4.

Product Data
4.1 Geometrical Data

Requirements and Methods to Manage

Geometrical Data for Engineering
Applications
 Overview
• Geometric Modeling
– Applications
– Historical roots of Technical Modeling
• Overview of Geometrical Models
– Criteria
– Wire-frame Models
– CSG
– Voxel/Octrees
– Triangle Meshes (Polygon Meshes)
– B-Rep
• The Boundary Representation-Model
– Basics
– Primitives and Basic Data Structures
– Modeling Kernels and File Formats
– B-REP Data in Databases

Schallehn: Data Management for Engineering Applications

 Historical Roots of Technical Modeling
• Historical roots date back more than 2000 years
– Sketches and informal drawings used in ancient Egypt and Greece (Euclid, 300BC) to
medieval times
– Move from agricultural to industrial age increased importance of sharing information
for technical development and documentation
– Around the 19th century patents (protection of intellectual property) required
formalization of technical
• Manual Technical Drawing on paper standard way for technical modeling until the
1980s
– Formalized process with commonly used conventions for representing 3D geometries
represents “visual language”
– Projection methods (orthogonal, parallel, perspective) to map 3D geometries to 2D
– Data representing concrete measures with special syntax as dimension values or
parameters and legends
• First Computer Aided Design (CAD) developed in the 1960’s
– Became industrial practice in the 1980s
– Required digital representation of geometries

Schallehn: Data Management for Engineering Applications

 Historical Technical Drawings

Technical drawing describing details of Drawing of a US patent (cigarette

a helicopter by Leonardo da Vinci rolling machine)from 1881

[Source: wikipedia.org]
Schallehn: Data Management for Engineering Applications
 Technical Drawings

[Source: wikipedia.org]
Schallehn: Data Management for Engineering Applications
 CAD Systems

[Source: wikipedia.org]
Schallehn: Data Management for Engineering Applications
 Geometric Modeling
Geometric modeling refers to methods and data structures
suitable to represent the shape and topology of geometric
objects as data for computer applications.

• In Engineering also referred to as Solid Modeling highlighting

physical properties of solid objects
• Geometric model has central importance for product lifecycle
– Based on specifications and requirements from
– Result of initial design steps and step-wise refinements
– Input for further steps in product development like FEA, simulation,
mockups, production planning, manufacturing, etc.

Schallehn: Data Management for Engineering Applications

 Computer Graphics vs. CAD
• Computer Graphics
– General term for methods to create images from data
– Comprises geometric modeling methods (representing geometry) + rendering (creating
image)
– Geometric modeling focuses on efficiency of computations
– Two main branches
• Real-time rendering for fast graphics generation in interactive applications (games, virtual
worlds, CAD, etc.)
• Photo-realistic rendering for application requiring high (realistic) image quality (CGI in movies,
computer arts, etc.)

• Computer Aided Design

– Focus on formal representation of geometrical data
– Geometric models focus on expressiveness, completeness and correctness of geometry
– Uses methods from computer graphics (real-time rendering) for interaction
• CAD model needs to be mapped to rendering model

Schallehn: Data Management for Engineering Applications

 Geometry vs. Topology
• Terms for different aspects of representing local and global
properties of objects

Geometry: describes local features (dimensions, relations

between dimensions, primitive type, etc.) of each element of
an object.

Topology: describes transformations (position in space via

translation, rotation, etc.) of elements and how they are
connected to form complex shapes.

Schallehn: Data Management for Engineering Applications

 Classification of Geometric Modeling
• Over time and for different applications various
methods were developed, which differ regarding key
criteria

– Supported Dimensionality
– Supported Primitives (Geometry)
– Supported Construction (Topology)
– Supported Level of Detail/Approximation
– Intended Applications

Schallehn: Data Management for Engineering Applications

 Criteria: Dimensionality
• 2D
– Derived from paper 2D drafting
– Adapted in early CAD systems, today not often
practiced
– Commonly used in some applications
(electronic circuit design, architecture)
• 3D
– Constructions represented as 3D shapes
– Current standard in CAD

• 2½D
– Data represented as 2D + “3D interpretation”
– Sweeping methods for interpretaion, e.g.
• Extrusion (along a trajectory)
• Rotation

Schallehn: Data Management for Engineering Applications

 Criteria: Primitives
• Points and curves
– Points, lines, line loops
– Freeform curves like Splines and
NURBS
• Polygons
– Triangles of special importance as
most simple face type (used to
construct or approximate any
polygon or surface)
• 3D objects
– Basic 3D objects like prisms, spheres,
cuboids, etc.
• Space partitions
– Part of space according to a 3D grid
partitioning the room

Schallehn: Data Management for Engineering Applications

 Criteria: Construction Methods
• Basic topology
– Transformations: translation,
rotation, scaling, etc.
– Connection of vertexes, edges,
faces

• Set operations on primitives

– Union or intersection (symmetric)
– Relative complement/subtraction
(asymmetric)

• Sweeping
– 2½D construction methods like
extrusion, rotation etc.

Schallehn: Data Management for Engineering Applications

Criteria: Approximation/Level of Detail
• Exact Geometry
– Often can be represented applying
freeform curves (2D) and freeform
surfaces (3D), e.g. using NURBS
– Support for intuitive design

• Approximated Geometry
– Often implemented in terms of
tessellation: representing a complex
surface with simple polygons (e.g.
triangle meshes)
– Level of detail can be set according to
requirements
– More efficient for rendering
[Source: wikipedia.org]
Schallehn: Data Management for Engineering Applications
 Intended Applications
• Interactive work on geometries
– Requires intuitive methods and expressiveness
– Examples:
• CAD (development)
• Game content development
• CGI/special FX development
• Rendering
– Requires simple structures and efficient algorithms
– Examples
• Games
• CAD (interaction, display)
• Interactive virtual worlds
• Capturing real-word geometries
– As objects semantics are unknown, low-level representation required
– Examples
• 3D scanners or printers
• X-ray computed tomography (CT)
• Motion Capturing

Schallehn: Data Management for Engineering Applications

 Overview of Modeling Methods
• Wire-frame Models
• Constructive Solid Geometry (CSG)
• Voxel/Octrees
• Triangle Meshes (Polygon Meshes)
• B-Rep

Schallehn: Data Management for Engineering Applications

 Wire-Frame Models

Schallehn: Data Management for Engineering Applications

 Wire-Frame Characteristics
Supported Dimensionality • 2D
• 2½D
• 3D
Supported Primitives • Lines
• Curves (splines, ellipses etc.)
Construction Methods • Edges of physical objects represented by
lines and curves
Level of Detail • Precise description of edges possible
• Information about surfaces and volumes lost
Intended Application • Interactive Modeling: early CAD, simple
visualizations
• Rendering: early computer graphics

Schallehn: Data Management for Engineering Applications

 Wire-Frame Usage
• Rendering of 3D models applies
– Projection methods (perspective, orthogonal, parallel)
– Hidden edges can be removed, colored or dashed
(requires information about surfaces)
• Disadvantages
– Semantic loss (information about surfaces and volumes)
– Leads to ambiguities in interpretation

Schallehn: Data Management for Engineering Applications

 Constructive Solid Geometry (CSG)

Schallehn: Data Management for Engineering Applications [Source: wikipedia.org]

 CSG Characteristics
Supported Dimensionality • 3D
Supported Primitives • Parametrizable 3D basic shapes: cuboids,
spheres, prisms, etc.
• In some approaches: parametrizable free-form
shapes
Construction Methods • Boolean set operations
• Basic transformations: translation, rotation
Level of Detail • Many geometries can represented correctly
• If no free-form shapes are supported, according
geometries have to be approximated
Intended Application • Interactive Modeling: CAD, modeling for games
etc.

Schallehn: Data Management for Engineering Applications

 CSG Usage
• Rich semantics allow intuitive creation of 3D models
• Disadvantages
– Rendering and some verifications/evaluation of geometric models
require complex computations
– Construction of a shape by Boolean operators is ambiguous
– Some geometries (e.g. free-form surfaces such as used for automotive
design etc.) had to be approximated with basic primitives in early
approaches
• CSG combined with other methods (e.g. B-Rep) to support
interactive modeling
• Rarely used as stand-alone geometric model in CAD

Schallehn: Data Management for Engineering Applications

 Voxel-based Models

[Source: Tobias Wüstefeld, http://www.bilderzucht.de]

Schallehn: Data Management for Engineering Applications

 Voxel Characteristics
Supported Dimensionality • 3D
Supported Primitives • Volume elements = Voxel = “3D pixel”
• Space partitioned according to grid
Construction Methods -
Level of Detail • Approximation according to grid properties
(resolution)
Intended Application • Capturing or creating real-world geometries: 3D
scanners, 3D printers

Schallehn: Data Management for Engineering Applications

 Octrees
• Problem of voxel-based 3D data:
huge amounts of data for
reasonable resolution
• Could be solved by compression
techniques
• Alternative: Octrees
– Partition space hierarchically
– Starts with 8 sub-cubes (8=oct)
– Only cubes which are not completely
filled or completely empty are
furthermore subdivided recursively
– Creates (unbalanced tree)
– 3D equivalent to 2D- Quadtree
Schallehn: Data Management for Engineering Applications [Source: wikipedia.org]
 Triangle/Polygon Meshes

[Source: wikipedia.org]

Schallehn: Data Management for Engineering Applications

 Polygon Mesh Characteristics
Supported Dimensionality • 2D (less commonly used)
• 3D
Supported Primitives • Triangles
• Other polygons (quads, arbitrary)
Construction Methods • Basic topology of vertexes of polygons
• Combinations like triangle strips and triangle fans
for easier definition
Level of Detail • Requires approximation of all curved surfaces
(e.g. sphere, free-form) and edges
• Level of detail/approximation can be controlled
by number of triangles/polygons
Intended Application • Real-time rendering: fast computation of graphics
for interactive applications

Schallehn: Data Management for Engineering Applications

 Triangle Mesh Usage
• 3D real-time rendering
– Based on geometrical projection methods
– Addition of color, texture, lighting, etc.
• Approximation can be controlled to be below level of
perception → higher computation effort
• Specialized hardware (Graphic Processing Units) work on
triangle meshes
• More complex models like B-Rep, CSG, etc. are mapped to
triangle meshes for rendering/visualization

Schallehn: Data Management for Engineering Applications

 Polygon Meshes
• In general, any polygon can be
used to describe/approximate
surfaces
• Higher number of vertexes
– Allows easier modeling
– Introduces more complex
computations
– verification

[Source: wikipedia.org]

Schallehn: Data Management for Engineering Applications

 Boundary Representation (B-Rep)

Schallehn: Data Management for Engineering Applications

 B-Rep Characteristics
Supported Dimensionality • 3D
Supported Primitives • Surfaces modeled by
- Vertexes
- Edges as lines or diverse curves
- Faces as arbitrary planar polygons or
free-form surfaces
Construction Methods • Basic topology polygons (as for polygon meshes)
• Basic transformations: translation, rotation
• Boolean set operations (as in CSG)
• Sweeping to create 3D geometries from 2D
shapes
Level of Detail • Precise description of geometries possible
• Approximations possible for lower level of detail
Intended Application • Interactive work on geometries: CAD etc.

Schallehn: Data Management for Engineering Applications

 History of B-Rep
• Developed in the early 1970s
– Ian Braid developed basic concepts and first prototype of a modeling kernel
(ROMULUS) from CAD perspective
– Bruce Baumgart developed basic data structures and algorithms from a
computer graphics perspective
• ROMULUS became blueprint for current modeling kernels
– Parasolid
– ACIS
• B-Rep was extended over the years
– Free-form curves and surfaces
– Set operations (as in CSG)
– Sweeping
• Because of rich semantics and intuitive modeling capabilities became de
facto standard for CAD

Schallehn: Data Management for Engineering Applications

 B-Rep Basic Topology
• Hierarchical definition of
– Vertexes defined based on points (geometry)
– Edges defined based on vertexes of lines or curves
– Loops defined as sequence of closed edges
– Faces defined by loops
– Shell defined by enclosing faces, defines body

Schallehn: Data Management for Engineering Applications

 B-Rep Topology vs. Geometry

From [1]

Schallehn: Data Management for Engineering Applications

 B-Rep Data Structures
• Geometry and topological relationships can be
stored in simple lists
• More efficient algorithms (validation, rendering, etc.)
possible for advanced structures with some
redundancies → Winged Edge (also half edge)

Schallehn: Data Management for Engineering Applications

 B-Rep Simple Data Structures

Schallehn: Data Management for Engineering Applications

 Data Structure: Winged Edge /1
• Winged Edge typical data structure to represent polygon
networks defined by vertexes, edges, faces
• Allows fast traversal of surface by keeping (redundant)
connections of each edge to vertexes and faces, e.g.

Edge is stored in EdgeList as:

Identifier:
Defining vertexes:
Defined faces:
Neighbor edges:

Schallehn: Data Management for Engineering Applications

 Data Structure: Winged Edge /2
• Stores for each edge class WE_Edge {
– Defining vertexes WE_Vertex vert1, vert2;
– Defined faces WE_Face aFace, bFace;
WE_Edge aPrev, aNext, bPrev, bNext;
– Neighbor edges
// clockwise ordering
– Actual data (e.g. curve WE_EdgeDataObject data;
function) }

• Stores for each vertex class WE_Vertex {

List<WE_Edge> edges;
– Defined edges
WE_VertexDataObject data;
– Actual data (e.g. }
coordinates)
class WE_Face {
• Stores for each face List<WE_Edge> edges;
– Defining edges WE_FaceDataObject data;
}
– Actual data
[Source: Winged Edge at wikipedia.org]
Schallehn: Data Management for Engineering Applications
 Validation of B-Rep Models
• Construction methods alone do not guarantee a valid geometry
• Most common problem: is the described 3D shape closed by defining
surfaces?
• Simple geometries such as polyhedrons (plane edges and surfaces) can be
checked e.g. by Euler-Poincaré formula:

V–E+F =2*(S–R)+H (with holes)

V–E+F =2 (without holes)

with V = number of vertexes, E = number of edges, F = number of faces,

S = number of shells, R = number of rings (holes in body), H = number of
holes (in faces)
• Curves and curved surfaces may require more complex checks

Schallehn: Data Management for Engineering Applications

 B-Rep: Sweeping Methods
• Extrusion: extension of a 2D shape along an arbitrary vector
or depth orthogonal to face definition
• Protruding: extending an existing 3D geometry by extrusion
of a marked surface region
• Revolve (Rotation): space covered by revolving a 2D shape
around a specified rotation axis
• Blending: room covered by transition of one 2D geometry to
another
• General Sweeping: may extend 2D geometry along any
arbitrary curve (free-form)

Schallehn: Data Management for Engineering Applications

 CSG in B-Rep
• Set operations as defined in CSG can be used on any set of
shells created by basic topology or sweeping
• Alternative methods can be used to achieve same design

From [1]

Schallehn: Data Management for Engineering Applications

 B-Rep Modeling Kernels
• Early implementations: BUILD and Romulus
• Currently two dominating

ACIS (Alan, Charles, Ian's System)

• Developed by Spatial Corp. (part of Dassault) since 1986
• Used in, e.g., AutoCAD (derived kernel), CATIA (up to V5), SolidEdge
(before V5)
• Convergence Geometric Modeler (CGM) developed by same company for
recent versions of CATIA

Parasolid
• Now owned by Siemens PLM
• Used in, e.g. , SolidEdge (since V5), SolidWorks, Siemens NX

Schallehn: Data Management for Engineering Applications

ACIS Data
Model
From [1]

Schallehn: Data Management for Engineering Applications

 ACIS File Formats
• ACIS defines two file formats
– .sab Binary File format
– .sat Text File format
• Binary file contains identical information as text, but is more
compact
• Contain:
– 3 line header
– Core data according to defined entity types
– Optional: update history
• Supported as exchange formats for many CAD tools (even
with different kernels)

Schallehn: Data Management for Engineering Applications

 ACIS .sat File Example
700 0 1 0
9 Cobalt v4 16 ACIS 7.0 Unknown 24 Thu Jul 18 22:12:55 2002
25.39999999999999858 9.999999999999999547e-07 1.000000000000000036e-10
body $1 -1 $-1 $2 $-1 $3 #
rgb_color-st-attrib $-1 -1 $4 $-1 $0 0 1 0 #
lump $-1 -1 $-1 $-1 $5 $0 #
transform $-1 -1 1 0 0 0 1 0 0 0 1 0 0 0 0.03937007874015747977 no_rotate
no_reflect no_shear #
id_attribute-st-attrib $-1 -1 $-1 $1 $0 3 #
shell $-1 -1 $-1 $-1 $-1 $6 $-1 $2 #
face $7 -1 $-1 $8 $9 $5 $-1 $10 reversed single #
…

[Source: Paul Bourke

http://paulbourke.net/dataformats/]
Schallehn: Data Management for Engineering Applications
 Parasolid
• Similar to ACIS, defines file formats
– .x_t (also .xt) for text data
– .x_b (also .xb) for binary data
• “~45% CAD data worldwide is Parasolid format”
[John Juckes: XT B-Rep; Making it Real]

Schallehn: Data Management for Engineering Applications

 Parasolid .x_t File Example
**ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz***********
*
**PARASOLID !"#$%&'()*+,-./:;<=>?@[\]^_`{|}~0123456789************
**PART1;MC=osf65;MC_MODEL=alpha;MC_ID=sdlosf6;OS=OSF1;OS_RELEASE=
V4.0;FRU=sdl_parasolid_test_osf64;APPL=unknown;SITE=sdl-cambridgeu.
k.;USER=davidj;FORMAT=text;GUISE=transmit;DATE=29-mar-2000;
**PART2;SCH=SCH_1200000_12006;USFLD_SIZE=0;
**PART3;
**END_OF_HEADER***************************************************

T51 : TRANSMIT FILE created by modeller version 120000017 SCH_1200000_120060

12 1 12 0 2 0 0 0 0 1e3 1e-8 0 0 0 1 0 3 1 3 4 5 0 6 7 0 body
70 2 0 1 0 0 4 1 20 8 8 8 1 T list
13 3 3 0 1 0 9 0 0 6 9 shell
50 4 11 0 9 0 0 0 +0 0 0 0 0 1 1 0 0 plane
31 5 10 0 7 0 0 0 +0 0 0 0 0 1 1 0 0 1 circle
19 6 5 0 1 0 0 3 V region
16 7 6 0 ?10 0 0 5 0 0 1 edge
. . .

[Siemens PLM: Parasolid XT Format Reference]

Schallehn: Data Management for Engineering Applications

 STEP
• STandard for the Exchange of Product model data
• Developed since 1984 by international consortium
• Standardized since 1990s as ISO 10303
• Contains
– General methods for describing data and schemas
– Definitions of generic file formats
– Application-specific methods for engineering domains

Schallehn: Data Management for Engineering Applications

 STEP Parts relevant for Geometric Modeling

• Parts most relevant for Geometric Models:

– 10303-1x Description Methods, e.g.
• 10303-11 EXPRESS and EXPRESS-G
– 10303-2x Implementation Methods, e.g.
• 10303-21 STEP files
• 10303-23 C++ Language Binding
• 10303-28 STEP XML
– Further 10303-XX Integrated generic resources
• 10303-42 Geometric and topological representation
• 10303-52 Mesh-based topology
– 10303-2XX Application Protocols
• 10303-203 Configuration controlled 3D designs of mechanical
parts and assemblies
• 10303-214 Core data for automotive mechanical design
processes (mostly superset of 203)
Schallehn: Data Management for Engineering Applications
 AP214 EXPRESS-G Schema (Excerpt)

[Source: wikistep.org]

Schallehn: Data Management for Engineering Applications

 AP 214 EXPRESS Schema (Excerpt)
(* SCHEMA geometry_schema; *)

ENTITY cartesian_point
SUPERTYPE OF (ONEOF(cylindrical_point, polar_point, spherical_point))
SUBTYPE OF (point);
coordinates : LIST [1:3] OF length_measure;
END_ENTITY;

[Source: steptools.com]

Schallehn: Data Management for Engineering Applications

 Example AP214 .TSEP File
ISO-10303-21;
HEADER;
FILE_DESCRIPTION( ( '' ), ' ' );
FILE_NAME( 'pumpHousing.stp', '2004-04-13T21:07:11', ( 'Tim Olson' ), ( 'CADSoft Solutions
Inc' ), ' ', 'ACIS 12.0', ' ' );
FILE_SCHEMA( ( 'automotive_design' ) );
ENDSEC;
DATA;
. . .
#3716 = POINT_STYLE( ' ', #6060, POSITIVE_LENGTH_MEASURE( 1.00000000000000E-06 ), #6061 );
#3717 = CARTESIAN_POINT( '', ( -1.10591425372267, 3.05319777988191, 0.541338582677165 ) );
#3718 = CURVE_STYLE( '', #6062, POSITIVE_LENGTH_MEASURE( 1.00000000000000E-06 ), #6063 );
#3719 = LINE( '', #6064, #6065 );
#3720 = CURVE_STYLE( '', #6066, POSITIVE_LENGTH_MEASURE( 1.00000000000000E-06 ), #6067 );
#3721 = CIRCLE( '', #6068, 1.75849340964528 );
#3722 = CURVE_STYLE( '', #6069, POSITIVE_LENGTH_MEASURE( 1.00000000000000E-06 ), #6070 );
#3723 = CIRCLE( '', #6071, 0.540114611464642 );
#3724 = SURFACE_STYLE_USAGE( .BOTH., #6072 );
#3725 = FACE_OUTER_BOUND( '', #6073, .T. );
. . .
ENDSEC;
END-ISO-10303-21;

[Source: Paul Bourke

Schallehn: Data Management for Engineering Applications
http://paulbourke.net/dataformats/]
 IGES
• Initial Graphics Exchange Specification
• Created in the early 1980s by National Bureau of
Standards (American government organization for
standardization)
• Supported by many CAD tools as exchange format
• Numeric encoding of entity types inspired by “punch
cards” → despite text format, hardly human
readable

Schallehn: Data Management for Engineering Applications

 IGES Example
IGES Start section created by Claris Graphics Translator v1.0 S 1
1H,,1H;,12Hbox+line.IGS,12Hbox+line.IGS,8HCGT v1.0,8HIGES 3.0,16,8,24,8,G 1
24,12Hbox+line.IGS,1.0,1,4HINCH,32000,1.6,13H910305.183439,,,3HIBB,3HWCPG 2
,4,0; G 3
124 1 1 0 0 0 0 0 0 0 0 0D 1
124 0 0 1 0 0 0 D 2
124 2 1 0 0 0 0 0 0 1 1 0D 3
124 0 0 1 0 0 0 D 4
410 3 1 1 0 0 3 0 0 2 2 0D 5
410 0 0 1 0 0 0 D 6
124 4 1 0 0 0 0 0 0 0 0 0D 7
124 0 0 1 0 0 0 D 8
124 5 1 0 0 0 0 0 0 1 1 0D 9
…

[Source: Paul Bourke

http://paulbourke.net/dataformats/]

Schallehn: Data Management for Engineering Applications

 Further Important Formats
• DXF – Textual AutoCAD exchange format
• DWG – Binary AutoCAD format
• VRML (Virtual Reality Markup Language)
• X3D – XML-based follow-up to VRML
• 3DXML – Exchange Format of CATIA (set of zipped XML files)
• STereoLithography (STL)
• Collada – XML-based language for Automation including geometrical data
• JT – Jupiter Tesselation (Siemens PLM – possible follow up to XT)
• ...
• ...
• ...
• Many document (e.g. PDF) and pixel-based graphics formats (TIFF, PNG,
etc.) supported as export formats

Schallehn: Data Management for Engineering Applications

 X3D Example
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE X3D PUBLIC "ISO//Web3D//DTD X3D 3.2//EN"
"http://www.web3d.org/specifications/x3d-3.2.dtd">

<X3D profile="Interchange" version="3.2"

xmlns:xsd="http://www.w3.org/2001/XMLSchema-instance"
xsd:noNamespaceSchemaLocation="http://www.web3d.org/specifications/x3d-
3.2.xsd">
<Scene>
<Shape>
<IndexedFaceSet coordIndex="0 1 2">
<Coordinate point="0 0 0 1 0 0 0.5 1 0"/>
</IndexedFaceSet>
</Shape>
</Scene>
</X3D>

Schallehn: Data Management for Engineering Applications

 3DXML Example (snippet)
. . .
<Faces>
<Face strips="4 2 14 22,48 23 50 26,49 25 39 20,7 19 3 1,21 0 24
18,29 35 32 38,27 40 34 45,33 44 36 46,37 47 30
42,31 43 28 41,15 11 5 9 8 10 17 12 15 11">

<SurfaceAttributes>
<Color xsi:type="RGBAColorType" red="1" green="1" blue="1"
alpha="1"/>
</SurfaceAttributes>
</Face>
</Faces>
<Edges>
<LineAttributes lineType="SOLID" thickness="2">
<Color xsi:type="RGBAColorType" red="0" green="0" blue="0"
alpha="1"/>
</LineAttributes>
<Polyline vertices="0 0 0,360 0 0"/>
<Polyline vertices="0 500 0,360 500 0"/>
. . .
</Edges>
. . .

Schallehn: Data Management for Engineering Applications

 STL Example
solid
:
:
facet normal 0.0 0.0 1.0
outer loop
vertex 1.0 1.0 0.0
vertex -1.0 1.0 0.0
vertex 0.0 -1.0 0.0
endloop
endfacet
:
:
endsolid

[Source: Paul Bourke

http://paulbourke.net/dataformats/]

Schallehn: Data Management for Engineering Applications

 Geometry Data in Databases
• Motivation and open Problems
• CAD data and Relational Databases (RDBMS)
– CAD metadata
– CAD data as Binary Large Objects (BLOBs)
– CAD data in Database File Systems
– CAD data as structured data
• CAD data and advanced Database Concepts
– Object-relational Database (ORDBMS)
– Object-oriented Databases (ODBMS)
– XML Databases
– Cloud Databases
– Specialized DBMS
– Spatial data and Geographic Information Systems
• STEP SDAI
Schallehn: Data Management for Engineering Applications
 Motivation
• General idea: use advantages of DBMS
– Efficient access to huge data volumes
– Multi-user support
– Controlled consistency
• RDBMS are
– Not often used for geometry data
– Commonly used for PDM/PLM and other engineering applications
• ORDBMS and ODBMS theoretically provide suitable features,
but have other disadvantages
• Geometry still most often stored in files due to the following
problems →

Schallehn: Data Management for Engineering Applications

 Open Problems /1
• Data too complex for RDBMS and ORDBMS
– Many primitives, operations, etc. require many database types (or tables)
– Topological relationships require following connections between many
entities (objects, tuples) for even simple geometries
– Following references implemented in terms of expensive JOIN (combining
data from different tables) and SELF-JOIN (combining data within tables)
operations → bad performance
– ORDBMS allow richer modeling constructs and use of references, but data is
most often still stored in a fragmented way → bad performance
• Lack of standardization for ORDBMS and ODBMS
– SQL:1999 and SQL:2003 describe object-relational standard, but are
implemented by existing DBMS in varying ways (unlike the relational core
SQL-92)
– ODMG standard for object-oriented DBMS hardly implemented at all
– Limited portability and reusability of developed solutions

Schallehn: Data Management for Engineering Applications

 Open Problems /2
• Lack of acceptance for ORDBMS and ODBMS
– Object-relational concepts supported by many existing DBMS, but not always
used (due to problems with portability, technological complexity, etc.)
– ODBMS niche market with small companies: unstable product support
• Data exchange problematic
– Exchange requires neutral physical representation
– If DBMS is used, data has to be exported to exchange format anyway
– Requires application-specific export filters database-to-file
• Archiving CAD data problematic
– Archived data needs to be interpretable even after decades
– DBMS of limited use for long-time archiving due to proprietary physical data
structures
• …

Schallehn: Data Management for Engineering Applications

 Storing CAD Data using RDBMS
• 4 possible alternatives

1. Store only metadata in DBMS

2. Store CAD data in dedicated format as binary large
objects (BLOB)
3. Store CAD data (as BLOB) in database file system
4. Store CAD data as structured data in RDBMS

Schallehn: Data Management for Engineering Applications

 CAD Metadata in RDBMS
• Data about CAD data typically has simpler structures
– About design process: creator, dates, status, versions, etc.
– Relation of partial design (part) to overall product within product
structure
– References to other engineering data
• Suitable for storage in RDBMS
• Typically task of Product Lifecycle Management (PLM) system
• CAD systems often
– tightly integrated with PLM or
– offer own PLM-functionality (e.g. CATIA un V5 and V6 integrates
ENOVIA)

Schallehn: Data Management for Engineering Applications

 Example: Metadata in CATIA V5
[Dassault Systems:
CATIA Version 5
Documentation]

Schallehn: Data Management for Engineering Applications

 CAD Data as Binary Large Objects (BLOB)
• Since SQL:1999 DBMS support arbitrary binary data
• Often used to store files in tables of database (e.g. from PLM system)
• Semantics of BLOB
– unknown to DBMS → no functionality except for reading and writing as one
value
– Typically conforms to proprietary or standard format of used CAD system

CREATE TABLE construction_part (

part_id INT PRIMARY KEY,
name VARCHAR(100),
responsible INT FOREIGN KEY
REFERENCES engineer(eng_id),
dxf_file BLOB,
. . .
);

Schallehn: Data Management for Engineering Applications

 Database File Systems
• DBMS provide storage facilities based on BLOBs that
externally can be used as any file system
– BLOB storage can be mounted as virtual file system
– (CAD) files stored in this file system are physically stored in and
controlled by the DBMS
– Allows access via file or database interfaces
• Advantages from CAD perspective
– Flatly structured metadata can be easily linked with complex CAD data
– Accesses and consistency to some degree controlled by DBMS
mechanisms
– Transparent integration with file based activities
– Advanced recovery mechanisms of DBMS can be used
– Similar functionality as network/distributed file systems

Schallehn: Data Management for Engineering Applications

 Example: Oracle SecureFiles
[Oracle® Database SecureFiles and
Large Objects Developer's Guide
11g Release 2]

Schallehn: Data Management for Engineering Applications

 CAD Data as Relational Data
• Theoretically possible to create tables from types
defined in
– Modeling kernels
– STEP standard
• Implemented in several research prototypes and few
commercial systems
• No common practice due to the disadvantages
mentioned before
– Poor performance due to complex data
– Problems with archiving and exchange

Schallehn: Data Management for Engineering Applications

 CAD Data as Relational Data

From [5]

Schallehn: Data Management for Engineering Applications

 CAD Data with ORDBMS
• Rich semantic modeling of potential benefit
– Type system and specialization (inheritance) allows 1:1
implementation of CAD schemas (e.g. modeling kernel or
STEP AP214) in SQL
– References and nested tables (NF2 = Non-First-Normal-
Form) allow creation of complex types/objects
• No common practice due to previously mentioned
disadvantages, mainly
– Still poor performance because of fragmented storage
– Lack of acceptance

Schallehn: Data Management for Engineering Applications

 Example: Winged Edge in SQL:2003 (excerpt)
. . .
CREATE TYPE we_vertex_type UNDER geometry_type (
edges REF(we_edge_type) MULTISET,
coordinates FLOAT ARRAY(3),
);

CREATE TYPE we_edge_type UNDER geometry_type (

vertex1 REF(we_vertex_type),
vertex2 REF(we_vertex_type),
aface REF(we_face_type),
bface REF(we_face_type),
neighbours REF(we_edge_type) ARRAY(4),
);

CREATE TABLE edge OF we_edge_type;

CREATE TABLE vertex OF we_vertex_type;
. . .

Schallehn: Data Management for Engineering Applications

 Example: NF2

From [5]
Schallehn: Data Management for Engineering Applications
 CAD Data with ODBMS
• Allow implementation of CAD schemas according to data models of C++,
Java, C#, etc. (see STEP SDAI ISO 10303-22 →)
• Objects of classes can be persisted with special mechanisms
– Schema consisting of persistence-capable classes can be created from source
code (using pre- or post-processor tools)
– Named objects (lookup of single object via unique name)
– Object networks with entry points (named root objects)
• Persistence by reachability: store objects along with objects “reachable” via
references (transitive closure)
– Collections (sets, lists, multi-sets) of objects
• May provide query interface
• Provide good performance and easy development → some CAD systems
(mainly in the 1990s) used ODBMS
• Today no common practice due to mentioned disadvantages, mainly lack
of acceptance and standardization

Schallehn: Data Management for Engineering Applications

 ODBMS Example: ODMG Java Binding
…

class WE_Edge {
WE_Vertex vert1, vert2;
WE_Face aFace, bFace;
WE_Edge aPrev, aNext, bPrev, bNext;
WE_EdgeDataObject data;

public static void main(String args[]) {

…
Database db = odmg.newDatabase();
…
WE_Edge e1 = new WE_Edge(…);
db. Bind(e1,”myEdge1”);
…
}
}

Schallehn: Data Management for Engineering Applications

 CAD and ODBMS: Architecture Example
[Kim, Han: Encapsulation of geometric
functions for ship structural CAD
using a STEP database as native storage.
Computer-Aided Design, 2003]

Schallehn: Data Management for Engineering Applications

 Special Functionality in ODBMS
• Because ODBMS rather popular in engineering some systems
implemented specific functionality, e.g.
– Long/design transactions: check out/check in mechanism
– Workspaces: store data of one user or group in separate location
during long transactions
– Support for versioning and variants on data model level: create
sequential (versions) and parallel (variants) manifestations of one
object
– Nested transactions: allow transaction within transaction to support
complex design process
– Database file systems: as in some RDBMS (↑)
– …

Schallehn: Data Management for Engineering Applications

 Further DBMS Types
XML DBMS
• Allow storage of XML data (documents, document collections), i.e. useful
complementary to XML CAD file formats
• Allow querying via specialized
• Similar problems regarding performance as ORDBMS

Cloud DBMS
• So-called no-SQL systems for simple storage of weakly structured data of
possibly huge amounts (keyword Big Data) on the Web/in the Cloud
• Requirements and usefulness for engineering applications is topic of
current research

Schallehn: Data Management for Engineering Applications

 Specialized DBMS for CAD Data
• No existing system easily fulfills all requirements
• Several attempts in industrial and academic research
to develop tailor-made DBMS
• Often based on concepts of object-oriented DBMS

Schallehn: Data Management for Engineering Applications

 Specialized DBMS Example: PRIMA

[Härder, Meyer-Wegener et al. PRIMA - a DBMS Prototype

Supporting Engineering Applications. VLDB 1987]
Schallehn: Data Management for Engineering Applications
 Spatial Data
• Related: geometrical data in Geographic Information
Systems (GIS) is strongly supported by DBMS
– Less complex than CAD data (few primitives, less flexible
topologies)
• Standardization from GIS and DBMS community
– Open Geospatial Consortium (OGC), e.g. Simple Feature
Model for 2D
– SQL/MM Spatial
• Numerous implementations, e.g. Oracle Spatial

Schallehn: Data Management for Engineering Applications

 Simple Feature Model (OGC)

[Brinkhoff: Geodatenbanksysteme in
Theorie und Praxis. Wichmann, 2013]
Schallehn: Data Management for Engineering Applications
 Example: Oracle Spatial

[Brinkhoff: Geodatenbanksysteme in
Theorie und Praxis. Wichmann, 2013]

Schallehn: Data Management for Engineering Applications

 STEP SDAI
• Standard Data Access Interface ISO 10303-22 defines
standard bindings to languages (C, C++, Java) for STEP data
access
• Similar to an API for an RDBMS (ODBC, JDBC) or ODBMS
defines basic functionality such as
– Sessions
– Database connectivity
– Data dictionary
• Defines mappings of EXPRESS types to language constructs, e.
• Not specific to geometrical data → used more often for other
applications

Schallehn: Data Management for Engineering Applications

 Literature / Further Readings
[1] Ali K. Kamrani, Emad Abouel Nasr: Engineering Design and
Rapid Prototyping. Springer, 2010.
ISBN 978-0-387-95862-0
[2] Ian Stroud: Boundary Representation
Modelling Techniques. Springer 2006.
ISBN-10: 0-387-84628-312-4
[3] M. M. M. Sarcar, K. Mallikarjuna Rao,K. Lalit Narayan :
Computer Aided Design and Manufacturing
ISBN 978-8-120-33342-0
[4] ACIS Documentation
http://doc.spatial.com/index.php/Main_Page
[5] A. Kemper, M. Wallrath: An Analysis of Geometric Modeling in
Database Systems. ACM Comput. Surv. 19(1): 47-91 (1987)

Schallehn: Data Management for Engineering Applications