- Email: [email protected]
Fatigue Behavior of Dissimilar Aluminum Alloy Spot Welds
Available online at www.sciencedirect.com
ScienceDirect Procedia Engineering 114 (2015) 149 – 156
1st International Conference on Structural Integrity
Fatigue Behavior of Dissimilar Aluminum Alloy Spot Welds Jidong Kanga*, Yuhua Chena, David Siglerb, Blair Carlsonb, David S. Wilkinsonc a CanmetMATERIALS, 183 Longwood Road South, Hamilton L8P 0A5, Caanda General Motors Global R&D Center, 30500 Mound Road, Warren 48090-9055, U.S.A. c McMaster University, 1280 Main Stresst West, Hamilton L8S 4L7, Canada
b
Abstract With the broader utilization of a variety of aluminum alloys in the automotive industry for structural lightweight applications, the need for resistance spot welding (RSW) of dissimilar aluminum alloys is increasing. General Motors (GM) has developed a proprietary RSW process using a multi-ring, domed electrode geometry that significantly improves the performance of the aluminum resistance spot welds. In addition, to enhance structural performance, epoxy adhesives are also often applied prior to RSW to obtain weld-bonded joints. As a contribution, the load-controlled fatigue behavior of dissimilar aluminum alloy spot welds made of 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet with and without the addition of adhesive prior to welding was studied. The same GM proprietary resistance spot welding electrode and current schedule was applied to both welding conditions leading to a larger nugget size when using adhesive, but both weld configurations presented similar maximum load in tension-shear testing. X-ray computed tomography was used to detect internal welding discontinuities such as voids and also to follow the damage evolution and fatigue crack initiation and growth during interrupted fatigue testing of the spot welds. The results show that the main fatigue crack initiates at the edge of the nugget and penetrates through the Aural2 die casting sheet in the thickness direction. Using the structural stress concept, it was also found that the structural stress-fatigue life curve for AA5754 to Aural2 aluminum spot welds with and without adhesive falls into a master curve indicating that the nugget size which corresponds to the tensile and bending strength dominates fatigue life. © 2015 2015Published The Authors. Published byisElsevier Ltd. article under the CC BY-NC-ND license © by Elsevier Ltd. This an open access (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering. Peer-review under responsibility of INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering
Keywords: Fatigue; resistance spot welding; AA5754; Aural2; X-ray computed tomography, Adhesive
* Corresponding author. Tel.: +001-905-645-0820; fax: +001-905-645-0831. E-mail address: [email protected]
1877-7058 © 2015 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering
doi:10.1016/j.proeng.2015.08.053
150
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
1. Introduction Facing the challenges of fuel economy and vehicle lightweighting, aluminum alloys are being used increasingly in automotive structural applications. Among the commonly used joining techniques, resistance spot welding (RSW) is desirable due to its inherently low cost, high speed and accessibility. However, resistance spot welding of aluminum is sparsely used due to the perceived lack of robustness, the wide flange requirements, and the lower fatigue strength of aluminum spot welds relative to steel welds [1-3]. Recently, General Motors has developed a proprietary RSW process using a multi-ring, domed (MRD) electrode geometry that significantly improves the performance of the aluminum resistance spot welds [3-5]. In addition, to enhance structural performance adhesive is also often applied prior to RSW to obtain weld-bonded joints [6]. In the present study, the same GM proprietary RSW electrode and current schedule was used to weld 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet with and without the addition of adhesive prior to welding. We then studied the load-controlled fatigue behavior of these dissimilar aluminum alloy spot welds. X-ray computed tomography was used to detect the resulting internal welding defects such as voids as well as to follow the damage evolution and fatigue crack initiation and growth during interrupted fatigue testing of the spot welds. We also discussed the feasibility of using the structural stress concept to rationalize the load-fatigue life curves for AA5754 to Aural2 dissimilar aluminum spot welds with and without the addition of adhesive.
Nomenclature d Fy Mx t V(SS)
diameter of the weld nugget maximum tensile force bending moment sheet thickness structural stress
2. Materials and experimental procedures The materials used in the present study were 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet. The nominal chemical composition of the two materials is shown in Table 1. Table 1 Nominal chemical composition of Aural-2 and AA5754 (wt %) Element
Si
Mn
Mg
Fe
Ti
Sr
Cu
Cr
Zn
Al
ˉ
ˉ
ˉ
Bal.
≤ 0.1
≤ 0.3
≤ 0.2
Bal.
Aural-2
9.5 - 11.5
0.3 - 0.6
0.1 - 0.4
≤ 0.25
≤ 0.1
0.010.018
AA5754
≤ 0.4
≤ 1.0
2.6-3.6
≤ 0.4
≤ 0.15
ˉ
The same GM proprietary RSW electrode and current schedule was used to fabricate lap-shear joints made of 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet with and without the addition of epoxy adhesive prior to welding. The specimens were machined to final dimensions of 20 mm in width and 90 mm in total length using the water-jet cutting technique. Some of the specimens were sectioned through the cross-section perpendicular to the length direction and mounted for microhardness (Hv) measurements and macrostructure observations following a standard metallographic procedure.
151
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
Tensile shear testing was carried out to determine the maximum tensile load capacity of the welds and the corresponding maximum load levels for the load controlled fatigue testing. The load-controlled fatigue tests were conducted at stress ratio of 0.1 at five maximum load levels of 3483N, 2322N, 1744N, 1450N and 1161N. All fatigue testing specimens were analyzed using X-ray computed tomography (XCT) scans prior to testing to obtain the 3D models of the size and distribution of damage (e.g. pores, welding defects, etc.) within the weld nuggets. All the XCT scans were carried out using a Hytec Edge CT machine at Jesse Garant &Associates. The prefilter used is 0.5mm Molybdenum. The tube voltage is 165kV and the tube current is 300 PA. The source to object distance is 210 mm and the source to detector distance is 880 mm. In order to follow damage evolution during fatigue testing, interrupted fatigue testing was conducted for 2 specimens each for specimens with and without addition of adhesive prior to welding at maximum load level of 2322N every 20,000 cycles until fatigue failure. During each interruption, specimens were taken out from the fatigue testing machine for XCT scans to obtain 3D models of the damage evolution. 3. Results and discussion The macrostructures of the weld nuggets for the AA5754 to Aural2 spot welds with and without addition of adhesive are shown in Fig. 1. From Fig. 1, it is seen that both the penetration and nugget diameter are smaller in the AA5754 side compared to the Aural2 side. The addition of adhesive enhanced the penetration in both sides and enlarged the nugget size (Fig. 1b). Note that the nugget diameter on the AA5754 side is smaller than that on the Aural2 side, which results in the notch root being located at the perimeter of the AA5754 side of the nugget (red arrows). a
b
AA5754
20
AA5754
30
19
14
Aural2
Aural2
31
30
20
31
14
31
31
19
Fig. 1. Macrostructure and microhardness of AA5754 to Aural resistance spot welds (a) without adhesive and (b) with adhesive.
When looking at the microhardness measurements (Fig. 1), it is seen that microhardness of the weld nuggets is higher than the other regions of the weld, i.e. heat affected zone and base metal under both wedling conditions. However, the microhardness is lower in the centre of the weld nugget with more variation when welding with the addition of adhesive indicating porosity exists within the nugget that will reduce nugget strength.
152
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
Despite the differences in nugget size and penetration, the maximum tensile load from the tensile shear testing is slightly higher in the spot weld without the addition of adhesive (Table 2). This observation confirms that the nugget strength is lower in the dissimilar spot welds with the addition of adhesive. This was most likely caused by the fracture mode, i.e., nugget pullout, which would initiate at the notch root and propagate into the nugget volume on the Aural2 side of the weld. Additional porosity in the weld nugget with adhesive then results in reduced strength.
Table 2 Tensile shear test results of the spot joint of 2mm AA5754 to 3mm Aural2 2mm AA5754 to 3mm Aural2 Bare Adhesive
Maximum Load, N 6166 5987 5941 5952
Average, N 6077 5947
The fatigue testing results are shown in Fig. 2. From Fig. 2, it is shown that for a given maximum load level, fatigue life of the dissimilar aluminum spot welds with the addition of adhesive is longer than those without addition of adhesive.
Fig. 2. Fatigue test results for aluminum alloy spot weld joints with and without adhesive.
XCT scan results (Fig. 3) confirm that there is some evidence of damage growth and coalescence within the nuggets (see Step 2-4 in Fig. 3a and 3b) but the main fatigue crack initiates at the edge of the nugget (see the red circles in Step 4 in Fig. 3a and 3b) and penetrates through the Aural2 die casting sheet in the thickness direction. This is consistent with the observations of all the fatigue fractured specimens. The results we obtained so far present an apparent contradiction. On one hand, the dissimilar aluminum spot welds with the addition of adhesive have larger nugget sizes, one would expect longer fatigue life in these spot welds
153
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
according to the earlier research stating the nugget size controls the fatigue life [1-2]. On the other hand, the two aluminum dissimilar welds have the same tensile-shear load capacity indicating that one would expect similar fatigue properties. Since the spot welds with the addition of adhesive are indeed lower in nugget strength, one can even expect that their fatigue properties could be also be reduced. a
Step 1
Step 2
Step 3
Step 4
Step 2
Step 3
Step 4
b
Step 1
Fig. 3. X-ray computed tomography scans of damage evolution in aluminum alloy spot weld joints (a) with and (b) without adhesive.
To help understand this contradiction, we can rationalize the fatigue results of these dissimilar aluminum resistance spot welds using the structural stress concept [8-11]. Following the structural stress concept (Fig. 4) developed by Rupp et al [7], we have
154
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
V max (SS)䠙V max ( M x ) V max ( Fy ) k(
where k
0.6 t , M x
1.872M x dt
2
)
(1)
Fy
Sdt
2.5Fy ; V max䠄 SS䠅 , d, Fy, Mx, t is maximum structural stress, diameter of the weld nugget,
maximum tensile force, bending moment and sheet thickness of Aural2, respectively. Eq. (1) indicates that the diameter of the spot welds plays a competitive role with sheet thickness in determining the structural stress. A larger nugget diameter is beneficial to lower the structural stress level that usually leads to longer fatigue life.
Fig. 4. Concept of structural stress in aluminum spot welds [7, 10].
The nugget size of the spot welds used in the present study were measured and shown in Table 3. Table 3 The average nugget size of the spot welds with and without adhesive Load(N)
Without adhensive
With adhensive
3483
6.8
8.3
2322
6.2
7.9
1744
6.0
8.0
1450 1161
6.2 6.7
7.6 8.0
Using Eq. (1) and the data from Table 2, we re-plotted the fatigue test results using the maximum structural stress as shown in Fig. 5. From Fig. 5, it is illustrated that the maximum structural stress-fatigue life curve for AA5754 to Aural2 dissimilar aluminum spot welds with and without adhesive falls into a master curve. At relatively higher structural stress levels, i.e. 126 MPa and above, the fatigue life of the spot welds without the addition of adhesive is longer than the ones with the addition of the adhesive. Below this structural stress level of 126 MPa, both the structural
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
155
stress and fatigue life of these two spot welds are indeed almost the same. These results suggest that the nugget size which corresponds to the tensile and bending strength dominates fatigue life in the two aluminum dissimilar spot welds used in the present study.
Fig. 5. Fatigue test results for aluminum alloy spot weld joints with and without adhesive using maximum structural stress.
4. Conclusions We studied the load-controlled fatigue behavior of a dissimilar aluminum alloy spot weld made of 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet using the same GM proprietary resistance spot welding process with and without the addition of adhesive prior to welding. X-ray computed tomography was used to detect the resulting internal welding defects such as voids as well as to follow the damage evolution during interrupting fatigue testing. The following conclusions can be drawn from the present study, 1) The nugget size of the dissimilar aluminum spot welds is larger in the spot welds with the addition of adhesive but the maximum load capacity in tension-shear testing is the same as the ones without the addition of adhesive. 2) There is evidence of damage growth and coalescence within the nuggets, but the main fatigue crack initiates at the edge of the nugget and penetrates through the Aural2 die casting sheet in the thickness direction. 3) The maximum structural stress-fatigue life curve for AA5754 to Aural2 dissimilar aluminum spot welds with and without adhesive falls onto a master curve indicating that the nugget size which corresponds to the tensile and bending strength dominates fatigue life.
156
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
Acknowledgements The authors are grateful the financial support from Canadian federal government interdepartmental Program on Energy R&D and General Motors Canada. We thank Mr. Jie Liang at CanmetMATERIALS for help in fatigue testing and Dr. Hong-Tae Kang at University of Michigan, Dearborn in structural stress analysis. In-kind contribution from Jesse Garant and Associates Metrology Center for X-ray computed tomography scanning is also acknowledged. References [1] P.H. Thornton, A.R. Krause, R.G. Davies, The aluminum spot weld, Welding Journal, 75 (1996) 101s-108s. [2] A. Gean, S. A. Westgate, J. C. Kucza, J. C. Ehrstrom, Static and fatigue behavior of spot-welded 5182-0 aluminum alloy sheet, Welding Journal, 78 (1999) 80s-86s. [3] D. R. Sigler, J. G. Schroth, M. J. Karagoulis, D. Zuo, New electrode weld face geometries for spot welding aluminum. Conference Proceedings, AWS Sheet Metal Welding Conference XIV, Livonia, May 11–14, 2010, pp. 1–19. [4] D.R. Sigler, J.G. Schroth, Karagoulis, M. J. Weld electrode for attractive weld appearance. U.S. Patent 8,222,560. July 17, 2012. [5] D. R. Sigler, Blair E. Carlson, Paul Janial, Improving aluminum resistance spot welding in automotive structures, Welding Journal, 92 (2013) 64-72 [6] P. C. Wang, S. K. Chisholm, G. Banas, F. V. Lawrence, The role of failure mode, resistance spot weld and adhesive on the fatigue behavior of weld-bonded aluminum, Welding Journal, 74 (1995) 41s-47s. [7] G. Wang, M. E. Barkey, Investigating the spot weld fatigue crack growth process using X-ray imaging, Welding Journal, 85 (2006) 84s90s. [8] A. Rupp, K. Stijrzel, V. Grubisic, Computer aided dimensioning of spot-welded automotive structures. SAE Paper # 950711, 1995 [9] P. Dong.. A structural stress definition and numerical implementation for fatigue analysis of welded joints. International Journal of Fatigue, 23 (2001) 865–876. [10] H-T Kang, P. Dong, J.K. Hong. Fatigue analysis of spot welds using a mesh-insensitive structural stress approach. International Journal of Fatigue, 29 (2007) 1546–1553. [11] H-T. Kang, M.E. Barkey, Y.Lee. Evaluation of multiaxial spot weld fatigue parameters for proportional loading. International Journal of Fatigue, 22(2000) 691–702.
ScienceDirect Procedia Engineering 114 (2015) 149 – 156
1st International Conference on Structural Integrity
Fatigue Behavior of Dissimilar Aluminum Alloy Spot Welds Jidong Kanga*, Yuhua Chena, David Siglerb, Blair Carlsonb, David S. Wilkinsonc a CanmetMATERIALS, 183 Longwood Road South, Hamilton L8P 0A5, Caanda General Motors Global R&D Center, 30500 Mound Road, Warren 48090-9055, U.S.A. c McMaster University, 1280 Main Stresst West, Hamilton L8S 4L7, Canada
b
Abstract With the broader utilization of a variety of aluminum alloys in the automotive industry for structural lightweight applications, the need for resistance spot welding (RSW) of dissimilar aluminum alloys is increasing. General Motors (GM) has developed a proprietary RSW process using a multi-ring, domed electrode geometry that significantly improves the performance of the aluminum resistance spot welds. In addition, to enhance structural performance, epoxy adhesives are also often applied prior to RSW to obtain weld-bonded joints. As a contribution, the load-controlled fatigue behavior of dissimilar aluminum alloy spot welds made of 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet with and without the addition of adhesive prior to welding was studied. The same GM proprietary resistance spot welding electrode and current schedule was applied to both welding conditions leading to a larger nugget size when using adhesive, but both weld configurations presented similar maximum load in tension-shear testing. X-ray computed tomography was used to detect internal welding discontinuities such as voids and also to follow the damage evolution and fatigue crack initiation and growth during interrupted fatigue testing of the spot welds. The results show that the main fatigue crack initiates at the edge of the nugget and penetrates through the Aural2 die casting sheet in the thickness direction. Using the structural stress concept, it was also found that the structural stress-fatigue life curve for AA5754 to Aural2 aluminum spot welds with and without adhesive falls into a master curve indicating that the nugget size which corresponds to the tensile and bending strength dominates fatigue life. © 2015 2015Published The Authors. Published byisElsevier Ltd. article under the CC BY-NC-ND license © by Elsevier Ltd. This an open access (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering. Peer-review under responsibility of INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering
Keywords: Fatigue; resistance spot welding; AA5754; Aural2; X-ray computed tomography, Adhesive
* Corresponding author. Tel.: +001-905-645-0820; fax: +001-905-645-0831. E-mail address: [email protected]
1877-7058 © 2015 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of INEGI - Institute of Science and Innovation in Mechanical and Industrial Engineering
doi:10.1016/j.proeng.2015.08.053
150
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
1. Introduction Facing the challenges of fuel economy and vehicle lightweighting, aluminum alloys are being used increasingly in automotive structural applications. Among the commonly used joining techniques, resistance spot welding (RSW) is desirable due to its inherently low cost, high speed and accessibility. However, resistance spot welding of aluminum is sparsely used due to the perceived lack of robustness, the wide flange requirements, and the lower fatigue strength of aluminum spot welds relative to steel welds [1-3]. Recently, General Motors has developed a proprietary RSW process using a multi-ring, domed (MRD) electrode geometry that significantly improves the performance of the aluminum resistance spot welds [3-5]. In addition, to enhance structural performance adhesive is also often applied prior to RSW to obtain weld-bonded joints [6]. In the present study, the same GM proprietary RSW electrode and current schedule was used to weld 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet with and without the addition of adhesive prior to welding. We then studied the load-controlled fatigue behavior of these dissimilar aluminum alloy spot welds. X-ray computed tomography was used to detect the resulting internal welding defects such as voids as well as to follow the damage evolution and fatigue crack initiation and growth during interrupted fatigue testing of the spot welds. We also discussed the feasibility of using the structural stress concept to rationalize the load-fatigue life curves for AA5754 to Aural2 dissimilar aluminum spot welds with and without the addition of adhesive.
Nomenclature d Fy Mx t V(SS)
diameter of the weld nugget maximum tensile force bending moment sheet thickness structural stress
2. Materials and experimental procedures The materials used in the present study were 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet. The nominal chemical composition of the two materials is shown in Table 1. Table 1 Nominal chemical composition of Aural-2 and AA5754 (wt %) Element
Si
Mn
Mg
Fe
Ti
Sr
Cu
Cr
Zn
Al
ˉ
ˉ
ˉ
Bal.
≤ 0.1
≤ 0.3
≤ 0.2
Bal.
Aural-2
9.5 - 11.5
0.3 - 0.6
0.1 - 0.4
≤ 0.25
≤ 0.1
0.010.018
AA5754
≤ 0.4
≤ 1.0
2.6-3.6
≤ 0.4
≤ 0.15
ˉ
The same GM proprietary RSW electrode and current schedule was used to fabricate lap-shear joints made of 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet with and without the addition of epoxy adhesive prior to welding. The specimens were machined to final dimensions of 20 mm in width and 90 mm in total length using the water-jet cutting technique. Some of the specimens were sectioned through the cross-section perpendicular to the length direction and mounted for microhardness (Hv) measurements and macrostructure observations following a standard metallographic procedure.
151
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
Tensile shear testing was carried out to determine the maximum tensile load capacity of the welds and the corresponding maximum load levels for the load controlled fatigue testing. The load-controlled fatigue tests were conducted at stress ratio of 0.1 at five maximum load levels of 3483N, 2322N, 1744N, 1450N and 1161N. All fatigue testing specimens were analyzed using X-ray computed tomography (XCT) scans prior to testing to obtain the 3D models of the size and distribution of damage (e.g. pores, welding defects, etc.) within the weld nuggets. All the XCT scans were carried out using a Hytec Edge CT machine at Jesse Garant &Associates. The prefilter used is 0.5mm Molybdenum. The tube voltage is 165kV and the tube current is 300 PA. The source to object distance is 210 mm and the source to detector distance is 880 mm. In order to follow damage evolution during fatigue testing, interrupted fatigue testing was conducted for 2 specimens each for specimens with and without addition of adhesive prior to welding at maximum load level of 2322N every 20,000 cycles until fatigue failure. During each interruption, specimens were taken out from the fatigue testing machine for XCT scans to obtain 3D models of the damage evolution. 3. Results and discussion The macrostructures of the weld nuggets for the AA5754 to Aural2 spot welds with and without addition of adhesive are shown in Fig. 1. From Fig. 1, it is seen that both the penetration and nugget diameter are smaller in the AA5754 side compared to the Aural2 side. The addition of adhesive enhanced the penetration in both sides and enlarged the nugget size (Fig. 1b). Note that the nugget diameter on the AA5754 side is smaller than that on the Aural2 side, which results in the notch root being located at the perimeter of the AA5754 side of the nugget (red arrows). a
b
AA5754
20
AA5754
30
19
14
Aural2
Aural2
31
30
20
31
14
31
31
19
Fig. 1. Macrostructure and microhardness of AA5754 to Aural resistance spot welds (a) without adhesive and (b) with adhesive.
When looking at the microhardness measurements (Fig. 1), it is seen that microhardness of the weld nuggets is higher than the other regions of the weld, i.e. heat affected zone and base metal under both wedling conditions. However, the microhardness is lower in the centre of the weld nugget with more variation when welding with the addition of adhesive indicating porosity exists within the nugget that will reduce nugget strength.
152
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
Despite the differences in nugget size and penetration, the maximum tensile load from the tensile shear testing is slightly higher in the spot weld without the addition of adhesive (Table 2). This observation confirms that the nugget strength is lower in the dissimilar spot welds with the addition of adhesive. This was most likely caused by the fracture mode, i.e., nugget pullout, which would initiate at the notch root and propagate into the nugget volume on the Aural2 side of the weld. Additional porosity in the weld nugget with adhesive then results in reduced strength.
Table 2 Tensile shear test results of the spot joint of 2mm AA5754 to 3mm Aural2 2mm AA5754 to 3mm Aural2 Bare Adhesive
Maximum Load, N 6166 5987 5941 5952
Average, N 6077 5947
The fatigue testing results are shown in Fig. 2. From Fig. 2, it is shown that for a given maximum load level, fatigue life of the dissimilar aluminum spot welds with the addition of adhesive is longer than those without addition of adhesive.
Fig. 2. Fatigue test results for aluminum alloy spot weld joints with and without adhesive.
XCT scan results (Fig. 3) confirm that there is some evidence of damage growth and coalescence within the nuggets (see Step 2-4 in Fig. 3a and 3b) but the main fatigue crack initiates at the edge of the nugget (see the red circles in Step 4 in Fig. 3a and 3b) and penetrates through the Aural2 die casting sheet in the thickness direction. This is consistent with the observations of all the fatigue fractured specimens. The results we obtained so far present an apparent contradiction. On one hand, the dissimilar aluminum spot welds with the addition of adhesive have larger nugget sizes, one would expect longer fatigue life in these spot welds
153
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
according to the earlier research stating the nugget size controls the fatigue life [1-2]. On the other hand, the two aluminum dissimilar welds have the same tensile-shear load capacity indicating that one would expect similar fatigue properties. Since the spot welds with the addition of adhesive are indeed lower in nugget strength, one can even expect that their fatigue properties could be also be reduced. a
Step 1
Step 2
Step 3
Step 4
Step 2
Step 3
Step 4
b
Step 1
Fig. 3. X-ray computed tomography scans of damage evolution in aluminum alloy spot weld joints (a) with and (b) without adhesive.
To help understand this contradiction, we can rationalize the fatigue results of these dissimilar aluminum resistance spot welds using the structural stress concept [8-11]. Following the structural stress concept (Fig. 4) developed by Rupp et al [7], we have
154
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
V max (SS)䠙V max ( M x ) V max ( Fy ) k(
where k
0.6 t , M x
1.872M x dt
2
)
(1)
Fy
Sdt
2.5Fy ; V max䠄 SS䠅 , d, Fy, Mx, t is maximum structural stress, diameter of the weld nugget,
maximum tensile force, bending moment and sheet thickness of Aural2, respectively. Eq. (1) indicates that the diameter of the spot welds plays a competitive role with sheet thickness in determining the structural stress. A larger nugget diameter is beneficial to lower the structural stress level that usually leads to longer fatigue life.
Fig. 4. Concept of structural stress in aluminum spot welds [7, 10].
The nugget size of the spot welds used in the present study were measured and shown in Table 3. Table 3 The average nugget size of the spot welds with and without adhesive Load(N)
Without adhensive
With adhensive
3483
6.8
8.3
2322
6.2
7.9
1744
6.0
8.0
1450 1161
6.2 6.7
7.6 8.0
Using Eq. (1) and the data from Table 2, we re-plotted the fatigue test results using the maximum structural stress as shown in Fig. 5. From Fig. 5, it is illustrated that the maximum structural stress-fatigue life curve for AA5754 to Aural2 dissimilar aluminum spot welds with and without adhesive falls into a master curve. At relatively higher structural stress levels, i.e. 126 MPa and above, the fatigue life of the spot welds without the addition of adhesive is longer than the ones with the addition of the adhesive. Below this structural stress level of 126 MPa, both the structural
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
155
stress and fatigue life of these two spot welds are indeed almost the same. These results suggest that the nugget size which corresponds to the tensile and bending strength dominates fatigue life in the two aluminum dissimilar spot welds used in the present study.
Fig. 5. Fatigue test results for aluminum alloy spot weld joints with and without adhesive using maximum structural stress.
4. Conclusions We studied the load-controlled fatigue behavior of a dissimilar aluminum alloy spot weld made of 2mm thick AA5754 wrought sheet and 3mm thick Aural2 die casting sheet using the same GM proprietary resistance spot welding process with and without the addition of adhesive prior to welding. X-ray computed tomography was used to detect the resulting internal welding defects such as voids as well as to follow the damage evolution during interrupting fatigue testing. The following conclusions can be drawn from the present study, 1) The nugget size of the dissimilar aluminum spot welds is larger in the spot welds with the addition of adhesive but the maximum load capacity in tension-shear testing is the same as the ones without the addition of adhesive. 2) There is evidence of damage growth and coalescence within the nuggets, but the main fatigue crack initiates at the edge of the nugget and penetrates through the Aural2 die casting sheet in the thickness direction. 3) The maximum structural stress-fatigue life curve for AA5754 to Aural2 dissimilar aluminum spot welds with and without adhesive falls onto a master curve indicating that the nugget size which corresponds to the tensile and bending strength dominates fatigue life.
156
Jidong Kang et al. / Procedia Engineering 114 (2015) 149 – 156
Acknowledgements The authors are grateful the financial support from Canadian federal government interdepartmental Program on Energy R&D and General Motors Canada. We thank Mr. Jie Liang at CanmetMATERIALS for help in fatigue testing and Dr. Hong-Tae Kang at University of Michigan, Dearborn in structural stress analysis. In-kind contribution from Jesse Garant and Associates Metrology Center for X-ray computed tomography scanning is also acknowledged. References [1] P.H. Thornton, A.R. Krause, R.G. Davies, The aluminum spot weld, Welding Journal, 75 (1996) 101s-108s. [2] A. Gean, S. A. Westgate, J. C. Kucza, J. C. Ehrstrom, Static and fatigue behavior of spot-welded 5182-0 aluminum alloy sheet, Welding Journal, 78 (1999) 80s-86s. [3] D. R. Sigler, J. G. Schroth, M. J. Karagoulis, D. Zuo, New electrode weld face geometries for spot welding aluminum. Conference Proceedings, AWS Sheet Metal Welding Conference XIV, Livonia, May 11–14, 2010, pp. 1–19. [4] D.R. Sigler, J.G. Schroth, Karagoulis, M. J. Weld electrode for attractive weld appearance. U.S. Patent 8,222,560. July 17, 2012. [5] D. R. Sigler, Blair E. Carlson, Paul Janial, Improving aluminum resistance spot welding in automotive structures, Welding Journal, 92 (2013) 64-72 [6] P. C. Wang, S. K. Chisholm, G. Banas, F. V. Lawrence, The role of failure mode, resistance spot weld and adhesive on the fatigue behavior of weld-bonded aluminum, Welding Journal, 74 (1995) 41s-47s. [7] G. Wang, M. E. Barkey, Investigating the spot weld fatigue crack growth process using X-ray imaging, Welding Journal, 85 (2006) 84s90s. [8] A. Rupp, K. Stijrzel, V. Grubisic, Computer aided dimensioning of spot-welded automotive structures. SAE Paper # 950711, 1995 [9] P. Dong.. A structural stress definition and numerical implementation for fatigue analysis of welded joints. International Journal of Fatigue, 23 (2001) 865–876. [10] H-T Kang, P. Dong, J.K. Hong. Fatigue analysis of spot welds using a mesh-insensitive structural stress approach. International Journal of Fatigue, 29 (2007) 1546–1553. [11] H-T. Kang, M.E. Barkey, Y.Lee. Evaluation of multiaxial spot weld fatigue parameters for proportional loading. International Journal of Fatigue, 22(2000) 691–702.