StressLimitsi
Desig
Howlargeca
wepermitthestressestobeOrco
verselyHowlargemustapartbetowithsta
dagive
setofloadswhataretheoverallco
ditio
sorlimitsthatwilldetermi
ethesizea
dmaterialforapartDesig
limitsarebasedo
avoidi
gfailureoftheparttoperformitsdesiredfu
ctio
Becausediffere
tpartsmustsatisfydiffere
tfu
ctio
alrequireme
tstheco
ditio
swhichlimitloadcarryi
gabilitymaybequitediffere
tfordiffere
teleme
tsAsa
examplecomparethedesig
limitsforthefloorofahousewiththoseforthewi
gofa
airpla
eIfweweretodetermi
ethesizeofthewoode
beamsi
ahomesuchthattheysimplydid
otbreakwewould
otbeveryhappywiththemtheywouldbetoo‘spri
gy’Walki
gacrosstheroomwouldbelikewalki
gouto
adivi
gboardObviouslyweshouldbeco
cer
edwiththemaximum‘deflectio
’thatweasi
dividualsfi
dacceptableThislevelwillberathersubjectivea
ddiffere
tpeoplewillgivediffere
ta
swersI
factthesamepeoplemaygivediffere
ta
swersdepe
di
go
whethertheyarepayi
gfortheflooror
otA
airpla
ewi
gstructureisclearlydiffere
tIfyoulookouta
airpla
ewi
dowa
dwatchthewi
gduri
gturbule
t
fweatheryouwillseelargedeflectio
si
factyoumaywishthattheyweresmallerHoweveryouk
owthattheimporta
tissueisthatof‘structurali
tegrity’
otdeflectio
Wewa
ttobeassuredthatthewi
gwillremai
i
tactWewa
ttobeassuredthat
omatterwhatthepilota
dtheweatherdothatwi
gwillco
ti
uetoactlikeagooda
dproperwi
gI
factwereallywa
ttobeassuredthatthewi
gwill
everfailu
dera
yco
ditio
sNowthatisaprettytallorderwhok
owswhatthe‘worst’co
ditio
smightbeE
gi
eerswhoarerespo
sibleforthedesig
ofairpla
ewi
gstructuresmustk
owwithsomedegreeofcertai
tywhatthe‘worst’co
ditio
sarelikelytobeIttakesgreatpatie
cea
ddedicatio
forma
yyearstoassemblee
oughtestdataa
dfailurea
alysestobeabletopredictthe‘worst’caseThege
eralprocedureistodevelopstatisticaldatawhichallowustosayhowfreque
tlyagive
co
ditio
islikelytobee
cou
teredo
ceevery1000hoursoro
ceevery10000hoursetcAswesaidearlierourobjectistoavoidfailureSupposehoweverthataparthasfailedi
servicea
dweareaskedWhy‘Error’assuchca
comefromthreedisti
ctlydiffere
tsourcesa
yorallofwhichca
causefailure
f1Errori
desig
Wethedesig
ersorthedesig
a
alystsmayhavebee
abittoooptimisticMaybeweig
oredsomeloadsmaybeourequatio
sdid
otapplyorwere
otproperlyappliedmaybeweoverestimatedthei
tellige
ceoftheusermayweslippedadecimalpoi
t2Errori
ma
ufactureWhe
adevicei
volvesheavilystressedmemberstheeffectivestre
gthofthemembersca
begreatlyreducedthroughimproperma
ufactur