!

" #$%% &

432

FEATURE-BASED GEOMETRIC DIMENSION AND TOLERANCE

MODELING SYSTEM

Prof. Dr. M.M. Koura* Prof. Dr. I.M. Elewa** Prof. Dr. R. Gadh***
Dr. B.S. Prabhu+ Dr. K.A. Mohamed~
* Design and Production Engineering Department, Ain Shams Univ., Egypt
** Production Engineering and Mech. Design Dept., Mansoura Univ., Egypt
*** Henry Samuel School of Engineering, UCLA, Los Angles, CA, USA
+ Mech. and Aerospace Engineering Dept., UCLA, Los Angles, CA, USA
~ Engineering Science Dept., Suez Canal University, Egypt

ABSTRACT. This paper reports a complete system for modeling, representation, and validation of
geometric tolerances, using feature-based approach from boundary representation (B-Rep) based solid
models. The utility of feature based representation scheme for representing Geometric Dimensioning and
Tolerancing (GDT) information in geometric models is highlighted. Explicit information from B-Rep is
culled using feature extraction and recognition algorithms such as graph matching and form/design
feature rules.
The major objective is to develop a scheme necessary for interfacing the tolerance module with the solid
model. This includes the development of a point-and-click graphic user interface that allows a user to
select geometric or functional entities (face, axis, surface, datum, etc.). It validates the entity selection
based on the context of the type of tolerance being defined, collects the controlled and controlling feature
and numerical value(s) for the tolerance in question. This work also proposes, an apposite tolerancing
model representation as part of the feature based geometric model that can be easily used by the downline
manufacturing activities.
The proposed schema provides built-in checking of validity of tolerances defined during modeling. Also,
a scheme for over-toleranced and over/under dimensioning validity checking is introduced.

KEYWORDS: Tolerancing, Dimensioning, Tolerance Representation, Solid Model, B-Rep, Feature

Extraction, Feature Recognition,

1 INTRODUCTION
Geometric tolerances state the maximum allowable variation of a form or its position from the perfect
geometry implied in the drawing [1]. There are both ISO and ANSI standards for geometric tolerances [
2, 3, 4]. These two standards are slowly reconciling their differences, but still some remaining. Both
standards were deigned for use with engineering drawings, not CAD models. Consequently the intention
of the standards is to create a common language, by use of symbols, conventions, and rules so that the
person interpreting the tolerance on an engineering drawing will follow the intentions of the designer.
Since the standards such as ISO or ANSI were designed for human use and interpretation, it was not
necessary to base them on a formal theory. Instead, the standards are based on design practice.
Difficulties in modelling computer understandable tolerance specification and mathematical validation of
tolerances has provided the motivation for finding a way or theory of tolerances, on which future
standards can be based [5].
Today, solid modelling systems are widely used for various kinds of applications in mechanical
engineering. Even though a geometric model created by a solid modelling system includes all the
information about the part shape [6]. They do not represent tolerances in a form suitable for automation
of many design and manufacturing activities, such as manufacturing, assembly planning, inspection and
tolerance analysis. This deficiency is causing concern in both the research and industrial communities, as
it is preventing the efficient automation of many activities.
 !" #$%% &
433

2 PREVIOUS WORK AND LITERATURE REVIEW

Owing to the importance of tolerance representation and analysis in the field of CAD/CAM system, a
growing number of research activities in these areas have been observed in recent years. Tolerance
specification falls into three areas viz., mathematical, representational, and computational. Mathematical
theories provide the basic concepts to view an imperfect or variational part. Representation schemes
attempt to document tolerance specification for storage in a computer database. Computational methods
have been devised to understand the design and computational consequences of tolerance specification
[6]. When one examines the specific approaches presented one finds that it is often difficult to separate
the three aspects.

2.1 MATHEMATICAL METHODS

Several researchers have tried to establish mathematical bases with the intention of unifying solid model
based representation of different tolerance types.
The tolerance zone was defined by Requicha [7, 8, 9,] as a region that must lie within the object’s
boundaries. The maximal and minimal object volumes are obtained by offsetting the object by equal
amount on each side of the nominal. Srinivasan and Jayaraman [10, 11] state that neither the current
standards, nor Requicha theory are adequate for describing allowable variations from the nominal, when
the functional constrains are considered. They build on Requicha’ theory, and introduce virtual
boundaries. Etesami [12, 13] models tolerances using a very similar approach to that of Requicha. He
called his method offset features boundary solids.
Hoffman [14] characterizes tolerances as a set of inequalities of the Form L ≤ f (x) ≤ U
Where x is a variational parameter and L and U are the lower and upper limits of the tolerance function f.
Turner and Wozny [15, 16] formalized the above-mentioned concepts by considering the geometric
variation as a vector. A set of independent geometric variations was selected as basis vectors to define s
vector space. The tolerances define inequality constrains in this space. Collectively, the tolerance
constraints define an in-tolerance region of the space [17, 18, 19].
Although mathematical theories of tolerances have been proposed with a view to unify their
representation in a computer-based model, such theories have abandoned important tolerance concepts
such as size, feature of size, Taylor’s principle, etc. Thus, the ANSI standard as it current stands defies
mathematical representation because it is not based on a mathematical model. Secondly, the use of such
theories and their possible acceptance depends on the development of efficient tolerance specification
and evaluation algorithms. Such algorithms do not presently exist and may be difficult to derive in future
as observed by many authors. Therefore, it may be difficult to implement tolerance representation system
based completely on mathematical tolerance theories, thus raising questions about their ultimate
acceptance by design and manufacturing engineers.

2.2 FRAMEWORK MODELS

Framework schemes may be dependent or independent on the data structure of the solid model. The
schemes reviewed here are: evaluated entity structure, Vgraph, variational constructive geometry, and
constraint-based face graphs.
In essence, the method amounts to define explicit features (as per Pratt and Wilson terminology [20]) and
then defining tolerances on them. Features were treated as a set of faces limited to planar, cylindrical,
conical and spherical types.
A multi-layer scheme was proposed and implemented by Ranyak and Wilson [21]. Tolerance features
were placed at the lowest level of the hierarchy and treated as unique because they were built from
primitives that can have only one tolerance value in each of the tolerance categories. Higher-level
features could be built from these to meet the needs of applications.
The specification of tolerances requires references to evaluated geometric entities. Requicha has devised
a scheme for overcoming this problem [22]. The method is based on a structure called “VGraph” and 2D
intersection set operators that allow one to reference only a portion of a face (VFace) by intersecting the
object with a virtual object. Attributes, such as tolerances, can be attached to VFace. A conceptual
framework for tolerance representation and analysis based on CSG is also presented in [23], somewhat
along the lines of Requicha, but without a mechanism to refer to partial faces.
Several researchers, proposed remarkably similar schemes for tolerancing that may be described as
constraint based face graphs. The researchers include Faux [24], Gossard, Zuffante, and Sakurai [25], and
 !" #$%% &
434
Roy and Liu [26]. Features are built from faces or sets of faces (single planar face, profile, pair of parallel
faces, etc.) structured in a binary tree where the interior nodes are Boolean set operators [27]. The
dimension relationships are represented by graph whose nodes represents the nominal faces of the part,
and the arcs represents the nominal distance between the two faces. Roy et al [28, 29] developed linking
data structure to hold the tolerance information through constraint nodes. The face list is built up as the
features are added to the initial evaluated B-Rep.
Clement et al and Irani et al arrived at a tolerance representation model that is compatible with the
standards and which seems to be theoretically and mathematically sound [30, 31].
Using the set of displacement theory by [32], Clement et al have proven that there are only seven
elementary surface types. On this basis, a computer system can automatically propose tolerance types
employing an assembly model (geometric tolerances). Shah presents a method for an efficient model for
representing the dimensions of a part [33]. The model is used to capture the designer’s GD& T scheme
on a feature based design model, and validate its completeness. Bernstein and Preiss represented
dimensions in a solid modeler by using a direct graph or constraint network [34]. Bernstein did not clarify
how tolerance arcs are represented in such scheme.
The tolerance representation framework is compatible with tolerance standards ANSI Y14.5M, but still
there are shortcomings. One of them is, they do not deal with the problem of determination of virtual
boundaries (VB) for the mechanical parts, their model contains data about material conditions but there is
not a method to get the VB. Also, these methods do not introduce a proper method for tolerance model
verification and validation.

2. 3 COMPUTATIONAL METHODS
Many researchers have adopted a vector-based description of tolerance. Mathematical models are used to
describe the specified geometry of a workpiece. Vector-based tolerancing utilizes 3D tolerances as an
integrated part of the Vectorial product model where the different geometrical features are treated
separately. This method supports automatic generation of NC-codes and inspection [35, 36, 37, 38].
Turner and Wozny propose a model they introduce three types of variables: design, model and tolerance.
From this point on, the problem of tolerancing is reduced to establish relations between tolerance and
design variables through the use of a vector space. Wirtz [39] adopted a completely different approach to
that to Turner [40, 41]. Wirtz defines the position, direction and dimension of an element or a surface by
vectors established with respect to a unique reference frame. Tolerance intervals are then associated with
each component of these vectors. His method has been the basis of a proposal for a new Swiss coordinate
measuring standard [42].
The Computational model treats tolerance as a vector that means the tolerance has magnitude and
direction and at the same time has a vector for a variation. But, this method also still suffers from the
ANSI Y14.5M non-compatibility standards.
Also, current representation schemes do not provide enough flexibility for inputting, validating, retrieving
and manipulating data required for CAD/CAM integration achieving computer integrated manufacturing
systems (CIM).
To overcome the deficiency of current tolerance representation models, a new scheme for tolerance
representation is needed. Hence the aim of this work is to develop a model that: -
• Is compatible with tolerance standards (ANSI Y14.5M)
• Is capable of computational ability for tolerance data handling.
• Combines framework tolerance representation and computational methods for tolerance
representation in a manner that eliminates the disadvantages described earlier.
• Includes a method for tolerance validation

3 FEATURE EXTRACTION AND RECOGNITION

Dimensioning and Tolerancing (D&T) representation is a feature-based representation; therefore it needs
proper geometric features identified from the solid modellers for its application. Two related types of
features need to be defined:
• Lower-level-features (such as points, lines, arcs, splines or surfaces)
• Higher-level features, which are the combination of low-level features (or the combination of other
higher-level features)
 !" #$%% &
435
Lower-level features have unique geometrical entities and are well defined, whereas higher-level features
are design specific and their choice depends on the function and the context of the application. Therefore,
for D&T implementation, information on these features must be extracted or recreated from the solid
model after the design process ends.
In Boundary Representation (B-rep) based solid model, each geometric entity of the object has its own
identity [43]. Therefore, it is possible to provide the same kind of geometric tolerancing information for
each of them. Some of the entities (i.e. median planes, axes, etc.) which are not really generated in the
boundary representation need to be incorporated in the B-rep data structure so that all kinds of geometric
as well as conventional tolerancing can be represented.
B-Rep suffers from the lack of information regarding the higher level features and their spatial
constraints. Feature extraction and recognition methods are used to determine this information.
B-Rep based feature recognition approach are either rule-based or graph-based [44]. In the rule-based
approach [45, 46], a set of rules is described to specify the topological and geometric information of
predefined features. However, it impossible to define all rules for all features, and new sets of rules are
required to define features with slight adjustments. In addition to this, the search of the solid model can
be highly complex when the parts are large. Gadh [47] have developed a technique to address this
problem, which he calls Differential Depth Filter. This filter reduces the search space to just the
silhouette edges obtained by looking the object from several viewpoints.
In graph-based approach, the face and edge adjacency relationships of a part are transformed into nodes
and arcs in a graph. Feature patterns are also defined by graphs. Recognition is done by searching the
graph of the part to find sub graphs matching one of the features pattern subgraph. The graph-matching
approach [48, 49] uses the graph nature of a B-Rep solid model.
Curvature Regions (CR) approach are presented for feature recognition in solid model [44, 50, 51].
Primitive features are found by converting B-Rep of the CAD model to a higher level of representation
called Curvature Region Representation (CR-Rep).
The graph-based recognition approach has an advantage over the others due to the graph nature of B-Rep
based solid models. The graph structure is most suitable for representing geometric and topological
information for the features. Each feature class has its own graph representation for efficient and direct
matching processes. So, a system that recognizes features automatically using Graph-Based Method will
be adopted in this paper.

3. 1 THE GRAPH-BASED METHOD

The graph representation is special type of representation that defines the shape or features using its
topological relationship. That includes the geometrical information (planes, lines and their points), and
the topological information between surfaces, edges and their vertices. The feature recognition system
comprises two major modules. The first one is for feature extraction (dealing with extracting faces from
the boundary models that are feature members). This module uses a loop-based method and a face-based
method. The second module is for feature recognition using graph-based feature recognition. In other
words, after the feature members are found by searching face edge graph, by, matching these feature
member sub-graphs with pre-defined sub-graph patterns it is possible to recognize the features.
Figure 1 illustrates a simple part with a feature and the procedure for feature extraction and recognition.
The edges are scanned to extract which faces constitute a feature. Each edge is common to two connected
groups of edges called loops. If all the edges in the loop are concave, the loop is categorized as being
concave. Loops may also be convex or hybrid. Then, the extracted sub-graphs are compared with pre-
defined sub-graph patterns of features, for recognition.. Figure 2 illustrates the module used for feature
extraction and recognition.

4 TOLERANCE MODEL
The proposed new model helps the designer to input and validate tolerance data, and to detect
over/under-toleranced features. It is also used to model tolerance data in such a way to be compatible
with tolerance standards ANSI Y14.5M. This representation aids tolerance analysis and synthesis.
Selecting suitable manufacturing process is possible by integrating separate process specific modules.
The model delivers inputs to inspection and quality control.
Figure 3 illustrates the relation between these models and how they interface with each other. From
Figure 3, the solid model of 3D parts is constructed and then the tolerance model mechanism manages the
 !" #$%% &
436
different tolerance models. First, the datum system and its validation is constructed. Then the tolerance
type is selected and the required data for the deference tolerance type (datums, features, faces, edges and
points) are obtained. The input tolerance is then validated with the selected entities, and the degree of
freedom (DOF) validation system is used to detect whether there is any over-tolerance. The tolerance
data is stored in records, using pre-defined templates for each tolerance type, which is called the tolerance
data structure (TDS). A model for manufacturing process selection is introduced based on a knowledge
base system. Rules for production process selection based on production cost and process capability are
presented.

F1

F2 F4

F3
Extracted Feature
X= Convex, V= Concave

-1- -2- -3- -4-

Fig. 1 Graph Based Extraction and Recognition

 !" #$%% &
437
Solid Model Data
File

Retrieve B- B-Rep Data Analysis

Rep Data (Topological Analysis)

Surfaces, Edges,
Feature Extraction
Vertices Information

Get Related
Surfaces
(RS)

I=0

I= I+1

System Knowledge-
Recognition Feature
Base
Yes

If RS=pre- No
If I <= No. RS
defined Feature

Yes

Getting Feature No

End Feature
Definition

Fig. 2 Feature Extraction and Recognition Module

 !" #$%% &
438

Solid Model

Tolerance Data Tolerance

Tolerance Model Structure Data
construction Model Records

Datum Reference
Tolerance
construction&
Validation Model
Validation Model

Manfacturing
Selection Model

Manufactring processes
Cabapility and Cost-
Tolerance Knowledge
Base

Fig. 3 Developed Tolerance models

5 TOLERANCE DATA STRUCTURE (TDS)

The current tolerance standard is ASME Y14.5M (ASME, 1994a). The shape and size of part features
can vary in several possible ways. The variations are classified into six standard tolerance classes: size,
form (flatness, straightness, circularity, cylindricity), orientation (parallelism, perpendicularity,
angularity), position (location, concentricity), run out (circular, total), and profile (line, surface).
All categories of tolerance specification have been studied to establish their data structures. Tolerance
specification involves specifying a datum reference frame (DRF), and the relationships among five kinds
of tolerance shape elements: vertex, edge, surface, axis and median plane. There are three kinds of
features to which a geometric characteristic is applicable:
1. Individual feature. A single surface, edge, or size feature that relates to a perfect geometric
counterpart of itself as the desired form; no datum is proper nor is one used.
2. Related feature. A single surface or element feature that relates to a datum, or datums, in form and
orientation; or a size feature (e.g. hole, slot, pin, shaft) that relates to a datum, or datums in form,
attitude (orientation), runout, or location.
3. Undetermined feature. A single surface or edge feature (whose perfect geometric profile is
described) that may, or may not be related to a datum, or datums.
The information content representing the GD& T and other accuracy data is analyzed according to ANSI
standard Y14.5M and presented in table 1.

6. DATUMS AND DATUM ALLOCATION

The ANSI standard (ANSI Y14.5M) refers to portions of planar surfaces as features. If a feature is used
as a reference of measurement, it is called a datum feature. When selecting datum, for validation this
selection, the system has the ability to answer the following types of questions [35]:
1. Does the selected geometric entity satisfy the conditions to be a datum (planar or cylindrical,
mutually orthogonal, etc.)?
 !" #$%% &
439
2. Is the entity a feature or feature of size?
3. Are any of the GD&T specified conflicting, or redundant?
The first and second questions are determined automatically from the 3D part database by selecting the
datum and then retrieving its data from the system database and comparing it with the required datum and
produce a message if the selected one is a valid datum or not. The third question is checked by validating
the selected datum and the relationship between them. Also there is a module to check the selected
tolerance type and selected feature and the required datum to see if it is valid for feature control frame.
Table 1 illustrates tolerance type and the required number of datum. Figure 4 illustrates the algorithm
used for datum allocation and verification module.
Table. 1 Tolerance types and their associated data structure.

Tolerance Type Required Information Remarks

Size Fa_No, UpToLim, LoToLim Fa_No: A pointer referring to the face

Form Fa_No, TolType, TolValue UpToLim: upper tolerance limit
LoToLim: lower tolerance limit
Fa_No, TolType, TolValue,
Orientation TolType: type of applied tolerance
MatCon, Datum
TolValue: tolerance value
Fe_No, TolType, TolValue, MatCon: mathematical condition
Location Datum: An integer refer to a datum
MatCon, Datums
Datums : Integers refer to datums (2 or 3)

Select Face

Yes

The face
No
satisfy datum
requirments

Yes

The face
No
is feature
member

No Yes

Claculate the
Datums required data
A and B
Prep. to the No
face

No
It is datum A It is datum B
Yes

Yes No
Datum A
Establish the selected Yes Prep. to the
face as a datum
face
 !" #$%% &
440
Fig. 4. Datum Allocation and Verification Module

7 TOLERANCE MODEL VALIDATION

There are two levels of validation; the first one during the modeling; and the second one after the model
is constructed to check the over-tolerance.
The model provides built-in checking of legality of tolerances defined during modeling. Types of checks
include tolerance limiting value, entity types (planar, cylindrical, etc), and entity relationships like the
orthogonality of two planes. When an entity is picked for defining tolerances, the entity is checked to see
if it is the specific type required for the tolerance

7. 1 OVER-TOLERANCED VALIDITY CHECKING

A scheme for over-toleranced validity checking is introduced. This scheme depends upon the degree of
freedom (DOF) techniques and primitive geometric elements DOF defines the independent ways in which
an object’s geometry can change. In spatial geometry, DOF can be resolved into kinematics DOF: they
are divided into three transnational DOF (TDOF = t1, t2, t3) and three rotational DOF (RDOF = θ1, θ2,
θ3). Primitive geometric elements are equivalent to the theoretical reference frames meant to represent
the real surfaces. Moreover, it describes visually the degrees of invariance associated to a surface.
Dimensions and ANSI Y14.5M tolerance classes reference or target features-of-size. A list of all such
entities used is as follows: part faces, edges, center points, mid-planes of slots or pockets, width and
radius dimensions. A dimension parameter Dp and three geometric primitives can cover all of these:
point, straight line, and planar face. Non-linear shape can be defined in terms of these three as follows. A
circle or sphere is described by a center point and a radius (Dp); a cylinder by an axis (line) and a radius
(Dp); profiles and sculptured surfaces can be attached to a coordinate system (lines or planes) for the
purpose of relative positioning, as demonstrated in Kramer’s work [42].
A solid model can be composed into a collection of these geometric primitives; if the sizes, shapes and
locations of the constituent geometric primitives are fully controlled, the solid model is fully defined.
A degree of invariance can be seen as a translation or a rotation under which a given geometry remains
unchanged. For instance, the line or axis of a cylindrical surface can only be translated along or rotated
about it, and that leaves four degrees of invariance; two in translation and two in rotation.
The total number of DOF of the object is less than six. Table 2 shows the three basic geometry primitives
and their respective DOF. The entity local coordinate systems that will be used have been made
consistent with ASME Y14.5M.
Table 2 Geometric Tolerances and DOF For geometric primitives

Degree of freedom in local coordinate

Basic
Corresponding
geometric Symbol Location Orientation
Surface
elements Tolerance Tolerance
X Y Z X Y z
Point Sphere tx ty tz - - -

Line Cylinder tx ty - θx θy -

Plane Plane - - tz θx θy -

8. MANUFACTURING PROCESS SELECTION MODEL

The importance of the availability of manufacturing process information during the design phase has two
major benefits. First, this piece of information is used in tolerance analysis and synthesis using statistical
methods [52]. Second, to make cost analysis at design phase, which is so important to reduce the total
product cost. The manufacturing process selection depends on two aspects. The first one is process
capability, which means the production process being able to produce the part with the required accuracy;
when there are several manufacturing processes that have the ability to produce the part with the required
 !" #$%% &
441
process. The second one depends on the manufacturing process cost, so the process with less
manufacturing cost is selected.
A knowledge base using the tolerance-production cost relationship is built for manufacturing process
selection. It uses rules such as
If (There is more than manufacturing process able to produce the product) then:
Select the one that has less production cost.
Also, rules that combine the production cost and required tolerance are built using rules for processes
capability and manufacturing process selection such as:
If (the required tolerance is T, get its IT grade from system database) then:
Select the process that has the capability to produce this grade

9. COMPUTER MODEL AND IMPLEMENTATION

This paper is concerned about tolerance representation issues with 3D solid models. It introduces a GUI
for tolerance input process. It allows visualization of the 3D solid model for the mechanical part in a
friendly intelligent user interface It helps the designer to enter the tolerance type and value, and validates
the designer inputs using Degree of Freedom (DOF). The system utilizes the Feature Extraction and
Recognition technique to help in the input process, validation and production process selection. System
flow chart is illustrated in Figure 5, and can be summarized as follow:
• Load Sol File (load 3D solid model files)
• File Analysis (analyze the file to get geometrical and
topological data) Start
• Feature Extraction and Feature Recognition (extract
feature members and recognize them)
• Part Visualization (display the part in 3D graphic screen)
• Datum Allocation (specify the datum faces or features Load 3DSolidFile
and validate it)
• Tolerance Input (enter tolerance data that includes type
and value)
File Analysis
• Tolerance Validation (check the entered tolerance
according to ANSI standard rules and for over-tolerance)
Currently, tolerance representations must be built by
Feature Extraction&
programs that directly invoke the constructors. These require Recognition
information about the internal names of faces in a solid
model, which are very inconvenient to obtain. To overcome
that, a point-and click graphic interface is developed, that Part Visualization
allows a user to select the face and automatically generates
the data needed by the constructors.
The developed system has an intelligent 3D interactive user
interface, which includes a 3D-visualization system for the Datum Allocation
part to be toleranced. The 3D-visualization model uses Java
3D classes. Java 3D is an API used to create standalone 3D
applications and 3D applets [53, 54]. Java 3D is a standard Tolerance Input
extension to Java 2 Java developed kit (JDK) [55]. After the
3D file is loaded and analyzed, the 3D part is visualized
using Java 3D classes and the developed classes. Tolerance Validation
The system is developed using object oriented programming
paradigm Java.
Classes for tolerance construction are developed as follow:
Using System Tool
• Class for general tolerance representation.
• Classes for size, form, location, orientation, runout and
profile tolerances.
End
• Class for interfacing with solid model B-Rep data that
extracts required information for the tolerance model
Figure 5 System Flow Chart
such as the internal names of faces in a solid model. It
then automatically generates the data needed
 !" #$%% &
442
then automatically generates the data needed
• Class for store modify and retrieve tolerance data
Also, a knowledge base is developed for manufacturing process selection. The system is implemented
using many examples (Figures 6, 7) and proves that it is an effective tool for input and validation of
tolerance information. That includes successful feature extraction and recognition, detecting over-
tolerance cases, tolerance type validation during input process, flexible inputting, modifying and
retrieving tolerance data.
10 CONCLUSIONS
For complete tolerance representation in the CAD data model, and to overcome the deficiencies of
current tolerance representation models, a scheme for tolerance representation is introduced in this paper.
This scheme is dependent on developed algorithms for tolerance representation; CAD files based 3D
solid model database and friendly intelligent user interfaces (UI). The new models uses information
embedded in a B-Rep solid model. A linking data structure is created to allow the input of tolerance
information to a B-rep database.
Ensuring the validity of tolerances is crucial to the successful implementation and effective use of
tolerancing system. Therefore, there are many validation stages during the tolerance input and modeling
process that uses developed validation models through intelligent user interface. This includes a method
to detect over-tolerancing. Datum selection validation also is introduced to ensure that the faces or
feature selected to be a datum satisfies the datums requirements. This means that, the system prevent
having errors in the tolerance model and its associated data.
Moreover, the GD& T model is consistent with engineering practice and compatible with Y14.5M.
Therefore, this GD& T model can support seamless integration between CAD and CAM.
The procedure outlined here can be further integrated with other modules of CAD/CAM like automatic
process planning (APP). Thus the entire design/manufacturing data interchange can be simplified.
Furthermore, the software is created and used in a PC environment making the system inexpensive and
flexible.

Fig. 6 System implementation Example 1

Fig. 7 System implementation example 2.

11 REFERENCES
 !" #$%% &
443
1. Giesecke, F E, et al, “ Technical Drawing”, 10th ed, Simon& Schuster/A Viacom Company, Upper
Saddle River, NJ, 1997.
2. American National Standards Institute, Dimensioning and Tolerancing, ANSI/ASME Y14.5M, NY,
1994.
3. American National Standards Institute, Mathematical definition Dimensioning and Tolerancing,
ANSI/ASME Y14.5.1M, 1994.
4. ISO 1101, Technical drawings- Geometric tolerancing – Tolerance of form, orientation, location and
run out- Generalities, definations, symbols, indication on drawing, 1983.
5. Shah J. and Mantyla, M, “Parametric and feature-based CAD/CAM, concepts, techniques and
applications”, John Wiley son Inc., New York, NY, 1995.
6. Juster N P, “Modelling and representation of dimensions and tolerances: a survey”, Computer Aided
Design, Vol. 24, n 1, pp 3- 17, 1992.
7. Rossignac, J R, “Blending and offsetting solid models”, Technical Report TM-54 Production
Automation Project, University of Rochester, USA, 1985.
8. Rossignac, J R and Requicha G, “ Offsetting operations in solid modelling” Computer Aided
Geometry, Vol. 3, pp 129-148, 1986.
9. Requicha, G., ‘Mathematical definition of tolerance specification’, Manufacturing Review, Vol. 6 n
4, pp 269-274, 1993.
10. Jayaraman, R, and Sirnivasan, V “Geometrical Tolerancing 1: virtual boundary requirements” IBM
Journal Research& Development, Vol. 33, n 2, pp 105-124, 1989.
11. Jayaraman, R, and Sirnivasan, V, “Issues in conditional tolerances for CAD Systems” IEEE
international Conference of Roboticsand Automation, St Louis, Missouri, USA, March, 1985.
12. Etesami, F, “Tolerance verification through manufacturing part modeling”, Journal of Manufacturing
System, Vo. 7, n 3, pp 223-232, 1988.
13. Etesami, F, ‘A mathematical Model for Geometric Tolerances’, Journal of Mechnical Engineering,
Vo. 115, pp 81-97, 1993.
14. Hoffman, P, “Analysis of tolerances and process inaccuracies in discrete part manufacturing”,
Journal of CAD, v 14 n 2, p 83-88, 1982.
15. Turner, J V and Wonzy, M J, “ A mathematical theory of tolerances, in geometric modeling for CAD
application”, H W McLaughlin and J L Enacarnacao, Eds, Elsvier, New York, NY, 1988.
16. Turner, J V and Wonzy, M J, ‘The M-space theory of tolerances’, Proceeding of the ASME Design
Automation Conference, Chicago, IL, Vol. 23 n 1, pp 217-225, Sept. 17-19, 1990.
17. Guilford, J., Sethi, M., and Turner J., ‘Worst case and statistical tolerance analysis of the daughter
card assembly’, Computer in Engineering, Vol. 1, pp 343-350, 1992.
18. Roy, U. and Li, B., ‘Representation and interpretation of geometric tolerances for polyhedral objects-
I. Form tolerances’, Computer Aided Design, Vol. 30 n 2, pp 151-161, 1998.
19. Guilford, J., and Turner, J., ‘Advanced analysis& synthesis for geometric tolerances’, Manufacturing
Review, Vol. 6 n 4, pp 305-313, 1993.
20. “Description of NONAME” , Technical Report – Geometric Modelling User Manual 1 Geometric
Modeling Project, University of Leeds, UK, 1981.
21. Ranyak, P. and Wilson, P., ‘Features for tolerancing a solid model’, Conference Information:
Computers in Engineering Proceedings. San Francisco, CA, Proc. Publ by ASME, New York, NY,
pp 275-280, August, 1988.
22. Hillyard, R C and Braid, I C, “Characterizing non-ideal shapes in terms of dimensions and
tolerances”, Computer Graphic, Vo 12, n 3, pp 234-238, 1978.
23. Elgabry A K, “ A frame work for a solid tolerance analysis”, ASME Computers in Engineering
Conference, ASEM, New York, July 1986.
24. Faux, I D, “Reconciliation of design and manufacturing requirements for product description data
using function primitive part features”, CAM-I Report R86-ANC/GM/PP-01.1, 1986.
25. Gossard, D., Zuffante, D., and Sakurai, H., “Representation dimensions, tolerances and features in
MCAE System”, IEEE Computer Graphics and Applications, 1988.
26. Roy, U, and Liu, C, “Feature Based representational scheme of a solid modeler for providing
dimensioning and tolerancing information”, Robotics and Computer Integrated Manufacturing Vol. 4
n 3-4, pp 335-345, 1988.
27. Vijay S., and Rangarajan, J., ‘Geometric tolerancing: II. Conditional tolerances’, IBM-Journal of
Research and Development, Vol. 33 n 2, pp 105-124, 1989.
 !" #$%% &
444
28. Roy, U, Mantooth, K., Pollard, M.D., and Liu, C.R., ‘Tolerance representation scheme in solid
model: Part I’, Computer-Aided and Computational Design American Society of Mechanical
Engineers, Design Engineering Division, Vol. 19 n 1, pp 1-10, 1989.
29. Roy, U, Mantooth, K., Pollard, M.D., and Liu, C.R., ‘Tolerance representation scheme in solid
model: Part II’, Computer-Aided and Computational Design American Society of Mechanical
Engineers, Design Engineering Division, Vol. 19 n 1, pp 11-17, 1989.
30. Irani S.A. Mittal, R.O., and Lehtihert E.A., ‘Tolerance chart optimization,, International Journal of
Production Research, Vol. 27 n 9, pp 1531-1552, 1989.
31. Clement, A, and Desrochers, A, ‘A dimensioning and tolerancing assistance model for CAD/CAM
systems’, International Journal of Advanced Manufacturing Technology, Vol. 9, pp 352-361, 1994.
32. Clement A, Desrochers, A, and Riviere, A, “Theory and practice of 3-D tolerancing for
assembly”,CIRP International Working Seminar on Computer-Aided Tolerancing, Penn State
University, pp 25-55, 16-17 May 1991.
33. Shah J., Yan Y., and Zhang B., “Dimension and tolerance modeling and transformation in feature
based design and manufacturing”, Journal of Intelligent Manufacturing, Vol. 9, pp 475-488, 1998.
34. Bernstein, N S, and Preiss, K, “Representation of tolerance information in solid models”, ASME 15th
Design Automation Conference ASME, pp 37-48, 1989.
35. Gu, P.H. and Elmaraghy, H.A.,’ Expert tolerancing consultant for geometric Modelling’, Proceeding
of Manufacturing International Symposium on Product and Process Design, ASME, Vol. 1, pp 17-
22, 1988.
36. Turner, J U, “A vector space approach to the mathematical representation of tolerances on
mathematical parts”, IBM Technical Report, TR 00.3383, May 1986.
37. Turner, J, and Wozny, M, “Tolerances in computer-aided geometric design”, The Visual Computer,
Vol. 3, pp 214-226, 1987.
38. Martinsen, K., ‘Vectorial tolerancing for all types of surfaces’, Advances in Design Automation
American Society of Mechanical Engineers, Design Engineering Division-(Publication)-DE. Vol. 65
n 2, pp 187-198, 1993.
39. Wirtz A, “Vectorial tolerancing for production control and functional analysis in design”, CIRP
International Working Seminar on Computer-Aided Tolerancing, Penn State University, pp 77-84,
16-17 May 1991.
40. Desrochers, A. and Clement A., ‘A dimensioning and tolerancing assistance model for CAD/CAM
Systems’, International Journal of Advanced Manufacturing Technology, Vol. 9 n 6, pp 352-361,
1994.
41. Kramer G.A., ‘Solving geometric constraint system: a case study in knematics’, MIT Press, 1992.
42. Creveling, G.M., ‘Tolerance Design: A handbook for developing optimal optical specifications’,
Addison Wesley Longman, Inc., Massachusetts, 1997.
43. Roy U., and Li, B., ‘“ Representation and interpretation of Geometric Tolerances for Polyhedral
objects-1. For Tolerances’, Computer-Aided Design, Vol. 30, No. 2, pp. 151-161, 1998.
44. Sonthi, R., Kunjur, G. and Gadh R., Feature Recognition using curvature region for design- for-
manufacturability analysis, Technical papers of the North American Manufacturing Research
Institution Of SME, pp. 137-142, 1997.
45. Requicha, G. and Vandenbrande, J. H., ‘Spatial reasoning for the automatic recognition of
machinable features in solid models’. IEEE Transaction of Pattern Analysis and Machine
Intelligence, Vol. 12, pp 1269-1285, 1993.
46. Madurai, S. S. and Lin, L, 1992, ‘Rule-based automatic part feature extraction and recognition from
CAD data’, Computers Industrial Engineering, Vol. 22, pp 49-62, 1992.
47. Gadh, R., and Prinz, F.B., Recognition geometric forms using the differential depth filter, Computer
Aided Design, Vol. 24, no. 11, pp. 583-598, 1992.
48. Joshi, S. and Chang, T. C., ‘Graph-based heuristics for recognition of machined features from a 3D
solid model’, Computer Aided Design, Vol. 2, 58-66, 1988.
49. Garvankar, P. and Henderson, M. R., ‘Graph-based extraction of protrusions and depressions from
boundary representations’, Computer-aided Design, Vol. 22, pp. 442-450, 1990.
50. Sonthi, R., Kunjur, G. and Gadh, R., Shape Feature determination using the Curvature Region
Representation, Proceedings of FourthSymposium on Solid Modeling and Applications, Sponsored
by ACM SIGGRAPH, Atlanta, GA, 14–16, pp. 285–295, 1997.
 !" #$%% &
445
51. Sonthi, R. and Gadh, R., MMCs and PPCs as constructs of curvature regions for form feature
determination, Computer-Aided Design, Vol. 30, No. 13, pp. 997–1001, 1998
52. Mohamed, K, “Tolerance distribution for production processes integrated with CAM system” Ms
Thesis, Industrial Production Engineering, Faculty of Engineering, Mansoura University, Egypt,
1995.
53. Brown, K., and Petersen D., “ready-to-run Java 3D’, Wiley& Sons, Inc., Danvers, MA, 1999.
54. Barrilleaux, J., ‘3D user interfaces with Java 3D’, Manning publications Co., Greenwich, CT., 2001.
55. Sowizral, H., Rushforth, K., and Deering, M., ‘The Java 3D API specification’, Addison-Wesley
Pub Co, 2000.