Introduction

Nowadays, light-weighting is an in-evitable trend in the automotive in-dustry. Usually, the mass reduction ofvehicles is achieved either by the useof lighter, thinner, and stronger mate-rials and/or by optimization of designthroughout the vehicle structure.Thin-gauge steels might have some ad-

vantages over the use of aluminum interms of manufacturing cost (Ref. 1).However, resistance spot welding(RSW), which is still the predominantjoining technique in vehicle assembly,confronts two main difficulties in join-ing thin-gauge steels, especially theones thinner than 0.6 mm. These diffi-culties arise from primarily two rea-sons. First, the decrease in steel thick-ness leads to lower resistance of the

substrates, and thus the contact resist-ance at the faying interfaces accountsfor larger proportion of the total jouleheat generation. In other words, as thematerial gets thinner, the contactcharacteristics at the solid-solid inter-faces dominate the process (Ref. 2).Since the weld size strongly relates tothe contact status, it can be expectedthat the weldability in welding of thin-gauge steels would be poorer than thatof thicker gauge. Furthermore, an ex-traordinarily high temperature at theelectrode surface can be developed dueto the rapid heat transfer from theweld zone to electrode surface whenthe workpieces become thinner. Thisexcessively high temperature signifi-cantly accelerates the electrode degra-dation and eventually results in a re-duction in electrode life by 40–60%compared to ordinary gauge sheet(Refs. 3, 4).

Recently, resistance spot weldingwith metal strips/cover plates/processtapes inserted between the workpieceand electrode has been adopted to joinaluminum, magnesium, and ultrathingauge steels (Refs. 3, 5–11). The in-serted strip favors joule heat genera-tion during the welding process andmeanwhile shields the heat transferfrom the weld zone to the electrodesurface, which eventually reduces theelectrode tip temperature, and conse-quently prolongs the electrode life(Refs. 5, 6). Kolarik successfully joined2-mm-thick low-carbon steel to

WELDING RESEARCH

NOVEMBER 2014 / WELDING JOURNAL 411-s

SUPPLEMENT TO THE WELDING JOURNAL, NOVEMBER 2014Sponsored by the American Welding Society and the Welding Research Council

Effect of Inserted Strips on Electrode Degradationin Resistance Spot Welding

Test results indicated the insertion of a Cu55Ni45 metal strip showed the most promise for extending electrode life

BY Y. Y. ZHAO, Y. S. ZHANG, X. M. LAI, AND PEI-CHUNG WANG

ABSTRACT Recent trends toward economically fabricating lightweight vehicle structures whileensuring structural performance have led to the implementation of thin sheet steels inthe automotive industry. However, one of the main challenges in resistance spot weld­ing of ultra­thin steel (e.g., < 0.6 mm) is the extraordinarily short electrode life causedby the elevated electrode tip temperature. A method of inserting flexible strips be­tween the electrode and workpiece has been proposed to reduce the electrode tiptemperature, and consequently to prolong the electrode life.

In the present investigation, resistance spot welding of 0.4­mm­thick galvanized low­carbon steel with various inserted strips was experimentally investigated with a particularemphasis on the influence of inserted strips on the electrode degradation. Test resultsshowed that by inserting metal strips between the electrode and workpiece, the elec­trode life was prolonged by about 300%. The electrode face diameter was no longer aneffective indicator for the electrode degradation in resistance spot welding with insertedstrips. Surface alloying and recrystallization of the material near the electrode faceformed and played significant roles. Furthermore, the effects of the electro­thermal prop­erties and compositions of the inserted strips on the electrode tip temperature and de­gree of surface alloying were also evaluated. Among all the strips investigated in thisstudy, 0.12­mm­thick Cu55Ni45 metal strip exhibited the most promising results in allevi­ating the electrode degradation.

KEYWORDS • Resistance Welding • Thin­Gauge Steel • Automotive • Galvanized Steel

Y. Y. ZHAO (Zhangyansong@sjtu.edu.cn), Y. S. ZHANG, and X. M. LAI are with Shanghai Key Laboratory of Digital Manufacture for Thin­Walled Structures,Shanghai Jiao Tong University, Shanghai, PR China. PEI­CHUNG WANG is with Manufacturing Systems Research Lab, General Motors Research & DevelopmentCenter, Mound Road, Warren, Mich.

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 411

austenitic CrNi stainless steel by usinga DeltaSpot™ (Ref. 6) welding gun witha process tape (Ref. 7). Qiu and Abdofound the process tape/cover platetechnique to be effective in correctingthe heat imbalance and generating

heat with higherdensity in the alu-minum side during resistance spotwelding aluminum-to-steel (Refs. 8,9). Satonaka also reported the advan-tages of RSW with cover plates for

joining of magnesium alloy (e.g., acomparable welding current conditionto that for RSW of steel sheet) andfound a weld diameter that is largerthan the electrode diameter (Refs. 10,11). Zhao investigated the effect of theinserted strips on the electrode tiptemperature and weld quality duringRSW of ultrathin-gauge galvanizedsteels and found that the effect of theinserted metal strip strongly dependedon its material properties (i.e., resis-tivity and thermal conductivity) andthickness. With the proper selection ofthe metal strip, the electrode wear wasreduced significantly (Ref. 3).

To date, most of the publishedstudies focus on investigating the ef-fect of the inserted strips on heat gen-eration and temperature distribution,nugget formation, and joint strength,and there is little information regard-ing the role of the inserted strips onthe microstructure evolution at theelectrode surface.

The present study was undertaken toexperimentally study resistance weldingof 0.4-mm-thick galvanized low-carbonsteel with various inserted strips. Theobjective of this work was to gain a bet-ter understanding of the effect of thestrips on the electrode degradationfrom a metallurgical perspective, with a

WELDING RESEARCH

WELDING JOURNAL / NOVEMBER 2014, VOL. 93412-s

Fig. 1 — Resistance welding with inserted flexible strips. A —Schematic; B — experiment setup; C — configuration of the elec­trode (dimension in mm).

Fig. 2 — Function of number of welds during electrode weartesting of 0.4­mm­thick galvanized steel without inserted strips.A — Electrode face diameter. Carbon imprints of electrode faceunder the weld conditions of 1.8 kN, 5.7 kA, and 160 ms. B —After 200 welds; C — after 400 welds; D — after 600 welds.

A A

B

B C D

C

Table 1 — Chemical Compositions (wt­%) and Mechanical Properties of Galvanized Low­Carbon Steel(DC51D+Z from Bao Steel)

Chemical Composition Mechanical Properties

C Si Mn P S Fe Coating Wt. Yield Strength Tensile Strength Elongation(g/m2) (MPa) (MPa) (%)

0.04 0.01 0.23 0.01 0.001 balance 43–46 256 359 36

Table 3 — Mechanical and Electrical Properties of Strip Materials

Mechanical Properties AISI 304 Cu55Ni45 CuNi18Zn20 Copper

Tensile strength (MPa) 505 480 450 258Yield strength (MPa) 215 240 229 120Elongation (%) 70 45 40 48Modulus of elasticity (GPa) 193–200 170 133 120Shear modulus (GPa) 86 62 49 45Resistivity (μOhm cm) 68 55 29 2

Table 2 — Nominal Chemical Compositions of Strip Materials (wt­%)

Cu Cr Ni Zn Mn Fe

AISI 304 — 19 9 — <2 balanceCu55Ni45 balance — 44 — 0.5–2.0 <0.5CuNi18Zn20 balance — 18 20 <0.7 —Copper >99.0 — — — — —

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 412

particular emphasis on the electrodesurface alloying. The effects of the stripmaterial, thickness, and alloy elementon the electrode degradation were in-vestigated and discussed. Finally, basedon the test results, a preferable stripmaterial for resistance welding of 0.4-mm-thick galvanized low-carbon steelwas proposed.

Experimental Procedures

Materials

The steel selected in this study was0.4-mm-thick hot-dipped galvanizedlow-carbon steel (i.e., DC51D+Z fromBao Steel). Per manufacturer’s datasheet, chemical compositions and me-chanical properties are listed in Table 1.The metal strips used in this study are0.05-, 0.10-, 0.15-mm-thick AISI 304stainless steel, 0.10-mm-thick commer-cially available pure copper (hereinafterreferred to as “copper”) and CnNi18Zn

alloy, and 0.12-mm-thick Cu55Ni45 al-loy. Chemical compositions and me-chanical and electrical properties ofthese metal strips are listed in Tables 2and 3, respectively.

Welding Procedure

Electrode wear testing was per-formed using a mid-frequency DC weld-ing machine, and the welding setup isshown in Fig. 1A. Electrodes with a 5.0-mm-diameter flat tip made of Cr-Zr-Cualloy (compositions in wt-%: 0.2% Zr,0.5% Cr, 0.01% Al, and the rest is Cu),shown in Fig. 1B and C, were used. Testwelds were made on a 38 × 100-mmsheet of the material. The metal stripswith the same size, 38 × 100 mm, wereprepared. Table 4 lists the welding pa-rameters. Electrode wear tests were con-ducted under a single set of welding pa-rameters and at a welding rate of 30welds per minute. The test was termi-nated once the weld size fell below the

minimum weld nugget size (i.e., 2.5 mmof 0.4-mm thick sheet) (Ref. 12). Tomonitor electrode diameter and rate ofelectrode degradation, carbon imprints(Ref. 13) of electrode diameter were tak-en at the start of the electrode life test.Thereafter, imprints were taken at 100-weld increments. Peel specimens weregenerated at every 100-weld interval tomonitor the nugget diameters. Theweld nugget size was estimated bymeasuring the diameter of pullout but-tons during the peel tests.

Microscopic Analysis

To examine the changes at the elec-trode face during the degradationprocess, cross sections of the wornelectrodes were prepared and exam-ined. The polished samples wereetched in with a solution consisting of10-g iron (III) nitrate in 100 mL of dis-tilled water. The microstructures ofthe electrodes were examined by opti-cal microscope and scanning electronmicroscope (SEM) equipped with En-ergy-Dispersive Spectroscopy (EDS)analysis. Detailed experimental pa-rameters for SEM are listed in Table 5.

Results

Electrode life tests on RSW of 0.4-mm-thick galvanized low-carbon steel

WELDING RESEARCH

NOVEMBER 2014 / WELDING JOURNAL 413-s

Fig. 4 — Electrode face diameter as a func­tion of number of welds during RSW of 0.4­mm­thick galvanized steel with 0.1­mm­thick copper, CuNi18Zn20, and 304 stain­less steel strips.

Fig. 3 — Examination of the tested electrode after 600 resistance spot welds on 0.4­mm­thick galvanized low­carbon steel without strips. A — Cross section; B — enlarged view ofthe dashed box in A; C — SEM of tested electrode; D — EDS scan of the alloying layer shownin square C at the electrode surface.

A B

C D

Table 4 — Welding Parameters

Parameters Electrode Force (kN) Welding Current (kA) Squeeze Time (ms) Weld Time (ms) Hold Time (ms) Cooling Water Flow Rate (L/min)

Value 1.8 5.7 200 160 40 3

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 413

without and with various insertedstrips were conducted. There are threemain parts in the test results. First,the electrode degradation processwithout inserted strip was evaluated.This was followed by assessing the ef-fects of resistivity and thickness of theinserted strip on the electrode life. Fi-nally, we discuss the combined effectsof resistivity, thickness, and alloy con-tent of the inserted strip on the elec-trode degradation.

Electrode Degradation in RSWof Ultrathin Steel

Electrode life tests on RSW of 0.4-

mm-thick galvanized low-carbon steelwithout inserted strips were terminat-ed after 600 welds because the weldsize decreased to 2.4 mm, which is be-low the minimum acceptable size of2.5 mm (Ref. 12). Figure 2A presentsthe measured face diameters as a func-tion of weld number. As shown, elec-trode face diameter increased from 5to 6 mm after 600 continuous welds.The imprints of the electrode face af-ter 200, 400, and 600 welds are shownin Fig. 2B, C, and D, respectively. Care-ful examinations of the worn elec-trodes with 600 welds, shown in Fig.2D, indicated that severe cavitationdeveloped at the electrode face. Theseresults suggest that the electrodes had

degraded significantly. Microscopic examinations of the

worn electrodes after 600 welds wereperformed and the results are shownin Fig. 3. As shown in Fig. 3A, thecross section of the worn electrodesexhibited a rough surface profile. Afteretching, a dark layer, about 25~30 mmthick, was revealed at the electrodesurface, shown in Fig. 3B and C. EDSanalysis of the region labeled by asquare in Fig. 3C was performed andthe results are presented in Fig. 3D. Asshown, the layer contains about 86%iron, 11% copper, and 2% zinc (at.-%).Compared to the compositions of theas-received electrode material (i.e.,0.2% Zr, 0.5% Cr, 0.01% Al, and therest Cu), the difference in compositionis primarily caused by the sticking ofworkpiece and zinc coating onto theelectrode surface under the high tem-perature and the exertion of electrodeforce. Formation of this alloy layer andits high iron content indicated that se-vere electrode degradation occurred.

Electrode Degradation in RSWof Ultrathin Steel with InsertedStrip

To minimize the electrode degrada-tion, a metal strip was introduced inbetween the electrode and workpiece.It has been reported that the presenceof strip material strongly affected theheat generation and temperature dis-tribution during the welding process(Ref. 3). In this study, three strip ma-terials (i.e., commercially pure copper,CuNi18Zn20 alloy, and 304 stainlesssteel strips) that have lower, compara-ble, and higher electrical resistivitythan that of the workpieces were se-lected to assess their influences on thegrowth of electrode face diameter, re-crystallization, and surface alloying ofcopper electrode. To study the influ-ence of the strip materials on the elec-trode degradation process, the thick-ness of these strips was fixed at 0.10mm.

Electrode Face Diameter

Electrode life testing was performed,and the weld size was still acceptable af-ter 2000 welds by making use of the in-serted strips. Compared to the electrodelife of 600 welds in welding without

WELDING RESEARCH

WELDING JOURNAL / NOVEMBER 2014, VOL. 93414-s

Table 5 — Experimental Parameters for SEM

Parameters Accelerating Spot Size Detectors Working Distance (mm)Voltage (kV) (nm)

Value 5.00 4/5 SE 4.5~5.5

Fig. 5 — Electrode surfaces after 200, 400, and 600 welds in resistance welding of 0.4­mm­thick galvanized steel using 0.1­mm­thick copper, CuNi18Zn20, and 304 stainlesssteel strips.

Fig. 6 — Grain structure in resistance welding of 0.4­mm­thick galvanized low­carbonsteel with 0.10­mm­thick 304 stainless steel strip. A — Original copper electrode; B —recrystallized electrode.

A B

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 414

strips, it was prolonged by about 300%with using inserted strips.

The growth of electrode face diame-ter has been commonly used to assessthe extent of the electrode degrada-tion (Ref. 13). The changes in elec-trode face diameter are shown in Fig.4. For the purpose of comparison, testresults for the case without the stripare also included. As shown in Fig. 4,all electrode face diameters increasedwith increasing number of welds andreached a face diameter of about 5.3mm after 600 welds. Electrode face di-ameter for the case without a strip islarger than those with the presence ofthe strips by ~0.7 mm. The strip mate-rials used in this study had little influ-ence on the electrode face diameter. Tounderstand the effect of strip materialon the electrode degradation mecha-nism, the appearances of the wornelectrodes were analyzed.

Figure 5 presents the effect of thestrip material on the electrode appear-ances after 200, 400, and 600 continu-ous welds for copper, CuNi18Zn20, and304 strips. As shown, except the elec-trodes used with copper strip, all elec-trode faces exhibited little pitting orcavitation and mushrooming after 600welds. Previous studies suggested that anumber of different damage processesdeveloped during the degradation ofelectrodes when welding zinc-coatedsteel, namely: recrystallization of theelectrode material, mushrooming orgrowth of electrode face, and surface al-loying and pitting/cavitation (Ref. 14).However, careful examinations of theworn electrodes indicated that this isnot what we have observed in RSW ofthin steel with inserted strips. Test re-sults indicated that in RSW of thin steelwith inserted strips, the most promi-nent change in the electrode face wasthat a convex surface was formed at thecenter and expanded toward the edge.Since there is little difference in elec-trode face diameter, the worn electrodes

were cross-sectioned and examined mi-croscopically to see if they were recrys-tallized and surface alloyed during thewelding process. Test results are de-scribed next.

Recrystallization of Electrode Material

Published results (Refs. 15, 16)showed that recrystallization of theelectrode below the electrode surface

WELDING RESEARCH

NOVEMBER 2014 / WELDING JOURNAL 415-s

Fig. 7 — Effect of 0.10­mm strip material on the electrodes after 600 continuous resist­ance spot welds on 0.4­mm­thick galvanized low­carbon steel. A — Copper strip; B —CuNi18Zn20 strip; C — 304 stainless steel strip.

Fig. 8 — Electrode after 600 resistance spot welds on 0.4­mm­thick galvanized low­car­bon steel with 0.10­mm­thick strip. A — SEM observation with CuNi18Zn20 strip; B andC — EDS scan of the area shown in A.

A

A

B

B C

C

Table 6 — Effect of Strips on the Recrystallization and Surface Alloying of the Electrode in RSW of 0.4­mm­thick Galvanized Low­Carbon Steel

Strip Surface Alloy LayerMaterial Thickness (mm) Recrystallization Layer (m) Thickness (m) Composition (at.­%)

AISI 304 0.05 — — —0.10 252.9 6 87% Cu, 7% Fe, 6% Ni0.15 435.3 10 81% Cu, 16% Fe, 1.5% Ni

Copper 0.10 509.4 >100CuNi18Zn20 0.10 329.4 6.5 inner: 68% Cu, 31% Zn, 1% Ni

outer: 57% Cu, 40% Zn, 3% NiCu55Ni45 0.12 94.1 2 92% Cu, 8% Ni

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 415

frequently occurred due to high temper-ature, and it is usually accompanied by amarked reduction in the strength andhardness of the material. It was report-ed that a decrease in hardness from 176to 70–90 HV occurred within the re-crystallized zone (Ref. 4). Since elec-trode softening is one of the dominantdamage processes, the depth of recrys-tallization layer could be an effective in-dication of electrode degradation. Oncerecrystallization occurred, a pro-nounced change in grain structure wasobserved. Figure 6A and B present themicrostructures of the as-received andcopper electrodes after 600 welds in re-sistance welding of 0.4-mm-thick galva-nized low-carbon steel with 0.10-mm-thick 304 strip, respectively. As shown,while as-received copper electrode con-sisted of elongated grains, the recrystal-lized regions of the worn electrode werecomposed of primarily fine grains.

Due to this difference in mi-crostructures of as-received and re-crystallized electrodes, the thicknessof the recrystallized layer of the wornelectrodes was examined. Figure 7A, B,and C shows the recrystallized layersfor copper, CuNi18Zn20, and 304stainless steel strips, respectively. Asshown, while the thickness of the re-

crystallized layer ismore than ~500mm with a 0.1-mm-thick copperstrip, it reducedto ~330 and ~250mm for the 0.1-mm-thick CuNi18Zn20 and 0.1-mm-thick AISI 304 strips, respectively.These results inferred that the elec-trode tip temperature withCuNi18Zn20 strip is likely lower thanthat with copper strip but higher thanthat with 304 stainless strip.

Surface Alloying

During resistance welding, surfacecoatings often alloy with the electrodematerial due to the high temperatureexperienced at the electrode surface.This alloying makes a significant con-tribution to the overall electrode dam-age process (Ref. 4). To assess if thesurface alloying would form with thepresence of metal strips, SEM and EDSwere employed to investigate the mi-crostructure and chemical composi-tions of the alloying layer on the elec-trode surface and the results areshown in Figs. 8 and 9.

As shown in Fig. 8A, a ~6.5-mm-

thick alloying layer on the electrodesurface after 600 welds in RSW of 0.4-mm-thick galvanized low-carbon steelwith the presence of CuNi18Zn20strip was formed. As shown, this alloy-ing layer was composed of two sublay-ers. EDS analysis results of these sub-layers shown in Fig. 8B and C revealedthat they mainly contained copper,zinc, and a trace of nickel. These re-sults suggest that zinc in the strip ismost likely to alloy with the copperelectrode due to its lower meltingpoint and higher chemical activitycompared to nickel and copper.

Similar analyses were conducted onthe worn electrodes with the use of304 stainless steel strip and the re-sults are shown in Fig. 9. As shown inFig. 9A, the alloying layer with the useof 304 strip had a thickness of ~6.6mm, which is close to that (i.e., 6.5 mmthick) of using CuNi18Zn20 strip.However, the copper content in the al-loying layer with 304 strip is (~ 87%)

WELDING RESEARCH

WELDING JOURNAL / NOVEMBER 2014, VOL. 93416-s

Fig. 9 — Electrode after 600 resistance spot welds on 0.4­mm­thick galvanized low­carbon steel with 0.10­mm­thick strip. A —SEM with 304 stainless steel strip; B — EDS scan of the squareshown in A.

Fig. 10 — Effect of the thickness of 304 stainless steel strip on theelectrode degradation in resistance welding of 0.4­mm­thick gal­vanized low­carbon steel.

Fig. 11 — Effect of strip thickness on the recrystallization of theelectrode surface after 600 continuous resistance spot welds on0.4­mm­thick galvanized low­carbon steel with the presence of304 strips. A — 0.10 mm thick; B — 0.15 mm thick.

A

A B

B

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 416

far higher than that with CuNi18Zn20strip. Therefore, although the thick-nesses of the alloying layers are similarwith these two strips, the higher cop-per content in the alloying layer with304 strip indicates a lower extent ofsurface alloying.

Based on the aforementioned re-sults, we can conclude that pure copperis a less desirable choice than 304 steeland CuNi18Zn20 because of its extraor-dinary high thermal conductivity. More-over, a suitable strip should not containZn, owing to its low melting point andactive chemical property. Desired stripsincline to be copper-nickel based alloyand stainless steel. Within the selectedstrips studied here, due to the lowestdegree of electrode recrystallization andthe highest content of copper in the al-loying layer formed at the electrode sur-face by making use of 304 strips, 304stainless steel had the most promisingperformance in lowering electrode tem-perature and alleviating electrode degradation.

Effect of Strip Thickness

After 304 stainless steel was identi-fied as the preferred strip material, the

effect of its thick-ness on the elec-trode wear andsurface alloyingwas examinednext. The 304stainless steelwith the thick-nesses of 0.05,0.10, and 0.15mm was selected. Figure 10 presentsthe effect of the thickness of 304 stain-less steel strip on the changes to theelectrode surface. As shown, due to se-vere pitting occurring at the electrodesurface using a 0.05-mm-thick 304strip, tests were terminated after 200welds. Since a 0.05-mm-thick strip isquite thin, the asperities on the elec-trode face or workpieces occasionallyperforated through the strip under theexertion of electrode force and led to ex-treme current concentration, and conse-quently resulted in severe surface expul-sion and severe pitting on the electrodesurface. For 0.10- and 0.15-mm-thickstrips, a convex surface was formed atthe center of the electrode and expand-ed toward the edge. Comparisons of theworn electrodes shown in Fig. 10 indi-cated that the difference resulting from

using these two strip thicknesses wasthe diameter of the convexity beingsmaller for the thicker strip. Increasedstiffness of the thicker strip led to asmaller contact radius at theelectrode/strip and strip/workpiece in-terfaces, which further resulted in theconcentration of current density andheat generation at the contacting area.Therefore, the size of the convexity be-came smaller using thicker strips. More-over, the worn electrodes were exam-ined to assess the degree of recrystal-lization and surface alloying.

Figure 11A and B shows the crosssections of the electrodes with 0.10-and 0.15-mm-thick 304 strip after 600continuous welds, respectively. Asshown, recrystallized grains wereformed beneath the electrode face.While the thickness of the recrystal-lized layer was ~250 mm with a 0.10-

WELDING RESEARCH

NOVEMBER 2014/ WELDING JOURNAL 417-s

Fig. 12 — Electrode after 600 resistance spot welds on 0.4­mm­thick galvanized low­carbon steel with 0.15­mm­thick 304 stain­less steel. A — SEM observation; B — EDS scan of the regionshown in the square in A.

Fig. 13 — Resistance spot weld on 0.4­mm­thick galvanized low­carbon steel with 0.12­mm­thick Cu55Ni45 alloy strips after 600welds. A, B — Optical observation; C — SEM observation of theelectrode; D — EDS scan of the alloying layer.

Fig. 14 — Calculated R value of different strips in resistance spotwelding of ultrathin gauge steel.

A

B

BA

C D

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 417

mm-thick 304 strip, it increased to~435 mm when the strip was thickenedto 0.15 mm. These results suggestedthat the temperature at the electrodeface with a 0.15-mm-thick strip ishigher than that with a 0.10-mm-thickstrip. Thicker strip possesses higherelectrical bulk resistance, and there-fore is more favorable to heat genera-tion and especially since the resistivityof 304 stainless steel is much higherthan that of steel sheets. Therefore, ahigher temperature would develop atthe electrode face with thicker 304stainless steel strips.

SEM and EDS analysis results ofthe worn electrodes after 600 welds inRSW of 0.4-mm-thick galvanized low-carbon steel with 0.15-mm-thick 304steel, shown in Fig. 12, further validat-ed the above results. After etching, a10-mm-thick alloying layer emergedloosely at the electrode surface. Thisalloying layer contained ~81% copper,~16% iron, and traces of Ni and Cr(at.-%). Comparing the results shownin Figs. 9B and 12B indicates that theiron content in the alloying layer in-creased from ~7 to ~16% with an in-crease in thickness to the 304 stripfrom 0.10 to 0.15 mm, which suggest-ed a higher degree of surface alloying.

The increase in al-loying layer thick-ness and iron con-tent was likely at-tributed to the in-crease in joule heatgeneration with theuse of the thickerstrip. Combinationof the results of re-crystallization andsurface alloying

shown above indicated that 0.10-mm-thick 304 stainless steel strip is morefavorable than 0.15-mm-thick 304strip to prolong the electrode life inRSW of 0.4-mm-thick galvanized low-carbon steel.

Comparison of Strip Material

The aforementioned results of ef-fect of strip material on electrode wearindicated that the desired strips in-cline to be copper-nickel alloy andstainless steel. Therefore, it would beinteresting to compare copper-nickelalloy strip to 304 stainless steel. Tomake a fair comparison, Cu55Ni45was selected because its mechanicalproperties are similar to that of 304.The thickness of Cu55Ni45 alloy stripwas determined as 0.12 mm to ensurethat its electrical resistance is compa-rable to that of 0.10-mm-thick 304steel strip (i.e., resistivities of 304steel and Cu55Ni45 are 68 and 55mOhm/cm, respectively). Electrode lifetests were conducted and Fig. 13A andB presents the cross sections of theworn electrode after 600 continuousRSW of 0.4-mm-thick galvanized low-carbon steel with the presence of a

0.12-mm-thick Cu55Ni45 strip. Asshown, a layer of recrystallized grainswith ~100-mm-thickness was observedin the area near the electrode surface,which was far thinner than that (~250mm) of the electrodes with a 0.1-mm-thick 304 strip. Besides, the thicknessof the alloying layer shown in Fig. 13Cat the electrode surface decreased to~2 mm. EDS analysis results of the al-loying layer shown in Fig. 13D re-vealed that the composition of the lay-er is 91.9% Cu and the rest is Ni. Allthese results indicated a lower extentof electrode degradation with 0.12-mm-thick Cu55Ni45 strip.

To sum up all the test results pre-sented above, Table 6 lists the effect ofall strips used in this study on the re-crystallization and surface alloying inresistance welding of 0.4-mm-thick gal-vanized low-carbon steel. It can be seenthat the thicknesses of recrystallizedand alloying layers were the thinnest,and the alloying layer contained also thehighest copper content (~ 92%) withthe 0.12-mm-thick Cu55Ni45 stripamong all strips selected in this study.These results suggested that 0.12-mm-thick Cu55Ni45 is the most desiredstrip choice. And yet for all that, 0.10-mm 304 stainless steel exhibited thesecond-least electrode wear.

Discussions

Electrode Tip Temperature

Experimental results showed thatthe extent of the electrode degradationwith the presence of a metal strip wasquite different from that without themetal strip in resistance welding of 0.4-

WELDING RESEARCH

WELDING JOURNAL / NOVEMBER 2014, VOL. 93418-s

Fig. 15 — Binary phase diagrams (Ref. 20). A — Cu­Zn; B — Cu­Ni;C — Cu­Fe.

A

C

B

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 418

mm-thick galvanized low-carbon steel.The effect of the inserted strips on theelectrode tip temperature can be ex-plained from two aspects. One is “heat-ing effect” caused by the resistance ofthe strip itself and contact resistance ofadditional faying interfaces, and theother is “insulation effect” that the stripacts as a heat shield between the work-piece and electrode to insulate heattransfer from the weld zone to the cop-per electrode. Since it is difficult tomeasure the electrode tip temperatureexperimentally, an analytical model(Refs. 17, 18) was employed here to es-timate the electrode tip temperatureand analyze the effects of metal stripand its properties on the temperature atthe electrode-to-workpiece interface.This analytical analysis has the follow-ing assumptions:

1) Heat flow from the workpieceinto the electrodes is simplified to onedimension.

2) The peak temperature distribu-tion in the resistance spot weld can bedescribed by a sine wave half period,with the peak at the faying interface ofthe workpieces.

3) It is assumed that the thermalgradient in the electrode is linear, ex-tending from electrode tip tempera-ture E to the temperature of the cool-ing water (assumed to be room tem-perature for this analysis).

4) Thermal gradient in the strip islinear as well, extending from elec-trode tip temperature E to tempera-ture strip/workpiece interface temper-ature S.

5) The top and bottom electrodesare essentially straight sided.The function of temperature distribu-tion in the workpiece and boundarycondition can be summarized as

(1)

where P is the peak temperature inthe spot weld, x is the thickness ofthe workpiece, xE is the electrode facethickness (i.e., distance between theface and underside of the electrode), kEand kS are the thermal conductivitiesof the electrode material (Cu) andworkpiece, respectively, and x is thedistance from the weld faying surfacetoward the electrode face. kStrip andxstrip are the thermal conductivityand thickness of the inserted strip, re-spectively, and S is the temperatureat the workpiece/strip interface. Com-bining the function of temperaturedistribution and boundary conditions(i.e., in Equation 1), and then, the Eof RSW with the inserted strip can beestimated as

(2)In sum, the temperature developed atthe electrode surface E can be ex-pressed as

(3)To define an R for the ratio of P/E,Equation 3 can be rewritten as

(4)

The ratio R can quantitatively evaluatethe “insulation effect” of the insertedstrips, and a greater R value indicatesthe lower electrode tip temperature ifthe value of P is constant. Pluggingthe material properties (Ref. 19) intoEquation 4, the calculated results areshown in Fig. 14.

As shown, the temperature ratio Rincreases with the presence of thestrips. For a given sheet stack-up, thecalculated R value with copper strip is

slightly greater than that withoutstrip. These results suggest that thejoule heat generated at the weld zonecan readily transfer through the cop-per strip to the electrode surface, andconsequently result in high electrodetip temperature. The calculated R val-ues for 0.1-mm-thick 304 and 0.12-mm-thick Cu55Ni45 strips are the twohighest, and consequently lower elec-trode tip temperature can be expected.Test results showed that the electrodewear with 0.12-mm-thick Cu55Ni45was less than that with 0.1-mm-thick304, shown in Table 3, even thoughthe R value with 0.1-mm-thick 304 isgreater than that with 0.12-mm-thickCu55Ni45. This disagreement likely isattributed to two reasons. First, it isdifficult to estimate P quantitativelyby the analytical method due to a se-ries of assumptions and simplifica-tions made in the formulations. Fur-thermore, temperature is not the onlyfactor affecting the electrode wear;metallurgical factor is another impor-tant aspect and will be discussed in thesection below.

Alloying Element

In this part, the alloying compoundsformed between the electrode and in-serted strips and their influences on thehardness and electrical properties of thecopper electrode are discussed by refer-ring to binary phase diagrams. To ex-plain the effect of CuNi18Zn20 strip onthe electrode wear protection, the Cu-Zn binary phase diagram shown in Fig.15A is referred to. As shown, there areseveral intermetallic compounds (IMC)(e.g., a phase Cu3Zn and Cu9Zn solidsolution, b phase CuZn base solid solu-tion, and g phase Cu5Zn8 base solid so-lution) formed. With the presence ofCuNi18Zn20 strips, as shown in Fig. 8,the alloying compounds contained twosublayers that had 40 and 30% zinc, re-spectively. Based on EDS analysis re-sults of the composition in the alloyinglayer shown in Fig. 8 and the Cu-Znphase diagram, the outer layer shouldbe a +b phase while the inner layer iscomposed of a phase. Because b phaseIMC is quite soft (Ref. 20), it is unfavor-able for the electrode performance.

To assess Cu-Ni alloying compounds,see the Cu-Ni phase diagram in Fig.15B. As shown, Cu and Ni elements arecompletely soluble in each other and

=ΘΘ

=

+π

⎛⎝⎜

⎞⎠⎟

⎛⎝⎜

⎞⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

+⎛

⎝⎜

⎞

⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

+π

⎛⎝⎜

⎞⎠⎟

⎛⎝⎜

⎞⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

⎧

⎨

⎪⎪⎪⎪

⎩

⎪⎪⎪⎪

R

12 k

kxx

; traditional RSW

1k

k

x

x2 k

kxx

;

RSW with strips

P

E

E

S E

E

Strip

Strip

E

E

S

S

E

Θ =

Θ

+π

⎛⎝⎜

⎞⎠⎟

⎛⎝⎜

⎞⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

Θ

+⎛

⎝⎜

⎞

⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

+π

⎛⎝⎜

⎞⎠⎟

⎛⎝⎜

⎞⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

⎧

⎨

⎪⎪⎪⎪⎪

⎩

⎪⎪⎪⎪⎪

12 k

kxx

;

traditional RSW (Re fs. 17 , 18)

1k

k

x

x2 k

kxx

;

RSW with strips

E

P

E

S E

P

E

Strip

Strip

E

E

S

S

E

Θ =Θ

+⎛

⎝⎜

⎞

⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

+π

⎛⎝⎜

⎞⎠⎟

⎛⎝⎜

⎞⎠⎟

ΔΔ

⎛⎝⎜

⎞⎠⎟

1k

k

x

x2 k

kxx

EP

E

Strip

Strip

E

E

S

S

E

( )

( )

Θ = Θ −ΘπΔ

⎛⎝⎜

⎞⎠⎟ +Θ

Θ= = −

ΘΔ

Θ= Δ = −

Θ −ΘΔ

⎧

⎨

⎪⎪⎪⎪⎪⎪⎪⎪

⎩

⎪⎪⎪⎪⎪⎪⎪⎪

Δ

cos2 x

x

temperature distribution in workpiece

ddx

xk

k x

boundary condition at electrode /

strip int erface

ddx

x xk

k x

boundary condition at workpiece /

strip int erface

P S S

xE

Strip

E

E

Strip

s

S E

Strip

Strip

;

;

;

WELDING RESEARCH

NOVEMBER 2014 / WELDING JOURNAL 419-s

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 419

form a continuous solid solution. Thus,the alloying compounds would be a Cu-Ni solid solution with different propor-tions of the two elements. Similaranalysis was performed for Cu-Fe alloy-ing compounds with the Cu-Fe phase di-agram shown in Fig. 15C. As shown, thesolid solubility of iron in copper or cop-per in iron is extremely low (i.e., <2%),and there is no intermediate phaseformed between Cu-Fe. Therefore, Cu-Fe alloy with various components is allmade of the mixture of solid solutionsat both ends. In sum, Cu-Ni and Cu-Fesystems are not likely to form inter-metallic compounds, and therefore, thesurface alloying layer would affect littlethe hardness of the copper electrode.

Published results (Ref. 21) indicat-ed that the alloying element in copper-based solid solution exerted a signifi-cant influence on the electrical con-ductivity. Due to the extraordinaryelectrical conductivity of pure copper,addition of almost all other elementsinto copper would lead to a decrease inelectrical conductivity. It is noted thataddition of Fe into copper caused adramatic decrease in electrical conduc-tivity (Ref. 21). Therefore, Fe-rich al-loying layer deposited at the electrodesurface would significantly increasethe electrical resistance at the elec-trode surface, and consequently en-hance the joule heating, which wouldfurther exacerbate the electrode degra-dation. Therefore, considering the ef-fect of alloying on the electrical char-acteristics of copper electrode, an al-loying product between Cu-Ni alloystrip and the electrode provides theleast impact on the hardness and elec-trical conductivity of the electrode.This is consistent with the experimen-tal observations in this study.

Conclusions

1) The presence of inserted metalstrips alleviated the extent and rate ofelectrode degradation in resistance spotwelding of 0.4-mm-thick galvanizedlow-carbon steel. The process of elec-trode degradation was different fromthat of traditional resistance spot weld-ing. Alloying and recrystallization stilloccurred at the electrode surface, butthe growth in electrode surface diame-ter was significantly reduced.

2) The effects of the electrothermal

properties of the strips on the elec-trode tip temperature and surface al-loying were analyzed. Strips with com-paratively higher electrical resistivity(e.g., 304 stainless steel andCu55Ni45) lowered the electrode tiptemperature compared to the stripswith higher thermal/electrical conduc-tivity (e.g., commercially pure copper).

3) Analysis of the effect of the stripcompositions on the electrode surfacealloying revealed that the alloying com-pounds produced with the introductionof Cu-Ni alloy strip had the least impacton the hardness/strength and electricalconductivity of a copper electrode.

4) Among all strips investigated inthis study, 0.12-mm-thick Cu55Ni45alloy strip provided the best protec-tion of the electrode degradation in re-sistance welding of 0.4-mm-thick galva-nized low-carbon steel.

This research was supported by theGeneral Motors Collaborative ResearchLaboratory at Shanghai Jiao Tong Uni-versity. This research was also spon-sored by Shanghai Rising-Star Program(11QA1403600) and Project 51275304supported by National Natural ScienceFoundation of China.

1. Pecas, P., Henrique, M., Miranda, R.M., and Quintino, L. 1995. Laser welding oflow-thickness zinc-coated and uncoated car-bon steel sheets. Optical and Quantum Elec-tronics 27(12): 1193–1201.

2. Calva, C. M., and Eagar, T. W. 1990.Enhancement of the weldability in resist-ance spot welding. AWS Detroit Section SheetMetal Welding Conference IV, Southfield,Mich.

3. Zhao, Y. Y., Zhang, Y. S., Lai, X. M., andWang, P.-C. 2013. Resistance spot welding ofultrathin automotive steel. ASME Journal ofManufacturing Science and Engineering135(4): 021012.1–10.

4. Williams, N. T., and Parker, J. D. 2004.Review of resistance spot welding of steelsheets Part 2 Factors influencing electrodelife. International Materials Reviews 49(2):77–108.

5. Trommer, G. 2009. Resistance spotwelding using continuous tape. Welding Jour-nal 88(12): 20-s to 22-s.

6. Stieglbauer, W. 2006. Delta spot —Spot welding with process tape.

IIW Doc.III-1381-06.7. Kolarik, L., Sahul, M., Kolarikova, M.,

Sahul, M., Turna, M., and Felix, M. 2012. Re-sistance spot welding of dissimilar steels.Acta Polytechnica 52(3): 43–47.

8. Qiu, R. F., Iwamoto, C., and Satonaka,S. 2009. Interfacial microstructure andstrength of steel/aluminum alloy jointswelded by resistance spot welding with coverplate. Journal of Materials Processing Technol-ogy 209: 4186–4193.

9. Abdo, A., Harraz, M., Reisgen, U.,Schleser, M., and Schiebahn, A. 2013. Resist-ance spot welding of aluminum alloy 5182to zinc coated DP600 steel using processtape technique. International Journal for Sci-entific Research and Development 1(10):2212–2216.

10. Qiu, R. F., Satonaka, S., and Iwamoto,C. 2009. Mechanical properties and mi-crostructures of magnesium alloy AZ31Bjoint fabricated by resistance spot weldingwith cover plates. Science and Technology ofWelding and Joining 14(8): 691–697.

11. Satonaka, S., Iwamoto, C., Murakami,G.-i., and Matsumoto, Y. 2012. Resistancespot welding of magnesium alloy sheets withcover plates. Welding in the World 56(7-8):44–50.

12. American Welding Society. 2012. TestMethods for Evaluating the Resistance SpotWelding Behavior of Automotive Sheet SteelMaterials. AWS D8.9M:2012.

13. Zhang, X. Q., Chen, G. L., and Zhang,Y. S. 2008. Characteristics of electrode wearin resistance spot welding dual-phase steels.Materials & Design 29(1): 279–283.

14. Holliday, R. J., Parker, J. D., andWilliams, N. T. 1996. Relative contributionof electrode tip growth mechanism in spotwelding zinc coated steels. Welding in theWorld 37(4): 186–193.

15. Dong, S. J., and Zhou, Y. 2003. Ef-fects of TiC composite coating on electrodedegradation in microresistance welding ofnickel-plated steel. Metallurgical and Materi-als Transactions A 34(7): 1501–1511.

16. Chatterjee, K. L., and Waddell, W.1996. Electrode wear during spot welding ofcoated steels. Welding and Metal Fabrication64: 110–114.

17. Chuko, W., and Gould, J. E. 2002. De-velopment of appropriate resistance spotwelding practice for transformation-hard-ened steels — Phase 2: Evaluation of post-weld cooling rate techniques. Report to theAmerican Iron and Steel Institute.

18. Gould, J. E., Khurana, S. P., and Li, T.2006. Predictions of microstructures whenwelding automotive advanced high-strengthsteels. Welding Journal 85(5): 111-s to 116-s.

19. ASM Metals Handbook. 2001. Deskedition. ASM International.

20. ASM Metals Handbook. 1992. Vol. 3,Alloy Phase Diagrams. ASM International.

21. ASM Specialty Handbook: Copper andCopper Alloys. 2001. ASM International.

WELDING RESEARCH

WELDING JOURNAL / NOVEMBER 2014, VOL. 93420-s

Acknowledgments

References

Zhao 11-14_Layout 1 10/15/14 8:42 AM Page 420