packaging design, weld fatigue analysis and validation of
TRANSCRIPT
Mechanical Engineering Research; Vol. 8 No. 2; 2018 ISSN 1927-0607 E-ISSN 1927-0615
Published by Canadian Center of Science and Education
36
Packaging Design, Weld Fatigue analysis and Validation of Diesel Exhaust After-treatment System
Sonal Nareddiwar1, Tapobrata Dey1 & R. Sunilkumar2 1 Department of Mechanical Engineering, D Y Patil College of Engineering, Akurdi, 411044, Savitribai Phule Pune University, India 2 Cummins Technologies India Ltd., Pune, India Correspondence: Tapobrata Dey, Department of Mechanical Engineering, D Y Patil College of Engineering, Akurdi, 411044, Savitribai Phule Pune University, India. E-mail: [email protected] Received: August 9, 2018 Accepted: October 26, 2018 Online Published: November 30, 2018 doi:10.5539/mer.v8n2p36 URL: https://doi.org/10.5539/mer.v8n2p36 Abstract Diesel exhaust after treatment system is usually designed to meet stringent packaging constraints and emission norms. After treatment packaging has critical impact on the overall system efficiency and durability since many components in exhaust systems have welded joints. An after treatment inlet and outlet tube joints, connected to engine outlet and Original Equipment Manufacturer (OEM) tailpipe respectively are subjected to vibrations and bending moment leading to fatigue failure at the inlet/outlet welded joints. It has been observed over the years that the prevailing failure modes in after treatment systems are cracked welds at joints between inlet tubes and flanges, outlet tubes and connecting tailpipes. Fatigue failure is a complex and progressive form of local damage which occurs in welded components of exhaust after-treatment systems. Thus, this fatigue failure needs to be estimated accurately and at the early stage of design to save cost and time. But due to geometrical irregularities, compact packaging design and load transfer conditions, it becomes difficult to estimate accurate fatigue strength of the welded areas. Thus weld fatigue analysis, a high cycle fatigue test to validate inlet/outlet module of exhaust system against dynamic overturning bending moment and to calculate the location of minimum weld fatigue life within the inlet welded joints is performed. Weld fatigue analysis uses advanced fatigue assessment technique, BS 7608, Stress x Life (S x N) approach for accurate and precise estimation of welds. The present work deals with reducing the package volume of the after treatment system by applying different concepts, verifying design robustness by FEA simulation using ANSYS 18.2 and validating the structural durability of the system by testing. The objective of the present work is to estimate the fatigue life of the welded structures precisely and accurately, calculate the threshold bending moment to determine whether the design is robust to the bending moment loads seen over course of its life and make design modifications as per simulation result. Further the FEA and testing results of weld fatigue analysis are correlated. Keywords: Fatigue Analysis, After-treatment system, Fatigue life, Stress-based approach, packaging design 1. Introduction The diesel engine has the merits of high power output, low fuel consumption and excellent longevity. However, the fuel lean combustion also generates significant amount of hazard emissions, such as oxides of nitrogen and particle matters, exposure of which increases the risk of respiratory infection, heart problems, premature death and lung cancer. 2. Design Procedure Hence regulation on diesel engine emission has become more and more stringent, leading to a large volume of diesel aftertreatment catalysts. Catalyst packaging becomes a challenging task for diesel aftertreament system design. The criteria for an optimal packaging design are to: (1) fit large catalysts into the limited space of the vehicles, and (2) maintain the well utilization of packed catalysts. As emission regulations are getting stringent day by day, it is also leading to lighter and compact vehicles. The after treatment system is designed to be mounted either on engine or on chassis. Also while designing an after treatment system, it is needed to check the available space according to the mounting strategy. As per OEM space claim requirement, different concepts for compact packaging design are generated during brain storming sessions with the cross functional team. Using Pugh matrix, one best concept for
mer.ccsenet
compact pmodel for After desiinterferencfit check ocompared
3. Weld SThin stainincreasingstrength immost criticanalyzing assessing f2001). Thdistributioplastic behStructural concentratlocations s(Wei, YanFor analyzassessmenThe fatiguare classifprobabilitiThe curveequation (f
Where, N is the enS is the noKo is the cm is the sl∆ is a funcd is numbe
t.org
packaging is donewly selected
igning, OEM ce with the surron the basis ofto baseline de
tress and Fatinless steel comly used in exh
mprovement ancal failure movehicle’s stru
fatigue life of whe stress conce
n in welded johavior (Apurvastress methods
tions. The strusuch as weld t
ng, & Luo, 201zing the weldent approaches aue strength of wfied into severies of failure. Tes have a consfor the mean li
ndurance in cyominal stress raconstant for a pope of the S-N
ction of the staer of standard
own selected ad concept desifit check is drounding partsf feedback fromsign. The over
Figure
igue Life Assemponents with haust systems. nd characteristde occurring i
uctural durabilwelded compoentration in woints primarily a & Pande, 201s are now wideuctural stress ctoes or weld th4).
ed structures anare currently imwelded joints iral groups andThe figure 1 shstant slope betife of welded s
cles. ange. particular weld
N curve in log-andard deviatiodeviations from
Mechanical
and preliminarygn is shown as
done to verifys and keeping im OEM. Therrall system len
e 1. Preliminary
essment- BS76high heat resiBecause exha
tics and mechain the welded lity. However
onents, a reliabwelded joints d
depends on ge14).
ely used in fatigconcept is bashroats, and the
nd for fatigue mplemented. in steels structud for each grohows the S-N ctween 105 andstructure):
d classificationlog scale.
on of the fatigum the mean
Engineering Re
37
y CAD modelss in Figure 1. y whether the in mind the serre is considerangth has been r
y CAD model
608 Approachistance, high caust systems aranism involved
structures. Thwelded joints
le determinatiodominates the eometry, stress
gue life assessmsed on the assere are several
life estimation
ures is definedoup the characcurve defined f
d 107 cycles, w
𝑁 ∆
n
ue life
esearch
s are created u
system fits prviceability conable length redreduced by 12
for concept de
h corrosion resisre usually weld in welded joihus fatigue lifs are the weakon of fluctuatinfatigue behav
s concentration
ment of weldedsumption of a l variants of st
n of welds, ma
d by the Britishcteristic stress-for various cla
where the stres
sing Creo 2.0.
properly in thencern. The desiduction in the n%.
esign
stance and exclded structuresints (Oh, 2017fe assessment kest part of thng stress is reqvior of weldedn, loading case
d structures anglobal stress
tructural stress
any reliable and
h Standards 76- life curves aasses of weldedss-life relation
Vol. 8, No
The prelimina
e vehicle withign has passednew concept d
cellent workabs, more empha7). Fatigue is ois the critical he exhaust sy
quired (Alexand structure. The and material e
nd structures wdistribution a
s approaches a
d validated fat
608. In BS 760are defined ford joints.
nship is define
o. 2; 2018
ary CAD
hout any in OEM
design as
bility are asis is on ne of the step for
ystem. In dre et al., he stress elastic or
ith stress t critical available
tigue life
08, welds r various
d by the
(1)
mer.ccsenet
The variabBS approaTo accounassociatedthe mean v(BS7608,
Figu In the preswith a serion fatigue 4. Weld FWeld fatigdynamic oresulting sorder usedfatigue anamplitude model of tto simulatewithin the These strefurther cal
• The prfailure
• AThe passin
• Aa) In
Bb) Tc) Sd) T
t.org
ble d is termedach assumes ‘dnt for variabil
d probability ofvalue. Figure 1993).
ure 2. S-N curv
sent work, fatigies of classifiedbehaviour and
Fatigue Analysgue analysis is overturning bestructural stresd, as long as thenalysis is impo
in weld HAZthe welded bode the physical tstructure.
esses are then tlculates the weRequired Inpurimary input foe along with C Mome Numb Permis
Acceptance cring criterion forAnalysis assumn Fe-safe, fatig
BS7608 approaTemperature efSlip and separaThe effect of re
d as “design crd’ as a positiveity the weld f
f failure. The 2 shows weld
ves for Weld F
gue design andd weld S-N cud their mesh insis approach a high cycle f
ending momens calculations e overall geomortant because
Z is very less ady is built to mtesting. A stati
transformed ineld fatigue life uts: or weld fatigue
CAD models, lient Magnitude er of cycles = ssible probabiliteria: r acceptable anmptions: gue life calculaach with d=2 (2ffects are not cation between pesidual stress d
Mechanical
iterion” and ise if it is below tfatigue curvesstress-life (SN
d fatigue curve
atigue from B
d evaluation ofurves due to cansensitivity in s
fatigue test is unt. This analyswill be mesh-
metry of a compe weld strengas compared to
match the loadinic analysis is ru
nto the equivalbased on Mas
e analysis is moiners material p= 320Nm 1 million lity of failure =
nd safe design
ations are done2.2% of probabonsidered in thparts is assumedue to forming
Engineering Re
38
s expressed as the mean line
s given in varN) curves in BSes for a specifi
S 7608 for var
f welded jointsapability of constress determin
used to validatsis is used to insensitive, reponent is reaso
gth determineso ultimate tensng and constraun on each fin
lent structural ster S-N curve.
oment magnituproperties and
= 2.2 %
is 1e6 cycles a
e by considerinbility of failurhe simulation ed not to occurwill have little
esearch
number of sta(BS7608, 1993
rious weld fatiS7608 are definfic weld class
rious Probabili
s are primarily nsistently captunation at welds
te after treatmeevaluate life o
egardless of eleonably represes the assemblysile strength oaint conditionsite element mo
stress range an.
ude, number ofd boundary con
at 2.2% probab
ng steel materie).
r, justifying a le effect on me
andard deviatio3). igue methods
ned by standardwith varied pr
ities of Failure
carried out bauring the stress.
ent module inlof weld in termement size, eleented in a finitey strength andf the parent m
s of the physicaodel to determi
nd are fed as i
f cycles, perminditions.
bility of failure
ial properties o
linear model. an stress.
Vol. 8, No
ons from the m
are provided d deviations awrobabilities of
e (BS7608, 199
ased on structurss concentratio
lets and outletms of life cycement type, inte element modd also tolerab
material. Finite al tests, and thine the stresse
input in Fe-Sa
ssible probabil
e.
of weld class F
o. 2; 2018
mean life.
with an way from
survival
93)
ral stress n effects
s against cles. The tegration
del. Weld le stress element
en tested s present
fe which
lity of
F2 using
mer.ccsenet
e) V
f) Sh
g) S• I
In Fatiguethickness modelled uthe locatio
• WFatigue andrawings, welds. TheFigure 4 an
Table I. W
t.org
Variation in thi
Stress concentrave a negligib
Sensor bosses aImporting CA analysis, inletand extractingusing 4-node s
on of the weld
Weld details: nalyses of welthe details of
e weld betweend Table I sho
Weld Details of
ckness during
ations due to wle effect on da
are not weldedD geometry ant/outlet tube Cg mid surfacesshell elements.toes. Figure 3
Figure
ldments requirf different weldn inlet tube an
ows the weld d
Figur
f Inlet Tube JoiWeld Loca
Weld-1Weld-2Weld-3Weld-4Weld-5Weld-6Weld-7
Mechanical
forming opera
welded bosses amage. d in FE model. nd Solid to she
CAD model is is. The concern. The nodes ofshows the CA
e 3. CAD geom
re detailed knded areas are
nd surroundingdetails of inlet/o
re 4. Weld deta
int ation 1 W2 We3 4 5 6 Weld b7 Weld
Engineering Re
39
ation is neglec
and forces fro
Bosses share bell conversion:imported and cned welded arf the shell elem
AD geometry o
metry of inlet/o
owledge of thsought and ac
g components aoutlet tube join
ails of the inlet
Weld Weld between inleeld between inlet t
Weld between heWeld betweenWeld between
between inlet 2nd fd between inlet 2nd
esearch
ted.
om the accelera
bonded contac: converted intoreas whose fat
ments representf inlet tube ass
outlet sub-mod
he stress fieldsccordingly diffare modeled bynt.
t tube sub-mod
Name et tube and Marmotube and inlet flaneat shields at inlet flanges at inlet n Heat shields flange and 2nd head flange and outer
ation of wire, s
ct with parent b
o shell body bytigue life needting the weld asembly.
dule
s in critical referent shell boy using vertica
dule
on nges t
at shield r body
Vol. 8, No
straps, and insu
bodies.
y assigning appds to be evaluare offset (by t
egions. From todies are connal and inclined
o. 2; 2018
ulation
propriate uated are t/2) from
the weld ected by areas.
mer.ccsenet
• MIn Fatigueelements aTable II sh
Table II. D
• DContacts acontact typbodies bec
• It is assumand weld m
Table III: M
t.org
Meshing: e analysis, in thare less stiff anhows the mesh
Discretization d
Defining Contare defined so tpes, bonded cocause that allow
Assigning Mamed that the inlmaterial are as
Material Prope
he critical areand give realist
hing details.
details Element
3mm
tacts: that the bodies ontacts and Nows for rigid for
aterial Propertilet/outlet sub-msumed to be S
erties
Mechanical
as such as welic and accurat
Figure 5. Mes
Size Element m Quad4
Shell 1
stay in load trao-separation corce transfer. Fi
Figure
ies: module and all
SS409 and prop
ComponenElastic Modulus
Density, ƍ (kgYield Strength
Poisson’s r
Engineering Re
40
lded areas, verte results. Shel
shed inlet tube
type No. of No4 81
10726
ansfer path andontacts are of ligure 6 shows
e 6. Contact de
l other parts boperties are liste
nt Frame – Materia, E (MPa) g/m^3) h, (MPa) ratio
esearch
ry fine mesh oll 181 element
sub-module
odes No. of Ele3 13107
d offers realistilinear types wthe contact de
etails
oth are manufaed in Table III
al SS409 2.06e+5
7750 240 0.28
of quad elemets are used for
ements 73
ic force distribuwhich are used etails.
factured from S.
Vol. 8, No
ents is used. Ameshing. Figu
ution. Among for connecting
SS409. Both th
o. 2; 2018
lso quad ure5 and
different g various
he parent
mer.ccsenet
• 2 load stegeometricainvestigatemoment. Torthogonalsuperposedcycle. Fixed Supbody of thapplied to
Remote foshown in F
5. SimulatThe FEA r
t.org
Application oep approach: Mal configuratioe the influenceThe first step l input loadingd in Fe-safe to
ports: In Weldhe inlet subasse
constrain the e
orce: Also a reFig 8. Remote
tion Results results for the
f Loads and BMulti-axial strons. Thus, thee of the loads, in the Weld f
gs 1 m out fromo simulate the
d fatigue analyembly to simuexerting forces
emote force offorce is consid
1st iteration are
Mechanical
oundary condiress is always multi-axial etwo loading c
fatigue analysim the weld plloading of ea
ysis, fixed suppulate the forcess and fix the m
Figure 7. App
f 320N is appldered to be rig
Figure 8. App
e shown in Tab
Engineering Re
41
itions: s present in tffect on fatiguases are consiis is creating ane. The stres
ach element ov
ports are applis exerted by fix
model to avoid
plication of Fix
lied at 1000 mgid because of
plication of Re
ble IV.
esearch
the welded coue life should dered: axial tetwo load steps fields from t
ver the weld su
ied at 45° on txture at the timdeformation a
xed supports
mm distance tothe rigid conne
emote force
omponent withbe considered
ension load andps in ANSYS, these static solurface through
the top and botme of testing. at that location
o create momeections.
Vol. 8, No
h or without d properly. In d out of plane corresponding
lutions will behout the entire
ttom sides of tThe fixed suppas shown in F
ent on the inlet
o. 2; 2018
complex order to bending
g to two e linearly e loading
the outer ports are
Fig 7.
t tube as
mer.ccsenet
Table IV. FWe
Since weldthe accepta The Table Table V. F
We
•
Remove skincreased fThus, good
• Static analThe outpucreated byrespective bending m
t.org
FEA RESULTeld Location
Weld-1 Weld-2 Weld-3 Weld-4 Weld-5 Weld-6 WWeld-7 W
d 2 fails in 1st ance criteria.
V shows the 6
Fea results for eld Location
Weld-1 Weld-2 Weld-3 Weld-4 Weld-5 Weld-6 WWeld-7 W
Design modifkip weld and further, the wed to go with th
M-N curve plolysis results are
ut outlined in thy Fe-safe. This
number of fatimoment.
TS FOR 1st ITW
Weld between Weld between in
Weld betweeWeld betwWeld betw
eld between inlet Weld between inle
iteration, seve
6th iteration wi
6th iteration W
Weld between Weld between in
Weld betweeWeld betwWeld betw
eld between inlet Weld between inle
fications: apply full weleld fails. he iteration 6 d
ot: e read in Fe-sahis paper is a preveals the pe
igue life cycles
F
Mechanical
TERATION Weld Name
inlet tube and Manlet tube and inlet en heat shields at i
ween flanges at inlween Heat shields2nd flange and 2nd
et 2nd flange and o
eral design mod
ith design mod
Weld Name inlet tube and Ma
nlet tube and inlet en heat shields at i
ween flanges at inlween Heat shields2nd flange and 2nd
et 2nd flange and o
ld at weld 2 an
design.
afe for fatigue plot of stress ‘neak to peak stres is plotted in th
Figure 9. M-N
Engineering Re
42
Noarmon flanges
inlet let s
d heat shield outer body
difications we
difications pass
Noarmon flanges
inlet let s
d heat shield outer body
nd also the tub
life calculationnormal to the wess range expehe Fig 9 as M-N
Curve for Inle
esearch
o. of life cycles wi
re done and an
sing the accept
o. of life cycles wi
be length is in
n using BS 76weld toe’ over
erienced by a wN curve which
et sub-Module
ith 2.2 % probabil2.5 E+06
7.20 E+05 6.04 E+06
2.51 E+06 1.2 E+06 2.27 E+06 1.64 E+06
nalysis was run
tance criteria.
ith 2.2 % probabil3.17 E+06 1.18 E+06 5.12 E+06
2.87 E+06 1.2 E+06 2.07 E+06 1.64 E+06
ncreased by 6%
08 approach. r the course ofweld. The bendh is further used
Vol. 8, No
ity of failure (d=2
n until the wel
ity of failure (d=2
%. If the tube
f a single loadiding moment vd to calculate t
o. 2; 2018
2)
ld passes
2)
length is
ing cycle versus its hreshold
mer.ccsenet
• ThrlifeBen= 3333
• F
Fig 10 sho1E06 life cFatigue lifwelded joithe most crbetween hethese simulife cycles as per FEAmoment.
t.org
Calculation ofreshold bendine of the componnding moment33.74 Nm
3.74Nm bendinFatigue life plo
ows the fatigue cycles. fe plots for all wints of the currritical and proneat shield and f
ulation results iwithout failur
A results, inlet
f threshold benng moment is tnent. t required for 1
ng moment is tots of various
Figure 10
life plots for th
welded joints irent inlet tube dne to failure joflange having infer that with re. The probabitube design ha
Mechanical
nding momentthe moment ap
1 million life c
to be applied tcritical welds
(a)
(c)
(e)
0. Fatigue Life
he welded join
in the inlet subdesign have faints are weld 2fatigue life 1.297.8% confideility of three saas safe fatigue
Engineering Re
43
: pplied to get th
ycles of inlet t
to get 1 Millionat inlet tube jo
e plots for all w
nts in inlet sub
b module are shatigue life more2, at inlet tube a2 million meet ence interval (damples survivilife and is stru
esearch
he maximum f
tube = 27379 x
n life cycles. oint:
(b)
(d)
(f)
weld locations
module. All w
hown in Fig 10e than 1 millioand flange havthe acceptanced=2), all the wing 1.2 millionucturally durab
fatigue life cy
x ((1000000)(-0
at inlet
welds passes the
0 and summarion life cycles. Aving fatigue lifee criteria of 1 m
welds will survin life cycles durble and robust
Vol. 8, No
cles during the
0.319))
e acceptance c
ized in Table VAmong all these 1.18 million amillion life cycive more than 2ring test is 100to 333.74 Nm
o. 2; 2018
e service
riteria of
V. All the se joints, and weld les. Also 2 million 0%. Thus
bending
mer.ccsenet
6. ValidatThis is thehave adeqfatigue lifebut are nomanufactustrain to pr
• TIn order toto test a seof loadingpainted whtesting. ThThe bendinrotating shfrom a spinlocated beBefore startest is startRPM is ne Where, m is the wL is the moMb is the bR is the ra7. ValidatThe objecton the On according parametersexceeded twithout issThe test re
t.org
tion of Weld Fe standard test quate high cycle of aftertreatmot limited to lured using apprroduce high
F
Test Procedureo determine theeries of similar g cycles requirehite and all thhe after treatmng moment lo
haft. A spindlenning mass. D
etween the morting the test, ated. Post-test ineeded to be cal
weight of Rotatioment arm lenbending momedius of the rot
tion Results tive of this test and Off-Hwy ato Company
s for On- Higthe 1.2 millionsue. esults are show
Fatigue Analymethod used
le fatigue strenment inlet/outleeakage from wropriate produc fatigue failure
Figure 11. Test
e: e fatigue strengspecimens. Eaed to produce e welded jointent device is t
oad of 320N ise is attached toue to this, both
otor and the span exhaust inlenspection is colculated to run
ing mass ngth from centeent applied ating mass
is to validate tapplications. T’s standard teghway and Ofn cycle. Again t
wn as follows:
Mechanical
ysis to verify that tngth. Thus, thet welded tube within the tesction process. e. Figure 11 sh
setup for Weld
gth of a weldedach of the specfailure in each
ts are marked then installed o applied with
o the inlet/outleh rotating and bpindle that redet sub module ionducted and tat the bending
𝑅𝑃𝑀er of mass to f
the inlet designThe after treatmest procedure.ff-Highway apthe samples w
Engineering Re
44
the inlet or ouhe main purpos
joints. The outed inlet/outleThese samples
hows the test se
d Fatigue Ana
d joint configucimens is subjeh specimen is red in order to
on the Rotatingan eccentric met joint of an abending forces
duces vibrationis inspected fotest report is crg moment spec
𝑀 ∗first weld toe
n is robust to thment body joint The componpplications. A
were leak tested
esearch
utlet welded tuse of this test
utcome of this tet joint and ins are then subjetup for Weld
lysis of Inlet/O
uration under aected to consta
recorded. An o detect any dg bending test
mass which is after treatments are applied atn load from thr any leakagesreated. cified by the gi
∗ ∗
he bending momt components wnents were tes
fair amount d and inspected
ube joints of anis to determin
test will be failnternal cracks. ected to a loadFatigue Analy
Outlet of EGP
a given load coant amplitude lo
after treatmendeformation or
machine withconnected to t
t device and trat the inlet tube.
he motor into as or cracks. Tes
iven formula,
ment loads seewere tested forsted against tof component
d and continue
Vol. 8, No
n aftertreatmenne adequate hilure modes tha
The part needing that applieysis.
ondition, it is noading and thent system is cor cracks formeh the guillotinethe electric moansfers a mom. A vibration dafter treatmentst software is s
en over course or high-cycle fatthe recomments passed the d onto 2 millio
o. 2; 2018
nt device gh-cycle t include ds to be es a 360°
necessary e number ompleted d during
e clamps. otor by a
ment load damper is t device.
setup and
(2)
of its life tigue life
nded test test and
on cycles
mer.ccsenet
All three tor vibratiosystem durThe correlbetween si
t.org
test samples coons and showedring post inspelation of fatigimulation and
ompleted the 2d no internal aection. gue analysis ovalidation resu
Mechanical
Figure 12.
Figure 13.
Figure 14.
2.0 million testand external vi
of both by simults shows that
Engineering Re
45
Test data for s
Test data for s
Test data for s
t cycles at 320isible weld fati
mulation and vt design is robu
esearch
sample-1
sample-2
sample-3
0 N-m of test bigue crack in t
validation is sust and have s
bending load wthe Inlet sectio
shown in Tablafe fatigue life
Vol. 8, No
with no abnormon of the aftertr
le VI. The coe.
o. 2; 2018
mal jerks reatment
orrelation
mer.ccsenet.org Mechanical Engineering Research Vol. 8, No. 2; 2018
46
Table VI. Correlation of simulation and validation Weld Location Weld Name Simulation Results Validation Results
Weld-1 Weld between inlet tube and Marmon 3.17 E+06 2.0 E+06 Weld-2 Weld between inlet tube and inlet flanges 1.18 E+06 2.0 E+06 Weld-3 Weld between heat shields at inlet 5.12 E+06 2.0 E+06 Weld-4 Weld between flanges at inlet 2.87 E+06 2.0 E+06 Weld-5 Weld between Heat shields 1.2 E+06 2.0 E+06 Weld-6 Weld between inlet 2nd flange and 2nd heat shield 2.07 E+06 2.0 E+06 Weld-7 Weld between inlet 2nd flange and outer body 1.64 E+06 2.0 E+06
8. Conclusion The down selected concept for compact packaging design of after treatment system shows considerable length reduction by 12 % as compared to baseline design. The FEA results also states that with 97.8% confidence interval (d=2), all the welds will survive more than 1 million life cycles without failure with the design modification in tube length. It is also ensured about validation through FEA analysis of welds that the probability of three samples surviving 1.2 million life cycles during test is 100%. It is also inferred from the fatigue simulation that 333.74 N-m threshold bending moment is to be applied to achieve maximum service life of the inlet components. The validation results infer that all the three test samples have passed the safe design criteria of 1.2million life cycles without any cracks or failure. The correlation of Weld fatigue simulation and testing have 2% variation in both the results considering number of life cycles. Thus as per FEA results and testing, inlet tube design has safe fatigue life (reduced warranty claims) and is structurally durable and robust. Acknowledgment I express my sincere gratitude towards the faculty members. I am also thankful to Cummins India Ltd. for sponsoring and supporting in my project. References Oh, G. (2017). Bending Fatigue Strength and the Effect of Assembling Stress on Fillet Welded Joints of Catalyst
Muffler Flange Pipes. International Journal of Automotive Engineering, 8(2), 87-94. Alexandre, S. F., Camarão, A. F., Fernandes, A. D. O., Bezerra, M. A., Argentino, M. A., & Ripoli, R. R. (2001). A
Finite Element Approach for Fatigue Analysis of Welded Components (No. 2001-01-4077). SAE Technical Paper.
Apurva, A., & Pande, D. (2014). Mesh Insensitive Structural Stress Approach for Welded Components Modelled using Shell Mesh. International Journal of Engineering Research & Technology, 3(11), 1425-1434.
Wei, Z., Yang, F., & Luo, L. (2012). Fatigue life Assessment of Welded Structures with the Linear Traction Stress Analysis Approach. SAE international Journal of Material Manufacturing, 5(1), 183-193.
Code of practice for Fatigue design and assessment of steel structures, British standard, BS7608, 1993. Wei, Z., Hamilton, J., Yang, F., Luo, L., Lin, S., Kang, H., & Dong, P. (2013). Comparison of Verity and Volvo
Methods for Fatigue Life Assessment of Welded Structures (No. 2013-01-2357). SAE Technical Paper. Kyuba, H., & Dong, P. (2005). Equilibrium-equivalent structural stress approach to fatigue analysis of a rectangular
hollow section joint. International Journal of Fatigue, 27(1), 85-94. Kang, H. T., Dong, P., & Hong, J. K. (2007). Fatigue analysis of spot welds using a mesh-insensitive structural stress
approach. International Journal of fatigue, 29(8), 1546-1553. Gao, Y., Chucas, D., Lewis, C., & McGregor, I. J. (2001). Review of CAE fatigue analysis techniques for spot-welded
high strength steel automotive structures (No. 2001-01-0835). SAE Technical Paper. Stocco, D., Vilela, D., & Batalha, G. F. (2009). Spot weld fatigue durability performance evaluation through the use of
fea (No. 2009-36-0189). SAE Technical Paper.
mer.ccsenet.org Mechanical Engineering Research Vol. 8, No. 2; 2018
47
Zhang, S. (2001). Recent developments in analysis and testing of spot welds (No. 2001-01-0432). SAE Technical Paper.
Copyrights Copyright for this article is retained by the author(s), with first publication rights granted to the journal. This is an open-access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/).