Materials Today: Proceedings xxx (xxxx) xxx

Contents lists available at ScienceDirect

Materials Today: Proceedings

journal homepage: www.elsevier.com/locate/matpr

Prediction of spot weld fatigue life using finite element approach

Bhuvaneswaran S., Padmanaban R. ⇑
Department of Mechanical Engineering, Amrita School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore, India

a r t i c l e i n f o a b s t r a c t

Article history: Resistant spot welding is predominantly used in automotive industry for joining sheet metal. Mostly
Received 11 November 2019 automobile fatigue failure occurs in and around spot welding (80%). So, accurate prediction of the defor-
Received in revised form 9 December 2020 mation, stresses and fatigue life of spot weld is necessary. In this work finite element model of spot weld
Accepted 19 December 2020
and experimental study of coach peel test of spot welded low carbon steel sheets are presented. The prop-
Available online xxxx
erty change in various zones of spot weld is found by nano-indentation test followed by inverse finite
element modelling. Spot weld failure happens at the interface between nugget and heat affected zone
Keywords:
due to the geometric stress concentration and property change that occurs during the welding process.
Spot-weld
Heat affected zone
Solid elements are successful in capturing the geometric stress concentration, but they increase the sim-
Cold rolled steel ulation cost and hence a model with combined shell and solid elements is used. For nugget solid elements
Coach peel test are used and tore present all other parts of spot weld shell elements are used. Tensile and fatigue tests on
Finite Element Model coach peel specimen are done as per ASTM standardD903-98. The simulation results are found to be in
close agreement with experimental results.
Ó 2021 Elsevier Ltd. All rights reserved.
Selection and Peer-review under responsibility of the scientific committee of the International Mechan-
ical Engineering Congress 2019: Materials Science.

1. Introduction between breaking load and plug diameter. Min et al. [3] analyzed
the effect of parameters on the microstructural and mechanical
Resistant spot welding (RSW) is an important joining method characteristics of spot welded ferritic steel. Infrared camera was
used for sheet metal fabrication by automobile industry for several used to investigate the fracture phenomenon during tensile test.
decades. A typical car consists of about 5000 to 8000 spot welds. Ertas et al. [4] investigated the fatigue life of low carbon steel
RSW is a method that joins sheets together by applying pressure spot welds by conducting experiments and numerical studies. Fati-
and electric current to the metal sheets using copper electrodes. gue tests were performed using modified tensile shear specimens
The higher thermal conductivity and low electric resistance of cop- under varying load ranges. Finite element simulations used nonlin-
per compared to the work pieces, generates heat that melts the ear constitutive relations, plastic deformation and residual stresses
sheet metal and joining the sheets locally. The process is easy to to determine the stress and strain under cyclic loading. Khan et al.
operate and more suited for automation, qualifying as a good join- [5] report that the mechanical properties and microstructure of
ing method for mass production. Around 80% of automobile fatigue resistance spot welded advanced high strength steels (AHSS) are
failure occurs in and around spot weld. So, it is important to accu- superior to that of high strength steels. Harder microstructures
rately predict the stress and deformation due to spot weld (SW). were observed in the fusion zone of the AHSS. Chen et al. [6] resis-
Kong et al. [1] utilized numerical modeling of RSW combined with tance spot welded aluminium alloy AA6022 with zinc coated low
indentation tests to predict, the joint deformation, properties, ef- carbon steel and investigated the failure modes, mechanical
fect of nugget size and sheet thickness on the tensile shear strength strength and fracture mechanism. High welding current and short
of the spot welded joint. Han et al. [2] joined AA5754 using resis- welding time aided growth of aluminum nugget and also mini-
tance spot welding with different process parameters to produce mized intermetallic compound thickness improving mechanical
different joint stack-ups and varying weld quality. Joint strength properties. Huin et al. [7] resistance spot welded dual phase steel
was evaluated for three different test geometries and relationship and martensitic steel and investigated the failure modes and load
bearing capacity during cross tension and tensile shear tests. Sim-
ilar and dissimilar thickness configurations are studied. Sachin
⇑ Corresponding author.
et al. [8] investigated RSW of steel and developed an equivalent
E-mail address: [email protected] (R. Padmanaban).

https://doi.org/10.1016/j.matpr.2020.12.816
2214-7853/Ó 2021 Elsevier Ltd. All rights reserved.
Selection and Peer-review under responsibility of the scientific committee of the International Mechanical Engineering Congress 2019: Materials Science.

Please cite this article as: Bhuvaneswaran S. and Padmanaban R., Prediction of spot weld fatigue life using finite element approach, Materials Today: Pro-
ceedings, https://doi.org/10.1016/j.matpr.2020.12.816
Bhuvaneswaran S. and Padmanaban R. Materials Today: Proceedings xxx (xxxx) xxx

Nomenclature

E Elastic Modulus e Strain

d Diameter n Strain Hardening Exponent
q Density

simplified weld model using material characterization for mapping of shaft a palm kernel-cracking machine and investigated the effect
spot weld in a crash simulation. Jafari et al. [9] introduced FEA of of different loading. The evolution of mechanical properties and
RSW and reviews the governing equations, material property, microstructure of welded steel joints is influenced by inter-
boundary conditions and different numerical solutions for RSW. metallics phases, changes in HAZ, residual stresses, particle addi-
The electrical contact resistance is found to be the most influential tion and PWHT. Numerical simulation of spot weld will present
property. details like, thermal history, residual stress distribution, fatigue life
Song et al. [10] evaluated the load at the spot weld by numerical estimate etc., and aid in fast and economic design of spot welds.
simulation using finite element modeling, accounting residual Studies evaluating the fatigue life of spot welded joints by consid-
thermal stress and compared the results with crash analysis model ering the properties of different zones are not available. The aimof
results. Wu et al. [11] investigated the spatial distribution, this work is performing numerical simulation of resistance spot
microstructure, hardness and evolution of TMAZ in resistance spot welding and explore the material property of different zones, find
welded AA6111-T4 to discern the properties of different regions. the stress distribution in different zones and predict fatigue life of
Different fracture modes during the tensile- shear and coach-peel the spot weld. The fatigue life predicted is compared with experi-
tests are also studied. Dancette et al. [12] used finite element mental determined fatigue life.
method to simulate RSW of dual phase steels and determined weld
thermal cycles and influence of sheet thickness on the magnitude 2. Coach peel test
of heating and cooling rates. Experimental investigations were car-
ried and the microstructure and constitutive behavior were Coach peel test is an experimental method to understand the
explored. Rahman et al. [13] studied the fatigue life and location behavior and region of crack initiation in a spot weld. The test
of the weakest spot-welds in high strength steel due to variable was performed with a universal testing machine (UTM). For the
amplitude loading. The sheet thickness and the spot diameter were experiment, work piece with dimension according to ASTM stan-
found to be influencing the stress distribution near the spot welds. dard D903-98 as shown in Fig. 1 is used. The sheet thickness is
Morawski et al. [14] deals with the finite element discretization of 1.78 mm and material used is cold rolled steel with ayield strength
spot weld using simple solid elements. It was observed that using of 280 MPa. A displacement controlled UTM is used. The cross head
more number of solid elements capture the stress distribution sat- speed for the test is selected as 1 mm/minto capture the displace-
isfactorily. Shan et al. [15] combined micro-indentation tests and ment accurately. Finally, a graph is plotted between the displace-
finite element simulation to predict the local stress –strain ment and force as shown in Fig. 2.
response and microstructures in a high-pressure die cast AM60
Mg-alloy. Venkatesh et al. [16] examined the validity of indenta-
3. Fatigue test of spot weld
tion test for estimating elastic plastic properties, and employed
the test on a range of materials to examine the sensitivity and
Fatigue life of spot welded coach peel specimen is tested at
specified guidelines for extraction of material properties.
room temperature (strain controlled mode) using a closed loop
Prakash et al. [17] simulated room temperature indentation
servo-hydraulic fatigue test machine with digital controller unit.
testing of nickel thin film on steel using finite element method.
Low carbon steel specimen with dimension according to ASTM
The mechanical behavior of the specimen subjected to single and
D903-98 standard is used. The specimen is made up of two sheets
cyclic indentation is studied. Arivarasu et al. [18] laser welded
in L- shape and joined together by spot weld. The spot weld is done
Alloy 825 and report full penetration with joint very low heat
with a squeeze time of 99 cycles and weld time of 22 cycles, and a
input. Microstructure, microhardness and mechanical properties
hold a time of 90 cycles. The force applied by the electrodes is
of the joint are reported. Undamaged fusion zone has equiaxed
grain structure and higher hardness value. Akhtar et al. [19] char-
acterized various zones of HAZ for boron modified P91 steel using
Gleeble and report that the presence of boron influenced the nat-
ure and fraction of phases. PWHT resulted in hardness change
and shift in the type of failure to type-III. Akhtar et al. [20] simu-
lated sub-heat affected zone of P91B steel to study the phase trans-
formations occurring after welding and post weld heat treatment
(PWHT). Grain boundary embrittlement occurs after welding.
Grain size modification, presence of borides and boron enriched
particles produce phase transformation and enhance GB hardening
after PWHT. The life of a spot weld depends on the condition of the
base metal and effects introduced by various operations performed
before welding. Chinnaraj et al. [21] presents the effect of coiling
and uncoiling of sheets on the residual stress and fatigue life of
truck frame rail sections. Pan et al. [22] proposed a method based
of fatigue data and finite element method for the prediction of fati-
gue life of spot welds connecting sheets of different thickness. Afo-
labi et al. [23] utilized finite element method to predict fatigue life Fig. 1. Coach peel test specimen.

2
Bhuvaneswaran S. and Padmanaban R. Materials Today: Proceedings xxx (xxxx) xxx

Table 2
Spot weld setup.

Name Values
Squeeze time 99 cycles
Weld time 22 cycles
Hold time 90 cycles
Electrode force 6670 N (1500 lb)
Cap size, truncated 7.9 mm
Material Cold Rolled steel

Table 3
Fatigue testing.

Frequency 5 Hz
R- ratio (Pmax/Pmin) 0.1
Fig. 2. Force versus displacement plot. Load amplitude 457 N
Cycles to failure 17,000 cycles

6670 N. The specimen is secured by two hydraulic operated grip-

pers placed at the top and bottom of the specimen at a height of
30 mm. Constant amplitude sinusoidal wave load with the R-
ratio of 0 is applied and the operating frequency was 5 Hz. Failure
was noted near intersection of HAZ and nugget around 17,000
cycles. Material properties, details of spot welding, and fatigue
testing specifications are respectively given in Table 1, Table 2
and Table 3.

4. Results and discussion

4.1. Nano-indentation and inverse finite element modeling

Vickers hardness testing machine is used for measuring hard-

ness of spot weld. At various locations along the midline of the
weld, starting from BM (Base Metal), HAZ (Heat Affected Zone)
and Nugget Nano- indentation test is done as presented in Fig. 3.
Fig. 3. Indentation depth Vs indentation load.
Fig. 4 shows hardness distribution of different zones of spot weld-
ing. The hardness value is calculated from the diagonal length mea-
sured from the indentation. The indentation load is kept between 2
and 5 N. From the original base metal plastic property, the new
material property is calculated by slightly increasing the yield
value. The stress strain curve is finally found by comparing the
experimental test with finite element (FE) model with inverse FE
modelling technique as shown in Fig. 5. So, for Nugget, HAZ and
base metal different material properties are used as shown in
Fig. 6.
The discretization of spot-weld has been a challenging task. The
simplest finite element modeling technique for representing spot-
weld is using beam element as shown in Fig. 7. All the elements
corresponding to nugget are joined together by rigid body element.
Then the rigid body elements are joined together by beam element.
But, this method fails to capture stress concentration. By using
solid elements, stress concentration can be captured. But using
solid elements for the entire model will increase the simulation

Table 1 Fig. 4. Variation of hardness with distance.

Sheet metal properties.

Name Values cost. So, a combination of shell and solid elements are used to
Material name Cold Rolled Steel model the nugget.
Carbon 0.15 Fig. 8 represents the stress distribution obtained by different
Manganese 0.60 finite element modelling techniques. It can be seen that spot-
Sulphur 0.035
Phosphorus 0.050
weld with cluster solid nugget modelling techniques capture the
Yield Strength 200 MPa stresses better than other techniques for the same load case. The
UTS 306 MPa mesh convergence happened for 0.5 mm element size as shown
Sheet thickness 1.6 mm in Fig. 9. At 0.5 mm mesh with convergenceoccurring and devia-
Hardness 50–60 HRC
tion happening to be less than 5%.
3
Bhuvaneswaran S. and Padmanaban R. Materials Today: Proceedings xxx (xxxx) xxx

Fig. 5. Inverse finite element modelling.

4.2. Finite element simulation of coach peel test

A 3D finite element model is created and correlate with the

actual experiment. One side of the sheet metal is encastered and
for the other side except Y axis displacement, and other displace-
ments are arrested. Force is given as input and corresponding reac-
tion forceis got from the output. The FEM simulation is compared
with the actual tensile test and the results are plotted as shown
in Fig. 10.
When comparing results from experimental with modelling
results, force and displacement are almost same. Thus the shell
and solid cluster modelling for the representation of spot weld cor-
relates well with experimental results. Hence this model can be
used for fatigue analysis of spot weld. Fig. 11 shows the finite ele-
Fig. 6. Stress strain curves.

Fig. 7. Types of finite element model for spot weld.

4
Bhuvaneswaran S. and Padmanaban R. Materials Today: Proceedings xxx (xxxx) xxx

Fig. 8. Stress distributions in various models.

concentration on the heat affected zone causes crack initiation in

the spot weld. This model also gives the stress distribution in
and around the spot weld. The maximum stress is 201 MPa. This
finite element simulation is further used to predict the fatigue life
of spot-weld.

4.4. Fatigue life estimation

The finite element simulation for a single cycle, shown in Fig-13

is used for fatigue calculation. For the fatigue life calculation N-
Code design life software is used. The finite element simulation
indicates that the material has already reached its yield strength
and fatigue life cycle is less than 104cycles. Therefore, strain-life
low cycle fatigue approach is used to calculate the failure cycle
Fig. 9. Mesh convergence. of the spot weld. Fatigue crack initiation occurs along the plane
where maximum tensile stress occurs, and the fatigue life calcula-
tions are done using the stress and strain states in the specimen at
which the maximum tensile stress developed. Table 4 shows the
material property used for the fatigue life calculation of spot-
weld. For mean stress correction Morrow’s approach, as given in
Eq. (1) is used. This method is used to move the elastic part of
the material curve up and down according to the mean stress of
each cycle.
ðr0f  rm Þ
b
c
ea ¼  2Nf þ e0f  2N f ð1Þ
E
Whereea is strain amplitude, Nf is the number of reversal to failure, E
is Young’s Modulus, r’f and e’f are true strength and strain at frac-
ture, rm is mean stress, and b,c are material constants.

4.5. Comparison of predicted and measured fatigue life

The fatigue life calculation is done by strain based approach.

The red elements shown in Fig. 14 represent the crack initiation
Fig. 10. Comparison of experimental and simulation result.
zone at the failure cycle. The minimum life predicted during the
fatigue life calculation was 18,452 cycles. The numerical part of
the study was verified by comparing the predicted fatigue life
ment simulation of coach peel test and Fig. 12 shows the experi- model with the experimental one and a closer agreement was
mental coach peel specimen. obtained. The crack initiation is observed at HAZ zone which
matches with the experimental crack initiation zone.
4.3. Finite element simulation for predicting fatigue life
5. Conclusions
The fatigue life prediction models require accurate calculation
of stress and strain due to spot welding. For this purpose, commer- In this study spot weld is modelled with shell and solid ele-
cial FEA software ABAQUS is used. Residual stresses occurring dur- ments. This modelling technique is used to represent the spot weld
ing the formation of spot weld is included by changing the material for performing coach peel and tensile test. Fatigue test was con-
property of heat affected zone and nugget of spot weld. The mate- ducted by applying various loads. The numerical analysis corre-
rial change is captured by micro indentation test followed by the lates with the experimental test. A three dimensional finite
inverse FE modelling. In Fig. 13 stress concentrations between element model was created for the representation of spot weld
the nugget and heat affected zone is captured well. The high stress which includes the effect of mechanical load during the welding
5
Bhuvaneswaran S. and Padmanaban R. Materials Today: Proceedings xxx (xxxx) xxx

Fig. 11. Coach peel finite element simulations.

Fig. 12. Finite element model and test specimen. Fig. 13. Finite element simulation of spot weld.

Table 4
process. Design Life N-code software is used to calculate the fati- Fatigue properties and constants.
gue life of spot weld with strain-life (Low Cycle Fatigue) method.
Properties Value Expansion
By this study, we arrive at the following conclusions:
YS 200 Yield strength (Mpa)
UTS 306 Ultimate Tensile Strength (MPa)
 Shell and solid nugget modeling technique predicts the stress
E 2.06 e5 Elastic Modulus (N/mm2)
concentration in the Heat Affected Zone of spotweld. Me 0.29 Elastic Poisson’s Ratio
 The material property in the heat affected zone and nugget vary r’f 580 Fatigue strength Coefficient (MPa)
and the stress concentration at the interface between heat B 0.09 Fatigue strength Exponent
affected zone and nugget initiates failure. C 0.57 Fatigue ductility Exponent
e’f 0.51 Fatigue ductility coefficient
 The stress induced, deformationpatterns, crack initiation zone k’ 393 Cyclic strength coefficient (MPa)
and fatigue life are found by numerical simulation and experi- Ρ 7750 Density (kg/m3)
mental testing. N 0.11 Cyclic strain hardening
 Numerical fatigue life prediction closely correlates with exper-
imental fatigue life.
Further studies can be made by considering the effect of resid-
ual stress for the evaluation of fatigue life.

6
Bhuvaneswaran S. and Padmanaban R. Materials Today: Proceedings xxx (xxxx) xxx

References

[1] X. Kong, Q. Yang, B. Li, G. Rothwell, R. English, X.J. Ren, Mater. Des. Vol-29
(2008) pp. 1554-1561.
[2] L. Han, M. Thornton, D. Boomer, M. Shegold, J. Mater. Process. Technol. 211
(2011) pp.513-521.
[3] Ding Min, Liu Shi-Sheng, Hao Hong, Peng tao, zhang Pei-lei. Mater. Des. Vol-52
(2013) pp. 353-358.
[4] A.H. Ertas, O. Vardar, F.O. Sonmez, Z. Solim, J. Eng. Mater. Technol. 131 (1)
(2009) 011011.
[5] M.I. Khan, M.L. Kuntz, E. Biro, Y. Zhou, Mater. Trans., Vol-49 No 7 (2008)
pp.1629-1637.
[6] Nannanchen, Hui-Ping Wang, Blair E. Carlson R. Sigler, Min Wang, J. Mater.
Process. Technol., Vol 243(2017) pp.347-354.
[7] Thibaut Huin, Sylvain Dancette, Damein Fabregue, Thomas Dupuy, M.D.
PImetals. Vol- 6(2016) No 5,111.
[8] Sachin Patil. ‘‘Modeling, and characterization of spot weld material
configuration for vehicle crash analysis” Phd Dissertation2014,Wichita
StateUniv.
[9] Mehdi Jafarivardanjani, Jacek Senkara, AlirezaArayee., Przeglad Spawalnictwa
Vol-88 (2016).
[10] J.H. Song, H. Huh, H.G. Kim, S.H. Park, Int. J. Automotive Technol., Vol. 7No. 3
(2006) 329336.
[11] Wu Sainan, Bita Ghaffari, Elizabeth Hetrick, Mei Li, Qing Liu, Zhihong Jia, J.
Mater. Process. Technol. 249 (2017) 463–470.
[12] Sylvian Dancette, Veronique Massardier-Joudan, Damien Fabregue, Jacques
Merlin, Thomas Dupuy, Merlin Vouzekri., ISIJ Int., Vol-51(2011)99-107.
[13] Mustafizur Rahman, Rosli Abu Bakar, M.M. Noor, Shahrir Sani, Eur. J. Sci. Res.
22 (3) (2007) 444–456.
[14] Jon Morawski, Beat Lauterbach, Ilka Schwarzer, Laia Ramon-Villalonga, ANSA
& ETA Int. Conf. 12(2005)81-95.
Fig. 14. Contour plot of predicted fatigue life. [15] Z. Shan, A.M. Gokhale, Mater. Sci. Eng., Vol-361(2003) 267-274.
[16] T.A. Venkatesh, K.J. Van Vliet, A.E. Giannakopoloulos, S. Suresh, Scripta Mater.,
Vol- 42(2000)833-839.
[17] R.V. Prakash, S. Arunkumar, A. Ramesh, Int. Mech. Eng. Congress Exposition 3
CRediT authorship contribution statement (2011) 745–750.
[18] M. Arivarasu, P. Roshith, R. Padmanaban, S. Thirumalini, K.V. PhaniPrabhakar,
G. Padmanabham, Can. Metall. Quart. 56 (2) (2017) 232–244.
Bhuvaneswaran: Formal analysis,investigation and writing-
[19] M. Akhtar, A. Khajuria, V.S. Kumar, R.K. Gupta, S.K. Albert, Phys. Met. Metall.
original draft. Padmanaban: Methodology and writing - review 120 (7) (2019) 672–685.
& editing. [20] M. Akhtar, A. Khajuria, J.K. Sahu, J. Swaminathan, R. Kumar, R. Bedi, S.K. Albert,
Appl. Nanosci., Vol-8 No 7(2018), 1669-1685.
[21] K. Chinnaraj, M. Sathya Prasad, C. Lakshmana Rao, R. Padmanaban, J. Eng. Sci.
Declaration of Competing Interest Technol. 14 (3) (2019) 1161–1172.
[22] N. Pan, S. Sheppard, Int. J. Fatigue 24 (5) (2002) 519–528.
The authors declare that they have no known competing finan- [23] S.O. Afolabi, B.I. Oladapo, C.O. Ijagbemi, A.O. Adeoye, J.F. Kayode, J. Mater. Res.
Technol. 8 (1) (2019) 105–111.
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.