IEEE TRANSACTIONS ON COMPONENTS, PACKAGING AND MANUFACTURING TECHNOLOGY, VOL. 2, NO.

6, JUNE 2012 925

Workpiece Property Effect on Resistance

Spot Welding
Je-Ee Ho, Peng-Sheng Wei, Senior Member, IEEE, and Tzong-Huei Wu

Abstract— The weldability of resistance spot welding affected Heating a metal during welding affects its structure, deter-
by different thermal, physical, and metallurgical properties mined by the welding time, fusion zone shape, temperature,
is extensively investigated by realistically computing transient cooling rate, and the composition of the metal [9], [10].
mass, momentum, energy, species, and magnetic field intensity
transport in the alloy workpieces and electrodes. The properties Short time for welding of any material obviously reduces the
considered are the thickness, radius, equilibrium partition coeffi- extent of the heat-affected zone. It will increase the rate of
cient of workpieces, electrode-to-workpiece electrical conductivity cooling, and enhance the hardening of the weld, tendency
ratio, thermal conductivity ratio, and a joule heat-to-enthalpy to form shrinkage cracks, and porosity. Weld hardness can
change parameter. Resistance spot welding has been widely used increase rapidly with small increase in carbon content in
in joining thin workpieces in various electronic packaging and
manufacturing industries. Understanding of physical mechanisms steels, while porosity also increases with overheating. Long
for easily manipulating and controlling weld qualities in advance welding time, on the other hand, will reduce cooling rate
is important. This paper accounts for electromagnetic force, and, in some instances, the weld hardness, but may give
heat generations due to contact resistances at the electrode– rise to excessive indentation of the electrodes. Expulsion thus
workpiece interface and faying surface between workpieces, susceptibly occurs for high welding current and time [11].
and temperature-dependent bulk resistance of the workpiece.
The contact resistances are functions of hardness, temperature, The grain structures after solidification in welding are
electrode force, and surface condition. The computed results determined by not only cooling conditions but also solute
in general dimensionless expressions show that the welding is distributions. The direction of the columnar grains is always
feasible or onset time is shortened by decreasing thickness, radius, along the direction of heat flow during cooling. Since the
equilibrium partition coefficient of the workpiece, and electrode- cooling rate is the highest in the direction of electrode, the
to-workpiece electrical conductivity and thermal conductivity
ratios, and increasing the joule heat-to-enthalpy change para- directions of columnar grains are roughly perpendicular to the
meter. The corresponding heat transfer and species patterns are welded sheet metal [11]–[13]. Gould and Chang [13] observed
also presented. that in resistance spot welding of low carbon steels the
Index Terms— Nugget formation, phase change, physical solidification structure ranged from cellular at the fusion zone,
properties, resistance spot welding, thermal properties. heat-affected zone boundary, to cellular dendritic at the faying
surface, as a consequence of constitutional supercooling. The
finest cell/dendrite spacing was near the fusion zone–heat-
I. I NTRODUCTION affected zone interface, and the coarsest spacings were at the
weld centerline. Based on a 1-D thermal model accounting for
R ESISTANCE spot welding has been widely used in
joining thin and small-sized workpieces in electronics
packaging and manufacturing fields [1]–[4]. The materials to
melting, temperature-dependent material properties, effective
heat transfer in the liquid, electrode geometry and contact
be welded are brought together under pressure by a pair of resistance, and measured temperature, primary dendrite spac-
electrodes, and a heavy current is passed through them. In ing can be predicted. Wang et al. [12] summarized that the
view of electrical resistance, local heat at the faying surface nugget microstructure of most metals, such as steel, Mo, Ni,
causes the materials to become molten. The molten nugget Ti alloy, and Cu alloy, is a columnar dendritic structure and
grows until current flow is terminated. The joint is completed that of a few metals (i.e., Al alloy) is columnar grains around
via solidification due to cooling through electrodes [5]–[8]. periphery and equiaxed grains in the center. The nugget of
Mg alloy only consists of an equiaxed dendritic structure in
Manuscript received August 3, 2011; revised October 5, 2011; accepted the absence of columnar dendritic grains, which are different
November 1, 2011. Date of publication December 21, 2011; date of current from other materials. Some pores can also occur on the edges
version May 29, 2012. Recommended for publication by Associate Editor I. of weld nugget, and large pores were in the center of weld zone
Fidan upon evaluation of reviewers’ comments.
J.-E. Ho is with the Department of Mechanical and Electro-Mechanical in refractory alloy 50Mo–50Re (wt%) sheets, as observed by
Engineering, National Ilan University, Yilan 260, Taiwan (e-mail: Xu et al. [14]. The weld joints in alloys produced by powder
jeho@niu.edu.tw). metallurgy often contain voids produced by residual volatile
P.-S. Wei is with the Department of Mechanical and Electro-Mechanical
Engineering, National Sun Yat-Sen University, Kaohsiung 80424, Taiwan materials. In order to reduce shrinkage, solidification crack,
(e-mail: swei@mail.nsysu.edu.tw). and shrinkage cavities, improve grain structures, and enhance
T. H. Wu is with the Department of Mechanical Engineering, Yung tensile-shear strength of the weld nugget, Li et al. [15] and
Ta Institute of Technology and Commerce, Pingtung 316, Taiwan (e-mail:
wux0064@gmail.com). Shen et al. [16] imposed external magnetic field to achieve
Digital Object Identifier 10.1109/TCPMT.2011.2175226 good qualities.
2156–3950/$26.00 © 2011 IEEE

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 926 IEEE TRANSACTIONS ON COMPONENTS, PACKAGING AND MANUFACTURING TECHNOLOGY, VOL. 2, NO. 6, JUNE 2012

Welding current
Measurements of the temperature distribution and the cool-
ing rate are difficult during resistance spot welding due to L4 r0
its short duration and invisible nugget. Numerical simula- Coolant hole
tion therefore becomes an effective tool for understanding L3
L2 Electrode rc
resistance spot welding process in quantitative details. Based L1 rs
re
on a heat conduction model subject to Poisson’s equation z
Molten nugget Workpiece
for electric potential, temperature and potential fields can L
be effectively predicted [17]–[23]. More realistic studies of Faying surface rb
thermal and magneto-fluid dynamics can be found in Li et al.
0 r
[24], [25], and Wei and Yeh [26]. The latter further predicted
solute distributions after solidification during resistance spot
welding with a dc uniform current density and specified
dynamic electrical resistance as a linear decreasing function
of temperature [27], [28].
Since electric currents spread strongly on entering the work- Fig. 1. Sketch for resistance spot welding and coordinate system.
piece, the induced electromagnetic force gives rise to strong
convection in the molten nugget. Extending the model from workpieces and electrodes on weldability of resistance spot
Wei and Yeh [26], Wei et al. [29] accounted for electrical and welding. Good weldability also indicates if the nugget initiates
magnetic fields, and found the effects of electrical conductiv- earlier. This paper accounts for heat generations due to contact
ity, magnetic Prandtl number, electrode geometry and working resistances at the electrode–workpiece interface and faying
parameters on fluid and temperature fields in the liquid, solid, surface between workpieces, and temperature-dependent bulk
and mushy zones. Wang and Wei [30] further included relevant resistance of the workpieces. The contact resistances are func-
electrical contact and bulk resistance models to simulate tions of hardness, temperature, electrode force, and surface
realistic transport processes during resistance spot welding. condition. Taking into account transport processes induced by
Thermal properties for different alloys play important roles electromagnetic force, the nugget shape and solute concen-
in different nugget growth behavior. Feng et al. [31] proposed tration distribution after solidification and cooling rate, which
an axisymmetric finite element model for predicting temper- are essentially required for analyzing microstructures of weld
ature fields and nugget growth in resistance spot welding of nugget in metallurgical and materials fields, can be realistically
AZ31B Mg alloy. Also, the onset time for nugget formation presented. The fabrication, packaging, and manufacturing of
of Mg and Al alloys was about the same, which was less electronic components and their quality controls become more
that that of mild steel. The heat loss rate of magnesium alloy reliable.
from the weld region was smaller than that of aluminum alloy
and greater than that of mild steel. The welding current level II. S YSTEM M ODEL
required for magnesium alloy was slightly lower than that for
Resistance spot welding, as illustrated in Fig. 1, is in a cylin-
aluminum alloy, but higher than that for low carbon steel.
drical coordinate system with the origin at the intersection of
Wei et al. [32] also found that the nugget formation can be
the axisymmetric axis and lower electrode–workpiece surface.
delayed and heat transfer is reduced by increasing the solid-
The welding current is AC. Workpieces are binary alloys of
to-liquid thermal conductivity ratio and liquid-to-solid specific
radius rb and total thickness L. The electrode has a truncated
heat ratio.
cone face with radius re and outer radius ro , whereas the
Simulation of resistance spot welding should account for not
coolant hole has outer radius rc . The major assumptions [30]
only material properties, but also realistic ac current, dynamic
are as follows: 1) heat generations at the faying surface and
resistance, thermal convection, and magneto-hydrodynamics
electrode–workpiece interfaces are considered as volumetric
[30]. A major defect for dc spot welding is expulsion, which
heat sources; 2) thermal contact conductance is neglected due
is a consequence of incessant heat generation, and high
to its insignificant effect on nugget formation [34]; 3) the
temperature and liquid pressure [33]. The dynamic resistance
mushy zone is a porous medium in local phase equilibrium;
takes the sum of the bulk resistance of the two workpieces,
4) workpieces are mixtures in phases; and 5) properties
and contact resistances at the faying surface and two electrode–
are averaged in distinct phases within temperature ranges
workpiece interfaces. It should reflect an initial drop due to the
considered. In this paper, the model equations and results
rupture of the surface film, an increase of contact resistance
are expressed in dimensionless forms. This is because, when
and bulk resistance due to the heating of the workpiece, and
investigating resistance spot welding, providing dimensionless
the subsequent fall after the maximum resistance has been
quantities subject to independent working parameters is more
reached due to the reduction of the increasing rate of the bulk
general, systematic, and concise, without considering specific
resistance and decrease of the constriction resistance at the
materials.
contact surfaces [30]. Since electric currents spread strongly
on entering the workpiece, the induced electromagnetic force
gives rise to strong convection in the molten nugget [29]. A. Mathematical System Description
This paper is to apply the previous computer program With the above assumptions, dimensionless continuity,
[30] to investigate the effects of different properties of the momentum, energy, species, and magnetic field intensity

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 HO et al.: WORKPIECE PROPERTY EFFECT ON RESISTANCE SPOT WELDING 927

equations [30], respectively, become gβT (T̃ sol −T̃ e )r̃ 3o T̃0
Gr ≡ , θ0 ≡
∂ρ ν̃
2
T̃sol −T̃ e
+ ∇ · (ρVm ) = 0 (1)
∂τ βs f˜m,0
α
I 2 μ0 μr
Fo ≡ , Lo ≡
∂ρu m βT (T̃ sol −T̃ e ) ρ̃ α̃2
+ ∇ · (ρVm u m ) = Pr ∇ · (ρ∇u m )
∂τ c̃  ρ̃ η̃ ν̃
 C ≡ , ≡ 2 , Sc ≡ α (7)
Pr ρ(1 − g )2 ∂p c̃s r̃o h f D̃
− um − + Pr 2 Gr θ0 (T − Tsol )
Da 3
g ∂z
 σ   Lo ∂H where dimensionless parameters Pr, Pr m , Da, Gr, and Sc are
σ
+F0 f  − f ,Tsol − 2 μr H (2) the Prandtl, magnetic Prandtl, Darcy, Grashof, and Schmidt
π ∂z
numbers, K 0 the permeability constant, ν̃ , α̃ , D̃α , and η̃
∂ρv m the momentum, thermal, solutal, and magnetic diffusivities of
+ ∇ · (ρVm v m ) = Pr ∇ · (ρ∇v m ) the liquid, and βT and βs the thermal and solutal expansion
∂τ
coefficients, respectively. The last terms on the right-hand side
Pr ρ(1 − g )2 ∂p Lo μr ∂r H
− vm − − 2 H (3) of (2) and (3) stand for Lorentz force, whereas the last term on
Da g3 ∂r π r ∂r the right-hand side of (4) combines heat generation at contact
surfaces and in bulk workpieces. Determinations of distinct
∂ρh m
+ ∇ · (ρVm h m ) = C∇ · (K ∇h m ) + C∇ regions of the mushy zone, full liquid, and solid have been
∂τ presented in detail by Wei and Yeh [26]. Energy and magnetic
· [K ∇ (h s − h m )] − ∇ · [ρ (h  − h m ) Vm ] field intensity equations in electrodes are, respectively
     
Lo 1 1 1 ∂r H 2 ∂H 2   2
+ 2 + + (4) ∂ TE k E ks RC ∂ 2 TE 1 2 ∂ TE
π σc σ r ∂r ∂z = + + M
∂τ ρE RE ∂ζ 2 N2 ∂ξ 2
  
∂ρ f mα   Pr   ∂ 2 TE 1 d N 1 ∂ TE
+ ∇ · ρVm f mα = ∇ · ρ f ∇ f mα +2M + − 2M
∂τ Sc ∂ξ ∂ζ ξ N + rc dζ N ∂ξ
Pr  α      
+ ∇ · ρ f  ∇ f − f m −∇ · ρ f α − f mα Vm
α
(5) Lo ∂ HE ∂ HE 2
Sc + 2 M +
π ρE RE σE ∂ξ ∂ζ
   2 
∂μr H 1 HE 1 ∂ HE
+ ∇ · (μr H Vm ) = Pr m ∇ · ∇H + + (8)
∂τ σ ξ N + rc N ∂ξ
  
H ∂ 1 H μr Hvm   2
+Prm − 2 + (6) ∂ HE ∂ 2 HE 1 2 ∂ HE ∂ 2 HE
r ∂r σ σr r = Prm η E + + M + 2M
∂τ ∂ζ 2 N2 ∂ξ 2 ∂ξ ∂ζ
where transport variables Vm , u m , v m , h m , f mα , and H are,   
1 d N 1 ∂ HE HE
respectively, dimensionless velocity vector, velocity compo- + − 2M − (9)
nents in axial and radial directions, enthalpy, solute content, ξ N + rc dζ N ∂ξ (ξ N + rc )2
and magnetic field intensity. Subscripts m, , and s denote where subscript E represents electrodes, dimensionless elec-
mixture, liquid, and solid phases, whereas superscript α repre- trode temperature TE ≡ T̃E /T̃0 , spatial coordinates ξ ≡ [r −
sents solute species. All lengths and time in (1)–(6) are nondi- rc (z)]/[rs (z)−rc (z)] and ζ ≡ z − L, and geometrical functions
mensionalized by electrode outer radius r̃o , and time for ther- M ≡ −(drc /dζ + ξ d N/dζ )/N and N ≡ rs (ζ ) − rc (ζ ),
mal diffusion r̃o2 /α̃ , respectively. Dimensionless enthalpies of dimensionless electrode thermal and electrical conductivities
mixture, solid, and liquid are, respectively, h m ≡ h̃ m / h f , k E ≡ k̃ E /k̃s and σ E ≡ σ̃ E /σ̃ , and enthalpy R E = c̃ E T̃0 / h f .
h s = RT ≡ c̃s T̃0 T / h f , and h  = RC T + R(1 − C)Te + 1, Magnetic field intensity governing ac at the top of the upper
mixture thermal conductivity K ≡ g + gs ks , magnetic field electrode is
intensity H = H̃ π r̃o /I , and solute content f mα ≡ f˜mα / f˜m,0
α ,
α
solute content at the solidus temperature f,Tsol , solute mass 1 ξ 2 (1 − rc ) + 2rc ξ
HE = − sin 2πωτ (10)
fraction in the liquid f α , and solid fraction f s , where h f is 2(1 + rc ) ξ(1 − rc ) + rc
the latent heat at the eutectic point, c̃, the specific heat of
solid, ks ≡ k̃s /k̃ the dimensionless solid thermal conductivity, where dimensionless current frequency ω ≡ ω̃r̃o2 /α̃ .
T̃0 , T̃sol , and T̃e the dimensional initial, solidus, and eutectic Equation (10) is obtained by assuming uniform current across
temperatures, f˜mα and f˜m,0 α the solute content and initial solute the top surface. Magnetic field intensity in the coolant hole
content, g and gs the volume fraction of liquid and solid, vanishes due to negligible current. At the electrode–workpiece
μ0 and μr the free and relative magnetic permeabilities of interface, continuities of tangential electric and magnetic field
liquid, and I the welding current. Dimensionless parameters intensities are required. Magnetic field intensity at the surface
are, respectively, defined as follows: between workpiece surfaces and surroundings is satisfied by
negligible electric current escaped to the surroundings. Mag-
ν̃ η̃ Ko netic field intensity vanishes at the axisymmetric axis to avoid
Pr ≡ , Prm ≡ , Da ≡ 2
α̃ α̃ r̃o infinite current.

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 928 IEEE TRANSACTIONS ON COMPONENTS, PACKAGING AND MANUFACTURING TECHNOLOGY, VOL. 2, NO. 6, JUNE 2012

TABLE I
B. Electrical Contact Resistance
T YPICAL VALUES OF PARAMETERS
Electrical contact resistance is strongly affected by thermal
and mechanical effects due to temperature-dependent electrical Constriction resistance parameter, R1 30
properties and hardness [30], [35]. A local dimensionless Film resistance parameter, R2 30
electrical contact resistance for a single contact spot at Dimensionless electrode magnetic diffusivity, η E 0.154
the faying surface or electrode–workpiece interface can be Temperature ratio, θ0 1.0
simulated by [30] Effective thickness of heat source at contact surface, s 3.3 × 10−7
R1  sn Hv Solid-to-liquid thermal conductivity ratio, ks 1.0
Rc = n Hv + R 2 (11)
σ σf Parameter governing welding current, L o 4 × 107
Magnetic Prandtl number, Prm 3 × 104
where dimensionless parameters governing constriction and
film resistances are, respectively, defined as Dimensionless electrode tip radius, re 0.9
 Dimensionless maximum radius of coolant hole, rc 0.5
1 π H̃v0 r̃o H̃v0 Electrode enthalpy parameter, R E 0.59
R1 ≡ , R2 ≡ . (12)
2 R̃0 σ̃liq W W R̃0 σ̃liq Dimensionless Curie temperature, Tc 1.63
Dimensionless eutectic and melting temperatures, Te , Tm 5.0, 6.0
In (11) and (12), dimensionless electrical contact resistance
Solid-to-liquid permeability ratio, μrs /μr 3, 500
Rc ≡ R̃c / R̃0 , electrical conductivities of workpieces and oxide
Dimensionless electrode density, ρ E 1.28
or impurity film σ ≡ σ̃ /σ̃liq and σ f ≡ σ̃ f /σ̃liq , hardness
Hv ≡ H̃v / H̃v0, film thickness s ≡ s̃/r̃0 , and number of contact Electrical conductivity ratio, σsol 1.05
spot n at interfaces, σ̃liq and R̃0 the liquid electrical conductiv- Current frequency, ω̃ 60 Hz
ity and electrical contact resistance at the ambient temperature,
H̃v0 the hardness at the ambient temperature, and W the elec-
trode force, respectively. Choosing a load W = 3000 N, hard- to the total internal energy rise and heat convected to the
ness H̃v = 5 × 109 N/m2 , and the total number of contact spot electrodes, coolant hole, and surroundings.
n = 14940, an effective radius of a contact spot is found to be
 III. R ESULTS AND D ISCUSSION
W
ã = = 3.6 × 10−6 m (13) In this paper, the effects of workpiece and electrode prop-
πn H̃v erties on nugget shapes, heating and cooling rates and species
which is a reasonable comparison to the electrode radius concentration during resistance spot welding are realistically
of r̃o = 3 × 10−3 m [30]. Hardness, film resistance, and studied. A rather complete computer program with a successful
bulk resistivity in (11) are functions of temperature [30], model for simulating dynamic electric resistance developed by
determined after the energy (4) is solved. Contact resistance Wang and Wei [30] is used to predict transient mass, momen-
thus increases with decreasing electrode force and increasing tum, energy, species, and magnetic field intensity transport
hardness and number of contact spots at interfaces. Hardness in workpieces and electrodes. Dynamic electrical resistance
also decreases with increasing temperature. The size of takes the sum of temperature-dependent bulk resistance of
contact spots on contact surfaces is also a function of the workpieces and contact resistances at the faying surface and
electrode force, total number of contact spots, and hardness. electrode–workpiece interfaces. The contact resistance is com-
Interactions between these factors should reveal the facts that posed of constriction and film resistances, which are functions
an increase in temperature decreases film resistance, while of temperature as well as force, hardness, and surface condi-
increases constriction resistance in workpieces. tions. High heat generation thus occurs near the interfaces and
joule heat in bulk workpieces.
Instead of dealing with specific materials, this paper pro-
C. Numerical Solution Approach vides a general dimensionless investigation to study resistance
Equations (1)–(6), (8), and (9) are discretized by the control- spot welding of materials subject to realistic working para-
volume, implicit finite-difference scheme with staggered grids. meters. Dimensionless parameters governing the resistance
A grid system 50 × 53 in workpieces and 27 × 72 in the spot welding considered are dimensionless total thickness
electrode, time step 2 × 10−5 are selected. The maximum (L), radius (rb ), equilibrium partition coefficient (k p ) of
deviation of computed resistances by using grid systems of the workpieces, electrode-to-workpiece electrical conductivity
50 × 53 and 72 × 75 is less than 1%. The axial direction ratio (σ E ), thermal conductivity ratio (k E ), and the joule heat-
is divided into uniform spaces while the spacing ratio in the to-enthalpy change parameter (). The total workpiece thick-
radial direction is 0.95. The convergence tolerances for global ness also stands for the separation distance between the upper
energy and velocity components and enthalpy, concentration, and lower electrodes. The equilibrium partition coefficient
and magnetic intensity fields were 10−2 and 10−3 , respec- represents the ratio of solute concentrations of the solid and
tively. If solutions converged, temperature distributions were the liquid in equilibrium at the interface. Since the electrodes
used to calculate resistances. Computations then proceeded are often made of copper, the electrode-to-workpiece electrical
to the next time. The global energy balance was that heat conductivity ratio and thermal conductivity ratio also reflect
generation in the workpieces and electrodes should be equal electrical and thermal conductivities of workpieces. The joule

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 HO et al.: WORKPIECE PROPERTY EFFECT ON RESISTANCE SPOT WELDING 929

3 4
Right side: Solidus Right side:
Melting Liquidus Melting Solidus
τ = 0.543 3.5 τ = 0.523 Liquidus
2.5 δliq = 0.4L δliq = 0.2L 1.
06

1.45
Left side:
3 Left side:

1.47
Cooling Cooling
2 τ = 0.616
τ = 0.590 2.5
δliq = 0 δliq = 0

L = 1.6 1.29
1.5 2
z

z
1.45 2.06
1.47 1.82
1.5 2.97 4.48
1 2.21

4.48
2.37 4.48
5.10 1

5.71
2.99
4.9
0.5 10 9 6.03

4.90
5. 5 .4 0

5.10
4.08
0.5

3.65

2.97
3.31
3.31

4.08

3.65
4.08 5.49 3.65 4.90
0 0
−2 −1 0 1 2 −2 −1 0 1 2
r r
(a) (b)
3 3
Right side: Solidus Right side: Solidus
Melting Liquidus Melting Liquidus
τ = 0.771 τ = 0.544
δliq = 0.4L 2.5 δliq = 0.4L
Left side: Left side:
Cooling 2 Cooling

1.50
2
τ = 0.818 τ = 0.595
δliq = 0 δliq = 0
rb = 4.66 1.5 kp = 0.4
z

z
4
1.7
9 9 1.50
1.
1 1 2.21
1.74
3.76 4.93
4.16 4.90

4.93
4.36
3

3.76

2.98
3.30
0.5
3.30
3.76
4.95 5.99
2.90

4.16

2.90

.3
1.99
1.74

3 6.09
4.16

6.01
1.74
1.99

4.9
0 4.36 5.33
0 0
−5 −4 −3 −2 −1 0 1 2 3 4 5 −2 −1 0 1 2
r r
(c) (d)

Fig. 2. Dimensionless isothermals in workpieces and electrode at time when nugget thickness δliq = 0.4 L and zero for (a) L = 0.8, rb = 2.33, kp = 0.75,
 = 2.5 × 10−5 , σ E = 6.5, and k E = 11.4, (b) L = 1.6, (c) rb = 4.66, and (d) k p = 0.4.

heat-to-enthalpy change parameter can be revealed by the period is around 6.1 corresponding to 1830 K by choosing
equation ( j 2 /σ )/(∂ρh/∂t) ∼ (I 2 /σ̃ r̃04 )/(ρ̃ h f α̃ /r̃02 ) ≡ Lo. the initial temperature to be 300 K. Since temperature gradient
The parameters governing contact resistances at the faying is high, the cooling effect in the electrode during the heating
surface are R1 = 30 and R2 = 30 estimated by choosing period is also significant. The spatial variations of temperatures
typical values R̃0 = 10−4 ohm, H̃v0 = 108 N/m2 , W = in workpieces and electrode during the cooling period are less
1000 N, σ̃liq = 106 mho/m, and r̃o = 10−2 m. The working than those during the melting period.
parameters used in this paper are listed in Table I, based on As the dimensionless total thickness of workpieces increases
the properties of iron. Metallurgical properties are chosen as from 0.8 to 1.6, the isothermals at two times when nugget
a binary alloy of iron and manganese. thicknesses are 0.2L and zero are shown in Fig. 2(b), respec-
The isothermals in the workpieces and electrode at dimen- tively. In comparison to Fig. 2(a), a less time of 0.523 is
sionless times of 0.543 (or 60 current cycles = 1 s) and required to reach the same thickness of the liquidus line.
0.590 (or 66 cycles) are shown on the right- and left-hand Complete solidification, however, needs more time, because
parts in Fig. 2(a), respectively. In this case, dimensionless time span for complete solidification is 0.09 in comparison
parameters L = 0.8, rb = 2.33, k p = 0.75, = 2.5 × 10−5 , with 0.05, see Fig. 2(a). The nugget shape is similar to
σ E = 6.5, and k E = 11.4. The times are, respectively, at the previous figures by increasing the workpiece radius, as shown
instants when thicknesses of the liquidus line are 0.4L during in Fig. 2(c). The nugget grows more slowly, whereas the
the heating and melting periods, and zero during the cooling time required for complete solidification maintains nearly the
and solidification periods. Evidently, the molten nugget shape same. The effects of the equilibrium partition coefficient on
delineated by either the liquidus line or the solidus line is the nugget shape can be seen in Fig. 2(d). It can be seen that
approximately a rectangle. The highest temperature taking the mushy zone is wider for a low partition coefficient by
place near the center of the workpieces during the melting referring to Fig. 2(a).

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 930 IEEE TRANSACTIONS ON COMPONENTS, PACKAGING AND MANUFACTURING TECHNOLOGY, VOL. 2, NO. 6, JUNE 2012

5 20 5 20
Standard case (L  0.8) Standard case (rb  2.33)
4.5 L  1.6 4.5 rb  4.66
4 Standard case: 4 Standard case:
Melting Melting
τ  0∼0.543 15 15
3.5 Cooling
3.5 τ  0∼0.543
Cooling
3 τ  0.543∼0.590 3 τ  0.543∼0.590
L = 1.6: rb = 4.66:
2.5 Melting 2.5
τ  0∼0.523 10 Melting 10
I τ  0∼0.771 I
q

q
2 Cooling 2
Cooling
τ  0.523∼0.616 τ  0.771∼0.818
1.5 1.5
qz
1 qr 5 1 5
qz
0.5 qz 0.5 qr qz
qr qr
0 0
−0.5 0 −0.5 0
−1 −1
0 0.5 1 1.5 0 0.5 1 1.5
τ τ
(a) (b)
5 20 5 20
Standard case (σE  6.5) Standard case (kE  11.4)
4.5 σE  19.5 4.5 kE  7.62
4 Standard case: 4 Standard case:
Melting Melting
τ  0∼0.543 15 15
3.5 Cooling 3.5 τ  0∼0.543
Cooling
3 τ  0.543∼0.590 3 τ  0.543∼0.590
σE  19.5:
2.5 Melting 2.5 kE = 7.62:
τ  0∼0.738 10 Melting 10
I I
q

2 Cooling 2 τ  0∼0.362
τ  0.738∼0.784 Cooling
1.5 1.5 τ  0.362∼0.412
1 qz 5 1 qz qz 5
qz qr qr
0.5 qr qr 0.5
0 0
−0.5 0 −0.5 0
−1 −1
0 0.5 1 1.5 0 0.5 1 1.5
τ τ
(c) (d)

Fig. 3. Dimensionless axial and radial heat fluxes as functions of time for different (a) workpiece thicknesses, (b) workpiece radii, (c) electrode-to-workpiece
electrical conductivity ratios, and (d) electro-to-workpiece thermal conductivity ratios.

The corresponding heat fluxes in radial and axial directions The effects of workpiece thickness on dimensionless thick-
at the liquidus line on the faying surface and axisymmetric ness and width of the molten nugget as functions of dimen-
axis for different workpiece thicknesses are shown in sionless time are shown in Fig. 4(a). It can be seen that
Fig. 3(a). Axial temperature gradient can be estimated to be an increase in workpiece thickness delays the formation of
5 × 105 K/m for steels if an electrode radius of 3.5 mm and the molten nugget. The onset time increases from 30 to 45
initial temperature of 300 K are chosen. This value agrees well cycles (0.5 to 0.75 s) as dimensionless workpiece thickness
with the measured thermal gradient of steels [13]. Heat flux in increases from 0.8 to 1.6, since more energy is required
the axial direction is roughly two times higher than that in the to raise temperature in a thick workpiece. The growth rate
radial direction. As a consequence, columnar microstructure at the onset time is very high, and gradually reduces its
can be seen in the nugget, as mentioned previously. This is magnitude. The nugget then solidifies rapidly within a few
the reason why conventional 1-D thermal model can still lead cycles immediately after the welding current turns off. The
to relevant results [13], [27], regardless of 50% of error. Heat growth of thickness is much less than that of width. It is also
fluxes in both directions for a thick thickness are found to be found that the nugget maintains high growth rate, whereas the
much less than those for a thin thickness. This is attributed nugget diminishes gradually as workpiece thickness increases.
to the fact that more energy is required to raise temperature Therefore, even though nugget formation is delayed, the time
for a thicker workpiece. The effects of workpiece radius and to reach the same liquidus thickness is shorter by increasing
the electrode-to-workpiece electrical conductivity ratio on the workpiece thickness, as can be seen from Fig. 2(a) and (b).
heat fluxes, however, are insignificant, as shown in Fig. 3(b) With the same reason, an increase in workpiece radius delays
and (c), respectively. Heat fluxes in both directions increase nugget formation. Growth rates in both directions, however,
with the electrode-to-workpiece thermal conductivity ratio, as are insensitive to the variation of workpiece radius, as shown
shown in Fig. 3(d). in Fig. 4(b).

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 HO et al.: WORKPIECE PROPERTY EFFECT ON RESISTANCE SPOT WELDING 931

20 20
1 Standard case (L  0.8) 1
Standard case (rb = 2.33)
L  1.6 rb = 4.66
Standard case:
0.8 Standard case: 0.8 Melting
Melting τ  0∼0.543
τ  0∼0.543 15 15
Cooling
Cooling 0.6 τ  0.543∼0.590
0.6 τ  0.543∼0.590
rb  4.66:
L = 1.6: Melting
0.4 Melting 10 0.4 τ  0∼0.771 10
δ

δ
δr τ  0∼0.523 I I
Cooling Cooling
0.2 δz τ  0.523∼0.616 0.2 τ  0.771∼0.818
δr δr δr
δz δz δz
0 5 0 5

−0.2 −0.2

−0.4 0 0
−0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5
τ τ
(a) (b)
20 20
Standard case (kp  0.75) Standard case (  2.5 × 10−5)
1 1   3.0 × 10−5
kp  0.4
Standard case: Standard case:
0.8 Melting 0.8 Melting
τ  0∼0.543 15 τ  0∼0.543 15
Cooling Cooling
0.6 τ  0.543∼0.590 0.6 τ  0.543∼0.590
kp  0.4: Σ  3.0 × 10−5:
Melting Melting
0.4 τ  0∼0.544 10 0.4 τ  0∼0.269 10
δ

δ
Cooling I Cooling I
τ  0.544∼0.595 τ  0.269∼0.316
0.2 δr δr 0.2 δr δr
δz δz δz δz
0 5 0 5

−0.2 −0.2
0 −0.4 0
−0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5
τ τ
(c) (d)
20
1 Standard case (σE  6.5) 1 Standard case (kE = 11.4)
σE  19.5 kE = 7.62
Standard case: Standard case:
Melting 0.8 Melting
0.8 τ  0∼0.543 15 τ  0∼0.543 15
Cooling Cooling
τ  0.543∼0.590 0.6 τ  0.543∼0.590
0.6
σE  19.5:
Melting kE  7.62
0.4 τ  0∼0.738 10 0.4 Melting 10
δ

Cooling I τ  0∼0.362 I
τ  0.738∼0.784 Cooling
0.2 δr δr 0.2 δr δr τ  0.362∼0.412
δz δz δz δz
0 5 0 5

−0.2 −0.2

0 0
−0.4 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5
τ τ
(e) (f)

Fig. 4. Dimensionless nugget thickness and width as functions of time for different (a) total workpiece thicknesses, (b) workpiece radii, (c) equilibrium
partition coefficients, (d) joule heat-to-enthalpy change parameters, (e) electrode-to-workpiece electrical conductivity ratios, and (f) electrode-to-workpiece
thermal conductivity ratios.

Fig. 4(c) shows that a decrease in the equilibrium coefficient on nugget growth and diminution rates in both
partition coefficient shortens nugget formation. This is directions are insignificant. An increase in the joule heat-
attributed to the decrease in the solidus temperature and to-enthalpy change parameter shortens the time for nugget
early formation of the mushy zone between the pure liquid formation, as shown in Fig. 4(d). Its effect on nugget growth
and solid phases. The effects of the equilibrium partition is slight.

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 932 IEEE TRANSACTIONS ON COMPONENTS, PACKAGING AND MANUFACTURING TECHNOLOGY, VOL. 2, NO. 6, JUNE 2012

3
Right side: Right side:
1.5 Melting Melting
τ  0.543 τ  0.523
δliq  0.4L 2.5 δliq  0.4L
Left side: Left side:
Cooling Cooling
τ  0.590 2 τ  0.616
1 δliq  0 δliq  0
L  1.6
1.5
z

z
1.0000 1.0000 1.0000 1.0000
1.0009 0.
0.5 0.9857 0.9
93
1 0.9
99
99
986

0.9997
9
0.9992

1.0013 98

0.9992
1.0013 0.9 6

0.9999
0.9 1.0018 1.0018 0 44
93 00 .00 1.0044
9 1.0 .9857 9
1
0 99
0.5 0.9

0 0
−1 −0.5 0 0.5 1 −1 −0.5 0 0.5 1
r r
(a) (b)

Right side: Right side:

1.5 Melting 1.5 Melting
τ  0.771 τ  0.544
δliq  0.4L δliq  0.4L
Left side: Left side:
Cooling Cooling
τ  0.818 τ  0.595
δliq  0 δliq  0
1 1
rb  4.66 kp  0.4
z

1.0000 1.0000
1.0000 1.0000

0.9
0.5 0.5 2

989
0.9
76

1.0
0

1.000 14 1.0 989

00

0
97

0.9772

22 1.0 0.9
77

06
00

1.00
1.0

0.9922
0.

1.0028 0
6

0.9
0

922 1.0028 22

0
2 1.0060
1.00 92 2

06
4
0.9 01

1.0
1. 772
0.9

0.9
98

0.9922 0.9922
9

0 0
−1 −0.5 0 0.5 1 −1 −0.5 0 0.5 1
r r
(c) (d)

Right side:
1.5 Melting
τ  0.362
δliq  0.4L
Left side:
Cooling
τ  0.412
1 δliq  0
kE  7.62
z

1.0000 1.0000
0.9 0.9
97
0.5 0.9
977 1.0006 99
9
7
0.9999
0.9999

1.0012
1.0 1.0012
006 06 977
0.9 1.00 0.9
977

0
−1 −0.5 0 0.5 1
r
(e)

Fig. 5. Dimensionless concentration distributions in workpieces at times when dimensionless nugget thickness δliq = 0.4L and zero for (a) L = 0.8,
rb = 2.33, k p = 0.75,  = 2.5 × 10−5 , σ E = 6.5, and k E = 11.4, (b) L = 1.6, (c) rb = 4.66, (d) k p = 0.4, and (e) k E = 7.62.

Fig. 4(e) shows that nugget formation is delayed by increas- and enhances cooling rate in the electrode. The result can
ing the electrode-to-workpiece electrical conductivity ratio. also be interpreted as an increased magnetic diffusion in
This can be interpreted as an increase in the electrode-to- the workpiece, shortening the response time and reducing
workpiece electrical conductivity ratio reduces heat generation the jump of electric current density [36]. The effects of the

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 HO et al.: WORKPIECE PROPERTY EFFECT ON RESISTANCE SPOT WELDING 933

2.1
convection, and distributed electric current density are also
Experimental data, Gould (1987) responsible for induced errors.
1.8 1-D modelling, Gould (1987)
Nugget thickness (mm) This paper
1.5
IV. C ONCLUSION
1.2 Instead of dealing with specific materials, this paper pro-
0.9
vides a general dimensionless investigation of resistance spot
welding of materials with different properties subject to
0.6 Expulsion realistic working parameters. The conclusions drawn are the
Expulsion following.
0.3
1) Welding is comparatively uneasy or nugget formation
0 is delayed by increasing thickness, radius, equilibrium
5 10 15 20 partition coefficient of the workpiece, and electrode-to-
Welding current (kA)
workpiece electrical conductivity and thermal conduc-
Fig. 6. Comparisons of nugget thickness versus welding current in welding tivity ratios, and decreasing the joule heat-to-enthalpy
AISI 1008 steel between 1-D predications and measurements from Gould [27] change parameter.
and this paper. 2) In all cases, the nugget growth and solidification are
greater, whereas the heat transfer rate is lower in the
radial direction than those in the axial direction. The
electrode-to-workpiece thermal conductivity ratio on nugget growth and heat transfer rates gradually decrease their
growths in both directions are shown in Fig. 4(f). In view magnitudes. After the power is off, the molten nugget
of enhanced cooling effect, an increase in the electrode-to- is frozen within a few welding cycles, leading to a very
workpiece thermal conductivity delays nugget formation. high cooling rate in both directions.
The distribution of solute concentration plays an important 3) Even though nugget formation is delayed, the growth
role in properties and microstructures of weldments [37], [38]. rate becomes higher and duration time for complete
Solute distributions in the nugget at times corresponding to the solidification is longer by increasing workpiece thick-
liquidus line thicknesses of 0.4L and zero are shown in the ness. In contrast to the insignificant effect for workpiece
right- and left-hand parts of Fig. 5(a), respectively. It can be radius, heat transfer or cooling rate is significantly
seen that solute distributions near the side boundary of the reduced as workpiece thickness becomes thick.
nugget are strongly affected by convection. Solute accumu- 4) The deficit region of solute concentration is expanded
lated ahead of solidification front is convected in the counter- with strong convection near the side boundary of the
clockwise direction around the boundary to the interior of the molten nugget by increasing thickness and radius of
molten nugget [29], leading to a deficit of solute content in a the workpiece and electrode-to-workpiece thermal con-
thin layer around the boundary and an excess of solute content ductivity ratio. The mushy zone becomes wide as the
in the interior of the nugget after solidification. Regardless of equilibrium partition coefficient decreases.
strong convection, solute distributions at two instant times are
similar due to rapid solidification. Fig. 5(b) and (c) shows the
deficit region of solute near the side boundary is expanded as R EFERENCES
the thickness and radius of the workpiece increase. The mushy [1] J. H. Lau and S. W. R. Lee, Chip Scale Package, CSP: Design, Materials,
zone is clearly seen to be wide by reducing the equilibrium Process, Reliability, and Applications. New York: McGraw-Hill, 1999.
partition coefficient, as shown in Fig. 5(d). Fig. 5(e) shows [2] S. J. Dong, G. P. Kelkar, and Y. Zhou, “Electrode sticking during micro-
resistance welding of thin metal sheets,” IEEE Trans. Electron. Packag.
that the solute distribution becomes relatively uniform as the Manuf., vol. 25, no. 4, pp. 355–361, Oct. 2002.
electrode-to-workpiece thermal conductivity ratio decreases. [3] W. R. Bratschun and J. L. Leicht, “Welding of plated, dissimilar metals
This can be attributed to an early formation of the molten for RF/EMI shielding,” IEEE Trans. Compon. Hyb. Manuf. Technol.,
vol. 15, no. 6, pp. 931–937, Dec. 1992.
nugget, see Fig. 4(f). [4] A. Monnier, B. Froidurot, C. Jarrige, P. Testé, and R. Meyer, “A mechan-
This model is confirmed from a good agreement with a ical, electrical, thermal coupled-field simulation of a sphere-plane elec-
measured relationship between nugget thickness and welding trical contact,” IEEE Trans. Compon. Packag. Technol., vol. 30, no. 4,
pp. 787–795, Dec. 2007.
current provided by Gould [27] in welding AISI 1008, as [5] N. T. Williams and J. D. Parker, “Review of resistance spot welding of
shown in Fig. 6. In this case, solute concentration fˆm,0 α
= steel sheets. Part 1: Modelling and control of weld nugget formation,”
0.32 wt% of Mn in a Fe–Mn alloy, total thickness of the two Int. Mater. Rev., vol. 49, no. 2, pp. 45–75, 2004.
[6] S. K. Khanna and X. Long, “Review-residual stresses in resistance spot
workpieces L = 0.715, electrode tip radius re = 0.9, and welded steel joints,” Sci. Technol. Weld. Join., vol. 13, no. 3, pp. 278–
dimensionless parameter governing Lorentz force (or welding 288, Apr. 2008.
current) Lo = 6.17×108. This paper shows a better agreement [7] R. W. Messler and M. Jou, “Review of control systems for resistance
spot welding: Past and current practices and emerging trends,” Sci.
with the experimental result than the 1-D prediction from Technol. Weld. Join., vol. 1, no. 1, pp. 1–9, 1996.
Gould [27]. The former neglected heat transfer in the radial [8] J. C. Tan, S. A. Westgate, and T. W. Clyne, “Resistance welding of thin
direction so that nugget thicknesses are overestimated. Fur- stainless steel sandwich sheets with fibrous metallic cores: Experimental
and numerical studies,” Sci. Technol. Weld. Join., vol. 12, no. 6, pp. 490–
thermore, an idealized evaluation of contact resistance, which 504, Aug. 2007.
decreases linearly with temperature, ignorance of thermal [9] S. Kou, Welding Metallurgy. New York: Wiley, 1987.

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.
 934 IEEE TRANSACTIONS ON COMPONENTS, PACKAGING AND MANUFACTURING TECHNOLOGY, VOL. 2, NO. 6, JUNE 2012

[10] Y. R. Wang, Z. H. Mo, J. C. Feng, and Z. D. Zhang, “Effect of welding [33] P. S. Wei and T. H. Wu, “Effects of electrical current on transport
time on microstructure and tensile shear load in resistance spot welded processes in resistance spot welding,” Sci. Technol. Weld. Join., vol. 15,
joints of AZ31 Mg alloy,” Sci. Technol. Weld. Join., vol. 12, no. 8, pp. no. 6, pp. 448–456, Aug. 2010.
641–646, Nov. 2007. [34] Y. B. Li, X. M. Lai, and G. L. Chen, “The influence of interfacial
[11] Z. Han, J. E. Indacochea, C. H. Chen, and S. Bhat, “Weld nugget thermal contact conductance on resistance spot weld nugget formation,”
development and integrity in resistance spot welding of high-strength Adv. Mater. Res., vols. 97–101, pp. 3239–3242, Mar. 2010.
cold-rolled sheet steels,” Weld. J., vol. 72, pp. 209–216, May 1993. [35] S. S. Babu, M. L. Santella, Z. Feng, B. W. Riemer, and J. W. Cohron,
[12] Y. R. Wang, J. C. Feng, and Z. D. Zhang, “Microstructure characteristics “Empirical model of effects of pressure and temperature on electrical
of resistance spot welds of AZ31 Mg alloy,” Sci. Technol. Weld. Join., contact resistance of metals,” Sci. Technol. Weld. Join., vol. 6, no. 3, pp.
vol. 11, no. 5, pp. 555–560, Sep. 2006. 126–132, 2001.
[13] J. E. Gould and P. H. Chang, “Thermal modeling of spot welds,” J. [36] P. S. Wei and T. H. Wu, “Magnetic property effect on transport processes
Metals, vol. 38, pp. 39–42, May 1986. in resistance spot welding,” J. Phys. D Appl. Phys., vol. 44, no. 32, p.
[14] J. Xu, X. Jiang, Q. Zeng, T. Zhai, T. Leonhardt, J. Farrell, W. Umstead, 325501, 2011.
and M. P. Effgen, “Optimization of resistance spot welding on the [37] D. R. Attorre, J. D. Parker, and N. T. Williams, “Mechanism of formation
assembly of refractory alloy 50Mo-50Re thin sheet,” J. Nucl. Mater., of solute bands in resistance seam welds of mild steel,” Mater. Sci.
vol. 366, no. 3, pp. 417–425, 2007. Technol., vol. 12, no. 8, pp. 697–703, Aug. 1996.
[15] Y. B. Li, Q. Shen, Z. Q. Lin, and S. J. Hu, “Quality improvement [38] J. Heuschkel, “Some metallurgical aspects of carbon steel spot welding,”
in resistance spot weld of advanced high strength steel using external Weld. J., vol. 26, no. 10, pp. 560–582, 1947.
magnetic field,” Sci. Technol. Weld. Join., vol. 16, no. 5, pp. 465–469,
Jul. 2011.
[16] Q. Shen, Y. B. Li, Z. Q. Lin, and G. L. Chen, “Impact of external
magnetic field on weld quality of resistance spot welding,” ASME J.
Manuf. Sci. Eng., vol. 133, no. 5, pp. 051001-1–051001-7, Oct. 2011.
[17] K. Matsuyama, “Simulation techniques in resistance spot welding,” J.
Jpn. Weld. Soc., vol. 70, no. 5, pp. 596–609, 2001. Je-Ee Ho received the Ph.D. degree from the
[18] U. Dilthey, H.-C. Bohlmann, V. Sudnik, V. Erofeev, and R. Kudinov, Department of Mechanical Engineering, National
“Calculation of welding areas and numerical simulation of the resistance Taiwan University, Taipei, Taiwan, in 2009.
spot welding process,” Weld. Cut., vol. 52, no. 1, pp. E12–E16, 2000. He is currently an Associate Professor with the
[19] A. De and M. P. Theddeus, “Finite element analysis of resistance spot Department of Mechanical and Electro-Mechanical
welding in aluminum,” Sci. Technol. Weld. Join., vol. 7, no. 2, pp. 111– Engineering, National Ilan University, Yilan, Tai-
118, 2002. wan. His current research interests include laser
[20] A. De, M. P. Thaddeus, and L. Dorn, “Numerical modelling of resistance machineries and investigation of magnetic flow in
spot welding of aluminum alloy,” ISIJ Int., vol. 43, no. 2, pp. 238–244, microchannels via finite difference computations.
2003.
[21] A. De, L. Dorn, and K. Momeni, “Probing the role of instantaneous
current waveform in numerical modelling of resistance spot welding
process,” Sci. Technol. Weld. Join., vol. 10, no. 3, pp. 325–334, Jun.
2005.
[22] J. A. Khan, L. Xu, and Y. J. Chao, “Prediction of nugget develop-
ment during resistance spot welding using coupled thermal–electrical–
mechanical model,” Sci. Technol. Weld. Join., vol. 4, no. 4, pp. 201–207, Peng-Sheng Wei (SM’08) received the Ph.D. degree
1999. from the Department of Mechanical Engineering,
[23] J. A. Khan, L. Xu, Y. J. Chao, and K. Broach, “Numerical simulation University of California, Davis, in 1984.
of resistance spot welding process,” Numer. Heat Transf. Part A: Appl., He is currently a Professor with the Department
vol. 37, no. 5, pp. 425–446, Apr. 2000. of Mechanical and Electromechanical Engineering,
[24] Y. B. Li, Z. Q. Lin, S. J. Hu, and G. L. Chen, “Numerical analysis of National Sun Yat-Sen University (NSYSU), Kaohsi-
magnetic fluid dynamics behaviors during resistance spot welding,” J. ung, Taiwan. His current research interests include
Appl. Phys., vol. 101, no. 5, pp. 053506-1–053506-10, Mar. 2007. magneto-fluid dynamics and heat transfers during
[25] Y. B. Li, Z. Q. Lin, Q. Shen, and X. M. Lai, “Numerical analysis resistance spot welding, interfacial plasma physics,
of transport phenomena in resistance spot welding process,” ASME J. formation of the defects such as humping, rippling,
Manuf. Sci. Eng., vol. 133, no. 3, pp. 031019-1–031019-8, 2011. spiking and porosity encountered in lasers, and elec-
[26] P. S. Wei and F. B. Yeh, “Factors affecting nugget growth with mushy- tron beam welding.
zone phase change during resistance spot welding,” J. Heat Transf., Dr. Wei was a fellow of the American Society of Mechanical Engineers
vol. 113, no. 3, pp. 643–649, Aug. 1991. in 2000 and the American Welding Society (AWS) in 2007. He received
[27] J. E. Gould, “An examination of nugget development during spot the Outstanding Research Achievement Awards from the National Science
welding, using both experimental and analytical techniques,” Weld. J., Council in 2004, NSYSU in 1991, 2001, and 2004, and the Adams Memorial
vol. 66, no. 1, pp. 1–10, 1987. Membership Award from AWS in 2008. He held the Xi-Wan Chair Professor
[28] P. S. Wei and C. Y. Ho, “Axisymmetric nugget growth during resistance of NSYSU in 2009.
spot welding,” J. Heat Transf., vol. 112, no. 2, pp. 309–316, May 1990.
[29] P. S. Wei, S. C. Wang, and M. S. Lin, “Transport phenomena during
resistance spot welding,” J. Heat Transf., vol. 118, no. 3, pp. 762–773,
Aug. 1996.
[30] S. C. Wang and P. S. Wei, “Modeling dynamic electrical resistance
during resistance spot welding,” J. Heat Transf., vol. 123, no. 3, pp.
576–585, Jun. 2001. Tzong-Huei Wu received the Ph.D. degree from the Department of Mechan-
[31] J. C. Feng, Y. R. Wang, and Z. D. Zhang, “Nugget growth characteristic ical and Electro-Mechanical Engineering, National Ilan University, Yilan,
for AZ31B magnesium alloy during resistance spot welding,” Sci. Taiwan, in 2010.
Technol. Weld. Join., vol. 11, no. 2, pp. 154–162, Mar. 2006. He is currently an Assistant Professor with the Department of Mechanical
[32] P. S. Wei, T. H. Wu, and S. S. Hsieh, “Phase change effects on transport Engineering, Yung Ta Institute of Technology and Commerce, Pingtung,
processes in resistance spot welding,” J. Mech., vol. 27, pp. 19–26, Mar. Taiwan. His current research interests include computational fluid flows, heat
2011. transfers in welding processes, and resistance spot welding.

Authorized licensed use limited to: Malaviya National Institute of Technology Jaipur. Downloaded on January 18,2024 at 19:33:32 UTC from IEEE Xplore. Restrictions apply.