DSpace Arşivi :: by Yazar "Savran, Melih" değerine göre listeleniyor

Listeleniyor 1 - 1 / 1

Optimization of drilling parameters for minimizing delamination factor in biocomposites using multiple nonlinear neuro-regression and stochastic methods
(İdris Karagöz, 2025) Savran, Melih; Türkan, Mücahit Osman; Öncül, Mustafa; Aydın, Levent
In this study, a novel design optimization strategy is proposed to enhance the drilling performance of HDPE/Washingtonia fiber biocomposites, considering operational parameters such as drill diameter (d), feed rate (f), and spindle speed (N). A detailed investigation utilizing multiple nonlinear neuro-regression analyses is conducted to predict the delamination factor (Fd), based on a dataset obtained from a literature study. 14 candidate mathematical functions are suggested for modelling, and their accuracy is assessed through R2training, R2testing, and R2validation metrics, and boundedness check. The identified effective models are subsequently employed in optimization using modified version of four methods: Nelder-Mead (NM), Simulated Annealing (SA), Random Search (RS), and Differential Evaluation (DE). The optimum values for parameters f, N and d were determined to be 50 mm/min, 355 rev/min, and 5.00 mm respectively, resulting with the minimum Fd values as 1.11615.