Sunteți pe pagina 1din 8

A PROTOTYPE OF KBS FOR MATERIAL SELECTION IN

BUMPER BEAM DESIGN


Mohd Nizam Sudin*, Mohd Ruzi Harun, Ahmad Tajudin Hamzah and Shamsul
Anuar
Received: Jan 31, 2007; Revised: Jul 11, 2007; Accepted: Jul 16, 2007

Abstract
This paper describes the development of the Knowledge Based System (KBS) for material selection in
bumper beam design. The KBS is an expert system used to select the most suitable material by defining
constraint values into the system. The aim of this project is to select the most suitable material for an
automotive bumper beam. The selection method was based on the material selection requirements
such as mechanical properties, physical properties and economic value. The constraint values were
documented in the product design specification (PDS) derived from experimental work and design
calculation. The KBS will rank the materials in the form of a hierarchical graph for the best five of the
most suitable materials. The top ranking materials have the closest specification values to the user’s
defined value.
Keywords: Knowledge based system, material selection, bumper beam

Introduction
Computers science (CS) has become valuable system is one of the technology advantages
technology nowadays and has shown its available; by the combination of computers and
potential as a powerful tool for materials programming languages it can be used to solve
election. Some companies such as Furon, a problems easily, cheaply and with lower risk
manufacturer of structural bearings, has offered particularly in the material selection process
an expert system to its customers to select the compared to former approaches.
products, which best meet the customers Material selection process is an important
requirements including environmental consi- activity in product development process because
deration aspects and geometric compatibility. material plays an important role to ensure the
In the current demanding market, industry needs product meets the product requirements such as
a tool to overcome certain problems effectively lightweight, high specific strength, high specific
and rapidly. Therefore CS and Artificial Intelli- stiffness etc. In automotive research KBS has
gence (AI) communities have to turn this been developed to select material for engine
demand into opportunity. The KBS or expert components such as the piston, connecting

Faculty of Mechanical Engineering, University Teknikal Malaysia Melaka, Locked Bag 1200, Ayer Keroh,
75450, Melaka, Malaysia. E-mail: nizamsudin@utem.edu.my
*
Corresponding author
Suranaree J. Sci. Technol. 14(3):215-222
216 A Prototype of KBS for Material Selection in Bumper Beam Design

rod and piston ring (Sapuan et al., 2002). Sapuan frame-based system, which acts as a database
and Abdallah (1998) have developed an expert for material and contains all the required data
system for material selection of polymeric-based properties. Classification and reasoning of
composite material for an automotive pedal box material selection are carried out using a rule-
system. The system was developed to select a based system approach. The frame-based
suitable material for major elements in the pedal system of the KBS is very flexible, expressive
box system such as accelerator, clutch, brake and accessible for interpretation to the rest
pedal and mounting bracket. However most of of the system and it is used to describe the
the KBS were developed specifically only to a hierarchical relationship between the material
particular component. To increase the number and its attributes. Figure 1 and Figure 2 shows
of KBS available in automotive applications this the flow chart of the material selection process
project is aimed to develop a KBS for material and material selection hierarchy tree in the KBS
selection for an automotive bumper beam respectively.
design. In the next sections the factors, rules and
technique used for the KBS development will User Interface
be described.
User interface is an important part in the KBS
System Description system. It must be user-friendly for easy access
to the system. The user interface was intended
This system was developed based on heuristic to allow the designer to work quickly in the KBS
rules and the experience of the designer. It is a with a little familiarization. To interact with the

Figure 1. Flow chart of the material selection process


Suranaree J. Sci. Technol. Vol. 14 No. 3; July-September 2007 217

KBS the user just needs to use a mouse, menu, Td > Tmax : operation system is stop and go to
keyboard and a displayed window. The next material
displayed window will show either information Td < Tmax
from a knowledge-based or user option. Td = Tmax: the system continues to next constraint
The user interface was designed and
formulated using images such as buttons, text, The most important thing in this part is to
bitmaps and transcript. Figure 3 shows a user define the value of the constraints. The
interface of the KBS for material selection in constraints’ value is defined from the analysis
bumper beam design. or experiment result that has been done.

KBS Constraints User Input Data


In this project, material properties are utilized User input data is a dialogue platform between
as a rule to set the constraints in the KBS. Using the system and the user. In this part, the system
the material properties rule, the designer will will ask a few questions to the user about the
be able to examine whether the proposed constraint values of the material properties. The
materials are suitable as a candidate for the following is the dialogue between the user and
particular component. As an example, if the the KBS.
designer specifies the material with toughness
value, T d , the system will compare this Dialogue between the system and the user:
toughness, with the pre-defined toughness limit,
Tmax. If the value of the defined constraint is in What is the minimum value of compres-
the limit, the system will proceed to the next sive yield strength?
constraint, and if not, this material will be What is the maximum value of density?
eliminated for the next process. What is the minimum value of flexural

Figure 2. Material selection hierarchy tree


218 A Prototype of KBS for Material Selection in Bumper Beam Design

strength? object-oriented programming (OOP) language


What is the minimum value of Young’s offers flexibility, easing changes to programs
Modulus? and is widely used in a large-scale software
What is the minimum value of tensile yield engineering population. Furthermore, an OOP
strength? is easy to learn, simple to use and easy to
maintain, lending itself to more direct analysis
Database and Software to Develop coding and also has more understanding of the
KBS complex situations than other programming
methods. Today, there are many types of object-
First, a classification of material is defined oriented programming such as C++, Java,
and it will act as the main database for the Visual Basic, Kappa-PC, MS Access etc.
KBS. This is a very important part of the KBS In this project, the Visual Basic (VB)
development, mainly to reduce project cost and programming language was employed as a
to eliminate unwanted material at the early language in the KBS development. VB is a
development stage. The material classification simpler programming language and is easy to
was defined through a pre-material selection use, particularly to create a simple graphical user
process by considering the performance require- interface window and also for complex
ments of the automotive bumper beam. applications. It is a combination of visually
Finding a good programming language arranging components in the form of a shape,
is also important in developing the KBS. The specifying attributes and actions of components,
language must be easy to use and able to create and has the capability to write additional lines
a simple graphic user interface as well as to of codes for extra functions.
develop fairly complex applications.
The knowledge contained in a product Material Selection Process
model will act as an object, that it is belongs to
a “class” in the KBS. This means that the KBS In this system, two processes have been
is object-oriented and it was developed using employed to produce the relevant result for
an object-oriented programming language. An the bumper beam component. They are the

Figure 3. User interface of the KBS for material selection in bumper beam design
Suranaree J. Sci. Technol. Vol. 14 No. 3; July-September 2007 219

screening and ranking processes. The screening Rules:-


process purposely eliminates materials after
considering the constraints. The ranking process If (the compressive yield strength > or = 300
ranks in order the admissible solutions using MPa) and
performance metrics based on the objective. If (the flexural strength > or = 250MPa) and
In this section, rules have been imple- If (the tensile yield strength > or = 300
mented in the KBS for the selection of the MPa)
suitable material for the automotive bumper and
beam. For instance, when the conditions of If (the Young Modulus > or = 20GPa) and
a rule are satisfied then the conditions are If (the density < or = 2500 Kg/mc)
valid. Therefore, to select the material for the Then
automotive bumper beam, the following (This material is a candidate for the bumper
approach has been applied: beam component and goes to the next section)

If <X> Then <Y> Figure 4 shows the graph of the Young


Modulus (GPa) versus Density (Mg/m3). After
where X are the constraints defined by the the screening process the materials in the shaded
user and Y is the conclusion. Meaning that this area, that are Aluminum Alloys, GFRP, KFRP,
material fulfilled all constraints defined by the Glasses, Magnesium Alloys and others have
user. The process of material selection for the been chosen as candidates for the automotive
automotive bumper beam has been defined bumper beam. These materials will proceed to
through the following rules: the next process or ranking process for the

Figure 4. Graph of the Young Modulus versus Density (Ashby and Cebon, 1993a)
220 A Prototype of KBS for Material Selection in Bumper Beam Design

second stage in the selection process. The shaded


area in the graph indicated that all materials
satisfied all selection criteria defined by the user
(1)
and were considered as ‘passed’ the screening
process.
The material that has passed the screening In the programming language, the new
process is stored in the database for every performance index is seen as follows:
selection process and it can be examined at any
time. It is also possible to modify any selection tempRerult1 = ymRS ^ (1 / 2) / dRS
process to ensure the performance criteria tempRerult2 = tsRS ^ (2 / 3) / dRS
tempRerult3 = (ymRS ^ (1 / 2)) / (cRS * dRS)
are confirmed or unconfirmed until the most tempRerult4 = (tsRS ^ (2 / 3)) / (cRS * dRS)
suitable materials are found. The materials that M = tempRerult1 + tempRerult2 + tempRerult3 + tempRerult4
pass the screening process will be stored in
a disc file and compiled as the package. This Once the performance index of each
enables users continually to do a selection material in the database was computed, the
process and re-evaluate the selection criteria system will sort the material in rank based on
based on the input from other design informa- the score in the performance index. The
tion. It also forms a document in the selection material with the highest score in performance
process index will be ranked as the first and so on. The
score is to indicate the properties of the
Formulating Material Selection material closest to the material properties that
Process have been defined by the user in the system.
These materials become the most suitable
Based on the formulation developed by materials for the automotive bumper beam
Ashby (1993), the performance index that is a component.
combination of different material properties
has been employed in this system to select the Results and Discussion
material for the bumper beam in automotive
application. The detail of the performance The KBS for material selection has the
index is shown in Table 1. capability to select the most suitable material
In this project each performance index is for the automotive bumper beam. To select the
computed and summed, to produce the total most suitable material for the automotive
performance index M as in equation (1). This bumper beam the user needs to define the
total performance index is important to enable constraint value in the user dialogue box.
the KBS to process material in the database and Variations results will be produced if the user
select the material with the highest strength and varies the constraint values in the system. This
stiffness. This material also has shown a good revealed that the system was intelligent enough
result in terms of cost and weight. to manipulate a variation of input that has been

Table 1. Performance index


Function Objective Constraints Performance index
Beam Minimum weight Stiffness prescribed E / ρ 1/2
Beam Minimum weight Strength prescribed σy2/3 / ρ
Beam Minimum cost Stiffness prescribed E1/2 / Cmρ
Beam Minimum cost Strength prescribed σy2/3 / Cmρ
Suranaree J. Sci. Technol. Vol. 14 No. 3; July-September 2007 221

defined by the user in the system. smaller number of material candidates. From the
The material selection result in this top five ranks of material have been chosen 4,
system has been represented in Figure 5. In this which are aluminum alloys and could be the best
system, 5 types of materials have been selected material for automotive bumper beam.
to be the material candidates for the automotive During the past decade, aluminum alloy
bumper beam. If there are any changes of especially in 7xxx series has been used for
the defined constraint values, the material specific applications in the automotive industry,
candidates and the rankings will be varied such as hoods of the vehicles. Aluminum alloys
respectively. In this system if the user constraint were selected as the materials for the auto-
value cannot be found, the result will be motive bumper beam because they are a light-
presented as “NULL” in commands. weight material which is 40 - 60% lighter than
The result as shown in Figure 4 reveals steel. This may reduce vehicle weight and
that Norplex NP130HF Glass Fabric material is consequently reduce fuel consumption. The low
selected as the first rank, second is Aluminum elastic modulus of aluminum alloys is an asset
7178-T6; 7178-T651, third Aluminum 7178- when a structure is subjected to shock-loading
T76; 7178-T7651, fourth Aluminum 2048, and conditions. The aluminum alloys are able to
the last is Aluminum 7475-T7351. absorb energy three times higher than steel
The result has shown that the system before permanent damage occurs for an equal
has the capability to choose the best material moment of inertia and strength. Besides that
candidates for the automotive bumper beam. the aluminum alloys material has many
Norplex NP130HF Glass Fabric is the new advantages to make it suitable for automotive
material and the authors have yet found its application, such as:
application in the automotive industry since this Strong. The entire vehicle body can be
system was developed. The user may select aluminum.
another material from the top five-listed Durable. Good resistance to corrosion
material because this system intends to reduce and fatigue.
the huge number of material candidates to a Nontoxic.

Figure 5. Material selection result


222 A Prototype of KBS for Material Selection in Bumper Beam Design

Workable. It uses well-understood component with minimum effort and time. With
metal working processes. 100 materials and 6 rules for each material to
Recyclable. It is important to save select the best material in manual selection, it
energy and be a benefit to the environ- would take a longer time. By adopting the KBS
ment. as the material selection tool the process takes
Conductive. less time compared to the manual method. The
developed KBS supported the designer in
Conclusion selecting the best material for the bumper beam
component.
This project showed that the Knowledge-Based
System for material selection of the bumper References
beam is a new system that has been developed
to help design engineers to select the most Ashby, M.F. and Cebon, D. (1993a). Material
suitable materials. The results have shown that selection in mechanical design, [online].
the KBS in material selection gives a signifi- Available from: http://www.grantadesign.
cant impact in the material selection process. com/download/papers/university/gen1.
The selection process is based on the selection pdf. Accessed date: April 30, 2006.
criteria, which in turn in based upon the defined Ashby, M.F. (1993). Materials Selection in
constraint values from material engineers. If Mechanical Design. Pergamon Press,
the constraint values are satisfied, then those Oxford, UK, p. 65-83.
particular materials are selected as the best Sapuan, S.M. and Abdallah, H.S. (1998). A
materials. This revealed that the system was prototype knowledge based system for the
capable enough to manipulate the variation material selection of polymeric-based
of input. Visual Basic (VB) supported by composites for automotive components,
object-oriented programming and rule-based Composites A: Apply Science Manu-
reasoning was used to develop KBS for facturing, 29A(7):731-742.
material selection and was reliable and easy to Sapuan S.M., Jacob, M.S.D., Mustapha, F., and
use. The KBS works in full interactive mode Ismail, N. (2002). A prototype knowledge-
and gives feedback information to the user to based system for material selection of
follow the design procedures. ceramic matrix composites of automotive
The development of the KBS for material engine components. Material and Design,
selection enables designers to choose the 23(8):701-708.
best material candidate for the bumper beam

S-ar putea să vă placă și