Feature-Based Approach for a Process Supply System Design

Abstract

The authors are motivated to investigate an effective method for achieving the knowledge-driven design in order to address the efficiency drawback in common CAD applications. In this paper, a systematic method, to embed in-depth engineering knowledge and to realize smart design changes in an advanced feature-based design, is proposed. To proof the feasibility and the effectiveness of the proposed method, a process fuel and water supply system has been designed comprehensively in the conceptual design stage. The findings of this research work are presented with some critical discussions at the end of this paper. The authors believe that this approach is easy to be implemented and useful to improve the knowledge reusability and engineering design productivity.

Full text

Feature-Based Approach for a Process Supply System Design
Y.-S. Ma*, Q. Hadi
Department of Mechanical Engineering
University of Alberta
Edmonton T6G 2G8, Canada
*Email: yongsheng.ma@ualberta.ca
Tel: (+1)780.492.4443
Abstract—The authors are motivated to investigate an
effective method for achieving the knowledge-driven design in
order to address the efficiency drawback in common CAD
applications. In this paper, a systematic method, to embed in-
depth engineering knowledge and to realize smart design
changes in an advanced feature-based design, is proposed. To
proof the feasibility and the effectiveness of the proposed
method, a process fuel and water supply system has been
designed comprehensively in the conceptual design stage. The
findings of this research work are presented with some critical
discussions at the end of this paper. The authors believe that
this approach is easy to be implemented and useful to improve
the knowledge reusability and engineering design productivity.
Keywords—computer-aided design; knowledge modeling;
unified feature-based design; pressure vessel; parametric
modeling
I. INTRODUCTION
Computer-aided design (CAD) tools have been widely
used in engineering; but the problem that how effectively
the modern CAD tools are used to deal with the challenges
of engineering complexity and knowledge embedment has
been puzzling many companies. A company’s engineering
department very often finds that it is repeating some similar
design modeling effort and yet the engineering models are
not able to be reused. Although there had been temptations
for using a previous project design for a new one, but the
integrity of engineering model becomes unknown to too
tedious to check. Such concerns are particularly strong in
the system conceptual design stage. Hence, a systematic
feature based method, to embed in-depth engineering
knowledge and to realize smart design changes in an
advanced feature-based design, is proposed.
To proof the effectiveness of the proposed method, a
convincing case study is necessary. Via an in-depth project,
a pilot but useful software system has been developed to
create an efficient and reusable oil and water supply design
so that industrial contractors can enhance their practices by
adopting the proposed approach. The reason to study such a
system to proof the concept is due to the ever increasing
demand for electricity across the globe that has lead to a
rapid rise in power plant construction and expansion. The
fuel and water system design in power plants is essential in
engineering design as it has major technical, operational,
and economic impacts. Piping layout and equipment design
is constantly changing with the advancement in the design,
manufacturing, installation and construction of the plant due
to various unseen factors. Hence it becomes important that
the design details are updated in a mechanical model
according to the changes in process and instrumentation
diagrams (PIDs) or the layout design with minimum efforts.
This is where advanced feature-based design modeling can
play an effective and important role. Although the majority
of power plants which are being set up are either based on
coal or natural gas technology, a diesel fuel power plant is
studied due to its suitable engineering scale and the
limitation of the research resources. The authors believe that
the proven method can be equally applied to larger scale
projects.
This paper highlights the advanced design features that
have been identified via a real world research project,
illustrates the representation of them in a future feature
modeling and development effort, and demonstrates the
design feature implementation mechanisms based on
parametric modeling. The advantages and drawbacks of this
proposed feature-based method is discussed.
II. L
ITERATURE REVIEW
Traditionally, the feature concept was used for
manufacturing. For example, machining features are
traditionally defined as volumes of material removed in
machining operations [1]. However, since features can
represent engineering semantic patterns effectively, hence,
there have been many researchers proposing the expansion
of the concept and the related modeling schemes such that
feature-based knowledge representation and implementation
methods can be more coherently and consistently developed.
A thorough review related to new feature types and their
research state of art has been available in [2]. One attempt
was to make various types of features to be more unified
under a common scheme such that associative nature of
advanced features [3] [4], such as cooling channels in
plastic injection mould, can be modeled and supported for
productivity enhancement software tools with certain
information management automation [5]. Such evolvement
of feature definition has made advanced feature application
broaden and more convenient.
To design a process supply system with the advanced
feature-based approach, there has been some challenges.
First of all, the definitions of features in such design
projects are non-standard and have never been explicitly
identified. For example, the piping layout in a diesel fuel
supply sub-system could be similar to the definition of
cooling channels in mould design [5]. Such features related
to conceptual design are even more difficult because of the
2010 International Conference on Manufacturing Automation
978-0-7695-4293-5/10 $26.00 © 2010 IEEE
DOI 10.1109/ICMA.2010.11
38
2010 International Conference on Manufacturing Automation
978-0-7695-4293-5/10 $26.00 © 2010 IEEE
DOI 10.1109/ICMA.2010.11
38
2010 International Conference on Manufacturing Automation
978-0-7695-4293-5/10 $26.00 © 2010 IEEE
DOI 10.1109/ICMA.2010.30
1
2010 International Conference on Manufacturing Automation
978-0-7695-4293-5/10 $26.00 © 2010 IEEE
DOI 10.1109/ICMA.2010.30
1

abstract nature and further detailing in the down-stream [6]
[7]. Second, the design cycle involves multiple processes
including flow analysis, pressure vessel design and
geometry design simulation. Design changes occur
repetitively throughout the design cycles. The effective
mechanisms to implement, validate and consistently achieve
the downstream updates of changes are essential for
knowledge driven design approach [2] [8]. Hence, in-depth
case studies to identify, verify and prototype advanced
features are imperative to beef up the support for the
unification and diversification of feature-based knowledge
engineering approach.
III. M
ETHODOLOGY
Parametric modeling, as an enabling mechanism for
knowledge base engineering modeling, is one of the new
and smart modeling techniques being employed in
industries. Parametric modeling can be understood as a
CAD modeling method which supports expressions in the
CAD environment and enables design model dimensions
used to be changed by changing the expression values.
Parametric modeling is the supporting mechanism for
feature-based design (references) approach. Parametric
techniques depend on well defined mathematical relations in
such a way that if one entity or parameter is altered, its
impacts on all the other relative parameters can be updated
automatically. A well-defined parametric modeling
application can embed design knowledge into the design
models in a generic manner, and significantly increases the
productivity of design processes. The main advantage of
parametric modeling is that there is no need for the
generation of new models based on the changes of input
design parameters, which eventually leads to the savings of
time, effort and cost. However, without an appropriate
design scheme, simply using parametric modeling creates a
lot of CAD modeling parameters such that their semantic
meanings, relations, and rules applied to them are tedious to
manage and engineers will be overwhelmed by the “hidden”
relations among parameters and dimensions if the design
task becomes reasonably complex, or if the engineer is not
or no longer familiar to the hidden parameters and their
embedded constraints.
In order to make better use of the parametric design
mechanism such that, many engineers, who in these days are
dependent on CAD tools to deliver their design contents,
can easily operate the design steps with their common
engineering semantic patterns and their design intent can be
fully defined and managed throughout iterative design
cycles of revisions, feature technology is required. A feature
defines relations between different entities in a semantically
explicit and coherent manner. A review on the development
of feature technology and the new paradigm of concurrent
and collaborative engineering has been available [2]. Via
features, design entities are calculated based on complex
relationships and mathematical formulas according to
typical engineering design patterns. Advanced feature-based
approach basically entails flexible well-defined feature
definitions, constraint management and effective service
functions, where, usually, object-oriented software
engineering methodology is used. In the reported research
project, a unified feature concept is applied when initiating
the engineering activities.
Figure 1. Semantic definition of unified feature in UML format [9].
Figure 2. System design for the proposed method.
The definition of a unified feature has been given in [9]
and shown in Figure 1. Basically, the design intent can be
expressed as a set of common, flexible, and well-defined
data structures, i.e. unified features, where the engineering
conceptual patterns, e.g. the pipeline layout, a piece of
equipment, or a key design code to be checked in the
context of a process supply system, are represented
generically in a set of geometric or topologic entities,
associated driving parameters, constraints, and attributes.
Unified features can be used to define the engineering
patterns at different semantic detail levels, and a set of lower
level unified-features can be used to define a higher level
feature. UML (Unified Modeling Language) representation
is standardized software modeling design method for
1
IAttribute
IConstraint
0..*
0..*
0..*
0..*
Unified
Feature
Attribute
Topological
Entity
0..*
0..*
Feature
Model
0..*
0..*
0..*
Parameter
1
1..*
0..*
Other
Constraints
Geometric
Constraint
Self-
described
attribute
Association
attributes
Algebraic
Constraint
dependency
generalization
aggregation
composition
Constraint
Priority
Variables
CAE
modules
CAD Modules
Excel Modules
CAD
expressions
CAD 3D
models
CAD
wireframe
3D layout
models
Process design
parameters
CAE 3D mesh
models
CAE
analysis
model
Design code
checking
and
calculation
expressions
CAD 2D
PID
models
User interface
Drawing
parameters
Engineering
code related
parameters
Excel
process
Users
CAE
analysis
results
393922

illustrating class property data and process definitions and
their relations in object-oriented programming methodology
[10].
This work is not about the coding aspect to achieve
unified feature modeling via software. Rather, it is a
verification and application of the proposed concept of
unified feature through a real example. The research tries to
answer two questions. The first is how unified feature
properties can be corresponding to a real application, and
the second is if the definition of the unified feature can be
generic enough for those similar projects so that the
engineering approach with it can serve as an effective
engineering information management method.
IV. S
YSTEM ARCHITECTURE
From application point of view, the project was also
aimed to create an efficient and reusable supply system
design models for some common industrial applications,
such as a power plant, by developing a feature-based CAD
parametric design model. Typically, there are three sub-
systems are considered for a power plant supply system, i.e.
diesel fuel, portable water, and fire water. Figure 2 shows
the system design of the proposed method.
V. E
NGINEERING DESIGN PROCESS FLOW
In this work, to begin with, a process engineer (a user)
needs to interactively input the basic process parameters and
select the common process elements via an user interface of
Excel software template. Figure 3 shows a partial screen
snap of the input page.
Figure 3. Partial input page implemented with Excel.
Then, the detailed attributes to generate a 2D a process
and instrument diagrams (PID) for each sub-system is then
automatically created in a page. Based on his input, the 2D
PID diagram is parametrically generated in the CAD
system. Figure 4 shows the PID for the fire water supply
sub-system. Similarly, the PIDs for portable water and fire
water systems are also created parametrically. Such PIDs
are in fact a form of conceptual process design features.
They have to be validated by basic process calculation
formulas which are in turn a set of constraints to be satisfied
in the implementation of a feature-based system in the
future.
Next, those attributes of the newly-created PIDs is used
as the design inputs and the piping layout and pressure
vessel designs attributes were created semi-automatically
via necessary interactions between the user (now, it could be
a mechanical designer) and the template pages of the
modeling system.
Figure 4. PID diagram for the fire water supply sub-system.
Figure 5. Fully generated conceptual 3D design in the CAD system.
Note that professional code regulations have been fully
considered at this stage by validate the design inputs via a
set of built-in code checking templates specially tailored
according to design rules. In this case, regulation codes,
such as ASME pressure vessel design codes under Section
VIII, have been implemented. Such codes can be understood
404033

as design constraints and they have to be satisfied for the
detailed design parameters in order to validate the pressure
vessel conceptual design feature. Analytical calculation was
implemented in the Excel model.
Following the above design attribute generation step,
Autodesk Inventor was used for the development of the 3D
models and the driving parameters are controlled by Excel
spread sheet templates. Figure 5 shows a fully generated
conceptual design for the system. Figure 6 and 7 show the
fire water supply system model in more detail. Because of
the built-in software integration capability, this system
design (see Figure 2) supports that any changes can be
implemented into the 3D mechanical models with minimum
efforts via parametric modeling. Internal relations for each
design block, such a pressure vessel design dimensions, are
implemented into related formulas of Excel.
Figure 6. Fire water storage tanks and piping layout.
Figure 7. A closer view of the fire water supply sub-system.
Finally, all the analytical calculation done was further
validated with 3D finite element models which are created
based on the mesh output generated from the
aforementioned parametric CAD modeling step and
structural analysis was then carried out. Figure 8 shows the
stress analysis results for the diesel fuel storage tank with
the consideration of the saddles’ effect.
VI. C
ONSTRAINT IMPLEMENTATION
Design codes involved are modeled as a set of
constraints. The three sub systems are essential for the
operation of a power plant. Other than the diesel fuel sub-
system that is obviously necessary to provide constant
energy whereas the portable water and fire water system is
meant to supply water to the boiler, generator and the
incinerator plant building respectively. In the incineration
plant building the solid waste and liquid organic wastes are
treated. In all the systems, centrifugal pumps are used to
maintain the required pressure of systems. In the fire water
system jockey pump is also used so that if the pressure
drops due to any sort of leakage, or in emergency the
pressure will keep on dropping.
Figure 8. Maximum and minimum Von Mises stress values.
In order to maintain the required discharge pressure the
jockey pump will come into play. Jockey pumps are
attached in such a way that they come into act when drastic
pressure drop takes place. Once the pressure reaches up to
the minimum required pressure it is automatically switched
off.
The main design codes and standards implemented are:
x ASME Section VIII, Division 1 - Boiler & Pressure
Vessel Design Code.
x API 650 – Storage Tank Design.
x ASME B 31.1 - Power Piping Code.
x ASME B 16.10 – Face to Face dimensions of Valves.
x ASME B 16.34 – Valves – Flanged, Threaded and
Welding End.
414144

x ASME B 16.9 – Butt Welded Fittings.
x ASME B 16.5 – Pipe Flange and Flange Fittings.
x ASME B 36.10 – Welded and Seamless Wrought
Steel Pipe.
In the Excel calculation, modular approach is used. For
the project, in additional to the major pressure vessel design
code checking module, the following modules are
implemented as well: flow calculation, pressure loss
calculations, pipe wall thickness calculation, fire water tank,
nozzle reinforcement calculation, dike wall calculation, area
reinforcement, and stress verification for saddles. It was
found that it was quite convenient to cluster calculations
according to the rules corresponding to the relater regulation
sources and their calculation sections. In such a way, code
verification and validation are made easier to be conducted
either automatically or manually by the designers.
Figure 9 shows a partial code checking template in Excel
for the pressure vessel design. Figure 10 shows the partial
implementation of constraints used for saddle design stress
verification.
To automatically interface with the CAD models created
in the CAD software, a dedicated page of model-related
attributes, corresponding to those driving CAD model
parameters, is developed as shown in Figure 11. The
interface page is constructed according to the parametric
modeling requirement of Autodesk Inventor form so that the
integration between Excel and Inventor can be fully
supported and parametric modeling is then readily achieved.
Figure 9. Constraints implemented for pressure vessel design module.
VII. CHANGE MANAGEMENT
One advantage for feature-based modeling is the
parametric change management. Instead of changing
individual parameters, a set of them are change at a time. By
managing changes in groups, the consistency can be well
kept than updating parameters one-by-one because there
could be a lot of intermediate updating conflicts arise from
the in compatible values of a feature pattern.
Table 1 shows a set of parameters with the current and
new values to be assigned. Figures 12(a) and (b) shows
change effect on the horizontal water tank before and after
the change update. For the pressure vessel shown in Figure
12(a) there is only one rib in the middle of the support and
also there is no reinforcement pad shown, but for the
pressure vessel shown in Figure 12(b) there are two ribs and
also as per the nozzle reinforcement calculations the
reinforcement pad is updated automatically around the
nozzle connection.
Figure 10. Constraints implemented for the saddle of a horizontal storage
tank.
Figure 11. Partial list of attributes related to CAD model features.
424255

VIII. CONCLUSIONS
This paper reports a new application case of the unified
feature definition [2] and supports the effectiveness of
applying advanced conceptual design features in real
industrial engineering, i.e. a process supply system
consisting of diesel fuel, portable water and fire water sub-
systems. Under the guiding principles of unified feature, the
implementation was carried out in a schematic manner but
mainly enabled by parametric modeling between CAD and
Excel interactions. Although the software coding has not
been done due to the resources constraint, however, it has
been clear that the unified feature definition has been useful
for the identification and organization of engineering design
patterns in this traditional but well regulated application
domain. More research on the development of a reusable
design software toolkit is expected as the future work.
A
CKNOWLEDGMENTS
This research effort was partially supported by a Canada
NSERC Discovery grant (No. 355454-09) and the University
of Alberta.
TABLE
1
FEATURE-BASED CHANGE MANAGEMENT TESTING
PARAMETERS AND VALUES
Parameter
Original
Changed
Internal pressure of vessel (bar)
5
3
Height of the vessel from ground level (ft)
10
7
Outside diameter of the vessel shell (inches)
120
96
Distance between horizontal vessel (m)
7.5
7
Net capacity of the vessel (m
3
) 100
50
Saddle dimensions a (inches)
104
83.5
B (inches)
69
57
C (inches)
9
9
D (inches)
24
18
E (inches)
40
32
No of ribs (ul)
2
1
G (base) (inches)
1
1
H (web flange ribs) (inches)
0.75
0.5
K (wear) (inches)
0.5
0.375
Fire water tank height of each course (m)
2.625
2
Diameter of the tank (m)
8.53
8
Pressure vessel nozzle nominal wall thickness
of nozzle (inches)
0.375
0.5
Exterior projection (inches)
3
2
Interior projection (inches)
1
0
Fillet size (inches)
0.5
0.3
(a)
Pressure Vessel model using original parameter values
(b) Pressure Vessel model using new parameter values
Figure 12. Partial list of attributes related to CAD model features.
REFERENCES
[1] International Organization for Standardization (ISO), Industrial
Automation Systems and Integration – Product Data Representation
and Exchange – Part 224: Application Protocol: Mechanical Product
Definition For Process Planning Using Machining Features, ISO
10303-224:1999, ISO: Geneva, Switzerland,1999.
[2] Y. –S. Ma, G. Chen, and G. Thimm, “Paradigm shift: unified and
associative feature-based concurrent and collaborative engineering”,
Journal of Intelligent Manufacturing, vol. 16, 2008, pp. 625–641.
[3] W. F. Bronsvoort, and A. Noort, “Multiple-view feature modeling for
integral product development”, Computer-Aided Design, vol. 36, no.
10, 2004, pp. 929-946.
[4] R. B. Gottipolu, and K. Ghosh, “A simplified and efficient
representation for evaluation and selection of assembly sequences”,
Computers in Industry, vol. 50, no. 3, 2003, pp. 251–264.
[5] Y. –S. Ma, and T. Tong, “An object oriented design tool for
associative cooling channels in plastic injection mold”, Int. J. of Adv.
Manuf. Tech. vol 23, 2004, pp. 79 - 86.
434366

[6] K. –Y. Kim, Y. Wang, O. S. Muogboh, and B. O. Nnaji, “Design
formalism for collaborative assembly design”, Computer-Aided
Design, vol. 36, no. 9, 2004, pp. 849–871.
[7] Y. –H. Han, and K. Lee, “A case-based framework for reuse of
previous design concepts in conceptual synthesis of mechanisms”,
Computers in Industry, vol. 57, no. 4, 2006, pp. 305–318.
[8] S. Subramani, and B. Gurumoorthy, “Maintaining associativity
between form feature models”, Computer-Aided Design, vol. 37, no.
13, 2005, pp. 1319–1334.
[9] G. Chen, Y. –S. Ma, G. Thimm, and S. –H. Tang, “Associations in a
Unified Feature Modeling Scheme”, ASME Transactions Journal of
Computing & Information Science in Engineering, vol. 6, no. 2, 2006,
pp. 114-126.
[10] I. Jacobson, G. Booch, and J. Rumbaugh, The Unified Software
Development Process, Addison Wesley: Longman, 1998.
4444
77