Prediction and Optimization of Drilling Outputs in C355 Alloy with ANN Integrated CRITIC-WASPAS-COCOSO Method

ALPARSLAN C., Ozhabiboglu C., Vijayananth K., BAYRAKTAR Ş.

SILICON, vol.18, no.10, pp.3441-3461, 2026 (SCI-Expanded, Scopus)

  • Publication Type: Article / Article
  • Volume: 18 Issue: 10
  • Publication Date: 2026
  • Doi Number: 10.1007/s12633-026-03733-0
  • Journal Name: SILICON
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Chemical Abstracts Core, Compendex, INSPEC, Materials Science & Engineering Collection (ProQuest), Technology Collection (ProQuest)
  • Page Numbers: pp.3441-3461
  • Keywords: Al-Si based alloy, ANN, CRITIC-WASPAS- COCOSO, Mechanical properties, Microstructure
  • Recep Tayyip Erdoğan University Affiliated: Yes

Abstract

Structural, mechanical and machining properties of the as-cast and heat-treated Al-5Si-1Cu-Mg alloy produced by the permanent mold casting were experimentally investigated. Uncoated & Oslash;8 mm high-speed steel drills were used in the drilling tests. Drilling parameters were selected as five different cutting speed (V-m/min), feed rate (f-mm/rev) and constant depth of cut (DoC). The thrust force (Fz) and torque (Mz) parameters resulting from the drilling were evaluated with statistical analyses depending on the V and f variables. In this context, experimental results were analyzed using the CRITIC-WASPAS-COCOSO and Artificial Neural Networks (ANN) methods. Microstructure analyses revealed that as-cast alloy contained alpha-Al matrix, eutectic Si, acicular beta-Fe (beta-Al5FeSi) and script-like pi-Fe(pi-Al8Mg3FeSi6) intermetallic phases. In the heat-treated alloy, it was determined that the beta phase, apart from these phases, transformed into the theta(Al7FeCu2) phase under the effect of heat treatment. With CRITIC-WASPAS-COCOSO analysis, optimum drilling parameters were determined as 125 m/min and 0.05 mm/rev for V and f, respectively. Appropriate network structures were created with the ANN. The results obtained revealed that both CRITIC-WASPAS-COCOSO and ANN method can successfully predict the experimental data.