AI-Driven Prediction of Square-Hole Laser Trepanning Performance in AA7075/15%SiC/15% Glass Fiber Hybrid Composites Using Taguchi–ANOVA and Deep Neural Networks

Year : 2026 | Volume : 14 | Special Issue 01 | Page : 1932 1943
    By

    Parthiban K.,

  • Pazhanivel K.,

  • Dinesh S.,

  • Ganesh Kumar A.,

  • Arul M.,

  • Sharmila S.,

  1. Professor, Department of Mechanical Engineering, Government College of Engineering, Bargur, Tamil Nadu, India
  2. Professor, Department of Mechanical Engineering, Thiruvalluvar College of Engineering and Technology, Vandavasi, Tamil Nadu, India
  3. Associate Professor, Department of Mechanical Engineering, Dhanalakshmi College of Engineering, Chennai, Tamil Nadu, India
  4. Assistant Professor, Department of Mechanical Engineering, Thiruvalluvar College of Engineering and Technology, Vandavasi, Tamil Nadu, India
  5. Assistant Professor, Department of Mechanical Engineering, ARM College of Engineering and Technology, Chennai, Tamil Nadu, India
  6. Assistant Professor, Department of Computer Applications, Karpagam Academy of Higher Education, Eechanaari Coimbatore, Tamil Nadu, India

Abstract

Hybrid AA7075 composites reinforced with 15% silicon carbide (SiC) and 15% glass fiber were fabricated via the stir casting technique to improve machining and structural performance. The addition of dual reinforcements into the aluminum matrix was aimed at enhancing hardness, thermal stability, and surface quality during non-traditional drilling operations. Square-hole drilling was performed using a laser trepanning process, and the key responses—hole size accuracy, surface roughness, and taper angle—were systematically investigated. Using a Taguchi L25 orthogonal array integrated with ANOVA, trepanning speed was identified as the most significant factor influencing hole size and taper angle, while laser power was the primary determinant of surface roughness. Among the composites tested, AA7075 reinforced with 15% SiC and 15% glass fiber exhibited superior drilling performance compared with lower reinforcement levels, highlighting the synergistic effect of fiber–particle hybridization. To enhance predictive capability, artificial intelligence (AI) and deep learning models were applied to the experimental data. Random Forest regression and artificial neural networks (ANN) demonstrated close alignment with observed results, while deep neural networks (DNN) achieved the highest predictive accuracy, with regression coefficients above R² > 0.95 and minimized error indices. This integration of Taguchi–ANOVA optimization with AI-based prediction constitutes the novelty of the present work, bridging experimental design and intelligent modeling. The findings present a robust pathway for intelligent process planning and adaptive control in precision machining of hybrid composites, particularly in aerospace and structural applications.

Keywords: AA7075 hybrid composites, artificial intelligence, deep neural networks, glass fiber, laser trepanning, predictive machining, SiC, square-hole drilling, stir casting, Taguchi–ANOVA.

[This article belongs to Special Issue under section in Journal of Polymer & Composites (jopc)]

How to cite this article:
Parthiban K., Pazhanivel K., Dinesh S., Ganesh Kumar A., Arul M., Sharmila S.. AI-Driven Prediction of Square-Hole Laser Trepanning Performance in AA7075/15%SiC/15% Glass Fiber Hybrid Composites Using Taguchi–ANOVA and Deep Neural Networks. Journal of Polymer & Composites. 2026; 14(01):1932-1943.
How to cite this URL:
Parthiban K., Pazhanivel K., Dinesh S., Ganesh Kumar A., Arul M., Sharmila S.. AI-Driven Prediction of Square-Hole Laser Trepanning Performance in AA7075/15%SiC/15% Glass Fiber Hybrid Composites Using Taguchi–ANOVA and Deep Neural Networks. Journal of Polymer & Composites. 2026; 14(01):1932-1943. Available from: https://journals.stmjournals.com/jopc/article=2026/view=240424


References

  1. Saravanan G, Bhaskar GB. Influence of silicon carbide reinforcements on mechanical and wear behavior of AA7075/15% glass fiber hybrid composites. In: Editor, editor. Silicon. 1st edition. City, Country: Publisher; 2022. pp. 1451–1464.
  2. Kannan VS, Lenin K, Srinivasan D, Raj Kumar D. Investigation on laser square hole drilling of AA7475/SiC/ZrSiO₄ composites. In: Editor, editor. Silicon. 1st edition. City, Country: Publisher; 2022. pp. 4557–4574.
  3. Wang CS, Hsiao YH, Chang HY, Chang YJ. Process parameter prediction and modeling of laser percussion drilling by artificial neural networks. In: Editor, editor. Micromachines. 1st edition. City, Country: Publisher; 2022. pp. 529.
  4. Kuan AS, Fang TH, Chang CS, Huang JC. Multi-objective optimization of Nd: YAG laser micro-drilling using Taguchi L25 and grey relational analysis. In: Editor, editor. Int J Adv Manuf Technol. 1st edition. City, Country: Publisher; 2013. pp. 1947–1957.
  5. Yüce BE. Multi-response optimization in machining processes using Taguchi, ANOVA, and grey relational analysis methods: a comprehensive review. In: Editor, editor. J Mater Process Technol. 1st edition. City, Country: Publisher; 2022. pp. 117361.
  6. Shan X, Li J, Zhou J, Liu B, Wang T, Feng J. Experimental study and Random Forest prediction of surface roughness in laser-based manufacturing. In: Editor, editor. Metals. 1st edition. City, Country: Publisher; 2024. pp. 1148.
  7. Sharma A, Patel V, Bhatt P, Singh R, Mehta S. Machine learning approach for surface roughness prediction using Random Forest: application and validation. In: Editor, editor. Sci Rep. 1st edition. City, Country: Publisher; 2025. pp. 2234.
  8. Courtier AF, Zhu H, Eyres L, Cox M, Jones M, Fleming S. Modelling of fibre laser cutting via deep learning. In: Editor, editor. Opt Express. 1st edition. City, Country: Publisher; 2021. pp. 36487–36506.
  9. Mills B. Lasers that learn: the interface of laser machining and machine learning. In: Editor, editor. IET Optoelectron. 1st edition. City, Country: Publisher; 2021. pp. 273–288.
  10. Tewari R, Singh P, Gupta R, Sharma S. Parametric optimization of laser drilling of microwave absorber composites: hole quality and process modelling. In: Editor, editor. Opt Laser Technol. 1st edition. City, Country: Publisher; 2021. pp. 106641.
  11. Desai AA, Patel H, Solanki K, Mehta A. Multi-objective optimization of laser machining parameters in Nd: YAG laser micro-drilling of graphite/epoxy composites using Taguchi-based grey relational analysis. In: Editor, editor. Compos Struct. 1st edition. City, Country: Publisher; 2024. pp. 117901.
  12. Mani N, Subbiah D, Moorthy A, Arunagiri A. Optimization and prediction of machining parameters in nanoparticle-reinforced FMLs using AI techniques. In: Editor, editor. Matéria (Rio J.). 1st edition. City, Country: Publisher; 2025. pp. e20240645.
  13. Ramamoorthy KP, Balasubramani S, Dinesh S, Arul M. Intelligent hybrid optimization of WEDM for Ti-6Al-4V using Taguchi, deep learning and ANOVA for enhanced surface integrity and prediction accuracy. In: Editor, editor. Mater Technol. 1st edition. City, Country: Publisher; 2025. pp. 443–450.
  14. Subbiah D, Mani N, Arunagiri A, Kupusamy P. Tailoring mechanical properties of fiber metal laminates with BaSO₄ nanoparticle-infused epoxy systems. In: Editor, editor. Matéria (Rio J.). 1st edition. City, Country: Publisher; 2024. pp. 29.
  15. Wang Y, Zhang L, Chen X, Liu J, Zhao P, Li Y. Preparation technology, microstructure, and mechanical behavior of SiC particle-reinforced aluminum matrix composites: a systematic study. In: Editor, editor. J Alloys Compd. 1st edition. City, Country: Publisher; 2023. pp. 169698.
  16. Gutema EM, Alemu YT, Getachew SG, Gebreslassie AA. Drilling performance of Al-SiC metal matrix composites using Taguchi-ANOVA methods. In: Editor, editor. J Compos Sci. 1st edition. City, Country: Publisher; 2024. pp. 356.
  17. Antosz K, Gola A, Szybicki D, Kłos S. Machine learning applications for surface roughness prediction in milling operations: review and dataset analysis. In: Editor, editor. Sensors. 1st edition. City, Country: Publisher; 2025. pp. 1120.
  18. Yang H, Chen Q, Li S, Xu W, Zhang T. Artificial intelligence in machining: methods, challenges, and applications for surface quality prediction. In: Editor, editor. J Manuf Sci Eng. 1st edition. City, Country: Publisher; 2024. pp. 110801.
Special Issue Subscription Original Research
Volume 14
Special Issue 01
Received 19/09/2025
Accepted 24/09/2025
Published 21/04/2026
Publication Time 214 Days
7

Login


My IP

PlumX Metrics