Ak
owledgebasedsystemformaterialsselectio
i
mecha
icale
gi
eeri
gdesig
SMSapua
AbstractThispaperstudiesvariousworko
thedevelopme
tofcomputerizedmaterialselectio
systemTheimporta
ceofK
owledgebasedsystemKBSi
theco
textofco
curre
te
gi
eeri
gisexplai
edThestudyofKBSi
materialselectio
i
a
e
gi
eeri
gdesig
processisdescribedThedevelopme
ti
materialsdatabaseswhichsometimesserveasmaterialselectio
packagesisalsodiscussedTheuseofKBSi
materialselectio
a
dtheapplicatio
i
thedomai
ofpolymericbasedcompositearechose
astypicalexamples2001PublishedbyElsevierScie
ceLtd
1I
troductio
Materialsselectio
isatask
ormallycarriedoutbydesig
a
dmaterialse
gi
eersGutteridgea
dWaterma
1describedtheaimofmaterialsselectio
astheide
tificatio
ofmaterialswhichafterappropriatema
ufacturi
goperatio
swillhavethedime
sio
sshapea
dproperties
ecessaryfortheproductorcompo
e
ttodemo
strateitsrequiredfu
ctio
atthelowestcostForthepurposeofmaterialselectio
thousa
dsofdatawouldbe
eededtocharacterizeallthegradesof
fmaterialsMa
yselectio
systemsareavailabletohelpdesig
e
gi
eerstochoosethemostsuitablematerialsAtthemostbasicleveldesig
e
gi
eerscouldusetablesofmaterialpropertiesi
databooksHoweverdatasheetsarei
completea
do
cepublishedtheyaredifficulttoupdateDodda
dFairfull2describedhowi
formatio
aboute
gi
eeri
gmaterialsca
bedividedi
totwomai
categoriesiedataa
dk
owledgeDataisdefi
edastheresultsofmeasureme
tswhereask
owledgereprese
tstheco
ectio
sbetwee
itemsofdatathesourceofthisk
owledgewhichco
tributestoa
u
dersta
di
goftheresultsBoththecomputerizeddatabasea
dtheKBSofmaterialselectio
willbedescribedi
thefollowi
gsectio
s
2Thematerialdatabaseformaterialsselectio
21Supporti
gdataa
drigorouslogicforthehypothesisa
dobseruatio

I
rece
tyearsatte
tio
isbei
gmadetotheuseofcomputersystemstostorea
dprocessdataregardi
gthepropertiesofmaterialsIte
ablesthedesig
erstoachievelargecapacitya
drapidretrievalfromacomputerdatabasetoprovideeasyaccesstothematerialsdataAccordi
gtoPrasad3materialsarethecommo
eleme
tse
umerati
gpartsi
aproductrealizatio
processNowadaysdesig
e
gi
eers
ormallyrelyo
thematerialsthattheyarefamiliarwithHoweverwhe
desig
requireme
ts
fexceedtheco
strai
tsofsuchmaterialsorexceedtheco
strai
tso
materialpropertiesco
curre
te
gi
eeri
gteamsmustco
sideralter
ativematerialsWithdirecto
li
eaccesstoamaterialsdatabasetheco
curre
te
gi
eeri
gteamscouldselectmaterialsthatarelighterstro
gera
dloweri
costAssumi
gthattheimpactofsuchsubstitutio
sca
bea
alyzedorsimulatedtheteamscouldeasilymakea
optimumselecr