Modeling and Parametric Optimization of Gas Metal Arc Welding for Inconel 625 and UNS S31603 Using RSM and ANN

·        RSM models achieved high predictive accuracy with R² of    95.14% (DOP), 98.78% (BW), and 95.40% (BH), confirming strong statistical relationships between WAAM process parameters and bead geometry. ·        ANOVA identified two-way interaction effects as dominant contributors. WFR × GFR was most influential for BW (38.06%), Voltage × TS most affected DOP (21.08%), and Travel Speed was the primary linear factor for BH (29.73%). ·        The ANN model (4-10-3 feedforward, Levenberg–Marquardt) achieved overall R = 0.989 and showed superior prediction accuracy for Bead Width due to its ability to capture complex nonlinear parameter interactions. ·        Both RSM and ANN proved effective as complementary modeling tools. RSM provides interpretable statistical insight while ANN excels at nonlinear relationships. Hybrid use of both is recommended for WAAM process optimization.

[1]   Dinovitzer, M., Chen, X., Laliberte, J., Huang, X., & Frei, H. (2019). Effect of WAAM process parameters on bead geometry and microstructure. Additive Manufacturing, 27, 138–146. [2]  Yu, R., Zhao, Z., Bai, L., & Han, J. (2020). Prediction of weld reinforcement based on vision sensing in GMA additive manufacturing process. Metals, 10(8), 1041. [3]    Wacker, C., Köhler, M., David, M., et al. (2021). Geometry and distortion prediction of multiple layers for WAAM using artificial neural networks. Applied Sciences, 11(10), 4694. [4]   Lee, C., Seo, G., Kim, D. B., et al. (2021). Defect detection AI model for WAAM using HDR images. Applied Sciences, 11(16), 7541. [5]   Le, V. T., Doan, Q. T., Mai, D. S., et al. (2022).Prediction and optimization of WAAM parameters for 316L stainless steel. Journal of the Brazilian Society of Mechanical Sciences and Engineering, 44(9).  [6]    Vora, J., Parmar, H., Chaudhari, R., et al. (2022). Mechanical properties of GMAW-based WAAM UNS S31630. Journal of Materials Research and Technology, 20, 2748–2757. [7]    Vora, J., Parikh, N., Chaudhari, R., et al. (2022). Optimization of bead morphology in WAAM steel. Applied Sciences, 12(10), 5060.