented in the ANSYS Commands Reference. Use these commands only for extremely small models with few solution iterations (otherwise, they'll generate an excessive amount of output). /TRACK and /DEBUG are described in detail below. Two other useful "undocumented" commands are OUTEQ and /NERR. The command OUTEQ,on can be used to output results from all equilibrium iterations. The command /NERR,,,-1 causes errors to be reported as before, but the run continues anyway, normally terminating with either a) system abort or b) incorrect answers. 1.1.14.1 Tracking the Path of Program Logic The /TRACK command issues a message when the program logic enters and leaves some of the higher level subroutines. Subroutines TrackBegin and TrackEnd (see Chapter 6) set up the /TRACKcommand. Then, issue the command using the format below
/TRACK,MonLevel,PrintLevel,SumLevel

MonLevel is the level for timing monitoring. PrintLevel is the level for enter/exit printout, and SumLevel is the level at which the timing sum is output. Each of these arguments can be any value between 0 and 9 (default is 0). You can use the /TRACK command to identify which section of code is causing the program to abort. For example, to flag up to eight levels of subroutines to determine when the program logic enters and leaves them, you would issue the command /TRACK,,8.

1.1.14.2 Debugging Elements and Solutions The /DEBUG command generates debugging at various points in the output. You can specify one of three formats for /DEBUG: solution debug format, element debug format, and general debug format. 1.1.14.2.1 Solution Debug Format Issue the command using this format:
/DEBUG,-1,F1,F2,F3,F4,F5,F6,F7,F8,F9

For This Argument F1 F2

Choose One of These Values 1 (provides basic solution control debugging) 1 (provides transient debugging using Newmark constants) 2 (provides transient debugging using velocities and accelerations) 1 (provides element matrix debugging and prints matrix + load vectors)

F3

2 (provides element matrix debugging with load vectors only) 3 (provides element matrix debugging with matrix diagonals and load vectors) 1 (provides auto time stepping debugging) 1 (provides multifield debugging) 1 (provides arc-length debugging) 1 (provides basic Newton-Raphson debugging) 2 (provides Newton-Raphson debugging and prints out-of-balance forces or incremental displacement or each DOF) 3 (provides Newton-Raphson debugging and prints applied loads and n-r restoring force for each DOF) 1,2 (provides displacement vector debugging with displacement pointers)

F4 F5 F6

F7

F8

2 (provides displacement vector debugging with incremental displacement) 3 (provides displacement vector debugging with contact database) 1 (provides temporary programmer debugging)

F9

1.1.14.2.2 Element Debug Format Issue the command using this format:
/DEBUG,-3,G1,G2,G3,G4,G5,G6,G7,G8,G9

For This

Choose One of These Values

Argument G1 G2

1 (provides basic element pass debugging) 1 (provides element displacement and coordinate debugging) 1 (provides element matrix debugging and prints matrix + load vectors) 2 (provides element matrix debugging with load vectors only) 3 (provides element matrix debugging with matrix diagonals and load vectors) 1 (provides element load information debugging) 1 (provides element real constant debugging) 1 (provides element saved variable debugging) 1 (provides element material property debugging with linear material properties) 2 (provides element material property debugging with nonlinear properties) 1,2 (provides element nonlinear debugging with plasticity)

G3

G4 G5 G6 G7

G8

2 (provides element nonlinear debugging with large deformation) 3 (provides element nonlinear debugging with contact database) 1 (provides temporary programmer debugging)

G9

1.1.14.2.3 General Debug Format Issue the command using this format:
/DEBUG,H1,H2,,H4,H5

For This Argument

Choose One of These Values 1 (provides file header record information)

H1

2 (provides input line (character)) 3 (provides input line (decoded)) 1 (provides wavefront reordering and element checking debugging)

H2 2 (provides meshing debugging) 1 (provides nodal coordinate system transformation debugging) H4 H5 2 (provides displacement updating debugging) 1 (provides pre-element debugging, element characteristics debugging, and element field load debugging)

1.1.15 Other Useful Commands

Two other ANSYS commands, NSVR and /UCMD, can help you implement UPFs. (Neither command has an equivalent GUI path.) Use the NSVR command to define the number of extra variables that need to be saved for user programmable element options, such as user plasticity. Issue the /UCMD command to make a user routine into a custom command. For more information, see section 1.9 later in this chapter.

1.1.16 Generating Output

You can generate output controlled by the /OUTPUT command by using the FORTRAN write statement. The output unit for this statement is usually called IOTT. IOTT may be defined with the functionwrinqr. See the discussion on the function wrinqr in Chapter 6 for more details.

1.2 Reading Large Data Files More Rapidly

When files containing ANSYS-related data are large, loading them into the ANSYS program or writing them out to an external file can be a slow process. For example, consider an ANSYS problem file which contains nearly 462,000 lines, 150,000 of which contain nodal data and 97,383 of them containing data for elements. Because many of the lines in this file are in command format, the ANSYS program spends a lot of time reading it. You can shorten the time ANSYS takes to read such files by including two commands in your programs, UPFs, or macros: EBLOCK and NBLOCK. The NBLOCK command converts nodal data into fixed format data blocks (which ANSYS can read more quickly than commands). The EBLOCK command places element data into a fixed format block, one line per element. These commands also compress displacement constraint data to one line per constraint node.

1.2.1 EBLOCK Command Format

The EBLOCK command requires that you specify a valid format statement for five element attributes followed by the node numbers for each element. The command has the following format:
EBLOCK,Nnodes

You must follow the EBLOCK command with a valid format statement defining five element attributes, followed by Nnodes. Nnodes are the node numbers associated with

this element. Use commas to separate the node numbers, (don't use blanks) and enclose the string of node numbers in parentheses. The element attributes are:
Elnum The element number. Mat The material number. Real The real constants associated with this element. Type The element type number. Esys The system of coordinates this element uses.

The EBLOCK command can indicate multiple element lines in the file. The value EBLOCK returns will either be 0 (if the command executes successfully), or a negative element number if an error occurs.

1.2.2 NBLOCK Command Format

The NBLOCK command has this format:
NBLOCK,Ncoord

Ncoord is the number of coordinates defined for this node. The default for Ncoord is 3 and the maximum value is 6. The default coordinate values are the X, Y, and Z coordinates, and the other values are THXY, THYZ, and THZX. Normally, you specify one NBLOCK command for each line of nodal data. The value NBLOCK returns will be either 0 (if the command executes successfully) or a negative node number if an error occurs.

1.3 Routines for Creating New Elements

The next few pages describe the user routines and supporting subroutines you use to create new elements. Using these routines, you can create new element types, add them to the ANSYS element library, and use them as "regular" elements. You can create up to six independent element types (names USER100 - USER105). For demonstration purposes, example copies of the routines for MASS21, the structural mass element, and LINK8, the 3-D spar element, are included on the ANSYS distribution medium as uel100 and uel101 respectively.

1.3.1 Input and Output Abbreviations

The descriptions of the routines or functions within this chapter describe both the input arguments and output arguments. Argument information includes the argument's type, size and intent.

Argument type is one of the following: int - integer dp - double precision log - logical chr - character dcp - double precision complex

Argument size is one of the following: sc - scalar variable ar(n) - array variable of length n func - functional return value

Argument intent is one of the following: in - input argument out - output argument inout - both an input and an output argument

1.3.2 User Routines

Routines uec100 through uec105 describe the element characteristics. Routine elccmt (on the distribution medium) describes the input for these routines in detail. You can use subroutines uex100 throughuex105 to override default logic. Routines uec100 through uec105 define parameters such as:

2-D or 3-D geometry Degree-of-freedom set Symmetric or unsymmetric matrix Number of nodes Number of body loads (for example, temperatures) Number of surface loads (for example, pressures) Number of real constants Number of variables to be saved Number of rows in element matrices Linear or nonlinear element.

Routines uel100 through uel105 calculate the element matrices (stiffness, specific heat, etc.), the element load vector (force, heat flow, etc.), and any element output quantities. The element printout also is generated, and the variables to be saved are calculated and stored in the results file. Other user routines available for manipulating element information include the following:

Routines uep100 through uep105 provide printed output of line elements. The general ANSYS postprocessor, POST1, calls the subroutines, or you can call them using uel100 through uel105. Routine usertr allows access to the nodal transformations. Routine userac describes some of the data handling.

1.3.3 Subroutine uec100 (Defining Characteristics of the usr100 Routine)

subroutine uec100 (elcdn,ielc,kerr) c ***** this subroutine defines the characteristics of user100. c #include "impcom.inc" external erhandler #include "echprm.inc" c external nminfo,erinqr integer erinqr integer ielc(*),i,kerr character*28 elcdn c C c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr siz=sc,ar(n) intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ielc (int,ar(IELCSZ),inout) - element characteristics c see include deck elccmt for details c kerr (int,sc,inout) - error flag up to this point. c (do not initialize to zero) c c output arguments: c variable (typ,siz,intent) description c elcdn (chr,sc,out) - name of element c ielc (int,ar(IELCSZ),inout) - element characteristics

c details c kerr (int,sc,inout) c note to programmers: here c

see include deck elccmt for - error flag (set to 1 if error) the validity of keyopt values may be checked

1.3.3.1 Subroutines uec101 through uec105 The input and output arguments for subroutines uec101, uec102, uec103, uec104, and uec105 is identical to the uec100 subroutine listed above.

1.3.4 Subroutine uex100 (Overriding Element Characteristic Defaults)

subroutine uex100 (ielc,kerr) c *** subroutine to override element characteristic defaults *** c *** hence, this routine is needed only in rare cases. c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c #include "impcom.inc" #include "echprm.inc" integer ielc(*),i,kerr c c *** input and output are the same as for uec100, except that this c *** logic is called after the defaulting logic is finished. c *** this defaulting is done in ansys subroutine echdft(not a upf). c *** as indicated above, this routine is rarely needed, but if it is c *** desired to see the effect of echdft, you may print out the ielc array c *** leaving uec100 and print it out again entering this routine. c c typ=int,dp,log,chr siz=sc,ar(n) intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ielc (int,ar(IELCSZ),inout) - element characteristics c see include deck elccmt for details c kerr (int,sc,inout) - error flag up to this point. c (do not initialize to zero) c c output arguments: c variable (typ,siz,intent) description c ielc (int,ar(IELCSZ),inout) - element characteristics c see include deck elccmt for details c kerr (int,sc,inout) - error flag (set to 1 if error) c *** standard defaults are taken. the final results are given with c *** the debug accessed with /debug,,, ,,1 c c *** dummy logic to use all variables to keep analyzer happy

*** this should be removed by user element programmer

1.3.4.1 Subroutines uex101 through uex105 The source code for subroutines uex101, uex102, uex103, uex104, and uex105 is identical to the code for subroutine uex100 listed above.

1.3.5 Subroutine uel100 (Computing Element Matrices, Load Vectors, and Results)
subroutine uel100 (elem,ielc,elmdat,eomask,nodes,locsvr,kelreq, x kelfil,nr,xyz,u,kelout,zs,zass,damp,gstif,zsc,zscnr,elvol,elmass, x center,elener,edindx,lcerst) --- general lumped mass is demonstrated -------------------------------*** primary function: 1. compute element matrices, load vectors, and results *** secondary functions: 2. maintain element solution data

c c c c c c c *** user programmable functions may not be used in parallel processing *** c c *** Notice - This file contains ANSYS Confidential information *** c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c elem (int,sc,in) - element label (number) c ielc (int,ar(IELCSZ),in) - array of element type characteristics c (IELCSZ = array size, defined in echprm) c elmdat (int,ar(10),in) - array of element data c eomask (int,sc,in) - bit pattern for element output c (see outpcm) c nodes (int,ar(nnod),in) - array of element node numbers c (nnod = number of nodes; 1 in this case) c locsvr (int,sc,in) - location of the saved variables c on file .esav for this element c kelreq (int,ar(10),in) - matrix and load vector form requests c (indices for kelreq are given with output c arguments below) c kelfil (int,ar(10),in) - keys indicating incoming matrices and c load vectors (indices for kelfil are the c same as given for kelreq with output c arguments below) c nr (int,sc,in) - matrix and load vector size c xyz (dp,ar(6,nnod),in) - nodal coordinates (orig) and rotation angle

c u (dp,ar(nr,5),in) c c output arguments: c kelout (int,ar(10),out) c c below, c later) c zs (dp,ar(nr,nr),inout)(kelreq(1)) c zass (dp,ar(nr,nr),inout)(kelreq(2)) c damp (dp,ar(nr,nr),inout)(kelreq(3)) c gstif (dp,ar(nr,nr),inout)(kelreq(4)) c zsc (dp,ar(nr),out) (kelreq(5)) c zscnr (dp,ar(nr),out) (kelreq(6)) c (kelreq(7)) c elvol (dp,sc,out c elmass (dp,sc,out) c center (dp,ar(3),out) c elener (dp,ar(5),out) c edindx (int,ar(20),out) c lcerst (int,sc,inout) c

element nodal solution values keys indicating created matrices and load vectors (indices for kelout are the same as for kelreq as well as kelin and kelout stiffness/conductivity matrix mass matrix damping/specific heat matrix stress stiffness matrix applied f vector n-r restoring f vector or imaginary f vector element volume element mass centroid location element energies element result data file indexes position on result file

1.3.6 Subroutines uel101 through uel105

The input and output arguments for subroutines uel101, uel102, uel103, uel104, and uel105 is identical to subroutine uel100 listed above.

1.3.7 Subroutine uep100 (Printing Output for User Elements)

x x x subroutine uep100 (iott,ielc,elem,nodes,mat, kept,ept, kemn,emn, kems,ems, kens,ens, keel,eel, keth,eth, kenl,enl, kepl,epl, kecr,ecr)

c c *** primary function: produce printed output for user elements c used typically only for line elements c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c ********** this subroutine is provided for user information ********* c *** user programmable features may not be used in parallel processing ***

c c input arguments: c iott (int,sc,in) c ielc (int,ar(150),in) c elem (int,sc,in) c nodes (int,ar(2),in) c mat (int,sc,in) c kept (int,sc,in) c ept (dp,ar(*),in) c kemn (int,sc,in) c emn (dp,ar(*),in) c kems (int,sc,in) c ems (dp,ar(*),in) c kens (int,sc,in) c ens (dp,ar(*),in) c keel (int,sc,in) c eel (dp,ar(*),in) c keth (int,sc,in) c c eth (dp,ar(*),in) c kenl (int,sc,in) present c enl (dp,ar(*),in) c kepl (int,sc,in) c epl (dp,ar(*),in) c kecr (int,sc,in) c ecr (dp,ar(*),in) c c output arguments: c none

output unit number element characteristics element number node numbers material number key to print temperatures element pseudo-node temperatures key to print misc. non-summed data misc. non-summable data record key to print mmisc. summed data misc. summed data key to print nodal stresses nodal sresses at an element key to print nodal elastic strain element nodal elastic strains key to print thermal,initial, and swelling strains - nodal thermal strains at an element - key set if any nonlinear materials element nonlinear table key to print nodal plastic strains nodal plastic strains key to print nodal creep strains nodal creep strains

1.3.8 Subroutines uep101 through uep105

The source code for subroutines uep101, uep102, uep103, uep104, and uep105 is identical to subroutine uep100 listed above.

1.3.9 Subroutine usertr (Adjusting the Nodal Orientation Matrix)

subroutine usertr (node,tr) c *** primary function: adjust nodal orientation matrix c secondary function: study nodal orientation matrix c accessed with ielc(notran) = -100 c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description

c c c c c c c c c c c c c c

node tr

(int,sc,in) - node number being acted upon (dp,ar(32,32),inout) - nodal to global orientation matrix

output arguments: variable (typ,siz,intent) description tr (dp,ar(32,32),inout) - nodal to global orientation matrix tr is a matrix that is already defined based on the degrees of freedom selected. it does not normally need to be changed. it may be printed out here to study. its functional size is nr by nr, where nr is the number of degrees of freedom in the element

1.3.10 Subroutine userac (Accessing Element Information)

This subroutine is provided for demonstration purposes.
subroutine userac (elem) c *** primary function: To demonstrate user access of element information. c --- Given the element number, all information about the element is avaiable. c --- Starting with elmdat, one can get the element type, real constant number, c --- the material number, the element coordinate system number, as well as c --- the node numbers. Then, one can get more information about any or all c --- of these things. The below demonstrates getting the element type and c --- real constants. c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - element number c c output arguments: c none c

1.4 Supporting Subroutines for Element Creation

The subroutines described on the next few pages support the user routines used to create new elements.

1.4.1 Subroutine nminfo (Returning Element Reference Names)

subroutine nminfo (ielc,rname) c *** primary function: set element reference names c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c ielc (int,ar(150),inout) - element characteristic vector c rname (chr,sc,in) - 8 character reference name c c output arguments: c variable (typ,siz,intent) description c ielc (int,ar(150),inout) - element characteristic vector with c element name encoded c

1.4.2 Subroutine svgidx (Fetching the Index for Saved Variables)

subroutine svgidx (locsvr,svindx) c *** primary function: get the index for saved variables c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c locsvr (int,sc,in) - pointer to location of index c c output arguments: c svindx (int,ar(10),out) - the 10 word index of svr variables c 1-starting loc of this eles svr sets c 2-ending loc of this eles svr sets c 3:10-starting loc for each set c 3-structural svrs c 4-thermal/electric/fluid svrs c 5-magnetic svrs c 6-nonlinear svrs c 7-plasticity svrs c 8-creep svrs c 9-coupled svrs c 10-user svrs c

1.4.3 Subroutine svrget (Fetching Saved Variable Data for an Element)

subroutine svrget (svindx,nset,nsvr,svr) c *** primary function: get svr data set for an element c *** Notice - This file contains ANSYS Confidential information *** c input arguments:

c svindx svgidx) c nset c c c c c c c c c nsvr c c c

(int,ar(10),in) (int,sc,in)

- index for svr for this element (see - the set number in this index = 1 - structural svrs = 2 - thermal/electric/fluid svrs = 3 - magnetic svrs = 4 - nonlinear svrs = 5 - plasticity svrs = 6 - creep svrs = 7 - coupled svrs = 8 - user svrs - number of dp words expected in this set - number of dp words in this set - data in this set

(int,sc,inout)

output arguments: nsvr (int,sc,inout) svr (dp,ar(nsvr),in)

1.4.4 Subroutine svrput (Writing an Element's Saved Variable Set)

subroutine svrput (svindx,nset,leng,svr) c *** primary function: write out a svr data set for an element c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c svindx (int,ar(10),inout)- the index for svr for this element c nset (int,sc,in) - the set number in this index (see svrget) c leng (int,sc,in) - number of dp words in this set c svr (dp,ar(leng),in) - data in this set c c output arguments: c svindx (int,ar(10),inout)- updated index c

1.4.5 Subroutine svpidx (Writing the Saved Variable Element Index to a File)
subroutine svpidx (locsvr,svindx) c *** primary function: write the svr element index onto file c *** secondary functions: update the locsvr pointer to next element c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c locsvr (int,sc,inout) - pointer to start of svr for element c svindx (int,ar(10),in) - index to svr for this element c c output arguments: c locsvr (int,sc,inout) - pointer to start of svr for next element

1.4.6 Subroutine mreuse (Determining Which Element Matrices Can Be Reused)

c c c c c c c c Second note: this logic is essentially the same as the old c sfrm logic. Hopefully, further simplifications and enhancements c will be made in the future. (Especially in gap elements and in c multilayer elements) c the whole idea of kpheno, a holdover from the sfrm routines, c needs to be looked at and possibly eliminated. c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c kelrqq (int,ar(10),in) - request keys (needed for this analysis) c kelfil (int,ar(10),in) - keys indicating matrices on the file c elem (int,sc,in) - element number c ielc (int,ar(IELCSZ),in) - array of element type characteristics c kmasrt (int,sc,in) - does the mass matrix have rotational DOF? c 0 - no 1 - yes(with nlgeom, sfrm1n) c knlmg (int,sc,in) - nonlinear magnetic curve exists in this c element c 0 - no 1 - yes c kconve (int,sc,in) - key indicating existence of convections c in this element c 0,1 - no 2 or more - yes c must be input as 'i' if not used, as is c changed in this routine(for analyzer). c i = 0 must be used in calling routine c if kpheno = 1. c kpheno (int,sc,in) - key for type of phenomenon/level of check c 0 - structural like old sfrm1n,1s,3n,3s,fl c 1 - thermal like old sfrm1c,1t,2t,3t c 2 - electrical/magnetic like some of old c sfrmpo c 3 - general like old sfrmoo c kprop (int,sc,in) - key indicating which material properties c in the prop vector that need to be c checked (see below) c nprop (int,sc,in) - number of properties subroutine mreuse (kelrqq,kelfil,elem,ielc,kmasrt,knlmg,kconve, x kpheno,kprop,nprop,prop,propo,krvro,rvr,rvro,amodo,asymo, kelin) *** primary function: determine which Matrices can be REUSEd and which must be recomputed from iteration to iteration. Note: a few special elements have some supplementary logic to adjust these results further. No attempt as been made to include all such logic in these routines.

c prop (dp,ar(nprop),in) - current mat props c propo (dp,ar(nprop),inout)- previous material properties c krvro (int,sc,in) c = 0 - real constants are used by this element, and the old c values(rvro) have been saved; or the element does not c use real constants. Any change of real constants c causes all matrices to be reformed. c = 1 - real constants are used by this element and the old c values(rvro) have been saved. However, any change c of real constants will cause the run to terminate, c because the results would be too unpredictable. c (e.g. gap elements) c = 2 - element is nonlinear, so do not bother to check c = 3 - real constants are used by this element, and the old c values(rvro) have been saved. However, no checking is c done in this routine because of needed customized logic. c = 4 - real constants are used by this element, but the old c values(rvro) have not been saved because it was c decided not to use this much storage. therefore, no check c can be made to determine if matrices should be reformed. c (e.g. 100 layer elements) c = 5 - real constants are used by this element, but the old c values(rvro) have not been saved because the real c constants have no effect on matrix formulation. c (e.g. acoustic elements) c rvr (dp,ar(*),in) - current real constants c rvro (dp,ar(*),inout) - previous real constants c amodo (dp,sc,inout) - previous value of mode c asymo (dp,sc,inout) - previous value of isym c c output arguments: c propo (dp,ar(nprop),inout)- current material properties c rvro (dp,ar(*),inout) - current real constants c amodo (dp,sc,inout) - current value of mode c asymo (dp,sc,inout) - current value of isym c kelin (int,ar(10),out) - keys indicating matrices to form c

1.4.7 Subroutine subrd (Reading Element Load Data for a Substructure Generation Run)
subroutine subrd (iel,key,nd,vect,ka) c *** primary function: read element load data from file for substructure c generation run c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c iel (int,sc,in) - element number c key (int,sc,in) - type of load data c = 1 temperature c = 2 fluences c = 3 heat generation rates

c c c c c c c c c c c c c c c c c c c c

nd

(int,sc,in)

current densities pressures film coefficients bulk temperatures extra displacement shapes thermal strains(eptho in el42) thermal flux (as in el55) initial strains(epino in el01) magnetic virtual displacements calculated source field(hsn in el96) element load vector copy - do not scale(tempev in el42) first load step only - number of data items - array of load data - load activation key = 0 no load for this data = 1 load is active

= 4 =10 =11 =12 =13 =14 =15 =16 =17 =18 =20 =30

output arguments: vect (dp,ar(nd),out) ka (int,sc,out)

1.4.8 Subroutine subwrt (Writing an Element Load Vector to a File for a Substructure Generation Run)
subroutine subwrt (iel,nvect,key,nd,vect,ref) c *** primary function: write element load vect to file for substructure c generation run c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c iel (int,sc,in) - element number c nvect (int,sc,in) - number of load vectors c (current load step number) c key (int,sc,in) - type of load vect c = 1 temperature c = 2 fluences c = 3 heat generation rates c = 4 current densities c =10 pressures c =11 film coefficients c =12 bulk temperatures c =13 extra displacement shapes c =14 thermal strains(eptho in el42) c =15 thermal flux (as in el55) c =16 initial strains(epino in el01) c =17 magnetic virtual displacements c =18 calculated source field(hsn in el96) c =20 element load vector c =30 copy - do not scale(tempev in el42) c nd (int,sc,in) - number of vect items c vect (dp,ar(nd),in) - array of load data c ref (dp,sc,in) - reference value for zero load

c c

output arguments: none

1.4.9 Subroutine rvrget (Fetching Real Constants for an Element)

subroutine rvrget (iel,ireal,ielc,nrvr,rvr) c *** primary function: get the real constants for an element c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c iel (int,sc,in) - element number c ireal (int,sc,in) - real constant set number c ielc (int,ar(150),in) - elment type characteristics c c output arguments: c nrvr (int,sc,out) - number of real variables c rvr (dp,ar(*),out) - element real constants

1.4.10 Subroutine propev (Evaluating a Group of Material Properties)

subroutine propev (iel,mtr,lp,tem,prop,n) c *** primary function: to evaluate a group of material properties c c propev is used to pass two or more material property numbers c thru the lp array to determine which temperature dependent c material properties are to be evaluated. c thus, the 3 prope1 calls: c c c c call prope1 (elem,mat, 1,tem,e(1)) call prope1 (elem,mat,10,tem,alpha) call prope1 (elem,mat,13,tem,dens) should be combined as:

c integer lp(3) c data lp /1,10,13/ c call propev (elem,mat,lp(1),tem,prop(1),3) c c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: iel (int,sc,in) mtr (int,sc,in) lp (int,ar(n),in) - element number - material number(input quantity mat, mat comma - keys for which specific value is requested each group must be in ascending order (ex,ey,ez, etc)

c c c c c c c c c c c c c c c c c c c

if negative, a required property if zero, leave prop term unchanged EX = 1, EY = 2, EZ = 3, NUXY= 4, NUYZ= 5, NUXZ= 6, GXY = 7, GXZ = 9, ALPX=10, ALPY=11, ALPZ=12, DENS=13, MU =14, DAMP=15, KYY =17, KZZ =18, RSVX=19, RSVY=20, RSVZ=21, C =22, HF =23, EMIS=25, ENTH=26, LSST=27, PRXY=28, PRYZ=29, PRXZ=30, MURX=31, MURZ=33, PERX=34, PERY=35, PERZ=36, MGXX=37, MGYY=38, MGZZ=39, EGYY=41, EGZZ=42, TGXX=43, TGYY=44, TGZZ=45, SONC=46, (see chapter 2 of the elements volume of the user's manual for a detailed description)) tem n (dp,sc,in) (int,sc,in)

GYZ = 8 KXX =16 VISC=24 MURY=32 EGXX=40

- temperature at which to evaluate material - number of properties to be evaluated. (20 maximum) If n = 1, use prope1 instead. - values of material property

output arguments: prop (dp,ar(n),out)

1.4.11 Subroutine prope1 (Evaluating One Material Property)

subroutine prope1 (iel,mtr,icon,tem,prop1) c *** primary function: to evaluate one material property c (if multiple material properties are to c be evaluated, use propev) c *** secondary functions: to ensure that certain required props are present c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c iel (int,sc,in) - element number c mtr (int,sc,in) - material number c icon (int,sc,in) - key for which specific value is requested c (negative if property is required) c EX = 1, EY = 2, EZ = 3, NUXY= 4, NUYZ= 5, NUXZ= 6, GXY = 7, GYZ = 8 c GXZ = 9, ALPX=10, ALPY=11, ALPZ=12, DENS=13, MU =14, DAMP=15, KXX =16 c KYY =17, KZZ =18, RSVX=19, RSVY=20, RSVZ=21, C =22, HF =23, VISC=24 c EMIS=25, ENTH=26, LSST=27, PRXY=28, PRYZ=29, PRXZ=30, MURX=31, MURY=32 c MURZ=33, PERX=34, PERY=35, PERZ=36, MGXX=37, MGYY=38, MGZZ=39, EGXX=40 c EGYY=41, EGZZ=42, TGXX=43, TGYY=44, TGZZ=45, SONC=46, c c tem (dp,sc,in) - temperature at which to evaluate material c c output arguments: c prop1 (dp,sc,out) - value of material property c

1.4.12 Subroutine pstevl (Evaluating EX, NUXY, GXY, ALPX, and DENS at Element Temperature)
subroutine pstev1 (elem,matin,tem,prop) c *** primary function: to evaluate material properites for 1-d elements c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c elem (int,sc,in) - element number (for anserr) c matin (int,sc,in) - material reference number c if negative, no required properties c tem (dp,sc,in) - temperature for evaluation c c output arguments: c prop (dp,ar(5),out) - material properties: ex,nuxy,gxy,alpx,dens c

1.4.13 Function gtplop (Retrieving the Start Point for Plastic Data)
function gtplop (mat,plopt) c *** primary function: get the plasticity option c *** secondary function: return the virtual starting address for the data c (see nlpropcheck for similar logic) c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c mat (int,sc,in) - material reference number c c output arguments: c plopt (int,sc,out) - plasticity option: c 1 bkin c 2 mkin c 3 miso c 4 biso c 5 aniso c 6 dp c 7 anand c 12 melas c 100 user c gtplop (int,sc,out) - virtual starting position of the data c

1.4.14 Function gtcrop (Retrieving the Start Position for Creep Data)
function gtcrop (mat,cropt) c *** primary function: get the creep options

c *** secondary function: return the virtual starting address for the data c (see nlpropcheck for similar logic) c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c mat (int,sc,in) - material reference number c c output arguments: c cropt (int,sc,out) - creep option: c 10000*(irrad+1) + 1000*(secondary+1) + c (primary+1) c where: c primary - primary creep (constant C6) c secondary - secondary creep (constant C12) c irrad - irradation induced (constant C66) c see creep table documentation c (TB command) c gtcrop (int,sc,out) - virtual starting position of the data c

1.4.15 Function gtswop (Retrieving the Start Position for Swelling Data)
function gtswop (mat,swopt) c *** primary function: get the swelling option c *** secondary function: return the virtual starting address for the data c (see nlpropcheck for similar logic) c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c mat (int,sc,in) - material reference number c c output arguments: c swopt (int,sc,out) - swelling option: c 10000*(irrad+1) + 1000*(secondary+1) + c (primary+1) c where: c primary - primary swelling (constant C6) c secondary - secondary swelling (constant C12) c irrad - irradation induced (constant C66) c see swelling table documentation (TB command) c gtswop (int,sc,out) - virtual starting position of the data

1.4.16 Subroutine tbuser (Retrieving User Table Data)

subroutine tbuser (mat,numitm,tbprop) c *** primary function: return the tb data for the user table c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c mat (int,sc,in) - material property number c numitm (int,sc,in) - the number of data items requested c c output arguments: c tbprop (dp,ar(numitm),out) - array of tb data

1.4.17 Subroutine plast1 (Updating an Element's Plastic History)

subroutine plast1 (option,elem,intpt,mat,kstart,tem,dtem,e,ktform, x dens,flu,dflu,epel,eppl,statev,usvr,epeq,plwork,sigepl,sigrat,et) c *** primary function: to update the plastic history (for 1 component) c *** secondary functions: to compute the material tangent matrix if requested c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c option (int,sc,in) - plasticity option c elem (int,sc,in) - element number (label) c intpt (int,sc,in) - element integration point number c mat (int,sc,in) - material reference number c kstart (int,sc,in) - virtual starting address of the data table c tem (dp,sc,in) - temperature at the end of this substep c dtem (dp,sc,in) - temperature increment over this substep c e (dp,sc,in) - elastic modulus c ktform (int,sc,in) - request key for tangent matrix formation c dens (dp,sc,in) - material density c flu (dp,sc,in) - fluence at the end of this substep c dflu (dp,sc,in) - fluence increment over this substep c epel (dp,sc,inout) - modified total strain (trial strain) c eppl (dp,sc,inout) - plastic strain at previous substep c statev (dp,ar(6),inout) - state variables at previous substep c usvr (dp,ar(*),inout) - user-defined state variables (for userpl) c epeq (dp,sc,inout) - effective plastic strain at prev substep c plwork (dp,sc,inout) - accumulated plastic work at prev substep c c output arguments: c epel (dp,sc,inout) - elastic strain c eppl (dp,sc,inout) - updated plastic strain c statev (dp,ar(6),inout) - updated state variables c usvr (dp,ar(*),inout) - updated user-defined state variables c epeq (dp,sc,inout) - updated effective plastic strain c plwork (dp,sc,inout) - updated accumulated plastic work c sigepl (dp,sc,out) - stress value on stress-strain curve c sigrat (dp,sc,out) - ratio of trial stress to yield stress

c c c c

et

(dp,sc,out)

- tangent modulus - equivalent plastic strain increment

internal variables: deppl (dp,sc)

1.4.18 Subroutine creep1 (Updating an Element's Creep History)

subroutine creep1 (option,elem,intpt,mat,kstart,epel,e,epcrp, x statev,usvr,tem,dtem,fluen,dflu,sig) c *** primary function: to update the creep history for 1-d elements c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c option (int,sc,in) - creep option c elem (int,sc,in) - element number (label) c intpt (int,sc,in) - element integration point number c mat (int,sc,in) - material reference number c kstart (int,sc,in) - virtual starting address of the data table c epel (dp,sc,inout) - elastic strain c e (dp,sc,in) - elastic modulus c epcrp (dp,sc,inout) - creep strain at previous substep c statev (dp,ar(7),inout) - state variables at previous substep c usvr (dp,ar(*),inout) - user-defined state variables (for usercr) c tem (dp,sc,in) - temperature at the end of this substep c dtem (dp,sc,in) - temperature increment over this substep c fluen (dp,sc,in) - fluence at the end of this substep c dflu (dp,sc,in) - fluence increment over this substep c epel (dp,sc,inout) - elastic strain adjusted for creep increment c sig (dp,sc,inout) - stress (not really used) c c output arguments: c epcrp (dp,sc,inout) - updated creep strain c statev (dp,ar(7),inout) - updated state variables c usvr (dp,ar(*),inout) - updated user-defined state variables c sig (dp,sc,inout) - stress (recomputed if requested)

1.4.19 Subroutine swell1 (Updating an Element's Swelling History)

subroutine swell1 (option,elem,intpt,mat,kstart,epswel,epel,e, x fluen,dfluen,tem,dtem,usvr) c *** primary function: to update the swelling history for 1-d elements c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c option (int,sc,in) - swelling option

c elem (int,sc,in) c intpt (int,sc,in) c mat (int,sc,in) c kstart (int,sc,in) table c epswel (dp,sc,inout) c epel (dp,sc,inout) c e (dp,sc,in) c fluen (dp,sc,in) c dfluen (dp,sc,in) c tem (dp,sc,in) c dtem (dp,sc,in) c usvr (dp,ar(*),inout) usersw) c c output arguments: c epel (dp,sc,inout) c epswel (dp,sc,inout) c usvr (dp,ar(*),inout) c

element number (label) element integration point number material reference number virtual starting address of the data swell strain at previous substep elastic strain elastic young'S MODULUS fluence at the end of this substep fluence increment over this substep temperature at the end of this substep temperature increment over this substep user-defined state variables (for

- elastic strain adjusted for swelling inc - updated swelling strain - updated user-defined state variables

1.4.20 function nlget (Retrieving Material Non-Linear Property Information)

function nlget (mat,iprop,prop) c *** primary function: get a material non-linear property (TB) table. c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c mat (int,sc,in) - material number c iprop (int,sc,in) - property number (tbpnum in tblecm) c use 73 for tb,user c use 74 for tb,nl c c output arguments: c variable (typ,siz,intent) description c nlget (int,sc,out) - number of property values c prop (dp,ar(nlget),out) - vector of the property values c

1.4.21 Subroutine usereo (Storing Data in the nmisc Record)

subroutine usereo (elem,iout,nbsvr,bsvr,nnrsvr,nrsvr,npsvr,psvr, x ncsvr,csvr,nusvr,usvr,nnode,nodes,xyz,vol,leng,time, x timinc,nutot,utot,maxdat,numdat,udbdat) c c *** primary function: to call userou, which allows user to store c data in nmisc record c c *** Notice - This file contains ANSYS Confidential information *** c

c input arguments: c variable (typ,siz,intent) c elem (int,sc,in) c iout (int,sc,in) c nbsvr (int,sc,in) c bsvr (dp,ar(nbsvr),in) c nnrsvr (int,sc,in) c nrsvr (dp,ar(nnrsvr),in) c npsvr (int,sc,in) c psvr (dp,ar(npsvr),in) c ncsvr (int,sc,in) c csvr (dp,ar(ncsvr),in) c nusvr (int,sc,in) variables c usvr (dp,ar(nusvr),in) c nnode (int,sc,in) c nodes (int,ar(nnode),in) c xyz (dp,ar(6,nnode),in) c vol (dp,sc,in) c leng (dp,sc,in) c time (dp,sc,in) c timinc (dp,sc,in) c nutot (int,sc,in) c utot (dp,ar(nutot),in) c maxdat (int,sc,in) c c c output arguments: c variable (typ,siz,intent) c numdat (int,sc,out) udbdat c udbdat (dp,ar(maxdat),out) end c

description - element number - output unit number - number of basic element variables - basic element variables - number of nonlinear element variables - nonlinear element variables - number of plasticity element variables - plasticity element variables - number of creep element variables - creep element variables - number of user-supplied element user-supplied element variables number of nodes node numbers nodal coordinates and rotations (virgin) element volume (or area if 2-d) element length (beams,spars,etc) current time current sub step time increment length of dof solution vector utot solution vector size of user output array (3 x nnode) actually, = ielc(nmnmup)

description - number of user output items in array - user output items to be placed at the of the nmisc record

1.4.22 Subroutine eldwrt (Writing Element Data to a File)

subroutine eldwrt (ielem,edtype,lcerst,edindx,nval,value) c *** primary function: output element data to result file. c *** Notice - This file contains ANSYS Confidential information *** c input arguments: c ielem (int,sc,in) c edtype (int,sc,in) c lcerst (int,sc,inout) c edindx (int,ar(25),inout)c nval (int,sc,in) c c ielc(nmsmis), c as c c value (dp,ar(nval),in) element number element data type (see elparm) pointer to results file position index to results file data the total number of values if edtype = EDEMS, this should -always- be unless there is a variable number, in the layered shell elements. output values (real)

1.4.23 Subroutine eldwrn (Writing Element Nonsummable Miscellaneous Data to the Results File)
subroutine eldwrn (elem,ielc,lcerst,edindx,nudb,udbdat,nval,value, x ndval) c *** primary function: output element nonsummable miscellaneous data c to result file c *** Notice - This file contains ANSYS Confidential information *** c input arguments: c elem (int,sc,in) c ielc (int,ar(IELCSZ),in) c c lcerst (int,sc,inout) c edindx (int,ar(25),inout)c nudb (in,sc,in) c udbdat (dp,ar(*),in) c nval (int,sc,in) c c ielc(NMNMIS), c as c c value (dp,ar(ndval),in) c ndval (int,sc,in) than c element number element characteristic vector defined in elccmt pointer to results file position index to results file data size of what the user wants to add what the user wants to add the total number of values to be output(does not include nudb) this should -always- be unless there is a variable number, in the layered shell elements. output values dimension of value - must be no less ielc(NMNMIS) + ielc(NMNMUP)

1.4.24 Subroutine trrot (Computing the Rotation Vector)

subroutine trrot (tr,rot) c *** primary function: get the rotation vector from a transformation matrix c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c tr (dp,ar(3,3),in) - transformation matrix c c output arguments: c rot (dp,ar(3),out) - rotation vector c

1.4.25 Subroutine rottr (Computing the Transformation Matrix)

subroutine rottr (rot,tr) c primary function: compute transformation matrix from rotation vector *****

c *** Notice - This file contains ANSYS Confidential information *** c c ref(old): eqn. 20(transposed),rankin and brogan, jpvt,108(1986)165-174. c ref(new): eqn. (b.4), simo and vu-quoc, cmame, 58 (1986), 79-116 c c variable descriptions: c input: c rot - rotation pseudovector c output: c tr - transformation matrix corresponding to rot c

1.4.26 Subroutine xyzup3 (Updating an Element's 3-D Nodal Coordinates)

subroutine xyzup3 (nnod,u,nr,xyz,nx,xyzup) c *** primary function: update a 3-d ele nodal coords for large deformation c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c nnod (int,sc,in) - number of nodes c u (dp,ar(nr),in) - displacement vector c nr (int,sc,in) - size of the u vector c xyz (dp,ar(nx,nnod),in) - coordinates to be updated c nx (int,sc,in) - row size of xy c c output arguments: c xyzup (dp,ar(3,nnod),out) - updated coordinates c

1.4.27 Subroutine updrot (Updating the Rotation Pseudovector)

subroutine updrot (v2,w1) c primary function: update the rotation pseudovector for 3-d large rotations ***** c c *** Notice - This file contains ANSYS Confidential information *** c c the updating of the pseudovector uses the mathematics of quarternions c (ref: eqn. a5 of j. h. argyris, cmame, 32(1982)85-155). the c pseudovector uses the nomalization proposed by rankin and brogan (ref: c eqn. 15, jpvt, 108(1986)165-174). c c variable descriptions: c input: c v2 - rotation increment c w1 - previous rotation pseudovector c output: c w1 - updated pseudovector c

c c c c c

v1 = cos(v1/2) + 1/2*w1, w1 = 2*sin(v1/2)*e1 v2 = cos(v2/2) + 1/2*w2, w2 = 2*sin(v2/2)*e2 v21 = v2*v1 = cos(v21/2) + 1/2*w21 (quarternion multiplication) w1 =: v21 (w1 is updated)

1.4.28 Subroutine tmpget (Defining Current Temperature Loads)

x c c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c iel (int,sc,in) - element number c ielc (int,ar(IELCSZ),in) - array of element type characteristics c nnod (int,sc,in) - number of nodes in the nodes array c nodes (int,ar(nnod),in) - list of nodes c ref (dp,sc,in) - reference temperature c ndat (int,sc,in) - number of data items to get c begdat (dp,ar(ndat),in) - data at the beginning of this load step c c output arguments: c dat (dp,ar(ndat),out) - data at this time point c enddat (dp,ar(ndat),out) - data at end of this load step c iexist (int,sc,out) - flag if temperatures exist c = 0 - no temperatures c = 1 - yes temperatures c subroutine tmpget (iel,ielc,nnod,nodes,ref,ndat,begdat,dat,enddat, iexist) primary function: define the current temperature loads

1.4.29 Subroutine prsget (Defining Current Pressure Loads)

subroutine prsget (iel,ielc,nfac,ndat,begdat,dat,enddat,iexist) c primary function: define the current pressure loads c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c iel (int,sc,in) - element number c ielc (int,ar(IELCSZ),in) - array of element type characteristics c nfac (int,sc,in) - number of pressure faces c ndat (int,sc,in) - number of pressure values c begdat (dp,ar(ndat),in) - pressure at the beginning of load step c c output arguments:

c c c c c c

dat enddat iexist

(dp,ar(ndat),out) - pressures at this iteration (dp,ar(ndat),out) - pressure at end of this load step (int,sc,out) - flag if pressure exist = 0 - no pressure = 1 - yes pressure

1.4.30 Subroutine cnvget (Defining Current Convection Loads)

x c c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c iel (int,sc,in) - element number c ielc (int,ar(IELCSZ),in) - array of element type characteristics c nr (int,sc,in) - dimension of u (temperature) vector c u (dp,ar(nr),in) - most current temperatures c nfac (int,sc,in) - number of convection faces c ndat (int,sc,in) - number of convection values c beghc (dp,ar(ndat),in) - hcoef at the beginning of load step c begtb (dp,ar(ndat),in) - tbulk at the beginning of load step c c output arguments: c hc (dp,ar(ndat),out) - hcoef at this substep c tb (dp,ar(ndat),out) - tbulk at this substep c endhc (dp,ar(ndat),in) - hcoef at the end of this load step c endtb (dp,ar(ndat),in) - tbulk at the end of this load step c iexist (int,sc,out) - flag if convection exist c = 0 - no convection c = 1 - constant convection (with time) c does not require new element matrix c = 2 - changing convection (with time) c or deleted convection c requires new element matrix c subroutine cnvget (iel,ielc,nr,u,nfac,ndat,beghc,begtb, hc,tb,endhc,endtb,iexist) primary function: define the current convection loads

1.4.31 Subroutine hgnget (Defining Current Heat Generation Loads)

x c c c *** Notice - This file contains ANSYS Confidential information *** subroutine hgnget (iel,ielc,nnod,nodes,ndat,begdat,dat,enddat, iexist) primary function: define the current heat generation loads

c c c c c c c c c c c c c c c c c c c

typ=int,dp,log,chr,dcp

siz=sc,ar(n),func

intent=in,out,inout

input arguments: variable (typ,siz,intent) iel (int,sc,in) ielc (int,ar(IELCSZ),in) nnod (int,sc,in) nodes (int,ar(nnod),in) ndat (int,sc,in) begdat (dp,ar(ndat),in) output arguments: dat (dp,ar(ndat),out) enddat (dp,ar(ndat),out) iexist (int,sc,out)

description - element number - array of element type characteristics - number of nodes in the nodes array - list of nodes - number of data items to get - data at the beginning of this load step - data at this time point - data at end of this load step - flag if heat generation exist = 0 - no heat generation = 1 - yes heat generation

1.5 Routines for Modifying and Monitoring Existing Elements

The next few pages describe the user routines you use to modify or monitor existing ANSYS elements. These routines enable you to perform tasks including:

Computing load vectors for frequency domain logic Storing element output that users supply Modifying the orientation of material properties and stresses Modifying the orientation of material properties and stresses of layers within an element Performing a user-defined operation on a parameter for the COMBIN7 and COMBIN37 elements. Providing a user-defined initial thickness for SHELL181. Providing a user-defined initial strain for SHELL181. Providing a user-defined fictive temperature relationship for VISCO88 and VISCO99. Provides visco-elastic computation in the stiffness pass for VISCO88 and VISCO99.

1.5.1 Subroutine userfd (Computing the Complex Load Vector for Frequency Domain Logic)

subroutine userfd (nr,kpeak,kcbrm,kpfor,ktrsur,isur, x cd,cm,ct,cb,do,doext,aread,alenv,nuxy,faclen,conac,fluidt,visc, x dprtem,watmot,xyzup,tr,accel,puvel,u,zass, x forl,zsc,zsc2,pdyn,holdwv) c *** primary function: compute complex load vector for frequency domain logic c *** secondary functions: none c -- accessed with keyopt(12) = 2 c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c nr (int,sc,in) - matrix size c kpeak (int,sc,in) - keyopt for wave combinations c kcbrm (int,sc,in) - key for reduced matrices/cable option c kpfor (int,sc,in) - keyopt for wave printout c ktrsur (int,sc,in) - keyopt for surface treatment(unfinished) c isur (int,sc,in) - surface flag c cd (dp,sc,in) - normal drag coefficient (real constant) c cm (dp,sc,in) - inertial coefficient (real constant) c ct (dp,sc,in) - tangential drag coefficient (real constant) c cb (dp,sc,in) - buoyancy coefficient (real constant) c do (dp,sc,in) - outside diameter of pipe c doext (dp,sc,in) - outside diameter of insulation c aread (dp,sc,in) - area of displaced water c alenv (dp,sc,in) - length of element c nuxy (dp,sc,in) - poisson'S RATIO c faclen (dp,sc,in) - wetted fraction of pipe c conac (dp,sc,in) - added mass per unit length c fluidt (dp,sc,in) - fluid temperature c visc (dp,sc,in) - viscosity c dprtem (int,sc,in) - size of water motion table c watmot (dp,ar(1278),in) - water motion table c xyzup (dp,ar(3,2),in) - updated coordinates c tr (dp,ar(3,3),in) - local to global transformation matrix c accel (dp,ar(3),in) - acceleration vector c puvel (int,sc,in) - index for velocities in u matrix c u (dp,ar(nr,5),in - displacements and velocities c zass (dp,ar(nr,nr),in) - mass matrix c forl (dp,ar(12),inout) - force vector in element coordinates c zsc (dp,ar(nr),inout) - real load vector for frequency domain c zsc2 (dp,ar(nr),inout) - complex load vector for frequency domain c c output arguments: c forl (dp,ar(12),inout) - force vector in element coordinates c zsc (dp,ar(nr),inout) - real load vector for frequency domain c zsc2 (dp,ar(nr),inout) - complex load vector for frequency domain c pdyn (dp,ar(2),out) - dynamic pressure c holdwv (dp,ar(60),out) - wave information held for printout c

1.5.2 Subroutine userou (Storing User-Supplied Element Output)

subroutine userou (elem,iout,nbsvr,bsvr,nnrsvr,nrsvr,npsvr,psvr, x ncsvr,csvr,nusvr,usvr,nnode,nodes,xyz,vol,leng,time, x timinc,nutot,utot,maxdat,numdat,udbdat) c c *** primary function: store user supplied element output c in nmisc record c c in order to activate this user programmable feature, c the user must enter the usrcal command. c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c this routine is called by almost every element c the data is stored on the nmisc record. c warning: other data may be stored between the c documented data and this data. c in order to see the actual information on the nmisc c record, insert the command: c dblist,elp,elnum1,elnum2,elinc,11 c where elnum1 = the first element c elnum2 = the last element c elinc = the element increment number c after a set command in post1. c c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - element number c iout (int,sc,in) - output unit number c nbsvr (int,sc,in) - number of basic element variables c bsvr (dp,ar(nbsvr),in) - basic element variables c nnrsvr (int,sc,in) - number of nonlinear element variables c nrsvr (dp,ar(nnrsvr),in) - nonlinear element variables c npsvr (int,sc,in) - number of plasticity element variables c psvr (dp,ar(npsvr),in) - plasticity element variables c ncsvr (int,sc,in) - number of creep element variables c csvr (dp,ar(ncsvr),in) - creep element variables c nusvr (int,sc,in) - number of user-supplied element variables c (= nstv on the nsvr command) c usvr (dp,ar(nusvr),in) - user-supplied element variables c nnode (int,sc,in) - number of nodes c nodes (int,ar(nnode),in) - node numbers c xyz (dp,ar(6,nnode),in) - nodal coordinates and rotations (virgin) c vol (dp,sc,in) - element volume (or area if 2-d) c leng (dp,sc,in) - element length (beams,spars,etc) c time (dp,sc,in) - current time c timinc (dp,sc,in) - current sub step time increment c nutot (int,sc,in) - length of dof solution vector utot c utot (dp,ar(nutot),in) - solution vector

c maxdat (int,sc,in) c c output arguments: c variable (typ,siz,intent) c numdat (int,sc,out) udbdat c ielc(NMNMUP) c number c c udbdat (dp,ar(maxdat),out) end c c

- size of user output array (3 x nnode) description - number of user output items in array (maximum size of numdat is which is usually three times the of nodes. - user output items to be placed at the of the nmisc record

1.5.3 Subroutine useran (Modifying Orientation of Material Properties)

subroutine useran (vn,vref,elem,thick,xyzctr,bsangl) c user written routine to modify orientation of material properties c and stresses *************************** c applicable to: shell43,63,91,93,99, solid46,64 c accessed by keyopt c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c **** warning *** do not change any arguments other than bsangl. c if you do, your results are probably wrong. c c input(do not change)--c vn = vector normal to element c vref = unit vector orienting element, essentially edge i-j c elem = element number c thick = total thickness of element at this point (see note below) c xyzctr = location of element centroid or integration point c c output--c bsangl = output from this subroutine. it represents the angle(s) c between vref and the desired orientation. it may have c the default orientation coming in to useran. c This will be combined with the angles derived from c the ESYS command. c use 1 angle for 2-d elements and shells c use 3 angles for 3-d solids c

1.5.4 Subroutine usanly (Modifying Orientation of Material Properties and Stresses of Layers)

subroutine usanly (vn,vref,elem,thick,xyzctr,ln,bsangl) c c user written routine to modify orientation of material properties c and stresses of layers within an element ************************** c applicable to shell91,99, and solid46 c accessed with keyopt(4) c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c **** warning *** do not change any arguments other than bsangl. c if you do, your results are probably wrong. c input(do not change)--c vn = vector normal to element c vref = unit vector orienting element, essentially edge i-j c elem = element number c thick = thickness of layer ln at this point (see note below) c xyzctr = location of element centroid or integration point c ln = layer number c c output--c bsangl = output from this subroutine. it represents the angle(s) c between vref and the desired orientation of the layer. c it may have the default orientation coming in to useran. c c this angle does not affect the angle given in the printout. c users may want to add their own angle printout in usanly. c c also, this angle does not affect the angle shown using the c layplot or laylist commands c c use 1 angle for 2-d elements and shells c use 3 angles for 3-d solids c c the programmer of usanly can base their logic on the location c of the point in space, or an element attributes, such as thick c

1.5.5 Subroutine userrc (Performing User Operations on COMBIN7 and COMBIN37 Parameters)
c c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - element number c ireal (int,sc,in) - element real constant number subroutine userrc (elem,ireal,type,nusvr,usvr,parm,parmld, x c1,c2,c3,c4,fcon) primary function: user operation on parameter for combin7 and combin37 accessed with keyopt(9) = 1

c type (int,sc,in) - element type number c nusvr (int,sc,in) - number of user-supplied element variables c usvr (dp,ar(nusvr),inout) - user-supplied element variables c parm (dp,sc,in) - current value of the paramater c parmld (dp,sc,in) - value of the parameter at previous time ste c c1 (dp,sc,in) - real constant c1 c c2 (dp,sc,in) - real constant c2 c c3 (dp,sc,in) - real constant c3 c c4 (dp,sc,in) - real constant c4 c c output arguments: c variable (typ,siz,intent) description c usvr (dp,ar(nusvr),inout) - user-supplied element variables c may be sent .rst file with usereo c fcon (dp,sc,out) - result of calculation c c either c1 or c3 must be nonzero for this logic to be accessed, c

1.5.6 Function userpe (Calculating Rotation Caused by Internal Pressure)

function userpe (prs,rvrp,angle,ex,nuxy) c primary function: c c c calculate the rotation caused by internal pressure on an elbow element This function is only called by el18(pipe18) if keyopt(5) = 1

c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 *** ansys, inc. typ=int,dp,log,chr,dcp input arguments: variable (typ,siz,intent) prs (dp,ar(5),in) rvrp (dp,ar(11),in) angle (dp,sc,in) ex (dp,sc,in) nuxy (dp,sc,in) siz=sc,ar(n) intent=in,out,inout

description pressure vector real constants(see elements manual) subtended angle Young's modulus Poisson's ratio description rotation caused by internal pressure on elbow element

c output arguments: c variable (typ,siz,intent) c userpe (dp,sc,out) the c

1.5.7 Subroutine UElMatx (Accessing Element Matrices and Load Vectors)

x subroutine UElMatx (elem,nr,ls,zs,zsc,uelm,ielc,nodes,edof, elmdat,xyzang,lenu) User routine to access element matrices and load Needs to have USRCAL,UELMATX to be accessed. Called after the call to the element routine and before the solver. May be used to monitor and/or modify the element and load vectors.

c primary function: vectors. c c c c matrices c c c c

*** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 *** ansys, inc. typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout

c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - User element number c nr (int,sc,in) - number of rows in element matrix c ls (int,ar(nr),in) - Dof Index vector for this element matrix c zs (dp,ar(nr,nr,4),inout)- K,M,C,SS matrices for this element c zsc (dp,ar(nr,2),inout) - Element load vector and N-R correction vec c uelm (dp,ar(nr,5),in) - Nodal displacements for this element c ielc (int,ar(150),in) - Element type characteristics c nodes (int,ar(*),in) - Nodes for this element c edof (int,ar(*),in) - Dofs per node for this element c elmdat (int,ar(10),in) - Element data for this element c xyzang (dp,ar(6,*),in) - X,Y,Z,THXY,THYZ,THZX for each element node c lenu (int,sc,in) - Length of global displacement vector c output arguments: c zs (dp,ar(nr,nr,4),inout)- K,M,C,SS matrices for this element c zsc (dp,ar(nr,2),inout) - Element load vector and N-R correction vec c WARNING: any CHANGES to these (or any other) arguments will have a direc c impact on the solution, possibly giving meaningless results. The normal c usage of this routine is simply monitor what is happening.

1.5.8 Subroutine UTHICK (Getting User-defined Initial Thickness)

SUBROUTINE uthick (elemId, elemType, matId, realId,

$ numDomIntPts, curCoords, thickness) ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c c *** primary function: get the user defined thickness c c *** ansys(r) copyright(c) 1998 c *** ansys, inc. c c input arguments c =============== c Variable (type,sz,i/o) description c elemId (int,sc,i) element number c elemType (int,sc,i) element TYPE (181 etc.) c matId (int,sc,i) material number c realId (int,sc,i) real constant set number c numDomIntPts (int,sc,i) number of integration points c curCoords (dp,ar(3,numDomIntPts),i) c current coordinates c c output arguments c ================ c thickness (dp,ar(3,numDomIntPts),o) c thickness at the integration points c ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c c --- parameters c #include "impcom.inc" EXTERNAL erhandler c c --- argument list c INTEGER elemId, elemType, numDomIntPts, matId, realId DOUBLE PRECISION curCoords(3,numDomIntPts), $ thickness (numDomIntPts) c c --- local variables c INTEGER i cc COMMON /dummyIniThick/ errFlag ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c --- B E G I N C O D I N G c c --- author of UTHICK.F should remove the following error message c and any unnecessary flags. c CALL erhandler ('uthick', 5000, 2, 'ANSYS, Inc.-supplied $ version of UTHICK has been used. This $ works with the COMMENTED EXAMPLE input $ attached with uthick.F', 0.d0, ' ') c c --- linearly varying thickness for the cantilever model in the c commented example; thick(x=0) = 0.1, thick(x=10) = 0.05 c DO i=1,numDomIntPts thickness(i) = 0.1d0*(1.d0-curCoords(1,i)/20.d0)

END DO c 990 c RETURN END ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc ccccccccccccc c c C O M M E N T E D I N P U T F I L E T O B E U S E D W I T H c c T H E A N S Y S S U P P L I E D U T H I C K c ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc ccccccccccccc c!/batch,list c!/verify,ev181-57s c!/title,ev181-57s, Variable thickness shell with UTHICK c!/nolist c!/nopr c!/out,scratch c! c!/prep7 c! c!!!!!!! STANDARD USE OF SHELL 181 with Variable Thickness !!!!! c! c!*dim,thick,,10 c!thick(1) = 0.0975 c!thick(2) = 0.0925 c!thick(3) = 0.0875 c!thick(4) = 0.0825 c!thick(5) = 0.0775 c!thick(6) = 0.0725 c!thick(7) = 0.0675 c!thick(8) = 0.0625 c!thick(9) = 0.0575 c!thick(10)= 0.0525 c! c!*dim,lnodes,,11 c!*dim,unodes,,11 c! c!*do,i,1,11 c! x = (i-1)*1.0 c! n,i,x c! n,i+11,x,1.0 c! lnodes(i) = i c! unodes(i) = i+11 c!*enddo c! c!et,1,181 c!mp,ex,1,1000 c!mp,nuxy,1,0.3 c! c!*do,i,1,10 c! r,i,thick(i) c! real,i CONTINUE

c! e,lnodes(i),lnodes(i+1),unodes(i+1),unodes(i) c!*enddo c! c!!!!!!! USE OF SHELL 181 with Variable Thickness Through UTHICK !!!!! c! c!*dim,l2nod,,11 c!*dim,u2nod,,11 c!*do,i,1,11 c! x = (i-1)*1.0 c! n,i+22,x,2.0 c! n,i+33,x,3.0 c! l2nod(i) = i+22 c! u2nod(i) = i+33 c!*enddo c! c!et,2,181 c!keyopt,2,9,1 c!r,11 c!type,2 c!real,11 c! c!*do,i,1,10 c! e,l2nod(i),l2nod(i+1),u2nod(i+1),u2nod(i) c!*enddo c! c!/out c!elist c!etlist c!/out,scratch c! c!nsel,s,loc,x, c!d,all,ux c!d,all,uz c!d,all,rotx c!d,all,rotz c!d,all,roty c! c!nsel,s,loc,y c!d,all,uy c! c!nsel,s,loc,y,2.0 c!d,all,uy c! c!nsel,s,loc,x,10.0 c!f,all,my,0.004 c!alls c!fini c! c!/solu c!outres,all,all c!nlgeom,on c!solc,on c!nsubst,10,,5 c!eqslv,sparse c!solve c!fini c!

c!/post1 c!set,last c!nsel,s,loc,x,10.0 c!/out c!/com c!/com, ------------------------------------------------------c!/com, | Results of Node # 11 and 22 are from the standard use c!/com, | of SHELL181; the results of Node # 33 and 44 are c!/com, | obtained by using UTHICK (keyopt 9 = 1) and should c!/com, | match the results of node # 11 and 22 c!/com, ------------------------------------------------------c!/com c!prdi c!/out,scratch c!finish c! c!/delete,ev181-57s.emat c!/delete,ev181-57s.err c!/delete,ev181-57s.esav c!/delete,ev181-57s.osav c!/delete,ev181-57s.log c!/delete,ev181-57s.mntr c!/delete,ev181-57s.rst c!/delete,ev181-57s.full c!/golist c!/exit,nosave

| | | |

1.5.9 Subroutine USTRESS (Getting User-defined Initial Stress)

*deck,ustress parallel user smg SUBROUTINE ustress (elemId, elemType, matId, basis, $ curIntPt, curLayer, curSecPt, $ numDirect, numShear, curCoords, $ tLocation, iniData) ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c c *** primary function: get the user supplied stress state for c initial stress analysis c c *** ansys(r) copyright(c) 1998 c *** ansys, inc. c c input arguments c =============== c Variable (type,sz,i/o) description c elemId (int,sc,i) element number c elemType (int,sc,i) element TYPE (42, 181 etc.) c matId (int,sc,i) material number c basis (dp,ar(3,3),i) basis directions (for the element) c curIntPt (int,sc,i) current Integration point c curLayer (int,sc,i) current Layer number (for shell c and layered solids)

c curSecPt (int,sc,i) current point through the thickness in the c current layer c numDirect (int,sc,i) number of direct stress comp c (s_xx,s_yy,s_zz) c numShear (int,sc,i) number of shear stress comp c (s_xy,s_yz,s_zx) c curCoords (dp,ar(3),i) current coordinates c tLocation (dp,sc,i) normalized thickness location c (shell elements only) c -1 = bottom or first sec pt c +1 = top or last sec pt c 0 = mid surface c c output arguments c ================ c iniData (dp,ar(numDirect+numShear),o) c user input stress values c ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c c --- parameters c #include "impcom.inc" EXTERNAL erhandler c c --- argument list c INTEGER elemId, elemType, matId, curIntPt, curLayer, $ curSecPt, numDirect, numShear DOUBLE PRECISION curCoords(3), tLocation, basis(3,3), $ iniData(numDirect+numShear) c c --- local variables c INTEGER i,j,k DOUBLE PRECISION dummy (4,5,10) cc COMMON /dummyIniDat/ dummy,errFlag ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc c --- B E G I N C O D I N G c c --- author of USTRESS.F should remove the data statement, error msgs c and any unnecessary flags. The data you define here does not c necessarily have to be done with data statement c DATA (((dummy(i,j,k),i=1,4),j=1,5),k=1,2) / $ -0.1200E+02, -0.1249E+01, 0.0000E+00, -0.4003E-04, $ -0.7202E+01, -0.7495E+00, 0.0000E+00, 0.4515E-04, $ 0.1924E-04, -0.6599E-03, 0.0000E+00, 0.1729E-03, $ 0.7202E+01, 0.7481E+00, 0.0000E+00, 0.3007E-03, $ 0.1200E+02, 0.1247E+01, 0.0000E+00, 0.3859E-03, $ -0.1200E+02, 0.3359E+00, 0.0000E+00, 0.1337E-02, $ -0.7201E+01, 0.2019E+00, 0.0000E+00, 0.7683E-03, $ 0.1732E-03, 0.8891E-03, 0.0000E+00, -0.8421E-04, $ 0.7201E+01, -0.2001E+00, 0.0000E+00, -0.9367E-03, $ 0.1200E+02, -0.3341E+00, 0.0000E+00, -0.1505E-02/ DATA (((dummy(i,j,k),i=1,4),j=1,5),k=3,4) /

$ $ $ $ $ $ $ $ $ $

-0.1200E+02, -0.9424E-01, 0.0000E+00, -0.7202E+01, -0.5653E-01, 0.0000E+00, 0.4304E-03, 0.2657E-04, 0.0000E+00, 0.7202E+01, 0.5659E-01, 0.0000E+00, 0.1200E+02, 0.9429E-01, 0.0000E+00, -0.1200E+02, 0.3043E-01, 0.0000E+00, -0.7200E+01, 0.1827E-01, 0.0000E+00, 0.6953E-03, 0.3727E-04, 0.0000E+00, 0.7201E+01, -0.1820E-01, 0.0000E+00, 0.1200E+02, -0.3036E-01, 0.0000E+00, DATA (((dummy(i,j,k),i=1,4),j=1,5),k=5,6) / $ -0.1200E+02, -0.1066E-01, 0.0000E+00, $ -0.7200E+01, -0.6225E-02, 0.0000E+00, $ 0.8873E-03, 0.4321E-03, 0.0000E+00, $ 0.7201E+01, 0.7089E-02, 0.0000E+00, $ 0.1200E+02, 0.1153E-01, 0.0000E+00, $ -0.1200E+02, 0.4769E-02, 0.0000E+00, $ -0.7197E+01, 0.2935E-02, 0.0000E+00, $ 0.9262E-03, 0.1837E-03, 0.0000E+00, $ 0.7199E+01, -0.2568E-02, 0.0000E+00, $ 0.1200E+02, -0.4402E-02, 0.0000E+00, DATA (((dummy(i,j,k),i=1,4),j=1,5),k=7,8) / $ -0.1199E+02, -0.2508E-02, 0.0000E+00, $ -0.7195E+01, -0.1396E-02, 0.0000E+00, $ 0.7471E-03, 0.2729E-03, 0.0000E+00, $ 0.7197E+01, 0.1941E-02, 0.0000E+00, $ 0.1199E+02, 0.3054E-02, 0.0000E+00, $ -0.1198E+02, 0.3654E-02, 0.0000E+00, $ -0.7191E+01, 0.2247E-02, 0.0000E+00, $ 0.2508E-03, 0.1354E-03, 0.0000E+00, $ 0.7191E+01, -0.1976E-02, 0.0000E+00, $ 0.1199E+02, -0.3383E-02, 0.0000E+00, DATA (((dummy(i,j,k),i=1,4),j=1,5),k=9,10) / $ -0.1198E+02, -0.4337E-02, 0.0000E+00, $ -0.7186E+01, -0.2616E-02, 0.0000E+00, $ -0.6408E-03, -0.3430E-04, 0.0000E+00, $ 0.7185E+01, 0.2547E-02, 0.0000E+00, $ 0.1198E+02, 0.4269E-02, 0.0000E+00, $ -0.1197E+02, -0.1326E-02, 0.0000E+00, $ -0.7181E+01, -0.8506E-03, 0.0000E+00, $ -0.1444E-02, -0.1371E-03, 0.0000E+00, $ 0.7178E+01, 0.5763E-03, 0.0000E+00, $ 0.1196E+02, 0.1052E-02, 0.0000E+00, c i = 0 j = 0 k = 0 c

-0.1560E-02, -0.9405E-03, -0.1124E-04, 0.9180E-03, 0.1537E-02, 0.9459E-03, 0.5991E-03, 0.7890E-04, -0.4413E-03, -0.7882E-03/ -0.4662E-03, -0.3059E-03, -0.6545E-04, 0.1750E-03, 0.3353E-03, 0.2606E-03, 0.1814E-03, 0.6259E-04, -0.5620E-04, -0.1354E-03/ -0.3990E-04, -0.3183E-04, -0.1972E-04, -0.7612E-05, 0.4584E-06, 0.2321E-03, 0.1487E-03, 0.2362E-04, -0.1015E-03, -0.1849E-03/ 0.2315E-03, 0.1461E-03, 0.1810E-04, -0.1099E-03, -0.1953E-03, 0.7206E-03, 0.4404E-03, 0.2012E-04, -0.4002E-03, -0.6803E-03/

IF (curIntPt.EQ.1 .AND. curLayer.EQ.1 .AND. curSecPt.EQ.1) THEN CALL erhandler ('ustress', 5000, 2, 'ANSYS, Inc.-supplied $ version of USTRESS has been used. This $ works with the COMMENTED EXAMPLE input $ attached with ustress.F', 0.d0, ' ') END IF c iniData(1) = dummy(1,curSecPt,elemId) iniData(2) = dummy(2,curSecPt,elemId)

iniData(3) = dummy(3,curSecPt,elemId) iniData(4) = dummy(4,curSecPt,elemId) c 990 c RETURN END ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc cccccccccccccccc c c C O M M E N T E D I N P U T F I L E T O B E U S E D W I T H c c T H E A N S Y S S U P P L I E D U S T R E S S c ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc cccccccccccccccc c!/batch,list c!/PREP7 c! c!et,1,181 c!keyopt,1,10,1 c!r,1,1.0 c! c!mp,ex,1,1000 c!mp,nuxy,1,0.3 c! c!NBLOCK,6,SOLID c!(3i8,6e16.9) c! 23 0 0 0.000000000E+00 c! 24 0 0 9.90525175 0.000000000E+00 -1.18849741 c! 25 0 0 0.999933795 0.000000000E+00-1.162898000E-02 c! 26 0 0 1.99930664 0.000000000E+00-4.697524623E-02 c! 27 0 0 2.99754302 0.000000000E+00-0.106378219 c! 28 0 0 3.99406599 0.000000000E+00-0.189702267 c! 29 0 0 4.98830170 0.000000000E+00-0.296924624 c! 30 0 0 5.97967935 0.000000000E+00-0.427972294 c! 31 0 0 6.96762690 0.000000000E+00-0.582764249 c! 32 0 0 7.95157929 0.000000000E+00-0.761199170 c! 33 0 0 8.93097304 0.000000000E+00-0.963156159 c! 34 0 0 9.90524927 1.00000034 -1.18850710 c! 35 0 0 0.000000000E+00 0.999997689 c! 36 0 0 8.93097042 1.00000026 -0.963168610 c! 37 0 0 7.95157631 1.00000006 -0.761213374 c! 38 0 0 6.96762465 1.00000006 -0.582779590 c! 39 0 0 5.97967659 1.00000004 -0.427988151 c! 40 0 0 4.98830027 0.999999775 -0.296940984 c! 41 0 0 3.99406429 1.00000056 -0.189719078 c! 42 0 0 2.99754103 0.999999100 -0.106393013 c! 43 0 0 1.99930796 1.00000069 -4.699745205E-02 c! 44 0 0 0.999931339 1.00000098 -1.163438631E-02 c!N,R5.3,LOC, -1, c!N,R5.1,LOC, -1, c! c!EBLOCK,18,SOLID c!(18i7) CONTINUE

c! 1 1 1 1 0 0 0 0 4 0 23 25 44 35 c! 1 1 1 2 0 0 0 0 4 0 25 26 43 44 c! 1 1 1 3 0 0 0 0 4 0 26 27 42 43 c! 1 1 1 4 0 0 0 0 4 0 27 28 41 42 c! 1 1 1 5 0 0 0 0 4 0 28 29 40 41 c! 1 1 1 6 0 0 0 0 4 0 29 30 39 40 c! 1 1 1 7 0 0 0 0 4 0 30 31 38 39 c! 1 1 1 8 0 0 0 0 4 0 31 32 37 38 c! 1 1 1 9 0 0 0 0 4 0 32 33 36 37 c! 1 1 1 10 0 0 0 0 4 0 33 24 34 36 c! -1 c!EN,R5.1,ATTR, -1, c! c!nsel,s,loc,x, c!d,all,ux c!d,all,uz c!d,all,rotx c!d,all,rotz c!d,all,roty c! c!nsel,s,loc,y c!d,all,uy c!nall c! c!fini c! c!/solu c! c!nlgeom,on c!solcontrol,on c!nropt,full,,off c!lnsrch,off c!nsubst,1 c!outres,all,all c!cnvtol,f,,,,1.0e-4 c!cnvtol,m,,,,1.0e-4 c!eqslv,sparse c!solve c! c!/post1 c!set,last c!pres,s c!prdi ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc cccccccccccccccc

1.5.10 Subroutine UsrFictive (Providing User-defined Fictive Temperature Relationship)

subroutine UsrFictive (veinpt,tref,toffst,tem,ftl, ftc) c *** primary function: allow users to write their own c fictive temperature relationship c this logic is accessed with c5 = 11 on the tb,evisc table c *** secondary function: demonstrate the use of a user-written c fictive temperature relationship c c *** notice- this routine contains ansys,inc. confidential information *** c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (type,size,intent) description c c veinpt (dp,ar(95),in) - table from tb,evisc c tref (dp,sc,in) - reference temperature c toffst (dp,sc,in) - temperature offset from absolute zero c tem (dp,sc,in) - temperature at the end of this substep c ftl (dp,sc,in) - previous fictive temperature c c output arguments: c variable (type,size,intent) description c ftc (dp,sc,in) - fictive temperature c

1.5.11 Subroutine UsrViscEl (Performs Visco-elastic Computation)

subroutine UsrViscEl (veinpt,ncomp,tem,dtem,ex,gxy,eex,egxy,phil, x zil,g1l,g2l,hsm,hbm,smcm,bmcm,epsl,epst,ftl,dftl,tref,tvc,dvc, x dsig,rsig,cm,kerr,cml) ****** perform visco-elastic computation - stiffness pass

c c c ***** notice- this routine contains ansys, inc. confidential information **** c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c veinpt (dp,ar(95),in) - viscoelastic input data c ncomp (int,sc,in) - number of components (4 or 6) c tem (dp,sc,in) - temperature c dtem (dp,sc,in) - incremental temperature c ex (dp,sc,in) - bulk modulus at infinite time c gxy (dp,sc,in) - shear modulus at infinite time c smcm (dp,ar(ncomp,ncomp),in) - material matrix for shear modulus c bmcm (dp,ar(ncomp,ncomp),in) - material matrix for bulk modulus

c epsl (dp,ar(ncomp),in) - strain for previous iteration c epst (dp,ar(ncomp),in) - total strain for current iteration c tref (dp,sc,in) - reference temperature c c output arguments: c eex (dp,sc,out) - effective bulk modulus c egxy (dp,sc,out) - effective shear modulus c phil (dp,sc,inout) - previous shift factor c zil (dp,sc,inout) - previous pseudo time c g1l (dp,ar(ncomp,10),out) - recursive shear relaxation c g2l (dp,ar(ncomp,10),out) - recursive bulk relaxation c hsm (dp,ar(10),out) - recursive shear relaxation c hbm (dp,ar(10),out) - recursive bulk relaxation c ftl (dp,sc,inout) - previous fictive temperature c dftl (dp,ar(10),out) - incremental fictive temperature c tvc (dp,sc,inout) - total volume change c dvc (dp,sc,out) - incremental volume change c dsig (dp,ar(ncomp),out) - stress change c rsig (dp,ar(ncomp),out) - stress relaxation c cm (dp,ar(ncomp,ncomp),out)- total material matrix c kerr (int,sc,out) - error key c c argument of convenience: c cml (dp,ar(ncomp,ncomp),none)- no value (used only to avoid simplify c logic due to variable array sizes)

1.6 Routines for Customizing Material Behavior

This section describes the user routines you use to modify or monitor how materials behave. These routines enable you to perform tasks including:

Writing your own plasticity, creep, or swelling laws Updating the nonlinear strain history for a user-defined material Checking material data you've defined Computing the derivatives of the strain energy density "W" with respect to its invariants Controlling hygrothermal growth.

Note-If you write a material behavior routine using any of the ANSYS commands MPDATA, MPDELE, TB, or TBDELE, be aware that these commands interpret the string "_MATL" to mean the current active material when it appears in their MAT field. The "_MATL" is used in conjunction with the library (LIB) option of the MPREAD and MPWRITE commands. MPWRITE inserts "_MATL" in lieu of the specified material number as these commands are written to the material library file. This occurs only when you use the LIB option. When ANSYS reads a material

library file written in this format, it interprets "_MATL" to mean the currently active material (as defined by the MAT,MAT command). Don't use the "_MATL" string outside the scope of the MPREAD command.

1.6.1 Subroutine userpl (Writing Your Own Plasticity Laws)

subroutine userpl (elem,intpt,mat,ncomp,kfirst,kfsteq,e,nu,dens, x prop,d,ktform,timval,timinc,tem,dtem,toffst,flu,dflu,epel,eppl, x statev,usvr,epeq,plwork,sigepl,sigrat,depeq,dt) c c *** primary function: allow users to write their own plasticity laws. c this logic is accessed with tb,user. c the below demonstration logic is the same as using c tb,bkin, without adaptive descent (nropt,,,off). c Other plasticity rules may require internal c iterations and/or the more general definition of c plasticity theory, discussed in the Theory c Manual. c *** secondary function: demonstrate the use of user-written plasticity laws c in this routine: c a. update the nonlinear strain history c b. compute the material tangent matrix if requested c c *** Notice - This file contains ANSYS Confidential information *** c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (type,size,intent) description c c elem (int,sc,in) - element number (label) c intpt (int,sc,in) - element integration point number c mat (int,sc,in) - material reference number c ncomp (int,sc,in) - no. of stress/strain components (1,4 or 6) c 1 - x c 4 - x,y,z,xy c 6 - x,y,z,xy,yz,xz c kfirst (int,sc,in) - 1 if first time through, 0 otherwise c (useful for initializing state variables c to a non-zero value) c kfsteq (int,sc,in) - 1 if first equilibrium iteration of a c substep, 0 otherwise c c e (dp,sc,in) - average elastic modulus c nu (dp,sc,in) - average poisson ratio c dens (dp,sc,in) - current material density (mass/volume) c prop - linear material property array c (dp,ar(9),in) (ex,ey,ez, gxy,gyz,gxz, nuxy,nuyz,nuxz) c (dp,ar(1),in) if ncomp=1 (ex)

c d (dp,ar(ncomp,ncomp),in)- elastic stress-strain matrix c ktform (int,sc,in) - request key for tangent matrix formation c (=1, form tangent =0, do not form) c c timval (dp,sc,in) - current time value c timinc (dp,sc,in) - time increment over this substep c c tem (dp,sc,in) - temperature at the end of this substep c dtem (dp,sc,in) - temperature increment over this substep c toffst (dp,sc,in) - temperature offset from absolute zero c flu (dp,sc,in) - fluence at the end of this substep c dflu (dp,sc,in) - fluence increment over this substep c c epel (dp,ar(ncomp),inout)- modified total strain (trial strain) c epel = eptot - eppl - eptherm - ... c if a large strain analysis, epel is c rotation neutralized and is the hencky c (i.e. log) strain c eppl (dp,ar(ncomp),inout)- plastic strain from previous substep c c statev (dp,ar(ncomp,6),inout)- state variables from previous substep c usvr (dp,ar(nuval,nintp),inout)- additional state variables from c previous equilibrium iteration (saved c if the nsvr command is used) c c epeq (dp,sc,inout) - effective plastic strain from prev substep c plwork (dp,sc,inout) - accumulated plastic work from prev substep c c output arguments: c variable (type,size,intent) description c c epel (dp,ar(ncomp),inout)- elastic strain c eppl (dp,ar(ncomp),inout)- updated plastic strain c c statev (dp,ar(ncomp,6),inout)- updated state variables c usvr (dp,ar(nuval,nintp),inout)- updated additional state variables c c epeq (dp,sc,inout) - updated effective plastic strain c plwork (dp,sc,inout) - updated accumulated plastic work c c sigepl (dp,sc,out) - stress value on stress-strain curve at epeq c sigrat (dp,sc,out) - ratio of trial stress to yield stress c depeq (dp,sc,out) - increment in plastic strain (equivalent) c (used for auto time stepping - time step c is reduced if it exceeds .05) c c dt (dp,ar(ncomp,ncomp),out)- material tangent modulus c c fortran parameters (to be defined by the user): c variable (type) description c numinp (int) - number of data items in the user-defined c data table (tbdat commands) c nuval (int) - number of additional state variables per c integration point

c nintp (int) c c c note: nuval x nintp c c internal variables: c variable (type,size) c b (dp,ar(6,6)) c c (dp,ar(6,12)) c con (dp,sc) c deppl (dp,ar(6)) c dfds (dp,ar(6)) c dlamb (dp,sc) c ep (dp,ar(6)) c epshfo (dp,ar(6)) surf) c epshft (dp,ar(6)) c et (dp,sc) c h (dp,sc) strain) c n2 (int,sc) c seqtr (dp,sc) c sigtr (dp,ar(6)) c sigy (dp,sc) c vect (dp,ar(6)) c

- maximum number of integration points of an element to be used with this routine (14 is the maximum) = nstv(on nsvr command); cannot exceed 840! description - 2nd derivative of the yield function - part of deff - temporary variable - plastic strain increment - derivative of the yield function (normal) - plastic multiplier - shifted strain - initial shift strain (center of the yield - shift strain (center of the yield surface) - tangent modulus (stress/total strain) - plastic tangent modulus (stress/plastic ncomp squared, ncomp*ncomp equivalent (von mises) trial stress trial stress yield stress temporary vector

1.6.2 Subroutine usercr (Writing Your Own Creep Laws)

subroutine usercr (elem,intpt,mat,ncomp,kfirst,kfsteq,e,posn,d, x proptb,timval,timinc,tem,dtem,toffst,fluen,dfluen,epel,epcrp, x statev,usvr,delcr) c c *** primary function: allow users to write their own creep laws. c this logic is accessed with c6 = 100 c *** secondary function: demonstrate the use of user-written creep laws c c *** Notice - This file contains ANSYS Confidential information *** c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (type,size,intent) description c c elem (int,sc,in) - element number (label) c intpt (int,sc,in) - element integration point number c mat (int,sc,in) - material reference number c ncomp (int,sc,in) - no. of stress/strain components (1,4 or 6) c 1 - x c 4 - x,y,z,xy c 6 - x,y,z,xy,yz,xz c kfirst (int,sc,in) - 1 if first time through, 0 otherwise

c (useful for initializing state variables c to a non-zero value) c kfsteq (int,sc,in) - 1 if first equilibrium iteration of a c substep, 0 otherwise c c c e (dp,sc,in) - elastic young'S MODULUS c posn (dp,sc,in) - poisson'S RATIO c d (dp,ar(ncomp,ncomp),in)- elastic stress-strain matrix c proptb (dp,ar(72),in) - material properties input on tb commands c c timval (dp,sc,in) - current time value c timinc (dp,sc,in) - time increment over this substep c tem (dp,sc,in) - temperature at the end of this substep c dtem (dp,sc,in) - temperature increment over this substep c toffst (dp,sc,in) - temperature offset from absolute zero c fluen (dp,sc,in) - fluence at the end of this substep c dfluen (dp,sc,in) - fluence increment over this substep c c epel (dp,ar(ncomp),inout)- elastic strain c epcrp (dp,ar(ncomp),inout)- creep strain from previous substep c statev (dp,ar(ncomp*5+2),inout)- state variables from previous substep c usvr (dp,ar(nuval,nintp),inout)- additional state variables from c previous equilibrium iteration (saved c if the nsvr command is used) c c c output arguments: c variable (type,size,intent) description c c epel (dp,ar(ncomp),inout)- elastic strain adjusted for creep increment c epcrp (dp,ar(ncomp),inout)- updated creep strain c statev (dp,ar(ncomp*5+2),inout)- updated state variables c usvr (dp,ar(nuval,nintp),inout)- updated additional state variables c delcr (dp,sc,out) - equivalent creep strain increment (used c for creep ratio calculation) c c fortran parameters (to be defined by the user): c variable (type) description c nuval (int) - number of additional state variables per c integration point c nintp (int) - maximum number of integration points of c an element to be used with this routine c (14 is the maximum) c note: nuval x nintp = nstv(on nsvr command); cannot exceed 840! c c internal variables: c variable (type,size) description c con (dp,sc) - temporary variable c del (dp,ar(6)) - creep strain increments c epet (dp,sc) - equivalent elastic strain (before creep) c ept (dp,ar(6)) - total strain c eptot (dp,sc) - equivalent total strain, elastic + creep c sigen (dp,sc) - equivalent stress (before creep) c temabs (dp,sc) - temperature on the absolute scale

1.6.3 Subroutine usersw (Writing Your Own Swelling Laws)

subroutine usersw (option,elem,intpt,mat,proptb,ncomp,epswel, x epel,e,nuxy,fluen,dfluen,tem,dtem,toffst,timvll,timvnc,usvr) c c *** primary function: allow users to write their own swelling laws. c this logic is accessed with c72 = 10 c *** secondary function: demonstrate the use of user-written swelling laws c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c option (int,sc,in) swelling option c elem (int,sc,in) element number (label) c intpt (int,sc,in) element integration point number c mat (int,sc,in) material reference number c proptb (dp,ar(*),in) nonlinear material table (tb commands) c ncomp (int,sc,in) number of strain components (=1, 4, or 6) c 1 - truss or beam elements c 4 - 2-d solids and pipe elements c 6 - 3-d solids and most shells c epswel (dp,sc,inout) total accumulated swelling strain c before this substep c epel (dp,ar(ncomp),inout) elastic strain c e (dp,sc,in) elastic modulus c nuxy (dp,sc,in) poisson'S RATIO c fluen (dp,sc,in) total fluence (bf or bfe commands) c dfluen (dp,sc,in) increment in fluence for this substep c tem (dp,sc,in) temperature (bf or bfe commands) c dtem (dp,sc,in) increment in temperature for this substep c toffst (dp,sc,in) offset of temperature scale from absolute zero c (toffst command) c timvll (dp,sc,in) time at the end of this substep c timvnc (dp,sc,in) the increment of time this substep c usvr (dp,ar(*),inout) user-defined state variables(optional) c c output arguments: c epswel (dp,sc,inout) total accumulated swelling strain c after this substep c epel (dp,ar(ncomp),inout) elastic strain adjusted c for swelling increment c usvr (dp,ar(*),inout) updated user-defined state variables c

1.6.4 Subroutine uservp (Updating Nonlinear Strain History for Materials)

subroutine uservp (elem,mat,kfirst,kfsteq,g,prop,timval,timinc, x tem,toffst,delepn,svrn,svri,usvr,epeln,epeli,strn,stri,depdt, x cee,eta,norm) c c *** primary function: update the nonlinear strain history for a c user-defined material for the viscoxxx elements c *** secondary functions: compute the material tangent terms c accessed with tb,user and keyopt(1) = 1 c c *** Notice - This file contains ANSYS Confidential information *** c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (type,size,intent) description c c elem (int,sc,in) - element number (label) c mat (int,sc,in) - material reference number c kfirst (int,sc,in) - 1 if first time through, 0 otherwise c (useful for initializing state variables c to a non-zero value) c kfsteq (int,sc,in) - 1 if first equilibrium iteration of a c substep, 0 otherwise c c g (dp,sc,in) - shear modulus c prop (dp,ar(13),in) - linear material property array c (ex,ey,ez, gxy,gyz,gxz, nuxy,nuyz,nuxz, c alpx,alpy,alpz, dens) c c timval (dp,sc,in) - current time value c timinc (dp,sc,in) - time increment over this substep c c tem (dp,sc,in) - temperature at the end of this substep c toffst (dp,sc,in) - temperature offset from absolute zero c c delepn (dp,ar(3,3),in) - hencky strain increment over the substep c c svrn (dp,sc,in) - state variable from previous substep c svri (dp,sc,inout) - state variable from previous iteration c usvr (dp,ar(*),inout) - additional state variables from previous c iteration (saved if the nsvr command is c used) c c epeln (dp,ar(3,3),in) - elastic strain from previous substep c epeli (dp,ar(3,3),inout) - elastic strain from previous iteration c c strn (dp,ar(3,3),in) - stress from previous substep c stri (dp,ar(3,3),inout) - stress from previous iteration c c depdt (dp,sc,inout) - effective inelastic deformation rate c (d(deppl)/dt) from previous iteration c c output arguments: c variable (type,size,intent) description c

c c c c c c c c c c c c c c c c c c c c c c c

svri usvr epeli stri depdt cee eta norm

(dp,sc,inout) (dp,ar(*),inout) (dp,ar(3,3),inout) (dp,ar(3,3),inout) (dp,sc,inout) (dp,sc,out) (dp,sc,out) (dp,ar(3,3),out)

- updated state variable - updated additional state variables - updated elastic strain - updated stress - effective inelastic deformation rate - scalar linearization modulus (see below) - radial return factor (see below) - direction of plastic flow

fortran parameters (to be defined by the author): variable (type) description numinp (int) - number of data items in the user-defined data table (tbdat commands) nuval (int) - number of additional state variables per integration point nintp (int) - maximum number of integration points of an element to be used with this routine (8 is the maximum) note: nuval x nintp cannot exceed 840!

1.6.5 Subroutine userck (Checking User-Defined Material Data)

subroutine userck (curmat,ntb,tb) c *** primary function: check the user-defined material data, c input with the TB,user command. c *** secondary functions: none c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c curmat (int,sc,in) - current material number c ntb (int,sc,in) - dimension of tb c tb (dp,ar(ntb),in) - input table c c output arguments: c none c

1.6.6 Subroutine usrmooeny (Computing Derivatives of Strain Energy Density with Respect to Invariants)
subroutine usrmooney (mat,xnu,xiii,ndf,wi,wij,wini,pi,pij,prs,bm) c c *** primary function: c

c this is a user subroutine to compute the derivatives of the c strain energy density 'W' with respect to its invariants. c the user can modify this subroutine in order to incorporate c their own material law. c accessed with keyopt(7) = 1 and tb,mooney c c *** Notice - This file contains ANSYS Confidential information *** c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c *** comments: c c by definition, a material is said to be hyperelastic, if there c exists a strain energy density function, which is a scalar c function of one of the strain or deformation tensors, whose c derivative with respect to a strain component determines the c corresponding stress component. c c these strain energy density functions are functions of c invariants of strain tensors and a set of constants determined c in the laboratory tests. the user can specify their own form c of the strain energy density functions in terms of the invariants c of the cauchy-green tensors and a number of laboratory-determined c constants. c c in this subroutine the user is provided with (1) the invariants c 'XIII' of the cauchy-green strain tensor and (2) a flag 'NDF' c identifying the dimensionality of the problem. c c the subroutine first computes the distortion-only invariants c 'XJJJ'. the material constants that were specified by the user c through the 'TB' table are now put into local variables, e.g., c 'AI' and 'BI' in the mooney-rivlin example. the user then c defines the expression for the strain energy density function. c the term of 'W' involving 'BM' must remain intact; the c user will have their own expression for the first part of 'W', c e.g., terms involving constants 'AI' and 'BI' of the example of the c mooney-rivlin material. the expression for the distortion-based c pressure 'PRS' remains the same. 'PRS' along with 'BM' control c the degree of incompressibility of the material. c c the user will have to find the expressions for the derivatives c of 'W' (derivatives of 'PRS' remaining the same), i.e., 'WI' c and 'WIJ'. the subroutine returns the values of 'WINI', 'WI', c 'WIJ', 'PRS', 'PI', 'PIJ' and 'BM'. c c ---------------------------------------------------------------------c c input arguments: c xiii (dp,ar(3),in) - invariants of the cauchy-green tensor c mat (int,sc,in) - material reference number c xnu (dp,sc,in) - input possions ratio c ndf (int,sc,in) - dimensionality of the problem c (2 for plane strain) c (3 for 3d/axisymmetric) c

c output arguments: c wi (dp,ar(3),out) - first derivatives of strain energy density c with respect to invariants c wij (dp,ar(3,3),out) - second derivatives of strain energy density c with respect to invariants c wini (dp,sc,out) - strain energy density c c pi (dp,ar(3),out) - first derivatives of pressure c with respect to invariants c pij (dp,ar(3,3),out) - second derivatives of pressure c with respect to invariants c prs (dp,sc,out) - pressure c bm (dp,sc,out) - bulk modulus c c ---------------------------------------------------------------------c c definitions : c -------------------------------------------c wi(1) = d(w)/d(xiii(1)) c wi(2) = d(w)/d(xiii(2)) c wi(3) = d(w)/d(xiii(3)) c c wij(1,1) = d**2(w)/d(xiii(1))*d(xiii(1)) c wij(1,2) = d**2(w)/d(xiii(1))*d(xiii(2)) c wij(1,3) = d**2(w)/d(xiii(1))*d(xiii(3)) c c wij(2,2) = d**2(w)/d(xiii(2))*d(xiii(2)) c wij(2,3) = d**2(w)/d(xiii(2))*d(xiii(3)) c wij(3,3) = d**2(w)/d(xiii(3))*d(xiii(3)) c -------------------------------------------c pi(1) = d(prs)/d(xiii(1)) c pi(2) = d(prs)/d(xiii(2)) c pi(3) = d(prs)/d(xiii(3)) c c pij(1,1) = d**2(prs)/d(xiii(1))*d(xiii(1)) c pij(1,2) = d**2(prs)/d(xiii(1))*d(xiii(2)) c pij(1,3) = d**2(prs)/d(xiii(1))*d(xiii(3)) c c pij(2,2) = d**2(prs)/d(xiii(2))*d(xiii(2)) c pij(2,3) = d**2(prs)/d(xiii(2))*d(xiii(3)) c pij(3,3) = d**2(prs)/d(xiii(3))*d(xiii(3)) c -------------------------------------------c

1.6.7 Subroutine usermc (Controlling Hygrothermal Growth)

subroutine usermc (elem,time,msci,dmsci,mscr,mscra) c *** primary function: user control of hygrothermal growth c *** secondary functions: none c presently, called only by shell91 c c in order to activate this user programmable feature, c the user must enter the usrcal command.

c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c elem (int,sc,in) - element number c time (dp,sc,in) - time c msci (dp,sc,in) - moisture content c dmsci (dp,sc,in) - change of moisture content c mscr (dp,ar(3),in) - input material properties for growth c c output arguments: c mscra (dp,ar(3),out) - output material properties for growth c

1.6.8 Subroutine usermu (Supplying a Friction Coefficient for CONTAC48 and CONTAC49)
subroutine usermu (elem,type,mat,area,gap,loc,fn,nusvr,usvr, x time,timinc,vel,temp,stat,oldst,prevst,oldmu,prevmu,mu) c c *** primary function: user supplied coefficient of friction c called only by contac48,49 c accessed by keyopt(9) > 0 c c *** Notice - This file contains ANSYS Confidential information *** c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - element number c type (int,sc,in) - element type number c mat (int,sc,in) - material number c area (dp,sc,in) - element area (contac49) or length c (contac48) c gap (dp,sc,in) - present gap or penetration c loc (dp,ar(2),in) - location of the contact node projected c to the target surface - in dimenc sionless element coordinates c fn (dp,sc,in) - current contact force normal to the target c plane c nusvr (int,sc,in) - number of user-supplied element c variables c usvr (dp,ar(nusvr),in) - user-supplied element variables c time (dp,sc,in) - current time c timinc (dp,sc,in) - current substep time increment c vel (dp,ar(2),in) - contact node velocity components in c the target plane c temp (dp,sc,in) - average temperature c oldst (int,sc,in) - status of previous time step

c c c c c c c c c c c c c c c c c c

prevst stat oldmu prevmu

(int,sc,in) (int,sc,in) (dp,sc,in) (dp,sc,in)

(3,4: open; 2: closed and sliding; (1: closed and stuck) status of previous iteration (3,4: open; 2: closed and sliding; (1: closed and stuck) current status (3,4: open; 2: closed and stickslide status undetermined) friction coefficient used in the previous time step friction coefficient used in the previous iteration

output arguments: variable (typ,siz,intent) mu (dp,sc,out)

description - friction coefficient to be used this time step and iteration

1.6.9 Subroutine usrfc6 (Defining Custom Failure Criteria)

subroutine usrfc6 (elem,matlay,iott,keyer,tem,elim,slim,eps,sig, x fc6) c primary function: user subroutine for defining your own failure criterion c *** secondary functions: none c --- accessed with c6 = -1 c *** user programmable functions may not be used in parallel processing *** c this is currently only available with c shell99, the multilayer shell element c solid46, the multilayer solid element c shell91, the multilayer shell element c c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - element number c elim (dp,ar(9),in) - failure strains at the current temperature c (tbdata items 1-9) c slim (dp,ar(12),in) - failure stresses and coupling coefficients c at the current temperature c (tbdata items 10-21) c eps (dp,ar(6),in) - vector of strains c sig (dp,ar(6),in) - vector of stresses c tem (dp,sc,in) - temperature at this point in the model c matlay (int,sc,in) - material number c iott (int,sc,in) - unit number for writing c keyer (int,sc,in) - error flag (0 = ok, 1 = error) c c output arguments:

c variable (typ,siz,intent) c fc6 (dp,sc,out) solid46, c c #include "impcom.inc" external erhandler c

description - failure criterion to be processed by shell91, or shell99

external vmax,pplock,ppunlock integer elem,matlay,iott,keyer, loc double precision x eps(6),sig(6),fc6,tem,elim(9),slim(12), vmax,vect(9) c

1.6.10 Subroutines usrfc1 through usrfc5

The source code for subroutines usrfc1, usrfc2, usrfc3, usrfc4, and usrfc5 is identical to subroutine usrfc6 shown above.

1.6.11 Subroutine UserVisLaw (Defining Viscosity Laws)

subroutine UserVisLaw x (dudx,dudy,dudz, x dvdx,dvdy,dvdz, x dwdx,dwdy,dwdz, x u,v,w,x,y,z,kGeom, x Vis,Temp,Tref,Pres,Pref,Cf, x MFrac,DfNSpec,Time,VisNew) C C C C C C C C C C C C C C C C C C c c C Primary function: to provide a user defined viscosity relationship in terms of the following: pressure, temperature, position, time, velocity, & velocity-gradient In order to activate this subroutine the user must issue FLDA,PROT,VISC,USRV command. In addition the initial value of viscosity must be specified via FLDA,PROP,IVIS,value. This value is not available in this routine. Optionally the user may specify 4 additional coefficients which are available in this routine by the commands: FLDA,NOMI,VISC,value1 FLDA,COF1,VISC,value1 FLDA,COF2,VISC,value2 FLDA,COF3,VISC,value3 *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 *** ansys, inc.

C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C

input arguments: variable (typ,siz,intent) dudx (dp,sc,in) dudy (dp,sc,in) dudz (dp,sc,in) dvdx (dp,sc,in) dvdy (dp,sc,in) dvdz (dp,sc,in) dwdx (dp,sc,in) dwdy (dp,sc,in) dwdz (dp,sc,in) u (dp,sc,in) v (dp,sc,in) w (dp,sc,in) x (dp,sc,in) y (dp,sc,in) z (dp,sc,in) kGeom (int,sc,in) Vis (dp,sc,in) Temp (dp,sc,in) Tref (dp,sc,in) Pres (dp,sc,in) Pref (dp,sc,in) Cf (dp,ar(4),in) Mfrac (dp,ar(6),in) DfNSpec (int,sc,in) Time (dp,sc,in) output arguments: variable (typ,siz,intent) VisNew (dp,sc,out)

description velocity gradient component velocity gradient component velocity gradient component velocity gradient component velocity gradient component velocity gradient component velocity gradient component velocity gradient component velocity gradient component velocity component velocity component velocity component position component position component position component analysis type old viscosity temperature reference temperature pressure reference pressure input coefficients species mass fractions defined number of species time description new viscosity

1.6.12 Supporting Function egen

The function egen (kcomp,ep,nuxy) (function) combines kcomp strain components (ep) per:
function egen (kcomp,ep,posn) c primary function: combines strain components to give an "overall" strain c used in creep and plasticity calculations c secondary functions: none c c c c c c c formulation of overall value is by: ___________________________________________________________________ /1 2 2 2 1 2 2 2 \ / -*((ep - ep ) + (ep - ep ) + (ep - ep ) + -*(ep + ep + ep )) \/ 2 1 2 2 3 3 1 2 4 5 6 -----------------------------------------------------------------------(1 + posn)

c *** Notice - This file contains ANSYS Confidential information *** c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout

c c c c c c c c

input arguments: variable (typ,siz,intent) kcomp (int,sc,in) ep (dp,ar(6),in) posn (dp,sc,in) output arguments: egen (dp,func,out)

description - number of components of strain - the strain components - poisson's ratio - the combined strain value

1.7 Routines for Customizing Loads

This section describes the user routines you use to modify or monitor existing ANSYS elements. These routines enable you to perform tasks including:

Setting custom values for scalar fields (temperatures, heat fluxes, etc.) Changing element pressure information Changing information about element face convection surfaces Changing information about element face heat flux surfaces Changing information about element face convection surfaces Changing information about element face convection surfaces Changing information about element face charge density surfaces.

Activate these routines by issuing the USRCAL command or by choosing an equivalent menu path.

1.7.1 Subroutine usrefl (Changing Scalar Fields to UserDefined Values)

subroutine usrefl (key,iel,ielc,nnod,nodes,time,defalt,nd,dat) c *** primary function: change the scalar fields (temperatures, fluences, c heat generation, etc.) to what user desires. c *** secondary functions: none c c in order to activate this user programmable feature, c the user must enter the usrcal command. c c this routine is called at each substep of each load step c for which element or nodal temperatures(etc) are used. c it is called for each equilibrium iteration. c the call to get the standard ansys input element or nodal values c is made just before entering this routine. c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c *** Notice - This file contains ANSYS Confidential information *** c

c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c key (int,sc,in) - type of data desired c = 1 temperatures c = 2 fluences c = 3 heat generation rates c = 4 moisture contents c = 5 magnetic virtual displacements c iel (int,sc,in) - element number c ielc (int,ar(IELCSZ),in) - array of element type characteristics c nnod (int,sc,in) - number of nodes c nodes (int,ar(nnod),in) - list of nodes c time (dp,sc,in) - time of current substep c defalt (dp,sc,in) - default value (e.g. tunif) c nd (int,sc,in) - size of dat array c dat (dp,ar(nd),inout) - array of data as normally computed by element c as selected by key c c output arguments: c variable (typ,siz,intent) description c dat (dp,ar(nd),inout) - array of data passed back to element c this data represents values at the end c of the load step c c the input argument dat may be used in one of three ways: c 1. it may be simply passed thru c 2. it may be used as a flag(e.g. if dat(1) = -3.0, use c a certain set of logic) c 3. it may be completely ignored and instead defined with new logic c

1.7.2 Subroutine userpr (Changing Element Pressure Information)

subroutine userpr (ielc,ielem,time,ndat,dat) c *** primary function: change element pressure information. c c in order to activate this user programmable feature, c the user must enter the usrcal command. c c this routine is called at each substep of each load step for which c pressures are used. it is called for each equilibrium iteration. c it is called once per element. c the call to get the standard ansys input pressures is made just before c entering this routine. c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c

c input arguments: c variable (typ,siz,intent) description c ielc (int,ar(IELCSZ),in) - array of element type characteristics c ielem (int,sc,in) - element number for operation. c time (dp,sc,in) - time of current substep c ndat (int,sc,in) - number of pressure items for this element c dat (dp,ar(ndat),inout) - the element pressure vector c (has input values for each corner c of each face) c c output arguments: c variable (typ,siz,intent) description c dat (dp,ar(ndat),inout) - the element pressure vector c (defines input values for each corner c of each face) c c the input array dat may be used in one of three ways: c 1. it may be simply passed thru c 2. it may be used as a flag(e.g. if dat(1) = -3.0, use c a certain set of logic) c 3. it may be completely ignored and instead defined with new logic

1.7.3 Subroutine usercv (Changing Element Face Convection Surface Information)

subroutine usercv (elem,ielc,time,nr,u, ndat,hc,tb) c *** primary function: change element face convection surface info c c in order to activate this user programmable feature, c the user must enter the 'usrcal,usercv' command. c c *** ansys(r) copyright(c) 1998 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c elem (int,sc,in) - element number for operation. c ielc (int,ar(IELCSZ),in) - array of element type characteristics c time (dp,sc,in) - time of current substep c nr (int,sc,in) - number of nodal temperatures c of the element c u (dp,ar(nr),in) - vector of most recent values of the c temperatures c ndat (int,sc,in) - number of data points per element c hc (dp,ar(ndat),inout) - film coefficients c (has input values for each corner c of each face) c tb (dp,ar(ndat),inout) - bulk temperature c (has input values for each corner c of each face) c c output arguments:

c c c c c c c c c c c c c c c c c c c c c c c c the c c c

variable (typ,siz,intent) description hc (dp,ar(ndat),inout) - film coefficients (defines input values for each corner of each face) tb (dp,ar(ndat),inout) - bulk temperature (defines input values for each corner of each face) the input arguments hc and tb may be used in one of three ways: 1. they may be simply passed thru. 2. they may be used as a flag(e.g. if hc(2) = -3.0, use a certain set of logic). 3. they may be completely ignored. and instead redefined with new logic this routine is called during each substep of each load step. it is called for each equilibrium iteration. it is called once per element. it is called only during the heat flow load vector formulation stage, and not during the heat flow evaluation stage. the call to get the standard ansys input convection surfaces is made just before entering this routine, so this information is available to be modified, if desired. velocity-dependent film coefficients can be computed by inputting velocity as the input film coefficient or bulk temperature or by inputting the velocity as a function of location in space. routine could then compute the effective film coefficient. this

1.7.4 Subroutine userfx (Changing Element Face Heat Flux Surface Information)
subroutine userfx (ielc,ielem,time,nr,u, ndat,dat) c *** primary function: change element face heat flux surface info c c in order to activate this user programmable feature, c the user must enter the usrcal command. c c this routine is called during each substep of each load step. c it is called for each equilibrium iteration. c it is called once per element. it is called only during the heat c flow load vector formulation stage, and not during the heat flow c evaluation stage. c the call to get the standard ansys input heat flux surfaces c is made just before entering this routine, so this information is c available to be modified, if desired. c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (typ,siz,intent) description c ielc (int,ar(IELCSZ),in) - array of element type characteristics

c c c c c c c c c c c c c c c c c c c c c c c

ielem time nr u ndat dat

- element number for operation. - time of current substep - number of nodal temperatures of the element (dp,ar(nr),in) - vector of most recent values of the temperatures (int,sc,in) - number of data points per element (dp,ar(ndat),inout) - fluxes (has input values for each corner of each face)

(int,sc,in) (dp,sc,in) (int,sc,in)

output arguments: variable (typ,siz,intent) description dat (dp,ar(ndat),inout) - fluxes (defines input values for each corner of each face) the input argument dat may be used in one of three ways: 1. they may be simply passed thru. 2. they may be used as a flag(e.g. if dat(2) = -3.0, use a certain set of logic). 3. they may be completely ignored. and instead redefined with new logic

1.7.5 Subroutine userch (Changing Element Face Charge Density Surface Information)
subroutine userch (ielc,ielem,time,nr,u, ndat,dat) c *** primary function: change element face charge density surface info c c in order to activate this user programmable feature, c the user must enter the usrcal command. c c this routine is called during each substep of each load step. c it is called once per element. it is called only during the heat c flow load vector formulation stage, and not during the heat flow c evaluation stage. c the call to get the standard ansys input charge densities of surfaces c is made just before entering this routine, so this information is c available to be modified, if desired. c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c input arguments: c variable (typ,siz,intent) description c ielc (int,ar(IELCSZ),in) - array of element type characteristics c ielem (int,sc,in) - element number for operation. c time (dp,sc,in) - time of current substep c nr (int,sc,in) - number of nodal temperatures c of the element c u (dp,ar(nr),in) - vector of most recent values of the

c c c c c c c c c c c c c c

ndat dat

temperatures (int,sc,in) - number of data points per element (dp,ar(ndat),inout) - fluxes

output arguments: variable (typ,siz,intent) description dat (dp,ar(ndat),inout) - fluxes the input argument dat may be used in one of three ways: 1. they may be simply passed thru. 2. they may be used as a flag(e.g. if dat(2) = -3.0, use a certain set of logic). 3. they may be completely ignored. and instead redefined with new logic

1.7.6 Subroutine useracel (Modifying an Element's Acceleration/Angular Velocity Information)

subroutine useracel (key,elem,vect) c primary function: c c c c keywords: Modify the acceleration/angular velocity information for an element. Currently(5.3), only works with angular velocities used by spin softening.

user subroutine to return acceleration field

c object/library: c current - acel c *** Notice - This file contains ANSYS Confidential information *** c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c c c c c c c c c vc c c input arguments: key (int,sc,in) - Key 0 1 2 3 4 5 to desired acceleration components Translational acceleration Angular Velocity Angular Acceleration Angular Velocity around C.G. Angular Acceleration arount C.G. C.G. Location in global coord. system

NOTE: elem

Only key=1 for spin softening is currently implemented (int,sc,in) - Element number - Return value as given by key

output arguments: vect (dp,ar(3),out)

1.7.7 Examples of Using the Load Customization Routines

This section presents examples of how two of the load customization UPFs, the usercv subroutine and the useracel subroutine, can affect analysis results. 1.7.7.1 Example of Using usercv Below is an example of using the usercv subroutine. First, here is the ANSYS input for the example:
/bat,list /prep7 /title,test of velocity-dependent film coefficient /com default logic (triggered with usrcal,usercv) will set the /com film coefficient in SURF19 to the velocity in FLUID66 et,1,66 keyo,1,9,1 et,2,19,1,, ,1,1 keyo,2,8,2 et,3,55 r,1,1,1,,1 mp,dens,,1 mp,kxx,,11 mp,mu,,0.01 n,1 n,4,100 fill n,11, ,5 n,12,30,5 n,13,30,8 n,14, ,8 type,1 e,1,2 egen,3,1,-1 type,2 e,11,12,2 type,3 e,11,12,13,14 elist,all noor fini ! flag to save values ! temp DOF, no midside nodes, has extra node ! flag for convection surface

Now, look at the results from this example analysis (shown below). Note that the FLOW RATE magnitude for FLUID66 matches the HFILM for SURF19.
***** uelEMENT SOLUTION ***** LOAD STEP= 2 SUBSTEP = 1 TIME = CUM. ITER.= 2.0000 3

uel= 1 NODES= 1 2 0.00000E+00 FLOW RATE= -4.4721 HT COND RATE= 128.60 0.00000E+00 HT TRANSP RATE= 0.00000E+00 FRICTION FACTOR= 0.100000E-01 0.00000E+00 HFILM= 0.00000E+00 uel= 2 NODES= 2 3 0.00000E+00 FLOW RATE= -4.4721 HT COND RATE= -80.801 0.00000E+00 HT TRANSP RATE= 0.00000E+00 FRICTION FACTOR= 0.100000E-01 0.00000E+00 HFILM= 0.00000E+00 uel= 3 NODES= 3 4 0.00000E+00 FLOW RATE= -4.4721 HT COND RATE= -80.801 0.00000E+00 HT TRANSP RATE= 0.00000E+00 FRICTION FACTOR= 0.100000E-01 0.00000E+00 HFILM= 0.00000E+00

0 FLUID66

0 MAT=

1 PUMP HD= 0.00000E+00 0.00000E+00 0.00000E+00

RE=

HT CONV RATES(I,J)= CONV AREAS(I,J)=

0 FLUID66

0 MAT=

1 PUMP HD= 0.00000E+00

RE=

HT CONV RATES(I,J)= CONV AREAS(I,J)=

0.00000E+00 0.00000E+00

0 FLUID66

0 MAT=

1 PUMP HD= 0.00000E+00 0.00000E+00

RE=

HT CONV RATES(I,J)= CONV AREAS(I,J)=

0.00000E+00

*** WARNING *** CP= 0.800 TIME= 17:23:49 ANSYS,INC.-supplied version of coding for USERCV has been used. Using getvel66, the velocity from node 2 is 0.4472E+01 uel= 30.000 VOLUME= TAVG= 6.9801 4 SURFACE NODES= 11 12 EXTRA NODE= 2 MAT= 1 AREA=

30.000 8.7381

SURF19 HFILM= 4.4721 TBULK= 10.299 CONV. HEAT RATE= -209.40

4.4721 10.299 CONV. HEAT RATE/AREA= -

*** WARNING *** CP= 0.800 TIME= 17:23:49 ANSYS,INC.-supplied version of coding for USERCV has been used. uel= 5 NODES= -0.2753 11 13 14 MAT= PLANE55 -0.6346 0.6917 12 1 VOL= 90.00 3.029

TG(X,Y,SUM)= 6.980 7.609 1

TF(X,Y,SUM)=

1.7.7.2 Effect of useracel on Results

This example uses a two-dimensional model based on element PLANE42:

c c c c c Two Dimensional model (plane42) component rotation FAST OMGFAST SLOW OMGSLOW external cmpcheck, parevl logical cmpcheck double precision parevl integer kerr double precision subc(3) data subc / 3*0.0d0 / c c if (key .eq. 1) then --- see if element is a member of component FAST: if (cmpcheck('FAST ',elem)) then vect(2) = parevl('OMGFAST ',0,subc(1),1,kerr) --- see if element is a member of component SLOW: elseif (cmpcheck('SLOW ',elem)) then vect(2) = parevl('OMGSLOW ',0,subc(1),1,kerr) endif endif return end

The ANSYS input used for this model is as follows:

/bat,list /title, demonstrate /prep7 et,1,42 /nopr mp,ex,,12e6 mp,dens,,.004 n,1 n,11,50 fill ngen,6,20,1,11,1,,3 e,1,2,22,21 egen,10,1,1 egen,3,40,all d,1,all,,,101,20 d,all,uy esel,s,elem,,11,20 cm,FAST,elem eall esel,s,elem,,21,30 cm,SLOW,elem eall omega,0,0,0,1 OMGFAST = 1000 OMGSLOW = 500 ! note that first 10 finish /solu use of useracel

! define FAST component ! define SLOW component ! turn on spin softening ! define parameter for use in useracel ! define parameter for use in useracel elements will be stationary

/title, /title,

BASE CASE - STANDARD VERSION DEMONSTRATE USE OF useracuel(useracuel HAS BEEN LINKED IN AND USED)

Suppose that you ran this example problem using the standard ANSYS product. You'd receive the following output information:
*** FREQUENCIES FROM BLOCK LANCZOS ITERATION *** MODE FREQUENCY (HERTZ) 1 287.3798632612 2 287.3798632612 3 287.3798632612 4 869.2433944257 5 869.2433944257 6 869.2433944259

Now, suppose that you re-run the problem after linking in and using the useracel subroutine. Your new results should look like those shown below:
*** FREQUENCIES FROM BLOCK LANCZOS ITERATION *** MODE FREQUENCY (HERTZ) 1 239.2841196099 2 276.1423760135 3 287.3798632612 4 854.5488768013 5 865.5931519918 6 869.2433944258

1.8 Running ANSYS as a Subroutine

To call the ANSYS program, use the following:
program ansys

If you're running ANSYS on a UNIX system (but not under Windows 95 or Windows NT), you also can call the ANSYS program as a subroutine in a program you've written. To do so, use the following:
subroutine ansys

For multiple calls to subroutine ANSYS, you must open and close standard input in the calling routine. (Usually, input and output are FORTRAN units 5 and 6, respectively.) The calling routine can't use the database access routines. But, other user programmable features can use the database access routines freely.

1.9 Defining Your Own Commands

ANSYS, Inc. supplies a set of user routines, named user01 through user10, which you can use to define custom ANSYS commands. To do so, follow these steps: 1. Insert the code for the functions you want to perform into routine user01 (or user02, etc.).

2. Link the routine into the ANSYS program. 3. Issue the ANSYS command /UCMD to define a name for a custom command that calls and executes your routine. Use the command format shown below:
/UCMD,Cmd,SRNUM

Cmd is the name for your new command. It can contain any number of characters, but only the first four are significant. The name you specify can't conflict with the name of any ANSYS command or the names of any other commands or macros. SRNUM is the number of the routine your command should call; that is, a value between 01 and 10. For example, suppose that you create and link in a user routine for a parabolic distribution of pressure, and you name that routine user01. Issuing the command shown below creates a new command, PARB, that when issued calls your parabolic pressure distribution routine:
/UCMD,PARB,1

To make these "custom command" routines available in all your ANSYS sessions, include the /UCMD commands in your start-up file (START.ANS). You also can use /UCMD to remove a custom command. To do so, simply use a blank value for Cmd, as shown below:
/UCMD,,1

This command removes the PARB command. To list all user-defined command names, issue the command /UCMD,STAT.

1.9.1 Function user01

*deck,user01 user parallel ANSYS,INC c sid 5.1 copy of file s.user01.F last changed by jtm on 98/03/09 function user01 (intin,dpin,ch4in,ch8in) c *** primary function: user routine number 01 c c *** ansys(r) copyright(c) 1998 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c /*******************************************************************\ c | this is a user routine that may be used by users to include their | c | special coding. accesss to this routine is by the command usr1. | c | usr1 may be changed by the user using the command /ucmd. the | c | user may then use this routine to call his/her special routines. | c | ansys routines to access information in the ansys database may be | c | found in the "ansys programmer's manual", available from ansys,inc| c | see user02 for a simple example usage. | C | routines user03 to user10 are also available. |

C c c c c c c c c c c c c c c c c c c c c c c c c c

\*******************************************************************/ input arguments: variable (typ,siz,intent) intin (int,ar(12),in) dpin (dp,ar(12),in) ch4in (ch*4,ar(12),in) ch8in (ch*8,ar(12),in) description integers from command line double precision from command line upper case 4 characters from command as input 8 characters from command

If the usage had been set up with a /ucmd,aaaa,1 , user01 could be activated by the command aaaa,ln,6,all where ln = lab4in(2) 6 = intin(3) all = lab4in(4) output arguments: none user01 (int,sc,out) - result code (should be zero) (which is ignored for now)

************************************************************** intin(2), dpin(2), ch4in(2), and ch8in(2) are all representations of the contents of the second field on the command line (the field after the command label) ************************************************************** external integer wrinqr wrinqr

integer user01,intin(12), iott double precision dpin(12) character*4 ch4in(12) character*8 ch8in(12) c iott = wrinqr(2) c c c c c c c ***** user'S CODE IS INSERTED HERE ***** write (iott,2000) 2000 format (//' ***** CALL TO ANSYS,INC DUMMY USER01

*****'//)

***** do not return this result code in a real user routine user01 = -654321 ***** instead return a zero ***** user01 = 0 return end

1.9.2 Function user02 (Demonstrates Offsetting Selected Nodes)

*deck,user02 user ANSYS,INC c sid 5.1 copy of file s.user02.F last changed by jtm on 98/03/09 function user02 (intin,dpin,ch4in,ch8in)

c *** primary function: user routine number 02 c --- This demonstration offsets selected nodes with the command: c usr2,dx,dy,dz c *** ansys(r) copyright(c) 1998 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c c c c c c c c c /*******************************************************************\ | see user01 for additional information on user routines | \*******************************************************************/ input arguments: variable (typ,siz,intent) intin (int,ar(12),in) dpin (dp,ar(12),in) ch4in (ch*4,ar(12),in) ch8in (ch*8,ar(12),in) output arguments: none user02 (int,sc,out) description integers from command line double precision from command line upper case 4 characters from command as input 8 characters from command

- result code (should be zero) (which is ignored for now)

************************************************************** intin(2), dpin(2), ch4in(2), and ch8in are all representations of the contents of the second field on the command line (the field after the command label) ************************************************************** external integer wrinqr,ndinqr,ndgxyz,ndpxyz,erhandler wrinqr,ndinqr,ndgxyz

integer user02,intin(12), iott, maxnp,i,ksel double precision dpin(12),xyz(3) character*4 ch4in(12) character*8 ch8in(12) #include "ansysdef.inc" maxnp = ndinqr (0,DB_MAXDEFINED) do i = 1,maxnp ksel = ndgxyz (i,xyz(1)) if (ksel .eq. 1) then xyz(1) = xyz(1) + dpin(2) xyz(2) = xyz(2) + dpin(3) xyz(3) = xyz(3) + dpin(4) call ndpxyz (i,xyz(1)) endif enddo c ***** write to output file iott = wrinqr(WR_OUTPUT) write (iott,2000) 2000 format (/' NODE OFFSET COMPLETE ***** write to GUI window ***** '/) *****

x c

call erhandler ('user02',3000, 2,'NODE OFFSET COMPLETE',0.0d0,' ') ***** user02 = 0 return end required return value *****

1.9.3 Function user03 (Demonstrates Using Heap Memory)

*deck,user03 user ANSYS,INC c sid 5.1 copy of file s.user03.F last changed by jtm on 98/03/09 function user03 (intin,dpin,ch4in,ch8in) c *** primary function: user routine number 03. Gives example of c Heap usage c c *** ansys(r) copyright(c) 1998 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c /*******************************************************************\ c | see user01 for additional information on user routines | c \*******************************************************************/ c c input arguments: c variable (typ,siz,intent) description c intin (int,ar(12),in) - integers from command line c dpin (dp,ar(12),in) - double precision from command line c ch4in (ch*4,ar(12),in) - upper case 4 characters from command c ch8in (ch*8,ar(12),in) - as input 8 characters from command c c output arguments: none c user03 (int,sc,out) - result code (should be zero) c (which is ignored for now) c c ************************************************************** c intin(2), dpin(2), ch4in(2), and ch8in are all representations c of the contents of the second field on the command line c (the field after the command label) c ************************************************************** c #include "impcom.inc" external wrinqr,ndinqr,ndgxyz,ndnext,HeapAllocPtr, x HeapDealloc,erhandler integer wrinqr,ndinqr,ndgxyz,ndnext,HeapAllocPtr user03, intin(12), iott, i, ksel, xcptr, hXcptr, ycptr, hYcptr, zcptr, hZcptr, numnp, node double precision dpin(12),xyz(3), xmean, ymean, zmean, stdxyz(3), x sodx, sody, sodz x integer

character*4 character*8

ch4in(12) ch8in(12)

#include "ansysdef.inc" #include "stack.inc" c c c c c Get nodal xyz locations and calculate standard deviation of x coordinates, y coordinates, & z coordinates get number of currently selected nodes numnp = ndinqr (0,DB_NUMSELECTED) if (numnp .le. 0) go to 999 c allocate memory for x, y, & z coordinates of nodes xcptr = HeapAllocPtr(numnp,'XCoords ',HEAP_DOUBLE,hXcptr) ycptr = HeapAllocPtr(numnp,'YCoords ',HEAP_DOUBLE,hYcptr) zcptr = HeapAllocPtr(numnp,'ZCoords ',HEAP_DOUBLE,hZcptr) loop through all selected nodes i=1 node = 0 xmean = 0.0d0 ymean = 0.0d0 zmean = 0.0d0 node = ndnext(node) if (node .gt. 0) then c c get xyz coordinates ksel = ndgxyz (node,xyz(1)) store this node's dstack(xcptr + i) dstack(ycptr + i) dstack(zcptr + i) while xmean ymean zmean xyz coordinates = xyz(1) = xyz(2) = xyz(3)

10

we're looping, accumulate sums to calculate means = xmean + xyz(1) = ymean + xyz(2) = zmean + xyz(3)

c c

increment index i = i + 1 loop back up for next selected node goto 10 endif

c c

node = 0, at the end of node list calculate mean of xyz coordinates xmean = xmean / numnp ymean = ymean / numnp

zmean = zmean / numnp c calculate standard deviation for xyz coordinates sodx = 0 sody = 0 sodz = 0 do i=1, numnp sodx = sodx + (dstack(xcptr+i) - xmean)**2 sody = sody + (dstack(ycptr+i) - ymean)**2 sodz = sodz + (dstack(zcptr+i) - zmean)**2 enddo stdxyz(1) = sqrt(sodx / (numnp-1)) stdxyz(2) = sqrt(sody / (numnp-1)) stdxyz(3) = sqrt(sodz / (numnp-1)) c ***** write to output file ***** iott = wrinqr(WR_OUTPUT) write (iott,2000) stdxyz(1),stdxyz(2),stdxyz(3) 2000 format (/' STD FOR X COORDINATES:',G12.5,/, X ' STD FOR Y COORDINATES:',G12.5,/, X ' STD FOR Z COORDINATES:',G12.5) ***** write to GUI window ***** call erhandler ('user03',5000,2, x 'STD FOR X COORDINATES: %G %/ STD FOR Y x COORDINATES: %G %/ STD FOR Z COORDINATES: %G',stdxyz(1),' ') release dynamically allocated memory call HeapDealloc(hXcptr) call HeapDealloc(hYcptr) call HeapDealloc(hZcptr) 999 ***** required return value user03 = 0 return end *****

1.9.4 Functions user04 through user10

The source code for user routines user04, user05, user06, user07, user08, user09, and user10 is identical to function user01 shown above.

1.10 Supporting Subroutines

1.10.1 Function GetRForce (Getting Nodal Reaction Force values)
function user03 (intin,dpin,ch4in,ch8in)

c *** primary function: user routine number 03. Gives example of c Heap usage c c *** ansys(r) copyright(c) 1971,78,82,83,85,87,89,92,94-97 c *** ansys, inc. c *** Notice - This file contains ANSYS Confidential information *** c c /*******************************************************************\ c | see user01 for additional information on user routines | c \*******************************************************************/ c c input arguments: c variable (typ,siz,intent) description c intin (int,ar(12),in) - integers from command line c dpin (dp,ar(12),in) - double precision from command line c ch4in (ch*4,ar(12),in) - upper case 4 characters from command c ch8in (ch*8,ar(12),in) - as input 8 characters from command c c output arguments: none c user03 (int,sc,out) - result code (should be zero) c (which is ignored for now) c c ************************************************************** c intin(2), dpin(2), ch4in(2), and ch8in are all representations c of the contents of the second field on the command line c (the field after the command label) c ************************************************************** c

1.10.2 Function GetStackDisp (Getting Current Displacement Values)

function GetStackDisp (Node,Label,Value) c primary function: Get the displacement at a node from the disp vector c secondary functions: Return pointer for fast access c object/library: usr

c *** Notice - This file contains ANSYS Confidential information *** c Prolog is not CONFIDENTIAL INFORMATION c c c c c c c c c c typ=int,dp,log,chr,dcp input arguments: variable (typ,siz,intent) Node (int,sc,in) Label (ch*4,sc,in) output arguments: variable (typ,siz,intent) GetStackDisp (int,sc,out) siz=sc,ar(n) intent=in,out,inout

description - Node Number (User) - DOF Label (Upper Case) 'UX ','UY ','TEMP','VOLT','ROTY', etc description - status/pointer = 0 - data not valid

c access c c Value c

> 0 - dstack pointer to data for fast (dp,sc,out) see comments below - Solution value for Node,Label

example usage:

c external GetStackDisp c#include "stack.inc" (only if dstack(ptr) form is used) c integer GetStackDisp, ptr, Node2 c double precision Value c c c ptr = GetStackDisp (Node2,'UY ',Value)

later... Value = dstack(ptr)

1.10.3 Subroutine ElResultStrt (Getting Load Data from Analysis Results)

subroutine ElResultStrt (Label,Comp,LabAvg,TypeData,nVal,Loc) c *** primary function: (post1) Load data for later ElResultGet c *** Notice - This file contains ANSYS Confidential information *** c (prolog is not confidential) c c c c c c c c c c input arguments: variable (typ,siz,intent) Label (ch*4,sc,in) Comp (ch*4,sc,in) LabAvg (ch*4,sc,in) output arguments: TypeData (int,sc,out) nVal (int,sc,out) Loc (int,sc,out) description - Result Type - Result Component (8 char for ESTR) - 'AVG ' or 'NOAV' ('AVG ' default) - Code for data type - Number of values per point If 0, no data - Location of Comp in values

1.10.4 Subroutine ElResultGet (Getting Results Values at Selected Points)

subroutine ElResultGet (nPoints,ebest,elcord,TypeData,Loc, nVal,result) c *** primary function: (post1) Get results at selected points x c *** Notice - This file contains ANSYS Confidential information *** c (prolog is not confidential) c c input arguments: variable (typ,siz,intent) description

c c c c c c c c c c

nPoints

(int,sc,in) - Number of evaluation points *** from ElInterp *** ebest (int,ar(nPoints),in) - Element(s) containing points elcord (dp,ar(3,nPoints),in) - Element coordinates *** from ElResultStrt *** TypeData (int,sc,in) - Data type code Loc (int,sc,in) - Start of selected data nVal (int,sc,in) - Number of results per point output arguments: Result (dp,ar(nvar,nPoints),out) - Array of results

1.10.5 Subroutine ElInterp (Finding Element Coordinates)

x subroutine ElInterp (nPoints,xyzPoints,tolInside,tolOutside, ebest,elcord)

c primary function: Find element numbers containing xyz points c secondary functions: Find element coordinates of these points c object/library: upf c *** Notice - This file contains ANSYS Confidential information *** c (Prolog is not CONFIDENTIAL INFORMATION) c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout

c input arguments: c variable (typ,siz,intent) c nPoints (int,sc,in) group) c xyzPoints(dp,ar(3,nPoints),in)c tolInside(dp,sc,in) c c tolOutside(dp,sc,in) c 0.25) c output arguments: c variable (typ,siz,intent) c ebest (int,ar(nPoints),out) c elcord (dp,ar(3,nPoints),out)

description Number of points to find (do in one XYZ coordinates of each point Tolerance for point inside element (0.0d0 defaults to 1.0d-4) Maximum distance outside to be associated with an element (0.0d0 defaults to description - Best element number for each point - Element coordinates of the point

1.11 Access at the Beginning and End of Various Operations

You can access the logic just before an ANSYS run begins or just after a run ends, and at many other intermediate points, by using the ANSYS routines listed below. These routines can perform actions such as evaluating results or performing calculations. (None of the routines have input or output arguments.)

Issue the USRCAL command (or use an equivalent menu path) to activate or deactivate these routines.
User Subroutine Is Called UAnBeg At ANSYS start-up USolBeg Before solution ULdBeg Before a load step USsBeg Before a substep UltBeg Before an iteration UltFin After an iteration USsFin After a substep ULdFin After a load step USolFin After solution UAnFin At the end of an ANSYS run

Subroutines USSBeg, UItBeg, UItFin and USSFin default to reading a command macro file from the current working directory whose name is subroutine.mac (i.e., ussfin.mac is read by USSFin.F). No user action to relink the ANSYS program is required for the command macro to be read except that the calling subroutine must be activated by the USRCAL command. The design of the command reading ability of these subroutines is limited to APDL parameter setting commands (*GET, *SET, a=value, etc) and testing for general ANSYS commands is limited. Commands which are known to work include *DIM,*STAT. Commands which require another line (*MSG, *VWRITE) are not allowed. Other commands which are known to not work are the solution loading commands (D, F, SFE, etc). If these capabilties are required, the user will need to create a Fortran subroutine and link this subroutine into ANSYS, as described in Section 1.1. While parameter substitution into commands is not permitted, USSBeg, etc. were designed to be used in conjunction with dynamic tables and parameter substitution from the user subroutine. As an example, consider an ANSYS table defined as d5 = f(par1), If d5 contains values of displacement as a function of PAR1, then d5 may be used as a constraint, as
*dim,d5,table,10,1,1,PAR1 d5(1)=0,.1,.25, /solu d,5,ux,%d5%

Modify the value of PAR1 in USSBeg.MAC and the constraint on node 5, ux can then be modified in the middle of a load step.

The following is an example of a valid input that may be read by USSBeg, UItBeg, UItFin and USSFin.
/COM, SAMPLE ussfin.mac a=5 b=nx(1) *get,c,active,solu,Time,cpu *dim,array,,6 array(1) = 1 array(2) = 2 array(3) = 3 array(4) = 4 array(5) = 5 array(6) = 6 *vleng,3 *vfun,array(4),copy,array(1) *stat *stat,array(1) array(1)= nnode = ndinqr(0,14) *dim,array,,nnode *vget,array(1),NODE,1,NSEL *stat,array(1) array(1)= /eof

! *get function is ok ! *get is ok ! array parameters are ok

! vector operations are ok

/COM, COMMANDS BELOW THIS LINE ARE KNOWN TO NOT WORK p,1,6,2000 d,1,uy,.1 *msg,note THIS IS A TEST MESSAGE *vwrite,array(1) (/ b = ,f10.4) ! commands DO NOT work

1.12 Creating Your Own Optimization Routine

One way to customize the ANSYS program for your site's specific needs is to substitute your own external optimization for the standard ANSYS optimization algorithm. You can do so using either of these methods:

Link a user routine within the ANSYS program. Run a stand-alone optimization program using the ANSYS optimization "save" file, Jobname.OPT.

This chapter describes both methods. You can find additional information on design optimization techniques and procedures in Chapter 1 of the ANSYS Advanced Analysis Techniques Guide.

1.12.1 Linking a Custom Optimization Program to ANSYS

If you're performing iterative analyses for design optimization, and you wish to use the standard ANSYS optimizer, you simply choose GUI path Main Menu>Design Opt or issue the ANSYS command/OPT. (You must do either from the Begin level.) However, if you plan to use an optimization routine of your own design, you must do the following: 1. Define parameters for your external optimization routine, using either the OPUSER command or GUI path Main Menu>Design Opt>Method/Tool. (For more information about OPUSER, see theANSYS Commands Reference.) 2. Either issue the OPTYPE,USER command or choose menu path Main Menu>Design Opt>Method/Tool and specify user-defined optimization logic. 3. Issue the OPEXE command or choose GUI path Main Menu>Design Opt>Run. Issuing the OPTYPE,USER command (or its GUI equivalent) instructs the ANSYS program to bypass the standard ANSYS design optimizer logic and instead execute your custom optimization routine. Below is a simplified flow chart showing how a user-defined optimization routine interacts with the ANSYS program:

For information about the kopusr variable and the userop routine, see the next section.

1.12.2 Subroutine userop (Defining a Custom Optimization Routine)

Instead of creating your custom design optimization routine from scratch, you can simply tailor the userop routine available on your ANSYS distribution medium. Defined within this routine are a number of variables to which you assign the values you want. Below is a listing of the inputs and outputs for userop and the comments of common block cmopt:
c c c c c c c subroutine userop (iott,nterm) primary purpose: user-supplied optimization routine accessed with optype,user other input comes from the opuser command *** ansys copyright 1971,78,82,83,85,87,89,92,94,95,96 *** ansys, inc.

c**************************************************************************** **** c incoming arguments: * c * c * c Iott is the ansys output unit * c * c Nterm is passed back to routine opterm. You should set this variable * c as follows: * c nterm = 0 if optimization looping should stop * c nterm - 1 if optimization looping should continue * c * c**************************************************************************** **** c contents of optimization common block cmopt: c (only variables that are of interest or useful are documented) c c cmopt variable descriptions c --------------------------c c dopspr - (not used - spares) c dv(I) - design vector, I = 1 to maxdv c iopspr - (not used - spares) c kconv - convergence flag (kconv = 0 indicates convergence) c kopusr - flag choosing optimizer (0 = ansys, 1 = user-supplied opt) c (issuing the command optype,user sets kopusr = 1) c lopt - incrementing lopt iteration counter c maxdv - maximum number of design variables allowed in ansys c optimization (= 60) c maxpar - maximum number of parameters allows in ansys c optimization (= 400) c maxopt - maximum number of design sets allowed in optimization c (= 130) c nobj - flag for objective function defined (0=no, 1=yes) c ntdv - total number of design variables c ntsv - total number of state variables c numopt - total number of design sets c nsets - highest design set number c setnum(1) - the actual design set number of the ith design set c oplim(j,I) - Lower limit (j=1) and upper limit (j=2) of ith c optimization parameter (opvar). (I = 1 to 160.) c The state variables are stored first, then the c design variables, and then the objective function c optol(I) - tolerance of ith optimization parameter (opvar) c (I = 1 to 160.) c The state variables are stored first, then the

c c c c c c C c c c c

design variables. optvar(I,j) - The ith parameter of the jth design set (I = 1 to maxpar, j = 1 to maxopt) The state variables are stored first, then the design variables, then the objective function, and finally all remaining scalar parameter values. maxusr - user-specified, maximum allowed optimization iterations on the opuser command opus(I) - user-specified, double precision parameters on the opuser command (I = 1,8) objtol - objective function tolerance

When you finish customizing the userop routine, you relink it using either of the methods described in section 1.1.8 and in the ANSYS Installation and Configuration Guide for UNIX or in the ANSYS Installation and Configuration Guide for Windows . You must include the cmopt common block in your routine, to allow it to pass common data (such as design variables, state variables, and objective function values) to and from the ANSYS program.

1.12.3 Structuring Your Input

When your userop optimization routine finishes executing, program control returns to the ANSYS design optimizer, so that the ANSYS commands on Jobname.LOOP can execute for the next optimization loop. To use your own customization routine, you should issue the following ANSYS commands (in the order shown):
/OPT OPTYPE,USER OPUSER, OPVAR . . . OPEXE ! ANSYS optimization logic bypassed ! user-defined optimization to be done ! design and state variables & objective function values defined

! looping begins

1.12.3.1 Example of User Input for Optimization The commented example below shows you sample ANSYS command input for design optimization, done with a user-customized version of the userop routine:
c!!!---c!!! --c!!! --c!!! --c c c c c c minimize y = x1**2 + x2**2 subject to: 1 < x1 < 10 1 < x2 < 10 g = x1 - x2 > 1 ! ! ! ! ! ! use batch mode initial value for dv x1 initial value for dv x2 enter the ANSYS preprocessor define function y define function g

/batch,list x1=5 x2=5 /prep7 y = x1*x1 + x2*x2 g = x1 - x2

c c c c c c c c c c c

finish /opt opvar,x1,dv,1,10 opvar,x2,dv,1,10 opvar,y,obj,,,1 opvar,g,sv,1 optype,user opuser,10,1.1,2.0 opexe oplist,all finish

! ! ! ! ! ! ! ! ! !

finish prep finish preprocessing define first dv (range 1 to 10) define second dv (range 1 to 10) define objective function (tol = 1) define state variable (lower limit 1) user user-defined optimization logic define user opt. Modifiers (max iterations=10) execute userop routine list results

1.12.4 Using a Stand-Alone Optimization Program

If you can't link a customized version of the userop optimization routine, you can call a stand-alone optimization program that you've created. To do so, issue the ANSYS command /SYS (it has no equivalent menu path). Your external optimization program then uses the ANSYS optimization save file, Jobname.OPT, to pass data to and from the ANSYS program. You should structure your input to ANSYS as shown below, and you must run the program in batch mode:
/OPT OPTYPE,USER OPUSER, OPVAR . . . OPSAVE /SYS OPRESU OPEXE ! ANSYS optimization logic bypassed ! user-defined optimization to be done ! design and state variables & objective function values defined

! ! ! !

jobname.opt written for use by user-supplied program run the user-supplied program read results of user-supplied program from jobname.out looping begins

The OPSAVE and OPRESU commands write data to and read data from Jobname.OPT, and they also pass data to and from your custom optimization program.

1.12.5 What Does Jobname.OPT Contain?

The list below describes all of the relevant variables printed in Jobname.OPT. Note-The order of oplim, optol, and optvar is state variables, design variables, then objective function. In addition, optvar can contain any other ANSYS parameter.
Variable Description actopt Active optimization method andir Analysis file directory

anext annam bestsave bestset con1 con2 delfst delgr dvloop facexp facfst faclst factq fstfst fstpnt gradq(I) grdfst grdlst grdpnt grdtyp icongr

Analysis file extension Analysis file name Key to save the best database and results file Number of the current best optimization set Minimum parameter value Maximum parameter value Forward difference for first order optimization (delta on opfrst) Forward difference for gradient optimization tool (delta on opgrad) Indicates how to treat design variables on the loop file Fractional exponent for factorial optimization First design set number in the range of factored sets Last design set number in the range of factored sets Factor applied to penalty functions (first order optimization only) First design set number in the range of first order sets Reference design set number for first order optimization Vector saved for first order optimization First design set number in the range of gradient sets Last design set number in the range of gradient sets Reference design set number for gradient evaluation Reference design set type (best, last, or number) for gradient Key for selecting conjugate direction method Exponent that controls the resolution (sharpness) of the unconstrained objective ilamda function (in first order optimization) inopt Reserved for internal ANSYS use iobjxp Reserved for internal ANSYS use iopexe(I) Executable keys indicating which method or tool was run isvxp Reserved for internal ANSYS use jobnam Current name for file creation or use kconv Convergence flag (kconv = 0 indicates convergence) kfit1 Reserved for internal ANSYS use kfit2 Reserved for internal ANSYS use kopusr Flag choosing optimizer (0 = ansys, 1 = user optimization) kwght Reserved for internal ANSYS use labx(k) Current parameter names loopct Incrementing loop counter (for *GET command) lpot Maximum number of design loops (from OPRUN command) looptl Total number of optimization loops executed (using all methods) loptlp Saved value of lopt variable after each execution

maxfac Maximum allowed number of loops for factorial optimization analysis maxfst Maximum allowed number of iterations for first order optimization maxinf Maximum allowed sequential infeasible solutions (subp) maxrnd Maximum allowed number of iterations for random optimization maxsub Maximum allowed number of iterations for sub. prom. approx. mxpmrs Used for restart purposes (to reset mxparm in parmcm) nbest0 Best design set this first order evaluation nfeas Desired number of feasible solutions (random optimization) ninfes Counter for consecutive infeasible solutions nmswps The number of sweep evaluations to be made in each design variable dir. nobj Number of objective functions (must be 1) noqv Reserved for internal ANSYS use nrandm Reserved for internal ANSYS use nsqv Reserved for internal ANSYS use nsets Current number of sets that were made ntdv Total number of design variables ntfdf Finite difference counter for first order optimization ntline Line search counter for first order optimization ntmax Maximum number of parameters that have ever been defined ntsv Total number of state variables numopt Total number of design sets opam Random settings (related to routine VRAND) opdir Optimization file directory opext Optimization file extension opnam Optimization file name optlab All labels used in optimization (optimization labels and extras) optlab(I) Optimization labels optol(I) Tolerance optvar(I,j) Ith parameter of the jth design set opus(n) User parameters specified on OPUSER command (n = 1:8) psobj0 Best unconstrained objective value this first order evaluation rdfrom Indicates where to begin reading the analysis file (0 = top, 1 = after /Prep7) refobj Reference objective function for first order optimization savoutunit Saved output unit number seed Reserved for internal ANSYS use setnum(I) Set numbers sizeq Line search step size sq(j) Vector saved for first order optimization

stmapmax svparm swpfst swplst swppnt swptyp userp(n) vect(k)

Reserved for internal ANSYS use Indicates which type of parameters to save during looping First design set number in the range of sweep sets Last design set number in the range of sweep sets Reference design set number for sweep evaluation Reference design set type (best, last, or number) for sweep Reserved for internal ANSYS use Current scalar parameter values

1.13 Data Handling Routines

ANSYS provides UPF routines you can use for memory management and parameter processing.

1.13.1 Using the Memory Management Routines

ANSYS 5.4 allocates all dynamic memory in a large data space area called the Heap. Previous ANSYS releases (before the Heap was implemented) used a data space called STACK to allocate dynamic memory. For this reason, you access the Heap space using variables names ISTACK, DSTACK, CSTACK, and RSTACK, which are included via the STACK include deck. The ANSYS program allocates Heap space dynamically as it runs. User routines can use heap space to avoid adding large data sets (arrays) within these routines. Include the file stack.inc (on the ANSYS distribution medium) to allocate data space from the Heap. To do so, follow these steps: 1. Use the function HeapAllocPtr to notify the ANSYS program that space from the Heap is required. 2. Initialize the "handle" or "block number" to zero, as follows: iHandle = 0, for integers dHandle = 0, for double precision numbers cHandle = 0, for complex numbers rHandle = 0, for real numbers. 3. Allocate Heap space for an array or arrays, as follows: For integer numbers:
prti =HeapAllocPtr(ileng,'C8Label ',HEAP_INTEGER,iHandle)

For double precision numbers:

ptrdp =HeapAllocPtr(dleng,'C8Label ',HEAP_DOUBLE, dHandle)

For complex numbers:

ptrcp =HeapAllocPtr(cleng,'C8Label ',HEAP_COMPLEX,cHandle)

For real numbers:

ptrr =HeapAllocPtr(rleng,'C8Label ',HEAP_REAL ,rHandle)

The starting location for the array(s) in the Heap memory area is also returned. Where leng (argument 1) is the number of array words needed:

C8Label (argument 2) is a character*8 block name. HEAP_xxx (argument 3) is the word size based on integer words. (The HEAP_xxx definitions are parameters in the ansysdef.inc common). Handle (argument 4) is returned as the block number. The block number is used to deallocate the space. Ptr (the function result) is the starting location of the data space, minus 1.

You must include the ansysdef.inc common to get the parameter values of HEAP_INTEGER, HEAP_DOUBLE, HEAP_COMPLEX, and HEAP_REAL. 4. Use the arrays. You must include the common stack.inc in the routines accessing the Heap space. You can find stack.inc on the ANSYS distribution medium. The location of the first data entry for the array depends on the type of numbers in the array:
If the Array Contains The First Data Entry Begins at istack(ptri+1) Integer numbers Double precision numbers dstack(ptrdp+1) cstack(ptrcp+1) Complex numbers rstack(ptrr+1) Real numbers

5. If one routine allocates Heap space and another routine uses the space, you need to pass only the "handle" to the second routine. You also must include the stack common, just as you did in Step 4. TheHeapGetPtr function can retrieve the starting location minus 1, as follows:
ptri =HeapGetPtr(iHandle) ptrdp =HeapGetPtr(dHandle) ptrcp =HeapGetPtr(cHandle) ptrr =HeapGetPtr(rHandle)

This location is the same as the location described in Step 4. 6. Deallocate the stack space using the HeapDealloc subroutine, as follows: For integer numbers:
call HeapDealloc(iHandle)

For double precision numbers:

call HeapDealloc(dHandle)

For complex numbers:

call HeapDealloc(cHandle)

For real numbers:

call HeapDealloc(rHandle)

The next few pages provide input and output listings for the memory management routines. For an example, see the section "Function user03 (Demonstrates Using Heap Memory)," which appears earlier in this chapter.

1.13.2 Function HeapAllocPtr (Allocating Heap Space and Returning a Pointer)

function HeapAllocPtr (nValues, c16Label, flagAlloc, hHeapHandle) c primary function: Get A Block of Space from Heap and Return Pointer c keywords: integer function for heap allocate heap

c object/library:

c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c input arguments: nValues (int,sc,in) c16Label (ch*16,sc,in) flagAlloc(int,sc,in) output arguments: hHeapHandle (int,sc,out) HeapAllocPtr (int,sc,out) - Number of Values needed in Block - 16 character name for the Block - Flags describing this block (See ANSYSDEF and heapcom for more information) - Handle (Block number) for this block - Pointer to this data block (Use Ptr+i) for reference to ith value in the block

1.13.3 Subroutine HeapDealloc (Deallocating Heap Space)

subroutine HeapDealloc (hHeapHandle) c primary function: Unlock and Free a Heap Data Block c keywords: subroutine to free heap data heap

c object/library:

c *** Notice - This file contains ANSYS Confidential information *** c c c c Note: See heapcom.inc for information on use of Heap - Handle (Block number) for this block

input arguments: hHeapHandle (int,sc,out) output arguments: none

1.13.4 Function HeapGetPtr (Returns a Pointer for a Handle)

function HeapGetPtr (hHandle) c primary function: return pointer for an existing handle

c keywords: integer function for heap pointer return c object/library: c current - heap c *** Notice - This file contains ANSYS Confidential information *** c c c c c c NOTE: See heapcom.inc for documentation on heap usage - A heap handle - The corresponding pointer

input arguments: hHandle (int,sc,in) output arguments: HeapGetPtr (int,sc,out)

1.14 Parameter Processing Routines

The ANSYS distribution medium contains three routines you can use for parameter processing: pardim, parevl, and pardef.

1.14.1 Subroutine pardim (Creating a Dimensioned Parameter)

subroutine pardim (label8,labl4,nxyz,names) c *** primary function: create a dimensioned parameter c *dim,parm8,type,d1,d2,d3,name1,name2,name3

c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: label8 (chr*8,sc,in) labl4 (chr*4,sc,in) nxyz (int,ar(3),in) names (chr*8,ar(3),in) output arguments: none the name of the parameter to create 'TABL' or 'ARRA' or 'CHAR' the dimensions of the array Names for x,y,z directions in table

1.14.2 Function parevl (Finding and Evaluating a Parameter)

function parevl (label8,nval,subc,lvl,kerr) c *** primary function: find and evaluate a parameter c *** Notice - This file contains ANSYS Confidential information ***

c c c c c c c c c c c c

input arguments: label8 (chr*8,sc,in) nval subc lvl (int,sc,in) (dp,ar(3),in) (int,sc,in)

- the name of the parameter (must be upper case, left justified) - the number of subscripts (0,scaler) - values for the subscripts (if any) - 0,1 no error output 2, report error -1, set kerr flag with no anserr call - the value of the parameter (may be a packed character*8 - error flag (0,ok -1,output is packed 0=ok, 1=error, 2=error but TINY is used

output arguments: parevl (dp,sc,out) kerr (int,sc,out)

1.14.3 Subroutine pardef (Adding a Parameter)

subroutine pardef (label8,ctype,nval,subc,value,kerr) c *** primary function: add a parameter to parameter list c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c input arguments: label8 (chr*8,sc,in) - name of parameter label 8 is a character*8 variable that contains the name of the parameter that is to be defined. ctype (int,sc,in) - 0, dp 1,character ctype is an integer key which describes the type of data that the parameter data holds. This would also indicate the contents of "value" (arg 5). 0=double precision data 1=character data - number of subscripts nval is the number of subscripts that the "label8" (arg 1) contains. 1=single dimensioned variable (ex. x(10)) 2=double dimensioned variable (ex. y(10,3)) 3=triple dimensioned variable (ex. z(10,3,2)) -1=delete this parameter from the internal tables. - values of subscripts subc is a double precision vector that contains the subscripts of "label8" (arg 1). There should be enough values defined to match "nval" (arg 3). For example if "x" was dimensioned as "x(10,3,2)" and you wanted to set "x(5,1,1)=123.0", then "nval" (arg 3) should be set to 3, and "subc" should be set to 5.0, 1.0, 1.0, and "value" (arg 5) should be 123.0. Another example is if "y" was dimensioned to as "y(20,20)" and you were

nval

(int,sc,in)i

subc (dp,ar(3),in)

c should c c c c c The c c c c c c c c c was c c c c value (dp,sc,in) c c c c c c for c this c the c c c output arguments: c kerr (int,sc,out) c c c c

setting "y(5,8)=987", then "nval" (arg 3) be set to 2 and "subc" should be set to 5.0, 8.0, 0.0, and "value" (arg 5) should be 987.0. Remember subroutine "pardef" is only storing a data value of "label8" or "label8(x,y,z)". proper dimensions were set by a "*dim" command. Please note that although the values of "subc" should be double precision, subroutine "pardef" uses the intrinsic "nint" (nearest integer) function to get and use the integer equivalent. You should also note the "nval" (arg 3) and "subc" (arg 4) must fall within the range that set with a "*dim" or "*set" command or an error will occur. - the value for this parameter (should be a packed character*8 if ctype=1. To pack a char into a dp variable use "call chtodp(ch8,dp)". To unpack a dp variable into a char use "call dptoch(dp,ch8)" ) Value is the data value that is to be stored "label8" (arg 1). If "ctype=1" (arg 2) then value would be a "packed character" data from "chtodp" Ansys function. - error flag (0=ok, 1=error) kerr is an integer error flag that is returned to the calling subroutine. Any non zero number would indicate an error was detected in subroutine "pardef"

Note-For a demonstration of parameter processing, see the examples based on the useracel routine in this manual. You can find examples in section 1.7.7.

1.15 Miscellaneous Useful Functions

The ANSYS program has several miscellaneous functions you may find useful for working with UPFs:

The erhandler routine displays output messages (notes, warnings, and errors). The runcmd function lets you issue an ANSYS command from within a user routine.

The GetStackDisp routine retrieves current displacement values. The /UNDO command writes an "undo" file at critical points as a user routine executes. The /HOLD command allows you to synchronize multiple tasks in ANSYS. The /TRACK function enables you to do program tracing and timing.

For further descriptions of erhandler and /TRACK, see Chapter 3, "Subroutines for Users' Convenience." For details about the GetStackDisp function, see section 1.10.2.

1.15.1 Using Function runcmd

This function enables you to execute an ANSYS command from within a user routine. Inputs and outputs for runcmd are as follows:
function runcmd (command) c primary function: c object/library: upf c *** Notice - This file contains ANSYS Confidential information *** c c c c c input arguments: command (ch*128,sc,in) output arguments: none runcmd (int,sc,out) - A character string containing a valid ANSYS command, up to 128 char - An internally defined value, ignore Execute an ansys command from a user routine

1.15.2 Using the /UNDO Command

The "undo" file you create by issuing the /UNDO command is similar to the File.DB file created when you issue the SAVE command. The /UNDO command format is: Action is one of the following:

ON, to write the undo file OFF, to prevent the undo file from being written PROMPT, to have ANSYS ask permission before writing the file STATUS, to restore the file as it existed after executing the last command issued before the /UNDO command.

1.15.3 Using the /HOLD command

Issue the /HOLD command to synchronize tasks in ANSYS. The ANSYS program can synchronize tasks at the end of each results file set. The command format is:

Filename is the eight-character name of a message file. If the named file exists, the ANSYS program reads a command from the file and then deletes the file. TimeInterval is the length of time, in seconds, that ANSYS waits before trying to read the message file again. Timeout is the maximum length of time, in seconds, that ANSYS can wait between attempts to read the file.

Go to the beginning of this chapter

Chapter 2: Accessing the ANSYS Database

Go to the Next Chapter Go to the Previous Chapter Go to the Table of Contents for This Manual. Chapter 1 * Chapter 2 * Chapter 3 * Appendix A

2.1 What This Chapter Covers

This chapter describes how you can retrieve information in the ANSYS database (or store information in the database) by linking subroutines you create into the ANSYS program. You can use the database access routines with any of the user-programmable features. For example, you can create your own ANSYS commands and use them to execute database access routines (or have a database access routine call a user-defined command).

2.2 Inputs and Outputs for Database Access Routines

The descriptions of the database access routines or functions within this chapter describe both the input arguments and output arguments. Argument information includes the argument's type, size and intent.

Argument type is one of the following: int - integer dp - double precision log - logical chr - character dcp - double precision complex

Argument size is one of the following: sc - scalar variable ar(n) - array variable of length n func - functional return value

Argument intent is one of the following: in - input argument out - output argument inout - both an input and an output argument

2.3 Types of Database Access Routines

The rest of this chapter describes the functions and subroutines available for accessing information in the ANSYS database. The function and subroutine descriptions are grouped into sets: nodal information functions, element attribute information functions, etc.

2.4 Routines for Selecting and Retrieving Nodes and Elements

2.4.1 ndnext Function (Getting the Next Node Number)
function ndnext (next) c *** primary function: get the number of the next selected node c c *** Notice - This file contains ANSYS Confidential information *** c

c c c c c c c c c c c

typ=int,dp,log,chr,dcp input arguments: variable (typ,siz,intent) next (int,sc,in)

siz=sc,ar(n),func

intent=in,out,inout

description - the last node number used = 0 - use for initial value - the next selected node number = 0 - no more nodes

output arguments: ndnext (int,func,out)

2.4.2 ndprev Function (Getting the Number of the Previous Selected Node)
function ndprev (next) c *** primary function: get the number of the previous selected node c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c next (int,sc,in) - the next node number used c = 0 - use for initial value c c output arguments: c ndprev (int,func,out) - the previous selected node number c = 0 - no more nodes c

2.4.3 ndnxdf Function (Getting the Number of the Next Defined Node)
function ndnxdf (next) c *** primary function: get the number of the next defined node c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c next (int,sc,in) - the last node number used c = 0 - use for initial value c c output arguments: c ndnxdf (int,func,out) - the next defined node number c = 0 - no more nodes c

2.4.4 ndsel Function (Selecting, Unselecting, Deleting, or Inverting a Node)

c *** c *** c c *** c c c c c c c c c c c c c c node. c c c c subroutine ndsel (ndmi,ksel) primary function: to select, unselect, delete, or invert a node. secondary functions: none. Notice - This file contains ANSYS Confidential information *** typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout

input arguments: variable (typ,siz,intent) ndmi (int,sc,in)

ksel

(int,sc,in)

description - node number = 0 - all nodes < 0 - do not delete CPs and CEQNs (merge/offset/compress) - type of operation to be performed. ksel = 0 - delete node. = 1 - select node. =-1 - unselect node. = 2 - invert select status of

output arguments: none.

2.4.5 elnext Function (Getting the Number of the Next Element)

function elnext (next) c *** primary function: get the number of the next selected element c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c next (int,sc,in) - the last element number used c = 0 - use for initial value c c output arguments: c elnext (int,func,out) - the next selected element c = 0 - no more elements

2.4.6 elprev Function (Getting the Number of the Previous Selected Element)

function elprev (prev) c *** primary function: get the number of the previous selected element c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c prev (int,sc,in) - the last element used c = 0 - use for initial value c c output arguments: c elprev (int,func,out) - the previous selected element c = 0 - no more elements c

2.4.7 elnxdf Function (Getting the Number of the Next Defined Element)
function elnxdf (next) c *** primary function: get the number of the next defined element c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c next (int,sc,in) - the last element used c = 0 - use for initial value c c output arguments: c elnxdf (int,func,out) - the next defined element c = 0 - no more elements c

2.4.8 elsel Subroutine (Selecting, Unselecting, Deleting, or Inverting an Element)

subroutine elsel (ielei,ksel) c *** primary function: to select, unselect, delete, or invert an element. c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ielei (int,sc,in) - element number c = 0 - all elements

c c c c c c c c c

ksel

(int,sc,in)

- type of operation to be performed. = 0 - delete element. = 1 - select element. =-1 - unselect element. = 2 - invert select status for element

output arguments: none.

2.5 Node Information Routines

2.5.1 ndinqr Function (Getting Information About a Node)
The primary function of ndinqr is getting information about a node. You can also use this function to set the current node pointer to this node.
function ndinqr (node,key) c *** primary function: get information about a node. c *** secondary functions: set current node pointer to this node. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c Should be 0 for key=11, DB_NUMDEFINED, c DB_NUMSELECTED, DB_MAXDEFINED, and c DB_MAXRECLENG c key (dp,sc,in) - key as to information needed about c the node. c = DB_SELECTED - return select status: c ndinqr = 0 - node is undefined. c =-1 - node is unselected. c = 1 - node is selected. c = DB_NUMDEFINED - return number of defined nodes c = DB_NUMSELECTED - return number of selected nodes c = DB_MAXDEFINED - return highest node number defined c = DB_MAXRECLENG - return maximum record length (dp words) c = 2, return length (dp words) c = 3, c = 11, return void percent (integer) c = -1, c = -2, superelement flag c = -3, master dof bit pattern c = -4, active dof bit pattern c = -5, solid model attachment c = -6, pack nodal line parametric value c = -7, constraint bit pattern c = -8, force bit pattern c = -9, body force bit pattern c = -10,internal node flag

c c c c c c c c c c c c c c c c c c c c c on c

= -11, = -12, =-101, =-102, =-103, =-104, =-105, =-106, =-107, =-108, =-109, =-110, =-111, =-112, =-113, =-114, =-115, =-116,

pointer pointer pointer pointer pointer pointer pointer pointer pointer

to to to to to to to to to

node data record angle record attached elements attached couplings attacted constraint equations nodal stresses specified disp'S specified forces x/y/z record

pointer to nodal temperatures pointer to nodal heat generations pointer to calculated displacements

output arguments: ndinqr (int,func,out)

- the returned value of ndinqr is based setting of key.

2.5.2 getnod Function (Getting a Nodal Point)

subroutine getnod (node,v,kerr,kcrot) c *** primary function: get a nodal point c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c kerr (int,sc,inout) - message flag c = 0 - print no message if node is unselected c or undefined c = 1 - print message if node is undefined c = 2 - print message if node is undefined c or unselected c kcrot (int,sc,in) - output coordinates in this coordinate system. c if kcrot is negative, output theta and c phi coordinates in radians c c output arguments: c v (dp,ar(6),out) - Coordinates (first 3 values) and rotation c angles (last 3 values) c kerr (int,sc,inout) - select status c = 0 - node is selected c = 1 - node is not defined

c c

=-1 - node is unselected

2.5.3 putnod Function (Storing a Node)

subroutine putnod (node,vctn,kcrot) c *** primary function: store a node c *** secondary functions: display node if in immediate mode. c *** Notice - This file contains ANSYS Confidential information *** c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout

c input arguments: c variable (typ,siz,intent) c node (int,sc,in) c vctn (dp,ar(6),in) c c kcrot (int,sc,in) nodal c c c output arguments: none.

description - node number - array of 3 nodal coordinates and 3 nodal rotation angles. - local coordinate system in which the coordinates and angles are defined

2.5.4 ndgall Function (Getting the XYZ/Rotation Coordinates Vector for a Node)
function ndgall (node,xyz) c *** primary function: get x,y,z,rotx,roty,rotz vector for a node. c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c node (int,sc,in) - node number for operation. c c output arguments: c ndgall (int,sc,out) - status of node. c 0=node is undefined. c -1=node is unselected. c 1=node is selected. c xyz (dp,ar(6),out) - vector containing x,y,z,rotx,roty,rotz c

2.5.5 ndspgt Subroutine (Getting the Nodal Solution for a Node of an Element)
subroutine ndspgt (node,dofs,ndof,nrot,xyzang,nuvect,unode) c *** primary function: get the nodal solution for a node of an element c

c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c node (int,sc,in) - The node number c dofs (int,sc,in) - The dofs to retrieve for the node. c dof = degree of freedom c This is an integer representation using c bit pattern logic of the dof or the dof c set. This is normally done using the c parameters defined in echprm. Then c you could use dofs = UX + UY in the c calling routine, for example. c ndof (int,sc,in) - The number of node dofs (1, 2 or 3). c nrot (int,sc,in) - Key to rotate dofs from nodal to global c coordinate systems. c if 0, none. if 2, 2-d. if 3, 3-d c if > 0, dof set must include and only c include all terms of the vector (e.g. c UX,UY,UZ, or AX,AY,AZ). c xyzang (dp,ar(6),in) - The xyz virgin node coordinates c (including angles). Not used if c nrot = 0 or ndof < 2. c nuvect (int,sc,in) - Number of vectors to retrieve. Can vary c between 1 and 5. Normally 1 is what is c wanted. Other vectors include previous c values and/or velocities. See elucom for c all possibilites. Contents are analysis c type dependent. c c output arguments: c unode (dp,ar(ndof,nuvect),out) - Element nodal solution vectors in c the global coordinate system. c

2.6 Element Attribute Routines

2.6.1 elmiqr Function (Getting Information About an Element)
function elmiqr (ielem,key) c *** primary function: get information about an element. c *** secondary functions: set current element pointer to this element. c c *** Notice - This file contains ANSYS Confidential information *** c

c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ielem (int,sc,in) - element number c should be zero for key=11, DB_NUMDEFINED, c DB_NUMSELECTED, or DB_MAXDEFINED c key (int,sc,in) - information flag. c = DB_SELECTED - return select status: c elmiqr = 0 - element is undefined. c =-1 - element is unselected. c = 1 - element is selected. c = DB_NUMDEFINED - return number of defined elements c = DB_NUMSELECTED - return number of selected elements c = DB_MAXDEFINED - return maximum element number used c = DB_MAXRECLENG - return maximum record length c (int words) c = 2 - return length (int words) c = 3 - return layer number c (for cross reference files return number of entities) c = 4 - return address of first data word c = 5 - return length (in record type units) c = 6 - return compressed record number. c = 11 - return void percent (integer) c = 16 - return location of next record c (this increments the next record count) c = 18 - return type of file. c elmiqr = 0 - integer c = 1 - double precision c = 2 - real c = 3 - complex c = 4 - character*8 c = 7 - index c = 19 - return virtual type of file. c elmiqr = 0 - fixed length (4.4 form) c = 1 - indexed variable length (layer data) c = 2 - xref data tables c = 3 - bitmap data (for 32 data item packed records) c = 4 - data tables (three dimensional arrays) c = -1 - material number etc. (see elmcmx) c =-101 - pointer to element integers etc. (see elmcmx) c c output arguments: c elmiqr (int,func,out) - the returned value of elmiqr is based on c setting of key. c #include "impcom.inc" c external vminqr,virtrd integer vminqr c integer elmiqr, ielem, key, j, iloc(2) c

2.6.2 elmget Function (Getting an Element's Attributes and Nodes)

function elmget (ielem,elmdat,nodes) c *** primary function: get element attributes and nodes. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ielem (int,sc,in) - element number c c output arguments: c elmget (int,func,out) - status of element. c = 0 - element undefined c < 0 - number of nodes on unselected c element c > 0 - number of nodes on selected element c elmdat (int,ar(10),out) - element attributes. c elmdat(1) - material number c (2) - element type c (3) - real constant number c (4) - element number (obsolete, do not use) c (5) - coordinate system number c (6) - death flag c if 0 - alive c if 1 - dead c (7) - solid model reference c (8) - 100*shape + specific shape c (9) - reserved c (10) - p element include flag c if 0 - include c if 1 - exclude c c nodes (int,ar(*),out) - node numbers for element.

2.6.3 elmput Subroutine (Storing an Element)

subroutine elmput (ielem,elmdat,nnod,nodes) c *** primary function: store element attributes and node numbers. c *** secondary functions: set current element pointer to this element. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ielem (int,sc,in) - element number

c c c c c use) c c c c c c c c c c c c c c c c

elmdat

(int,ar(10),in) - element attributes. elmdat(1) - material number (2) - element type (3) - real constant number (4) - element number (obsolete, do not (5) (6) - coordinate system number - death flag if 0 - alive if 1 - dead (7) - solid model reference (8) - 100*shape + specific shape (9) - reserved (10) - p element include flag if 0 - include if 1 - exclude - number of nodes for this element. - node numbers for this element.

nnod nodes

(int,sc,in) (int,ar(*),in)

output arguments: none.

2.6.4 etyiqr Function (Getting a Data Item About an Element Type)

function etyiqr (itype,key) c *** primary function: get information about an element type. c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c itype (int,sc,in) - element type number c Should be 0 for key=11, DB_NUMDEFINED, c DB_NUMSELECTED, DB_MAXDEFINED, and c DB_MAXRECLENG c key (int,sc,in) - information flag. c = DB_SELECTED - return select status: c etyiqr = 0 - element type is undefined. c =-1 - element type is unselected. c = 1 - element type is selected. c = DB_NUMDEFINED - return number of defined element types c = DB_NUMSELECTED - return number of selected element types c = DB_MAXDEFINED - return highest element type number defined c = DB_MAXRECLENG - return maximum record length (int words) c = 2 - return length (int words) c = 3 - return layer number c (for cross reference files return c number of entities) c = 4 - return address of first integer word

c = 5 - return length (int words) c = 6 - return compressed record number. c = 11 - return void percent (integer) c = 16 - return location of next record c (this increments the next record count) c = 18 - return type of file. c etyiqr = 0 - integer c = 1 - double precision c = 2 - real c = 3 - complex c = 4 - character*8 c = 7 - index c = 19 - return virtual type of file. c etyiqr = 0 - fixed length (4.4 form) c = 1 - indexed variable length c (layer data) c = 2 - xref data tables c = 3 - bitmap data (for 32 data c item packed records) c = 4 - data tables (three c dimensional arrays) c = -n, return element characteristic n from etycom for element c type itype. c n is correlated to the parameter names in echprm. c see elccmt for definitions of element characteristics. c note- this will not overwrite the current setting of c etycom. c c output arguments: c etyiqr (int,func,out) - the returned value of etyiqr is based on c setting of key.

2.6.5 etyget Function (Getting Information About an Element Type)

function etyget (itype,ielx) c *** primary function: get element type data. c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c itype (int,sc,in) - element type number c c output arguments: c etyget (int,func,out) - status of element type. c = 0 - element type is undefined. c < 0 - number of data items on unselected

c c c c c c

ielx

(int,ar(*),out)

element type. > 0 - number of data items on selected element type. - element type data. see elccmt for description of data.

2.6.6 etyput Subroutine (Storing Element Type Data)

subroutine etyput (itype,n,ielx) c *** primary function: store element type data. c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c itype (int,sc,in) - element type number for operation. c n (int,sc,in) - length of data vector to store. c ielx (int,ar(*),in) - element type data. see elccmt for c description. c c output arguments: c none. c

2.6.7 echrtr Subroutine (Getting Information About Element Characteristics)

c c subroutine echrtr (iott,elcdn,ielc,kerr) primary function: collect all element characteristics based on ityp,jtyp, and keyopts

c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c iott (int,sc,in) - printout file c ielc (int,ar(150),inout) - input element characteristics c in positions 1 to 20. c (itype, jstif, keyopts, etc.) c c output arguments: c elcdn (chr,sc,out) - element descriptive name as character c string c ielc (int,ar(150),inout) - output element characteristics c in positions 21 to 150. c (kdim, ishap, idegen, etc.) c see elccmt for a full list

c c c c

kerr

(int,sc,out)

- error flag = 0 - no errors = 1 - errors

2.6.8 etysel Subroutine (Selecting, Unselecting, Deleting, or Inverting an Element Type)

subroutine etysel (itypi,ksel) c *** primary function: to select, unselect, delete, or invert an c element type. c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c itypi (int,sc,in) - element type number c = 0 - all element types c ksel (int,sc,in) - type of operation to be performed. c = 0 - delete element type. c = 1 - select element type. c =-1 - unselect element type. c = 2 - invert element type. c c output arguments: c none. c

2.6.9 mpinqr Function (Getting Information About a Material Property)

function mpinqr (mat,iprop,key) c *** primary function: get information about a material property. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c mat (int,sc,in) - material number c should be 0 for key=11, DB_NUMDEFINED, c DB_MAXDEFINED, and DB_MAXRECLENG c c iprop (int,sc,in) - property reference number: c if iprop = 0, test for existence of any material property with this c material number (with key = DB_SELECTED)

c EX = 1, EY = 2, EZ = 3, NUXY= 4, NUYZ= 5, NUXZ= 6, GXY = 7, GYZ = 8 c GXZ = 9, ALPX=10, ALPY=11, ALPZ=12, DENS=13, MU =14, DAMP=15, KXX =16 c KYY =17, KZZ =18, RSVX=19, RSVY=20, RSVZ=21, C =22, HF =23, VISC=24 c EMIS=25, ENTH=26, LSST=27, PRXY=28, PRYZ=29, PRXZ=30, MURX=31, MURY=32 c MURZ=33, PERX=34, PERY=35, PERZ=36, MGXX=37, MGYY=38, MGZZ=39, EGXX=40 c EGYY=41, EGZZ=42, TGXX=43, TGYY=44, TGZZ=45, SONC=46, c ORTH=54, CABL=55, RIGI=56 c HGLS=57, BM =58, QRAT=59, REFT=60, PLAS=61, CREE=62, FAIL=63, BH =64 c PIEZ=65, SWEL=66, WATE=67, CONC=68, PFLO=69, ANEL=70, ACOU=71, EVIS=72 c USER=73, NL =74, HYPE=75, NNEW=76, MOON=77, OGDE=78, SUTH=79, WIND=80 c (see mpinit for uncommented code) c c key (int,sc,in) - key as to the information needed c about material property. c = DB_SELECTED - return select status: c mpinqr = 0 - material prop is undefined. c = 1 - material prop is selected. c = DB_NUMDEFINED - return number of defined material properties c = DB_MAXDEFINED - return highest material property number defined c = DB_MAXRECLENG - return maximum record length (dp words) c = 2 - return length (dp words) c = 3 - return number of temp. values c = 11 - return void percent (integer) c c output arguments: c mpinqr (int,func,out) - returned value of mpinqr is based on c setting of key. c

2.6.10 mpget Function (Getting a Material Property Table)

function mpget (mat,iprop,temp,prop) c *** primary function: get a material property table. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c mat (int,sc,in) - material number c iprop (int,sc,in) - property reference number: c EX = 1, EY = 2, EZ = 3, NUXY= 4, NUYZ= 5, NUXZ= 6, GXY = 7, GYZ = 8

c GXZ = 9, ALPX=10, ALPY=11, ALPZ=12, DENS=13, MU =14, DAMP=15, KXX =16 c KYY =17, KZZ =18, RSVX=19, RSVY=20, RSVZ=21, C =22, HF =23, VISC=24 c EMIS=25, ENTH=26, LSST=27, PRXY=28, PRYZ=29, PRXZ=30, MURX=31, MURY=32 c MURZ=33, PERX=34, PERY=35, PERZ=36, MGXX=37, MGYY=38, MGZZ=39, EGXX=40 c EGYY=41, EGZZ=42, TGXX=43, TGYY=44, TGZZ=45, SONC=46, c ORTH=54, CABL=55, RIGI=56 c HGLS=57, BM =58, QRAT=59, REFT=60, PLAS=61, CREE=62, FAIL=63, BH =64 c PIEZ=65, SWEL=66, WATE=67, CONC=68, PFLO=69, ANEL=70, ACOU=71, EVIS=72 c USER=73, NL =74, HYPE=75, NNEW=76, MOON=77, OGDE=78, SUTH=79, WIND=80 c (see mpinit for uncommented code) c c output arguments: c mpget (int,func,out) - number of temperature values c temp (dp,ar(mpget),out) - vector of the temperature values c prop (dp,ar(mpget),out) - vector of the property values c

2.6.11 mpput Subroutine (Storing a Material Property Table)

subroutine mpput (mat,iprop,ntab,temp,prop) c *** primary function: store material property tables. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c mat (int,sc,in) - material number. c iprop (int,sc,in) - property reference number: c EX = 1, EY = 2, EZ = 3, NUXY= 4, NUYZ= 5, NUXZ= 6, GXY = 7, GYZ = 8 c GXZ = 9, ALPX=10, ALPY=11, ALPZ=12, DENS=13, MU =14, DAMP=15, KXX =16 c KYY =17, KZZ =18, RSVX=19, RSVY=20, RSVZ=21, C =22, HF =23, VISC=24 c EMIS=25, ENTH=26, LSST=27, PRXY=28, PRYZ=29, PRXZ=30, MURX=31, MURY=32 c MURZ=33, PERX=34, PERY=35, PERZ=36, MGXX=37, MGYY=38, MGZZ=39, EGXX=40 c EGYY=41, EGZZ=42, TGXX=43, TGYY=44, TGZZ=45, SONC=46, c ORTH=54, CABL=55, RIGI=56 c HGLS=57, BM =58, QRAT=59, REFT=60, PLAS=61, CREE=62, FAIL=63, BH =64

c PIEZ=65, SWEL=66, WATE=67, CONC=68, PFLO=69, ANEL=70, ACOU=71, EVIS=72 c USER=73, NL =74, HYPE=75, NNEW=76, MOON=77, OGDE=78, SUTH=79, WIND=80 c (see mpinit for uncommented code) c c ntab (int,sc,in) - number of entries in the table c (1 to 100) c tem (dp,ar(ntab),in) - temperature vector (ascending) c prp (dp,ar(ntab),in) - property vector c c output arguments: c none. c

2.6.12 mpdel Subroutine (Deleting a Material Property Table)

subroutine mpdel (mat,iprop) c *** primary function: delete material property tables. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c mat (int,sc,in) - material number. c iprop (int,sc,in) - property reference number: c (0 = all properties) c EX = 1, EY = 2, EZ = 3, NUXY= 4, NUYZ= 5, NUXZ= 6, GXY = 7, GYZ = 8 c GXZ = 9, ALPX=10, ALPY=11, ALPZ=12, DENS=13, MU =14, DAMP=15, KXX =16 c KYY =17, KZZ =18, RSVX=19, RSVY=20, RSVZ=21, C =22, HF =23, VISC=24 c EMIS=25, ENTH=26, LSST=27, PRXY=28, PRYZ=29, PRXZ=30, MURX=31, MURY=32 c MURZ=33, PERX=34, PERY=35, PERZ=36, MGXX=37, MGYY=38, MGZZ=39, EGXX=40 c EGYY=41, EGZZ=42, TGXX=43, TGYY=44, TGZZ=45, SONC=46, c ORTH=54, CABL=55, RIGI=56 c HGLS=57, BM =58, QRAT=59, REFT=60, PLAS=61, CREE=62, FAIL=63, BH =64 c PIEZ=65, SWEL=66, WATE=67, CONC=68, PFLO=69, ANEL=70, ACOU=71, EVIS=72 c USER=73, NL =74, HYPE=75, NNEW=76, MOON=77, OGDE=78, SUTH=79, WIND=80 c (see mpinit for uncommented code) c c output arguments: c none. c

2.6.13 rlinqr Function (Getting Information About a Real Constant Set)

function rlinqr (nreal,key) c *** primary function: get information about a real constant set c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c nreal (int,sc,in) - real constant table number c should be 0 for key=11, DB_NUMDEFINED, c DB_NUMSELECTED, DB_MAXDEFINED, and c DB_MAXRECLENG c key (int,sc,in) - information flag. c = DB_SELECTED - return select status c rlinqr = 0 - real constant table is undefined. c =-1 - real constant table is unselected. c = 1 - real constant table is selected c = DB_NUMDEFINED - return number of defined real constant tables c = DB_NUMSELECTED - return number of selected real constant tables c = DB_MAXDEFINED - return highest real constant table defined c = DB_MAXRECLENG - return maximum record length (dp words) c = 2 - return length (dp words) c = 3 - return layer number c (for cross reference files return number of entities) c = 4 - return address of first data word c = 5 - return length (in record type units) c = 6 - return compressed record number. c = 11 - return void percent (integer) c = 16 - return location of next record c (this increments the next record count) c c = 18 - return type of file. c rlinqr = 0 - integer c = 1 - double precision c = 2 - real c = 3 - complex c = 4 - character*8 c = 7 - index c = 19 - return virtual type of file. c rlinqr = 0 - fixed length (4.4 form) c = 1 - indexed variable length (layer data) c = 2 - xref data tables c = 3 - bitmap data (for 32 data item packed records) c = 4 - data tables (three dimensional arrays) c

c output arguments: c rlinqr (int,func,out) on c

- the returned value of rlinqr is based

2.6.14 rlget Function (Getting Real Constant Data)

function rlget (nreal,rtable) c *** primary function: get real constant data c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c nreal (int,sc,in) - real constant table number c c output arguments: c rlget (int,func,out) - number of real constant data obtained c rtable (dp,ar(*),out) - real constant data obtained c

2.6.15 rlsel Subroutine (Selecting or Deleting a Real Constant Set)

subroutine rlsel (nreai,ksel) c *** primary function: select or delete a real constant set c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c nreai (int,sc,in) - real constant table c = 0 - all real constant tables c ksel (int,sc,in) - type of operation to be performed. c = 0 - delete real constant table. c = 1 - select real constant table. c =-1 - unselect real constant table. c = 2 - invert real constant table. c c output arguments: c none c

2.6.16 csyiqr Function (Getting Information About a Coordinate System)

function csyiqr (ncsy,key) c *** primary function: get information about a coordinate system c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ncsy (int,sc,in) - coordinate system reference number c should be zero for key= DB_NUMDEFINED c or DB_MAXDEFINED c key (int,sc,in) - information flag. c = DB_SELECTED - return status: c csyiqr = 0 - coordinate system is not defined c -1 - coordinate system is not selected c 1 - coordinate system is selected c = DB_NUMDEFINED - number of defined coordinate systems c = DB_MAXDEFINED - maximum coordinate system reference c number used. c c output arguments: c csyiqr (int,func,out) - the returned value of csyiqr is based on c setting of key. c

2.6.17 csyget Function (Getting a Coordinate System)

function csyget (ncsy,csydpx,csyinx) c *** primary function: get a coordinate system c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c NOTE: As a time-saving device, this routine will not fetch the coordinate c system data from the database (an expensive operation) c if ncsy = csyinx(4), as this would indicate that the data is current. c If you wish to force this routine to fetch coordinate system data (in c the case of loading a local array, for example), you MUST set c ncsy != csyinx(4) before function call. c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description csycom name c ncsy (int,sc,in) - coordinate system number c csyinx(4) (int,sc,inout) - coordinate system number csyact c c output arguments: c csydpx (dp,ar(18),out)

c csydpx(1-9) - transformation matrix c (10-12) - origin (XC, YC, ZC) c (13-14) - coordinate system parameters cparm c cparm2 c (15) - spare c (16-18) - defining angles c csyinx (int,ar(6),out) c csyinx(1-2) - theta, phi singularity keys c (3) - coordinate system type icdsys c (csyinx(4) is inout) (4) - coordinate system number csyact c (5) - spare c (6) - spare c csyget (int,func,out) - status of coordinate system c = 0 - coordinate system exists c = 1 - coordinate system doesn't exist c

2.6.18 csyput Subroutine (Storing a Coordinate System)

subroutine csyput (ncsy,csydpx,csyitx) c *** primary function: store a coordinate system c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description csycom name c ncsy (int,sc,in) - coordinate system number c csydpx (dp,ar(18),out) c csydpx(1-9) - transformation matrix c (10-12) - origin (XC, YC, ZC) c (13-14) - coordinate system parameters cparm c cparm2 c (15) - spare c (16-18) - defining angles c csyinx (int,ar(6),out) c csyinx(1-2) - theta, phi singularity keys c (3) - coordinate system type icdsys c (4) - coordinate system number csyact c (5) - spare c (6) - spare c c c output arguments:

c c

none

2.6.19 csydel Subroutine (Deleting a Coordinate System)

subroutine csydel (ncsy) c *** primary function: delete a coordinate system c *** secondary functions: none c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c ncsy (int,sc,in) - coordinate system number c c output arguments: c none c

2.6.20 userac Subroutine (Demonstrates Use of Element Attribute Routines)

See Chapter 4 for an example that demonstrates how to use the userac subroutine to extract information about an element type and element real constants from the ANSYS database. You can find this subroutine on your ANSYS distribution medium.

2.7 Coupling and Constraint Routines

2.7.1 cpinqr Function (Getting Information About a Coupled Set)
function cpinqr (nce,key) c *** primary function: get information about a coupled set c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c nce (int,sc,in) - coupled set number c key (int,sc,in) - inquiry key: c should be zero for key=11, DB_NUMDEFINED, c DB_NUMSELECTED, DB_MAXDEFINED, and c DB_MAXRECLENG

c = DB_SELECTED - return select status c cpinqr = 1 - coupled set is selected c = 0 - coupled set in undefined c =-1 - coupled set in unseleted c = DB_NUMDEFINED - return number of defined coupled sets c = DB_NUMSELECTED - return number of selected coupled sets c = DB_MAXDEFINED - return the number of the highest numbered c coupled set c = DB_MAXRECLENG - return length of largest coupled set record c (max record length) c = 2 - return length (data units) c = 3 - return layer number c = 4 - return address of first data word c = 11 - return void percent (integer) c = 16 - return location of next record c c output arguments: c cpinqr (int,func,out) - the returned value of cpinqr is based on c setting of key c

2.7.2 cpget Function (Getting a Coupled Set)

function cpget (ncp,ieqn) c *** primary function: get a coupled set c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ncp (int,sc,in) - coupled set number c c output arguments: c cpget (int,func,out) - number of nodes in list c ieqn (int,ar(cpget+2),out) - coupled set info: c ieqn(1:cpget) - list of coupled nodes c ieqn(cpget+1) - set degree of freedom c ieqn(cpget+2) - number of nodes in list c (copy of return value) c

2.7.3 cpput Subroutine (Storing a Coupled Set)

subroutine cpput (ncp,n,ieqn) c *** primary function: store a coupling set c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout

c c c c c c c c c c c c c c c

input arguments: variable (typ,siz,intent) description ncp (int,sc,in) - coupled set number n (int,sc,in) - number of nodes in coupled set ieqn (int,ar(n+2),in) - info for storage ieqn(1:n) - list of coupled nodes ieqn(n+1) - degree of freedom label for set (ieqn(n+2) is inout) ieqn(n+2) - number of nodes in coupled set (copy of n) output arguments: ieqn(n+2) (int,sc,inout) - number of nodes in coupled set (another copy of n)

2.7.4 cpsel Subroutine (Selecting or Deleting a Coupled Set)

subroutine cpsel (ncpi,ksel) c *** primary function: select or delete a coupled set c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ncpi (int,sc,in) - coupled set number c ksel (int,sc,in) - select/delete flag c = 0 - delete coupled set c = 1 - select coupled set c output arguments: c none c

2.7.5 ceinqr Function (Getting Information About a Constraint Equation Set)

function ceinqr (nce,key) c *** primary function: get information about a constraint equation set c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c nce (int,sc,in) - constraint equation number c key (int,sc,in) - inquiry key: c should be zero for key=11, DB_NUMDEFINED, c DB_NUMSELECTED, DB_MAXDEFINED, and

c c c c c c c c c c set c c c c c c c c c c on c c

DB_MAXRECLENG - return select status ceinqr = 1 - equation is selected = 0 - equation is undefined =-1 - equation is unselected = DB_NUMDEFINED - return number of defined contraint equations = DB_NUMSELECTED - return number of selected contraint equations = DB_MAXDEFINED - return number of highest numbered constraint equation defined = DB_MAXRECLENG - return length of longest contraint equation = DB_SELECTED = 2 = 3 = 4 = 11 = 16 = -1 (max record length) return length (data units) return layer number address of first data word return void percent (integer) return location of next record return master dof for this eqn - the returned value of ceinqr is based setting of key

output arguments: ceinqr (int,func,out)

2.7.6 ceget Function (Getting an Constraint Equation)

function ceget (nce,ieqn,deqn) c *** primary function: get a constraint equation c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c nce (int,sc,in) - constraint equation number c c c c c c c c c output arguments: ceget (int,func,out) - number of dof in equation ieqn (int,ar(ceget+1),out) - integer info ieqn(1:ceget) - list of node*32+dof ieqn(ceget+1) - number of dof in equation (copy of return value) deqn (dp,ar(ceget+1),out) - dp info deqn(1:ceget) - list of coefficients deqn(ceget+1) - constant term

2.7.7 ceput Subroutine (Storing a Constraint Equation)

subroutine ceput (nce,n,ieqn,deqn) c *** primary function: store a constraint equation c *** secondary functions: none

c *** Notice - This file contains ANSYS Confidential information *** c input arguments: c nce (int,sc,in) - constraint equation set number c n (int,sc,in) - number of degrees of freedom in set c ieqn (int,ar(n+1),in) - integer info c ieqn(1:n) - node*32+dof for each dof in set c ieqn(n+1) - number of dof in set (copy of n above) c deqn (dp,ar(n+1),in) - dp info c deqn(1:n) - coefficients of each dof in set c deqn(n+1) - constant term c output arguments: none

2.7.8 cesel Subroutine (Deleting or Selecting a Constraint Equation)

subroutine cesel (ncei,ksel) c *** primary function: select or delete a constraint equation c *** secondary functions: none c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c ncei (int,sc,in) - constraint equation number c ksel (int,sc,in) - select/delete flag c = 0 - delete equation c = 1 - select equation c c output arguments: c none c

2.8 Nodal Loading Routines

2.8.1 disiqr Function (Getting a Information About Constraints)
function disiqr (node,key) c *** primary function: get information about constraints c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c

c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number for inquire. c key (int,sc,in) - key as to the information needed c = 1 - return constraint mask c = DB_MAXDEFINED, c DB_NUMDEFINED - return number of nodal constraints c NOTE: both DB_MAXDEFINED and c DB_NUMDEFINED produce the same c functionality c c output arguments: c disiqr (int,func,out) - the returned value of disiqr is based on

2.8.2 disget Function (Getting a Constraint from the Database)

function disget (inode,idf,value) c *** primary function: get a constraint from the data base c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c inode (int,sc,in) - node number (negative value for no c partabeval) c idf (int,sc,in) - reference number for the DOF: (1-32) c UX = 1, UY = 2, UZ = 3, ROTX= 4, ROTY= 5, ROTZ= 6, AX = 7, AY = 8 c AZ = 9, VX =10, VY =11, VZ =12 c PRES=19, TEMP=20, VOLT=21, MAG =22, ENKE=23, ENDS=24 c EMF =25, CURR=26 SP01=27, SP02=28, SP03=29, SP04=30, SP05=31, SP06=32 c (missing entries are spares) c c output arguments: c disget (int,func,out) - status of constraint. c = 0 - no constraint on this node c for this DOF c = 4 - this node has a constraint c defined for this DOF c value (dp,ar(4),out) - constraint values c value(1-2) - (real,imag) values of present settings c

2.8.3 disput Subroutine (Storing a Constraint at a Node)

subroutine disput (node,idf,value) c *** primary function: store a constraint at a node. c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c idf (int,sc,in) - reference number of DOF: (1-32) c UX = 1, UY = 2, UZ = 3, ROTX= 4, ROTY= 5, ROTZ= 6, AX = 7, AY = 8 c AZ = 9, VX =10, VY =11, VZ =12 c PRES=19, TEMP=20, VOLT=21, MAG =22, ENKE=23, ENDS=24 c EMF =25, CURR=26 (missing entries are spares) c c value (dp,ar(2),in) - (real,imag) values for constraint c c output arguments: c none. c

2.8.4 disdel Subroutine (Deleting a Constraint at a Node)

subroutine disdel (node,idf) c *** primary function: delete a constraint at a node c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number. c idf (int,sc,in) - reference number of DOF: (1-32) c UX = 1, UY = 2, UZ = 3, ROTX= 4, ROTY= 5, ROTZ= 6, AX = 7, AY = 8 c AZ = 9, VX =10, VY =11, VZ =12 c PRES=19, TEMP=20, VOLT=21, MAG =22, ENKE=23, ENDS=24 c EMF =25, CURR=26 (missing entries are spares) c c output arguments: c none.

2.8.5 foriqr Function (Getting Information About Nodal Loads)

function foriqr (node,key) c *** primary function: get information about nodal loads. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - number of node being inquired about. c should be 0 for key=DB_MAXDEFINED or c DB_NUMDEFINED c key (dp,sc,in) - key as to information needed c = 1 - return force mask for node c = DB_MAXDEFINED, c DB_NUMDEFINED - return number of nodal loadings c in model c NOTE: both DB_MAXDEFINED and DB_NUMDEFINED c produce the same functionality c c output arguments: c foriqr (int,func,out) - the returned value of foriqr is based on c

2.8.6 forget Function (Getting a Constraint from the Database)

function forget (inode,idf,value) c *** primary function: get a constraint from the data base c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c inode (int,sc,in) - node number (negative value for no c partabeval) c idf (int,sc,in) - reference number for the DOF: (1-32) c FX = 1, FY = 2, FZ = 3, MX = 4, MY = 5, MZ = 6, CSGX= 7, CSGY= 8 c CSGZ= 9, VFX =10, VFY =11, VFZ =12 c FLOW=19, HEAT=20, AMPS=21, FLUX=22, NPKE=23, NPDS=24 c CURT=25, VLTG=26 (missing entries are spares) c c output arguments: c forget (int,func,out) - status of constraint. c = 0 - no loading on this node for this DOF

c DOF c value c settings c settings

= 4 - this node has a loading for this (dp,ar(4),out) value(1-2) - (real,imag) values of present value(3-4) - (real,imag) values of previous

2.8.7 forput Subroutine (Storing a Nodal Load at a Node)

subroutine forput (node,idf,value) c *** primary function: store a nodal load at a node c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c idf (int,sc,in) - reference number for the DOF: (1-32) c FX = 1, FY = 2, FZ = 3, MX = 4, MY = 5, MZ = 6, CSGX= 7, CSGY= 8 c CSGZ= 9, VFX =10, VFY =11, VFZ =12 c FLOW=19, HEAT=20, AMPS=21, FLUX=22, NPKE=23, NPDS=24 c CURT=25, VLTG=26 (missing entries are spares) c c value (dp,ar(2),in) - (real,imag) values of force c c output arguments: c none. c

2.8.8 fordel Subroutine (Deleting a Nodal Load at a Node)

subroutine fordel (node,idf) c *** primary function: delete a nodal load at a node c *** secondary functions: none. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c idf (int,sc,in) - reference number for the DOF: (1-32) c FX = 1, FY = 2, FZ = 3, MX = 4, MY = 5, MZ = 6, CSGX= 7, CSGY= 8 c CSGZ= 9, VFX =10, VFY =11, VFZ =12 c FLOW=19, HEAT=20, AMPS=21, FLUX=22, NPKE=23, NPDS=24

c c c c c

CURT=25, VLTG=26 output arguments: none.

(missing entries are spares)

2.8.9 ntpiqr Function (Getting Information About a Nodal Temperature)

c *** c c *** c c c c c c c c c c c c c c c c c c on c function ntpiqr (node,key) primary function: get information about a nodal temperature Notice - This file contains ANSYS Confidential information *** typ=int,dp,log,chr,dcp input arguments: variable (typ,siz,intent) node (int,sc,in) key (int,sc,in) siz=sc,ar(n),func intent=in,out,inout

description - node number should be zero for key=2 - key for operation = 1 - return temperature status ntpiqr = 0 - node has no temperature constraint defined = 1 - node has a temperature constraint defined = 2 - return total number of nodal temperatures defined in model - the returned value of ndinqr is based

output arguments: ndinqr (int,func,out)

2.8.10 ntpget Function (Getting a Specified Nodal Temperature)

function ntpget (node,temp) c *** primary function: get specified nodal temperature. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c c output arguments: c ntpget (int,func,out) - temperature status of node. c = 0 - node has no temperature constraint

c c

temp

(dp,ar(2),out)

= 1 - node has a temperature constraint - the node temperature (new,old).

2.8.11 ntpput Subroutine (Storing a Nodal Temperature)

subroutine ntpput (node,temp) c *** primary function: store nodal temperature. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c temp (dp ,sc,in) - nodal temperature c c output arguments: c none.

2.8.12 ntpdel Subroutine (Deleting a Nodal Temperature)

subroutine ntpdel (node) c *** primary function: delete node temperatures. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c c output arguments: c none.

2.8.13 nhgiqr Function (Getting Information About Nodal Heat Generations)

function nhgiqr (node,key) c *** primary function: get information about nodal heat generations c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number

c should be 0 for key=2 c key (int,sc,in) - key for operation c = 1 - return whether node has a heat generation rate c defined c nhgiqr = 0 - no heat generation defined for node c = 1 - heat generation is defined for node c = 2 - return total number of nodal heat generation c rates defined in model c c output arguments: c nhgiqr (int,func,out) - the returned value of nhgiqr is based on c

2.8.14 nhgget Function (Getting a Nodal Heat Generation)

function nhgget (node,hg) c *** primary function: get specified nodal heat generation. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c c output arguments: c nhgget (int,func,out) - heat generation status of node. c = 0 - nodal heat generation undefined c = 1 - nodal heat generation is defined c hg (dp,ar(2),out) - the nodal heat generation (new,old).

2.8.15 nhgput Subroutine (Storing Nodal Heat Generation)

subroutine nhgput (node,hg) c *** primary function: store nodal heat generation. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c hg (dp ,sc,in) - nodal heat generation c c output arguments: none.

2.8.16 nhgdel Subroutine (Deleting a Nodal Heat Generation)

subroutine nhgdel (node) c *** primary function: delete nodal heat generations. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c c output arguments: none.

2.8.17 nfuiqr Function (Getting Information About Nodal Fluences)

function nfuiqr (node,key) c *** primary function: get information about nodal fluences c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c should be zero for key=2 c key (int,sc,in) - key for operation c = 1 - return status: c nfuiqr = 0 - node does not have a fluence constraint c = 1 - node has a fluence constraint c = 2 - return total number of nodal fluences defined on c model c c c output arguments: c nfuiqr (int,func,out) - the returned value of nfuiqr is based on c

2.8.18 nfuget Function (Getting a Nodal Fluence)

function nfuget (node,fluen) c *** primary function: get specified nodal fluence. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c c output arguments: c nfuget (int,func,out) - fluence status of node. c = 0 - node has no fluence constraint c = 1 - node has a fluence constaint c fluen (dp ,ar(2),out) - the nodal fluences (new,old).

2.8.19 nfuput Subroutine (Storing a Nodal Fluence)

subroutine nfuput (node,fluen) c *** primary function: store nodal fluence. c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c fluen (dp ,sc,in) - nodal fluence c c output arguments: none.

2.8.20 nfudel Subroutine (Deleting a Nodal Fluence)

subroutine nfudel (node) c *** primary function: delete node fluences. c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number c c output arguments: c none. c

2.8.21 ndciqr Function (Getting Information About Nodal Current Densities)

function ndciqr (node,key) c *** primary function: get information about nodel current densities c *** c c c c c c c c c c node c c c c c c c on c Notice - This file contains ANSYS Confidential information *** typ=int,dp,log,chr,dcp input arguments: variable (typ,siz,intent) node (int,sc,in) key siz=sc,ar(n),func intent=in,out,inout

description - node number should be zero for key=2 (int,sc,in) - key for operation = 1 - return nodal current status: ndciqr = 0 - no current density defined for this = 1 - node has a current density defined = 2 - total number of nodal current densities defined on model

output arguments: ndciqr (int,func,out)

- the returned value of ndciqr is based

Chapter 3: Subroutines for Users' Convenience

Go to the Next Appendix Go to the Previous Chapter Go to the Table of Contents for This Manual. Chapter 1 * Chapter 2 * Chapter 3 * Appendix A

3.1 What Subroutines Does This Chapter Describe?

This chapter describes ANSYS routines available to you for use in programming. Using these routines isn't required, but may make your life easier. These routines

include a set of general routines that perform utility-type functions, a set of routines supporting vector functions, a set of routines supporting matrix functions, and routines supporting message processing options.

3.1.1 Input and Output Abbreviations

The descriptions of inputs and outputs for the routines discussed in this chapter use the following abbreviations:

Argument type is one of the following: int - integer dp - double precision log - logical chr - character dcp - double precision complex

Argument size is one of the following: sc - scalar variable ar(n) - array variable of length n func - functional return value

Argument intent is one of the following: in - input argument out - output argument inout - both an input and an output argument

3.2 General Subroutines

3.2.1 dptoch Subroutine (Retrieve Eight Characters From a Double Precision Variable)
subroutine dptoch (dp8,ch8) c *** primary function: retreive 8 characters from a dp variable c *** Notice - This file contains ANSYS Confidential information *** c c c input arguments: dp8 (dp,sc,in) output arguments: - dp variable containing characters

ch8

(ch*8,sc,out)

- characters retreived from the dp word

3.2.2 wrinqr Function (Obtain Information About Output)

function wrinqr (key) c *** primary function: obtain information about output c c *** Notice - This file contains ANSYS Confidential information *** c c --- caution: the following variables are "saved/resumed". c --key=WR_COLINTER thru WR_SUPCOLMAX in "wrinqr/wrinfo" c --(data for "/fmt,/page,/header" commands). c --note that the whole common cannot be "saved/resumed". cwa c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description wrcom name c key (int,sc,in) c = WR_PRINT - print flag (kprint) prtkey c wrinqr = 0 - no output c = 1 - print c = WR_OUTPUT - current output unit number (iott) outfil c = WR_MASTEROUT - master output file frstot c = WR_COLINTER - interactive columns per page intcol c = WR_COLBATCH - batch columns per page batcol c = WR_LINEINTER - interactive lines per page intlin c = WR_LINEBATCH - batch lines per page batlin c = WR_CHARITEM - characters per output item chrper c = WR_CHARDECIMAL - characters past decimal chrdec c = WR_CHARINTEGER - characters in leading integer chrint c = WR_CHARTYPE chrtyp c wrinqr = 1 - using E format in output c = 2 - using F format in output c = 3 - using G format in output c = WR_SUPTITLE - tlabel supress key keyhed c = WR_SUPSUBTITLE - subtitle supress key keytit c = WR_SUPLSITER - ls,iter id supress key keyid c = WR_NOTELINE - note line supress key keynot

c = WR_SUPCOLHEADER - column header supress key keylab c = WR_SUPCOLMAX - column maximum supress key keysum c c output arguments: c wrinqr (int,func,out) - the value corresponding to key

3.2.3 erinqr Subroutine (Obtaining Information from the Errors Common)

function erinqr (key) c *** primary function: obtain information from errors common c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c key (int,sc,in) - item to be returned c 1=keyerr, 2=errfil, 3=numnot, 4=numwrn, c 5=numerr, 6=numfat, 7=maxmsg, 8=lvlerr c 9=mxpcmd,10=nercmd,11=nertim,12=nomore c 13=eropen, 14=ikserr, 15=kystat, 16=mxr4r5 c 17=mshkey, 19=opterr, 20=flowrn c 21=errhpi c c ---- below definitions copied from errcom 7/92 for user information c c *** key number= .......................... c (see ansysdef for parameter definitions) | c \/ c co keyerr - master error flag (ER_ERRORFLAG) co errfil - errors file unit number (ER_ERRORFILE) co numnot - total number of notes displayed (ER_NUMNOTE) co numwrn - total number of warnings displayed (ER_NUMWARNING) co numerr - total number of errors displayed (ER_NUMERROR) co numfat - total number of fatals displayed (ER_NUMFATAL) co maxmsg - max allowed number of displayed messages before abort(ER_MAXMESSAGE) co lvlerr - used basicly in solution (from cnvr command.) (ER_ERRORLEVEL) co -1=do not set keyerr for notes/errors/warnings. co -2=same as -1 but do not display message either. co mxpcmd - maximum number of messages allowed per command (ER_MAXCOMMAND) co nercmd - number of messages displayed for any one command (ER_NUMCOMMAND) co nertim - key as to how message cleared from u/i pop-up (ER_UICLEAR) co (as per rsg/pft 5/1/92 - only for "info" calls co -1=message is timed before removal

co 0=message needs pick or keyboard before removal co 1=message stays up untill replaced by another message co nomore display any more messages (ER_NOMOREMSG) co 0=display messages co 1=display discontinue message and stop displaying co eropen - 0=errors file is closed (ER_FILEOPEN) co 1=errors file is opened co ikserr - 0=if interactive do not set keyerr (ER_INTERERROR) c - 1=if interactive set keyerr (used by mesher and tessalation) co kystat - flag to bypass keyopt tests in the elcxx routines (ER_KEYOPTTEST) c associated with status/panel info inquiries. c 0=do not bypass keyopt tests c 1=perform all keyopt tests co mxr4r5 - mixed rev4-rev5 input logic (*do,*if,*go,*if-go) (ER_MIXEDREV) c (used in chkmix called from rdmac) c 1=rev5 found (*do,*fi-then-*endif) c 2=rev4 found (*go,:xxx,*if,....,:xxx) c 3=warning printed. do not issue any more. co mshkey - cpu intensive meshing etc. this will cause (ER_MESHING) c "nertim (11)" to be set to -1 for "notes", 1 for "warnings", c and 0 for "errors". checking of this key is done in "anserr". c 0=not meshing or cpu intensive c 1=yes, meshing or cpu intensive co syerro - systop error code. read by anserr if set. (18) co opterr - 0=no error in main ansys during opt looping (ER_OPTLOOPING) c 1=an error has happened in main ansys during opt looping co flowrn - flag used by "floqa" as to list floqa.ans (20) c 0=list "floqa.ans" c 1="floqa.ans" has been listed. do not list again. co espare - spare integer variables c c --- end of information from errcom c c output arguments: c erinqr (int,sc,out) - value corresponding to key c #include "impcom.inc" c integer key, erinqr

3.2.4 TrackBegin Subroutine (Beginning Tracking for a Subroutine Call)

subroutine TrackBegin (sub16) c *****function: mark beginning of track ansys call c *** Notice - This file contains ANSYS Confidential information ***

c c c c

input arguments: sub16 (char*(*),sc,in) output arguments: none

- name of subroutine being entered and left (16 characters max)

3.2.5 TrackEnd Subroutine (Ending Tracking for a Subroutine Call)

subroutine TrackEnd (sub16) c *****function: mark end of track ansys call c *** Notice - This file contains ANSYS Confidential information *** c c c c input arguments: subname (char*(*),sc,in) output arguments: none - name of subroutine being left (16 characters max)

3.2.6 erhandler Subroutine (Displaying ANSYS Errors)

subroutine erhandler (filein,msgid,msglvl,messg,dperr,cherr) c primary function: Display ANSYS error messages

c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c c c c c c c c c c c c c c typ=int,dp,log,chr,dcp input arguments: variable (typ,siz,intent) filein (ch*16,sc,in) siz=sc,ar(n) intent=in,out,inout

msgid

(int,sc,in)

msglvl

(int,sc,in)

messg dperr

(ch*256,sc,in) (dp,ar(*),in)

description - Filename used for character portion of message ID (this is the file name of the file which contains the source for this routine) - Numeric portion of the message ID 1 - 9999, unique for each erhandler call in the FILE. Recommend using a sequence, similar to format conventions, i.e., 3010, 3020, 3030 - level of error (same as anserr) 0=no label (used for u/i pop-ups) -1=no label (used for u/i pop-ups) timed as a note message 1=note, 2=warning, 3=error, 4=fatal - error message to display. use keywords of %i %g %c %/ for formating (same as anserr) - vector of data to display. contains both integer and double precision data.

c c cherr c characters c

(same as anserr) (ch*(*),ar(*),in) - vector of character data to display max length of character data is 32 vector element. (same as anserr)

3.2.7 intrp Subroutine (Doing Single Interpolation)

subroutine intrp (klog,kppx,kstpz,xval,ax,ay,yval,nmax,kyoff) c *** primary function: **** subroutine for single interpolation **** c (if double interpolation is needed, see intrpt) c c *** Notice - This file contains ANSYS Confidential information *** c c typ=int,dp,log,chr,dcp siz=sc,ar(n),func intent=in,out,inout c c input arguments: c variable (typ,siz,intent) description c klog (int,sc,in) - interpolation type c = 0 - use linear interpolation c = 1 - use log interpolation c kppx (int,sc,in) - X value end of table signal c = 0 - a repeated x-value will signal the end c of the table c = 1 - a repeated x-value will not signal the c end of the table c (only known use = c evaluation) c kstpz (int,sc,in) - Y value end of table signal c = 0 - a yval of zero will not signal the end c of the table (e.g. stress fitting) c = 1 - a yval of zero will signal the end of c the table (in general, material c properties (exception: alpx)) c c NOTE: the end of the table will be signaled thru c either of the above conditions, or more c commonly, that nmax values have been processed, c or that the present x table entry is less than c the previous one (ax(i) .lt. ax(i-1)). c evaluations done after the end of the table are c evaluated as if they were at the end of the c table. similarly, evaluations done before the c beginning of the table are done as if they were c done at the beginning of the table. c

c xval (dp,sc,in) table c ax (dp,ar(*),in) c ay (dp,ar(*),in) c nmax (int,sc,in) c c output arguments: c yval (dp,sc,out) table c kyoff (int,sc,out) c c c

- value of x with which to go into the - table of x values, in ascending order - table of y values - maximum table size allowed - value of y which comes back from the - xval status flag = 0 - xval in x range = 1 - xval less than minimum x = 2 - xval greater than maximum x

3.2.8 tranx3 Subroutine (Processing Geometry for 3-D Line Elements)

subroutine tranx3 (nnod,xyz,nx,tr) c *** primary function: geometric processor for 3-d line elements c with or without a 3rd node c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c nnod (int,sc,in) - number of nodes (2 or 3) c xyz (dp,ar(nx,*),in) - coordinates (x,y,z down) c nx (int,sc,in) - row dimension of xyz array c c output arguments: c tr (dp,ar(3,3),in) - transformation matrix c

3.2.9 systop Subroutine (Stopping an ANSYS Program Run)

subroutine systop (icode) c *** primary function: stop an ansys run c *** secondary functions: pass an error code to the system c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c icode (int,sc,in) - stop error code (0<icode<127) c 0 - normal exit c 1 - stack overflow error c 2 - stack level overflow c 3 - stack pop below zero c 4 - names do not match in stkpxp c 5 - command line argument error c 6 - accounting file error c 7 - auth file verification error c 8 - indicated error or end-of-run c 11 - error in user routine c 12 - macro stop command

c c c c c c c c c c c c c

untrapped xox error anserr fatal error possible full disk possible corrupted or missing file Error in VM routines (corrupt db?) unauthorized code section entered unable to open x11 server quit signal failure to get signal in max time (syhold) >32 - system dependent error output arguments: none

14 15 16 17 18 21 25 30 31

3.3 Vector Functions

3.3.1 vdot Function (Computing the Dot Product of Two Vectors)
function vdot (v1,v2,n) c *** primary function: compute dot product of vectors v1 and v2 c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: v1 (dp,ar(n),in) v2 (dp,ar(n),in) n (int,sc,in) output arguments: vdot (dp,sc,out) - vector v1 - vector v2 - length of vectors v1 and v2 - dot product of v1 and v2

3.3.2 vsum Function (Summing Vector Components)

function vsum (va,n) c *** primary function: sum the components of a vector c *** Notice - This file contains ANSYS Confidential information *** c c c c c input arguments: va (dp,ar(n),in) n (int,sc,in) output arguments: vsum (dp,sc,out) - vector va - length of vector va - vector sum

3.3.3 vmax Function (Retrieving the Maximum Vector Value at a Given Location)

function vmax (v,n,locmax) c *** primary function: return the max value and location in a vector c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: v (dp,ar(n),in) n (int,sc,in) output arguments: locmax (int,sc,out) vmax (dp,sc,out) - vector v - length of vector v - location of max value in vector v - max value in vector v

3.3.4 lastv Function (Retrieving the Position of the Last Nonzero Term in a Double Precision Vector)
function lastv (v,n) c *** primary function: find position of last non-zero term in a d.p. vector c c input arguments: c v (dp,ar(n),in) - vector V c n (int,sc,in) - the number of items in the vector C c output arguments: lastv (dp,sc,out) - the position of the last non-zero term

3.3.5 izero Function (Setting an Integer Vector to Zero)

function izero (ivect,n) c *** primary function: set an integer vector to zero c c input arguments: c v (int,ar(n),in) - vector V c n (int,sc,in) - the number of items in the vector

3.3.6 imove Function (Assigning Equal Values to Two Integer Vectors)

function imove (i1,i2,n) c *** primary function: move one vector to another vector c c input arguments: c i1 (int,ar(*),in) - vector i1 c n (int,sc,in) - length of vectors v1, v2 C c output arguments: v2 (int,ar(*),out) - vector v2

3.3.7 vzero Subroutine (Initializing a Vector to Zero)

subroutine vzero (v,n) c *** primary function: initialize a vector to zero c *** Notice - This file contains ANSYS Confidential information *** c c c c c input arguments: v (dp,ar(n),inout) n (int,sc,in) output arguments: v (dp,ar(n),inout) - vector to zero out - number of words to zero out - zeroed vector

3.3.8 vmove Subroutine (Moving One Vector into Another)

subroutine vmove (v1,v2,n) c *** primary function: move a vector into another vector c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c v1 (dp,ar(*),in) - vector v1 c n (int,sc,in) - length of vectors v1, v2 c c output arguments: c v2 (dp,ar(*),out) - vector v2

3.3.9 vimove Subroutine (Moving One Vector into Another Incrementally)

subroutine vimove (v1,inc1,v2,inc2,n) c *** primary function: move one vector into another c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c v1 (dp,ar(inc1,n),in) - vector v1 c inc1 (int,sc,in) - increment on vector v1 c inc2 (int,sc,in) - increment on vector v2 c n (int,sc,in) - number of items to be moved c c output arguments: c v2 (dp,ar(inc2,n),in) - vector v2

3.3.10 vinit Subroutine (Assigning a Scalar Constant to a Vector)

subroutine vinit (v,n,const) c *** primary function: initialize a vector to a constant c *** Notice - This file contains ANSYS Confidential information ***

c c c c c

input arguments: n (int,sc,in) const (dp,sc,in) output arguments: v (dp,ar(n),out)

- length of vector v - constant to set vector v to - vector v

3.3.11 viinit Subroutine (Assigning a Scalar Constant to a Vector Incrementally)

subroutine viinit (v,inc,n,const) c *** primary function: set the components of vector v to const by increments c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c inc (int,sc,in) - increment (first dimension) of vector v c n (int,sc,in) - length (second dimension) of vector v c const (dp,sc,in) - constant to set components of vector v to c c output arguments: c v (dp,ar(inc,n),out) - vector v

3.3.12 vapb Subroutine (Setting a Vector to Sum of Two Vectors)

subroutine vapb (a,b,c,n) c *** primary function: add vector a to vector b to get vector c c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: a (dp,ar(n),in) b (dp,ar(n),in) n (int,sc,in) output arguments: c (dp,ar(n),out) - a vector - b vector - length of vectors a,b,c - c vector

3.3.13 vapb1 Subroutine (Combining Two Vectors in One)

subroutine vapb1 (a,b,n) c *** primary function: add vector a to vector b and store in vector a c *** Notice - This file contains ANSYS Confidential information *** c c c input arguments: a (dp,ar(n),inout) b (dp,ar(n),in) - a vector - b vector

c c c

(int,sc,in)

- length of vectors a,b - a vector

output arguments: a (dp,ar(n),inout)

3.3.14 vapcb1 Subroutine (Multiplying a Vector to a Constant)

subroutine vapcb1 (a,b,n,const) c *** primary function: multiply vector b to constant, add to vector a, c and store in vector a c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c input arguments: a (dp,ar(n),inout) b (dp,ar(n),in) n (int,sc,in) const (dp,sc,in) output arguments: a (dp,ar(n),inout) - vector a - vector b - length of vectors a,b - constant to multiply b - vector a

3.3.15 vamb Subroutine (Gets a Third Vector by Subtracting One Vector from Another)
subroutine vamb (a,b,c,n) c *** primary function: subtract vector b from vector a to get vector c c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: a (dp,ar(n),in) b (dp,ar(n),in) n (int,sc,in) output arguments: c (dp,ar(n),out) - vector a - vector b - length of vectors a,b,c - vector c

3.3.16 vamb1 Subroutine (Subtracting One Vector from Another)

subroutine vamb1 (a,b,n) c *** primary function: subtract vector b from vector a and save in vector a c *** Notice - This file contains ANSYS Confidential information *** c c c input arguments: a (dp,ar(n),inout) b (dp,ar(n),in) - a vector - b vector

c c c

(int,sc,in)

- length of vectors a,b - a vector

output arguments: a (dp,ar(n),inout)

3.3.17 vmult Subroutine (Multiplying a Vector by a Constant)

subroutine vmult (v1,v2,n,const) c *** primary function: multiply a vector by a constant c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: v1 (dp,ar(n),in) n (int,sc,in) const (dp,sc,in) output arguments: v2 (dp,ar(n),out) - vector v1 - length of vectors v1, v2 - constant to multiply v1 - vector v2

3.3.18 vmult1 Subroutine (Multiplying a Vector by a Constant)

subroutine vmult1 (v1,n,const) c *** primary function: multiply a vector by a constant c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: v1 (dp,ar(n),inout) n (int,sc,in) const (dp,sc,in) output arguments: v1 (dp,ar(n),inout) - vector v1 - length of vector n - constant to multiply v1 - vector v1

3.3.19 vcross Subroutine (Defining a Vector via a Cross Product)

subroutine vcross (a,b,c) c primary function: calculate c = a x b c *** Notice - This file contains ANSYS Confidential information *** c c c typ=int,dp,log,chr,dcp input arguments: a (dp,ar(3),in) siz=sc,ar(n) intent=in,out,inout

- first vector to be cross-multiplied

c c c

(dp,ar(3),in)

- second vector to be cross-multiplied - resulting vector

output arguments: c (dp,ar(3),out)

3.3.20 vnorme Subroutine (Normalizing a Three-Component Vector)

subroutine vnorme (iel,v) c primary function: normalize a vector to unit length c this routine is to be called only from the elements. it is only c for a three component vector(i.e. processing geometry). c this routine also differs from vnorm in that an error message is called c if the vector length is zero. c *** Notice - This file contains ANSYS Confidential information *** c c c c c c input arguments: iel (int,sc,inout) v (dp,ar(3),inout) output arguments: iel (int,sc,inout) v (dp,ar(3),inout) - element number - vector to be normalized - if 0, vector has zero length - normalized vector

3.3.21 vnorm Subroutine (Normalizing a Vector to Unit Length)

subroutine vnorm (v,n) c *** primary function: normalize a vector to unit length c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c v (dp,ar(n),inout) - vector v c n (int,sc,inout) - dimension length of vector v c c output arguments: c v (dp,ar(n),inout) - normalized vector v c n (int,sc,inout) - n = 0 if error in operation c

3.3.22 ndgxyz Function (Getting the X,Y,Z Vector for a Node)

function ndgxyz (node,xyz) c *** primary function: get x,y,z vector for a node. c *** secondary functions: set current node pointer to this node. c

c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c node (int,sc,in) - node number for operation. c c output arguments: c variable (typ,siz,intent) description c ndgxyz (int,sc,out) - status of node. c 0=node is undefined. c -1=node is unselected. c 1=node is selected. c xyz (dp,ar(3),out) - vector containing x,y,z

3.3.23 ndpxyz Subroutine (Storing X,Y,Z for a Node)

subroutine ndpxyz (node,xyz) c *** primary function: store x,y,z vector for a node. c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c node (int,sc,in) - node number for operation. c xyz (dp,ar(3),in) - vector containing x,y,z c (vector should be in global system) c c output arguments: c none.

3.4 Matrix Subroutines

3.4.1 maxv Subroutine (Multiplying a Vector by a Matrix)
subroutine maxv (a,v,w, nr,nc) c *** primary function: multiply a vector by a matrix c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c input arguments: a (dp,ar(nr,*),in) v (dp,ar(*),in) nr (int,sc,in) nc (int,sc,in) output arguments: w (dp,ar(*),out) matrix vector number number a v of rows in matrix a of columns to multiply in matrix a

- product vector w

3.4.2 maxv1 Subroutine (Multiplying a Vector by a Matrix)

subroutine maxv1 (a,v, nr,nc) c *** primary function: multiply a vector by a matrix

c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c input arguments: a (dp,ar(nr,nc),in) - matrix a v (dp,ar(nc),inout) - vector v nr (int,sc,in) - number of rows in matrix a *** nr limited to 60 *** nc (int,sc,in) - number of columns to multiply in matrix a output arguments: v (dp,ar(nr),inout) - product, stored in vector v

3.4.3 matxv Subroutine (Multiplying a Vector by a Full Transposed Matrix)

subroutine matxv (a,v,w, nr,nc) c *** primary function: multiply vector by full transposed matrix c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c c c c c c c input arguments: a (dp,ar(nr,*),in) v (dp,ar(nv),in) nr (int,sc,in) - matrix a (first dimension must = nr) - vector v (nv must be greater or equal to nr) - first dimension and number of active rows of the untransposed matrix a (also the number of active rows of vector v) - number of columns of the untransposed matrix a (also the number of computed items in the product vector w) if negative, accumulate - product vector w

nc

(int,sc,in)

output arguments: w (dp,ar(na,*),out)

3.4.4 matxv1 Subroutine (Multiplying a Vector by a Full Transposed Matrix)

subroutine matxv1 (a,v, nr,nc) c *** primary function: multiply vector by full transposed matrix c *** Notice - This file contains ANSYS Confidential information *** c input arguments: c a (dp,ar(nr,*),in) c v (dp,ar(nr),inout) c nr (int,sc,in) c nc (int,sc,in) transposed) matrix vector number number a v of rows in matrix (un-transposed) of columns in matrix (un-

c c c output arguments: v (dp,ar(nc),inout)

*** nc limited to 60 *** - product, stored in vector v

3.4.5 matxb Subroutine (Transposing a matrix)

subroutine matxb (a,b,c, na,nb,nc, n1,n2,n3) c *** primary function: (a)t * (b) = (c) t means transpose c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c c c c c c input arguments: a (dp,ar(na,*),in) b (dp,ar(nb,*),in) na (int,sc,in) nb (int,sc,in) nc (int,sc,in) n1 (int,sc,in) n2 (int,sc,in) n3 (int,sc,in) matrix a matrix b number of rows in matrix a number of rows in matrix b number of rows in matrix c number of rows in matrix c to fill number of columns in matrix c to fill number of columns in matrix a and number of rows of matrix b to work with (the two need to be the same for the inner product)

output arguments: c (dp,ar(nc,*),out)

- product matrix c

3.4.6 maat Subroutine (Changing a Matrix Value via Addition, Multiplication, and Transposition)
subroutine maat(a,c, nc,n, con) c primary function: does con*a*at and sums the result onto c (a is a vector) c *** Notice - This file contains ANSYS Confidential information *** c c c c c c c c c typ=int,dp,log,chr,dcp input arguments: a (dp,ar(*),in) c nc n con siz=sc,ar(n) intent=in,out,inout

- vector to be multiplied by itself to generate an nxn square matrix (a by a-transposed) (dp,ar(nc,*),inout) - matrix to be accumulated onto (int,sc,in) - number of rows in the c matrix (int,sc,in) - size of square matrix (dp,sc,in) - multiplier on above square matrix

c output arguments: c c (dp,ar(nc,*),inout) - matrix to be accumulated onto c only the lower triangular matrix is done c Note: this routine is usually followed by matsym,

to do the complete matrix

3.4.7 matsym Subroutine (Filling the Lower Triangle from the Upper Triangle)
subroutine matsym (a,nd,n) c primary function: fill upper triangle from lower triangle c *** Notice - This file contains ANSYS Confidential information *** c c c c c c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout its lower triangular part upper triangular part of the a matrix to be processed

input arguments: a (dp,ar(nd,*),inout) - matrix to have copied to its nd (int,sc,in) - number of rows n (int,sc,in) - size of matrix

c output arguments: c a (dp,ar(nd,*),inout) - matrix that has its lower triangular part c copied to its upper triangular part

3.4.8 mctac Subroutine (Transposing a symmetric matrix)

c subroutine mctac (a,na,c,nc,nold,nnew) **** function: do a = c(transpose) * a * c , where a is symmetric **

c *** Notice - This file contains ANSYS Confidential information *** c input arguments: c a (dp,ar(na,na),inout) c c c na (int,sc,in) c c (dp,ar(nc,nnew),in) c nnew)) c nc (int,sc,in) c nold (int,sc,in) c 64 c nnew (int,sc,in) c size). c 64 matrix to be pre and post multiplied (part operated on must be square(nold x nold) and symmetric) first dimension of the a matrix matrix to pre and post multiply a by (part used may be rectangular(nold x first dimension of the c matrix size of part of 'A' matrix that is to be processed(input size). maximum = size of part of 'A' matrix that results from this operation(output maximum =

c output arguments: c a (dp,ar(na,na),inout) resulting matrix c (still square(nnew x nnew) and symmetric).

3.4.9 tran Subroutine (Transposing a matrix)

subroutine tran (zs,tr,nz,ntr,nrow,irot) c primary function: perform tr-transpose * zs * tr ************ c c *** Notice - This file contains ANSYS Confidential information *** c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout

c input arguments: c variable (typ,siz,intent) description c zs (dp,ar(nz,nz),inout) - matrix to be transformed c tr (dp,ar(ntr,ntr),in) - transformation matrix c nz (int,sc,in) - dimensioned size of zs matrix c ntr (int,sc,in) - dimensioned size of tr matrix c nrow (int,sc,in) - number of rows of zs matrix to transform c irot (int,sc,in) - block size to transform(size of tr matrix) c c c output arguments: variable (typ,siz,intent) description zs (dp,ar(nz,nz),inout) - transformed matrix

3.4.10 symeqn Subroutine (Solving Simultaneous Linear Equations)

subroutine symeqn (a,nd,n,nc) c primary function: solve a set of simultaneous linear equations c secondary functions: invert a matrix c this routine assumes that the matrix to be solved or c inverted is positive definite c *** Notice - This file contains ANSYS Confidential information *** c typ=int,dp,log,chr,dcp siz=sc,ar(n) intent=in,out,inout

c input arguments: c variable (typ,siz,intent) c a (dp,ar(nd,*),inout) c c c nd (int,sc,in) c n (int,sc,in) c nc (int,sc,in) c c c to c c c c

description - matrix to be solved or inverted second dimension must be at least: n + abs(nc) first dimension of the a matrix number of equations number of additional columns. if nc = +n or -n, invert n x n matrix and put result in the n+1 to 2xn columns. if nc is 0 or negative, nc will be reset n and then symeqn will set up identity matrix after the input matrix, where the result of the inversion will be put. if nc is positive and less than n, do a

partial inversion. see example 1 below.

c output arguments: c variable (typ,siz,intent) description c a (dp,ar(nd,*),inout) - results or inverted matrix. c starts in column n+1. c note: original information is destroyed. c c example 1: Solve three simultaneous linear equations: c call symeqn (a(1,1),3,3,1) c calling routine has a dimensioned as a(3,4) c each equation has its 3 coefficents in the first 3 columns, c and the constant term is in the fourth column. c solution is in fourth column. c c example 2: Invert a 3x3 matrix: c call symeqn (a(1,1),3,3,-3) c calling routine has a dimensioned as a(3,6) c input matrix was input in first 3 columns c output matrix in ouput in last 3 columns

3.5 Message Processing Routines

3.5.1 msgdi Subroutine (Conditionally Printing Messages, Scalars, Vectors, or Matrices)
subroutine msgdi (msg,i) c *** primary function: write text & integer to output if 10th debug key c is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c i (int,sc,in) - integer to print c c output arguments: c none

3.5.2 msgdic Subroutine (Conditionally Printing Messages and an Integer Vector)

subroutine msgdic (msg,k,n,inc) c *** primary function: write text & integer vector (with incrementing) to c output if 10th debug key is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c k (int,ar,in) - integer vector to print c n (int,sc,in) - size of vector c inc (int,sc,in) - increment c c output arguments: c none

3.5.3 msgdim Subroutine (Conditionally Printing Messages and Matrices)

subroutine msgdim (msg,k,n,m) c *** primary function: write text & integer matrix to debug file if 10th debug c key is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print

c c c c c c

k n m

(int,ar,in) (int,sc,in) (int,sc,in)

- integer matrix to print - row size of vector - column size of vector

output arguments: none

3.5.4 msgdir Subroutine (Conditionally Printing Messages, Integers, and Reals)

subroutine msgdir (msg,i,r) c *** primary function: write text, integer & real to output if 10th debug key c is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c i (int,sc,in) - integer to print c r (dp,sc,in) - real to print c c output arguments: c none

3.5.5 msgdit Subroutine (Conditionally Printing Messages and a Transposed Integer Matrix)
subroutine msgdit (msg,k,n,m) c *** primary function: write text & transposed integer matrix to output if c 10th debug key is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3

c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c k (int,ar,in) - integer matrix to print c n (int,sc,in) - row size of vector c m (int,sc,in) - column size of vector c c output arguments: c none

3.5.6 msgdiv Subroutine (Conditionally Printing Messages and an Integer Vector)

subroutine msgdiv (msg,k,n) c *** primary function: write text & integer vector to output if 10th debug key c is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c k (int,ar,in) - integer vector to print c n (int,sc,in) - size of vector c c output arguments: c none

3.5.7 msgdr Subroutine (Conditionally Printing Messages and Reals)

subroutine msgdr (msg,r) c *** primary function: write text & real to output if 10th debug key c is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program

c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c r (dp,sc,in) - real to print c c output arguments: c none

3.5.8 msgdrc Subroutine (Conditionally Printing Messages and Real Vectors)

subroutine msgdrc (msg,r,n,inc) c *** primary function: write text & real vector (with increment) to output if c 10th debug key is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c r (dp,ar,in) - real vector to print c n (int,sc,in) - size of vector c inc (int,sc,in) - increment c c output arguments: c none

3.5.9 msgdrm Subroutine (Conditionally Printing Messages and Real Matrices)

subroutine msgdrm (msg,r,n,m)

c *** primary function: write text & real matrix to output if 10th debug key c is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c r (dp,ar,in) - real matrix to print c n (int,sc,in) - row size of vector c m (int,sc,in) - column size of vector c c output arguments: c none

3.5.10 msgdrt Subroutine (Conditionally Printing Messages and Real Transposed Matrices)
subroutine msgdrt (msg,r,n,m) c *** primary function: write text & real transposed matrix to output if 10th c debug key is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c r (dp,ar,in) - real matrix to print c n (int,sc,in) - row size of vector c m (int,sc,in) - column size of vector c c output arguments: c none

3.5.11 msgdrv Subroutine (Conditionally Prints Messages and Real Vectors)

subroutine msgdrv (msg,r,n) c *** primary function: write text & real vector to output if 10th debug key c is greater than or equal to 3 c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c r (dp,ar,in) - real vector to print c n (int,sc,in) - size of vector c c output arguments: c none

3.5.12 msgdt Subroutine (Conditionally Prints Message Text)

subroutine msgdt (msg) c *** primary function: write text to output if 10th debug key c is greater than or equal to 3 c c *** secondary function: set debug level c c At source development, you may wish conditionally to print-out messages, c scalars, vectors or matrices to monitor the code. Once the program c runs OK, there is no need for these print-outs. This feature can be achieved c with the programmers debug utilities. c c To invoke this utility insert the following line in ANSYS input deck: c c /DEBUG,,, ,,, ,,, ,3 c c *** Notice - This file contains ANSYS Confidential information *** c c input arguments: c variable (typ,siz,intent) description c msg (ch*256,sc,in) - message to print c

c c

output arguments: none

Go to the beginning of this chapter