Abstract
The present investigation highlights the development of a suitable and novel deep neural network-based learning model for accurately predicting the weld quality and mechanical properties of difficult-to-join dissimilar materials. Optimized experimental welding parameters in the available literature were taken as input in the deep neural network (DNN). A feed-forward and back-propagated DNN was developed to apprehend the high non-complexity present in friction stir welding of dissimilar materials. Unlike most neural networks, activation functions were altered between layers, effectively capturing non-linearity. The developed model was used to design an experimental condition for dissimilar friction stir welding of aluminum and titanium. Microstructural characterization of the weld was performed to comprehend the influence of parameters on the quality of the joint produced. A close correlation between the machine-learning model and the experimental results was established. The coefficient of determination \(R^2\) between the predicted strength and the actual strength was 0.95 on the training dataset and 0.9 on the testing dataset. Similarly, \(R^2\) between the predicted strength and the actual strength for the experimental dataset was 0.91, thus making the model suitable for predicting experimental conditions and corresponding mechanical properties with the highest accuracy for any unknown dissimilar friction stir welds.
Similar content being viewed by others
References
G. Wang, Y. Zhao, and Y. Hao, J. Mater. Sci. Technol. 34(1), 73 (2018).
S.R. Nathan, A. Visvalingam Balasubramanian, G. Rao, T. Sonar, M. Ivanov, and K. Suganeswaran, Int. J. Lightweight Mater. Manuf. 6(4), 469 (2023).
C.B Smith, J.F Hinrichs, and W.A. Crusan. Robotic friction stir welding: the state of the art, in 4th Friction Stir Welding International Symposium, pp. 14–16, (2003)
K.J. Colligan, The friction stir welding process: an overview, in Friction Stir Welding. ed. by D. Lohwasser, and Z. Chen. Woodhead Publishing Series in Welding and Other Joining Technologies. (Woodhead Publishing, Cambridge, 2010), pp. 15–41
L.E. Murr, Friction-Stir Welding and Processing (Springer International Publishing, Cham, 2015), pp1–18.
Y. Mao, L.M. Ke, L. Fencheng, Q. Liu, C. Huang, and L. Xing, Mater. Des. 62, 334 (2014).
S. Amini, M. Amiri, and A.A. Barani, Int. J. Adv. Manuf. Technol. 76, 255 (2015).
Md. Reza-E-Rabby, and A. Reynolds, Procedia Eng. 90, 637 (2014).
P. Xue, D.R. Ni, D. Wang, B.L. Xiao, and Z.Y. Ma, Mater. Sci. Eng. A-struct. Mater. Prop. Microstruct. Process. 528, 4683 (2011).
M. Muthu, and V. Jayabalan, J. Mater. Process. Technol. 217, 105 (2015).
H. Bisadi, M. Akram Tavakoli, T. Sangsaraki, and K. Tour Sangsaraki, Mater. Des. 43, 80 (2013).
K.P. Mehta, and V.J. Badheka, Int. J. Adv. Manuf. Technol. 80, 2073 (2015).
M. Jayaraman, R. Sivasubramanian, V. Balasubramanian, and A.K. Lakshminarayanan, J. Sci. Ind. Res. 68, 36–43 (2009).
D.C. Montgomery, Design and Analysis of Experiments (John wiley & sons, Hoboken, NJ, 2017), pp1–31.
C. Rathinasuriyan and V.S.S. Kumar, Adv. Mater. Process. Technol. 7(4), 696 (2021).
S. Rajakumar, C. Muralidharan, and V. Balasubramanian, Proc. Inst. Mech. Eng., Part B: J. Eng. Manuf. 224(8), 1175 (2010).
S.M. Senthil, R. Parameshwaran, S. Ragu Nathan, M. Bhuvanesh Kumar, and K. Deepandurai, Struct. Multidiscip. Optim. 62(3), 1117 (2020).
G. Buffa, L. Fratini, and F. Micari, J. Manuf. Process. 14(3), 289 (2012).
R. Palanivel, I. Dinaharan, and R.F. Laubscher, J. Braz. Soc. Mech. Sci. Eng. 41(2), 1 (2019).
N.D. Ghetiya, and K.M. Patel, Procedia Technol. 14, 274 (2014).
H. Okuyucu, A. Kurt, and E. Arcaklioglu, Mater. Des. 28(1), 78 (2007).
S. Verma, M. Gupta, and J.P. Misra, MethodsX 5, 1048 (2018).
T.A. Shehabeldeen, M.A. Elaziz, A.H. Elsheikh, O.F. Hassan, Y. Yin, X. Ji, X. Shen, and J. Zhou, IEEE Access 8, 79896 (2020).
T. Medhi, S.A.I. Hussain, B.S. Roy, and S.C. Saha, Appl. Soft Comput. 104, 107190 https://doi.org/10.1016/j.asoc.2021.107190 (2021).
P.C. Pendharkar, IEEE Trans. Knowl. Data Eng. 17, 1379 (2005).
R.S. Mishra and Z.Y. Ma, Mater. Sci. Eng. R. Rep. 50(1–2), 1 (2005).
V. Msomi and S. Mabuwa, Adv. Ind. Manuf. Eng. 1, 100002 https://doi.org/10.1016/j.aime.2020.100002 (2020).
S. Mabuwa and V. Msomi, Mater. Today: Proc. 45, 5400 (2021).
H.H. Jadav, V. Badheka, D.K. Sharma, and G. Upadhyay, Mater. Today: Proc. 43, 84 (2021).
M. Gomathisankar, M. Gangatharan, and P. Pitchipoo, Mater. Today: Proc. 5(6), 14397 (2018).
S. Rathore, R.K.R. Singh, and K.L.A. Khan, Mater. Today: Proc. 43, 326 (2021).
M. Marichamy and S. Babu, Mater. Today: Proc. 39, 1527 (2021).
J. John, S.P. Shanmuganatan, and M.B. Kiran, Mater. Today: Proc. 5(11), 25458 (2018).
M. Makeshkumar, S.R. Surender, R. Madesh, M. Sasi Kumar, and K.P. Shankar, Mater. Today: Proc. 45, 8054 (2021).
M. Shunmugasundaram, A. Praveen Kumar, N.K. Amudhavalli, and S. Sivasankar, Mater. Today: Proc. 27, 1258 (2020).
K.T. Thilagham and S. Muthukumaran, Mater. Today: Proc. 27, 2260 (2020).
K.T. Thilagham and S. Muthukumaran, Mater. Today: Proc. 27, 2513 (2020).
R. Saravanakumar, T. Rajasekaran, and M. Dhanasekaran, Mater. Today: Proc. 47, 7006 (2021).
A. Kar, S. Suwas, and S.V. Kailas, Mater. Sci. Eng.: A 733, 199 (2018).
J.-W. Choi, H. Liu, and H. Fujii, Mater. Sci. Eng., A 730, 168 (2018).
A. Asmare, R. Al-Sabur, and E. Messele, Metals 10(11), 1480 (2020).
X. Peng, L. Canyi, Z. Yi, and H. Tang, IEEE Trans. Neural Netw. Learn. Syst. 29(1), 218–224 (2016).
S. Bhowmik, R. Panua, D. Debroy, and A. Paul, J. Energy Resour. Technol.-Trans. ASME 139, 042201 https://doi.org/10.1115/1.4035886 (2017).
A. Barakat, and P. Bianchi, SIAM J. Optim. 31(1), 244 (2021).
I. Sutskever, J. Martens, G. Dahl, and G. Hinton. On the importance of initialization and momentum in deep learning, in International conference on machine learning, pp. 1139–1147. PMLR, (2013)
D. Hunter, Y. Hao, M.S. Pukish III, J. Kolbusz, and B.M. Wilamowski, IEEE Trans. Ind. Inf. 8(2), 228 (2012).
Q. Wang, Y. Ma, K. Zhao, and Y. Tian, Ann. Data Sci. 9, 1 (2020).
K. He, X. Zhang, S. Ren, and J. Sun. Delving deep into rectifiers: surpassing human-level performance on imagenet classification, in Proceedings of the IEEE international conference on computer vision, pages 1026–1034, (2015)
V.R. Malik, and K.K. Saxena, Adv. Mater. Process. Technol. 8, 2156 (2022).
V. Malik, and S.V. Kailas, J. Mater. Process. Technol. 258, 80 (2018).
A. Kar, S.V. Kailas, and S. Suwas, Trans. Indian Inst. Met. 72, 1533 (2019).
A. Kar, S. Malopheyev, S. Mironov, R. Kaibyshev, S. Suwas, and S.V. Kailas, Mater. Charact. 171, 110791 (2021).
A. Kar, S. Suwas, and S.V. Kailas, Metall. Mater. Trans. A. 50, 5153 (2019).
A. Kar, D.K. Yadav, S. Suwas, and S.V. Kailas, Mater. Charact. 164, 110371 (2020).
Funding
Not applicable.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
There is no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1: Dataset 1
Sl. No. | Rotation speed (RPM) | Weld speed (mm/min) | Thickness of workpiece (mm) | Base material strength (MPa) |
---|---|---|---|---|
1 | 800 | 70 | 6 | 243 |
2 | 1600 | 180 | 3 | 210 |
3 | 800 | 70 | 6 | 327 |
4 | 1000 | 80 | 6 | 243 |
5 | 1000 | 90 | 3 | 210 |
6 | 1400 | 60 | 3 | 210 |
7 | 1100 | 60 | 6 | 94.1 |
8 | 800 | 40 | 3.42 | 450 |
9 | 1400 | 180 | 3 | 210 |
10 | 1000 | 90 | 3 | 210 |
11 | 1400 | 120 | 3 | 210 |
12 | 1600 | 90 | 3 | 210 |
13 | 1200 | 40 | 6 | 105 |
14 | 1200 | 150 | 3 | 210 |
15 | 900 | 37.5 | 5 | 310 |
16 | 1200 | 30 | 3 | 210 |
17 | 1600 | 135 | 6 | 408 |
18 | 670 | 60 | 6 | 408 |
19 | 1000 | 70 | 6 | 243 |
20 | 560 | 50 | 10 | 228 |
21 | 1000 | 80 | 6 | 327 |
22 | 800 | 75 | 3 | 465 |
23 | 2400 | 240 | 1.6 | 472 |
24 | 800 | 60 | 3 | 210 |
25 | 800 | 90 | 6 | 243 |
26 | 1600 | 150 | 3 | 210 |
27 | 1000 | 330 | 3 | 210 |
28 | 1400 | 150 | 3 | 210 |
29 | 1600 | 180 | 3 | 210 |
30 | 600 | 100 | 2 | 105 |
31 | 1000 | 100 | 6 | 408 |
32 | 1200 | 90 | 6 | 243 |
33 | 1000 | 120 | 3 | 210 |
34 | 900 | 50 | 10 | 228 |
35 | 1500 | 40 | 3 | 281 |
36 | 1100 | 20 | 6 | 572 |
37 | 1500 | 40 | 6.35 | 330 |
38 | 1600 | 90 | 3 | 210 |
39 | 1200 | 180 | 3 | 210 |
40 | 900 | 47.5 | 5 | 310 |
41 | 1000 | 60 | 6 | 408 |
42 | 1600 | 100 | 6 | 408 |
43 | 1000 | 180 | 3 | 210 |
44 | 1800 | 30 | 3 | 210 |
45 | 800 | 40 | 3.42 | 450 |
46 | 670 | 100 | 6 | 408 |
47 | 1200 | 120 | 3 | 210 |
48 | 1000 | 40 | 6.35 | 330 |
49 | 1600 | 30 | 3 | 210 |
50 | 710 | 63 | 6 | 310 |
Sl. No. | Rotation speed (RPM) | Weld speed (mm/min) | Thickness of workpiece (mm) | Base material strength (MPa) |
---|---|---|---|---|
51 | 800 | 180 | 3 | 210 |
52 | 1200 | 37.5 | 5 | 310 |
53 | 1000 | 90 | 3 | 465 |
54 | 800 | 40 | 3.42 | 450 |
55 | 670 | 135 | 6 | 408 |
56 | 1600 | 60 | 3 | 210 |
57 | 800 | 40 | 3.42 | 450 |
58 | 1000 | 60 | 3 | 210 |
59 | 1200 | 80 | 6 | 327 |
60 | 500 | 100 | 2 | 105 |
61 | 1800 | 60 | 3 | 210 |
62 | 1000 | 135 | 6 | 408 |
63 | 1800 | 120 | 3 | 210 |
64 | 1200 | 90 | 3 | 210 |
65 | 1200 | 120 | 3 | 210 |
66 | 1800 | 180 | 3 | 210 |
67 | 800 | 180 | 3 | 210 |
68 | 1200 | 80 | 6 | 243 |
69 | 1200 | 120 | 1.6 | 472 |
70 | 1200 | 70 | 6 | 243 |
71 | 1200 | 60 | 3 | 210 |
72 | 1200 | 90 | 3 | 210 |
73 | 700 | 60 | 6 | 572 |
74 | 450 | 40 | 6 | 310 |
75 | 800 | 90 | 3 | 210 |
76 | 1000 | 30 | 6.35 | 330 |
77 | 900 | 42.5 | 5 | 310 |
78 | 1600 | 120 | 3 | 210 |
79 | 1400 | 90 | 3 | 210 |
80 | 1000 | 120 | 3 | 210 |
81 | 1400 | 90 | 3 | 210 |
82 | 1400 | 47.5 | 5 | 310 |
83 | 700 | 30 | 10 | 228 |
84 | 1800 | 90 | 3 | 210 |
85 | 800 | 40 | 3.42 | 450 |
86 | 800 | 80 | 6 | 327 |
87 | 800 | 90 | 3 | 210 |
88 | 800 | 150 | 3 | 210 |
89 | 1200 | 90 | 6 | 327 |
90 | 900 | 30 | 10 | 228 |
91 | 700 | 50 | 10 | 228 |
92 | 800 | 150 | 3 | 210 |
93 | 560 | 40 | 6 | 310 |
94 | 1200 | 60 | 3 | 210 |
95 | 1000 | 60 | 6 | 408 |
96 | 1600 | 150 | 3 | 210 |
97 | 670 | 100 | 6 | 408 |
98 | 710 | 40 | 6 | 310 |
99 | 1500 | 20 | 6.35 | 330 |
100 | 560 | 30 | 10 | 228 |
Sl. No. | Rotation speed (RPM) | Weld speed (mm/min) | Thickness of workpiece (mm) | Base material strength (MPa) |
---|---|---|---|---|
101 | 1200 | 30 | 6.35 | 330 |
102 | 1800 | 180 | 1.6 | 472 |
103 | 1400 | 42.5 | 5 | 310 |
104 | 1200 | 30 | 3 | 210 |
105 | 1200 | 40 | 6 | 105 |
106 | 1000 | 70 | 6 | 327 |
107 | 1200 | 20 | 6.35 | 330 |
108 | 800 | 120 | 3 | 210 |
109 | 1400 | 30 | 3 | 210 |
110 | 560 | 20 | 6 | 310 |
111 | 1200 | 180 | 3 | 210 |
112 | 1800 | 30 | 3 | 210 |
113 | 710 | 20 | 6 | 310 |
114 | 800 | 80 | 4 | 497 |
115 | 1000 | 150 | 3 | 210 |
116 | 1800 | 150 | 3 | 210 |
117 | 900 | 40 | 10 | 228 |
118 | 450 | 63 | 6 | 310 |
119 | 1400 | 120 | 3 | 210 |
120 | 600 | 60 | 3 | 465 |
121 | 800 | 80 | 6 | 243 |
122 | 1400 | 60 | 3 | 465 |
123 | 1400 | 37.5 | 5 | 310 |
124 | 800 | 90 | 6 | 327 |
125 | 1000 | 100 | 4 | 497 |
126 | 800 | 120 | 3 | 210 |
127 | 800 | 100 | 2 | 105 |
128 | 700 | 100 | 2 | 105 |
129 | 700 | 40 | 6 | 572 |
130 | 1600 | 30 | 3 | 210 |
131 | 1800 | 90 | 3 | 210 |
132 | 670 | 60 | 6 | 408 |
133 | 1000 | 30 | 3 | 465 |
134 | 1200 | 150 | 3 | 210 |
135 | 700 | 40 | 10 | 228 |
136 | 1400 | 60 | 3 | 210 |
137 | 1000 | 180 | 3 | 210 |
138 | 900 | 40 | 5 | 186 |
139 | 1600 | 60 | 6 | 408 |
140 | 1800 | 120 | 3 | 210 |
141 | 1600 | 100 | 6 | 408 |
142 | 1600 | 120 | 3 | 210 |
143 | 1000 | 150 | 3 | 210 |
144 | 1800 | 180 | 3 | 210 |
145 | 1600 | 135 | 6 | 408 |
146 | 1600 | 60 | 3 | 210 |
147 | 1600 | 60 | 6 | 408 |
148 | 1400 | 30 | 3 | 210 |
149 | 800 | 60 | 3 | 210 |
150 | 1500 | 30 | 6.35 | 330 |
Sl. No. | Rotation speed (RPM) | Weld speed (mm/min) | Thickness of workpiece (mm) | Base material strength (MPa) |
---|---|---|---|---|
151 | 1800 | 150 | 3 | 210 |
152 | 1800 | 60 | 3 | 210 |
153 | 1200 | 47.5 | 5 | 310 |
154 | 1000 | 20 | 6.35 | 330 |
155 | 1200 | 40 | 6.35 | 330 |
156 | 1000 | 100 | 6 | 408 |
157 | 800 | 30 | 3 | 210 |
158 | 900 | 100 | 2 | 105 |
159 | 1200 | 42.5 | 5 | 310 |
160 | 1000 | 60 | 3 | 465 |
161 | 1000 | 90 | 6 | 243 |
162 | 1400 | 180 | 3 | 210 |
163 | 1000 | 60 | 3 | 210 |
164 | 1200 | 70 | 6 | 327 |
165 | 1000 | 30 | 3 | 210 |
166 | 800 | 45 | 3 | 465 |
167 | 670 | 135 | 6 | 408 |
168 | 450 | 20 | 6 | 310 |
169 | 1400 | 150 | 3 | 210 |
170 | 1000 | 90 | 6 | 327 |
171 | 800 | 30 | 3 | 210 |
172 | 560 | 40 | 10 | 228 |
Appendix 2: Dataset 2
Sl. No. | Rotation speed (RPM) | Weld speed (mm/min) | Thickness of workpiece (mm) | Base material strength (MPa) |
---|---|---|---|---|
1 | 1000 | 90 | 3.5 | 105 |
2 | 560 | 50 | 3.5 | 320 |
3 | 1000 | 90 | 3.5 | 310 |
4 | 800 | 40 | 3.5 | 105 |
5 | 1000 | 60 | 3.5 | 320 |
6 | 2400 | 240 | 3.2 | 105 |
7 | 1400 | 150 | 3.5 | 105 |
8 | 800 | 150 | 3.5 | 105 |
9 | 1000 | 150 | 3.5 | 320 |
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Maheshwari, S., Kar, A., Alam, Z. et al. Deep Neural Network-Based Approach for Modeling, Predicting, and Validating Weld Quality and Mechanical Properties of Friction Stir Welded Dissimilar Materials. JOM 75, 4562–4578 (2023). https://doi.org/10.1007/s11837-023-06121-w
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11837-023-06121-w