Prediction of thermal and mechanical properties of acrylate-based composites using artificial neural network modeling

Abstract

Poly(methyl methacrylate) (PMMA) has a broad spectrum of uses, especially in medical applications. The role of fine-grained alumina particles of PMMA composites was investigated in this study. The composites were based on PMMA modified with dimethyl itaconate (DMI) as a matrix and alumina particles (Al2O3) and alumina doped with iron (Al2O3-Fe) modified with 3-aminopropyl-trimethoxysilane (AM) and flax oil fatty acid methyl esters (biodiesel) as reinforcements. Three particle sizes were measured (~0.4, ~0.6 and ~1.2 μm). The highest thermal conductivity values were measured for the composite 5 wt.% Al2O3-Fe-AM. With the addition of 3 wt.% Al2O3-AM to the PMMA/DMI matrix, mechanical properties were improved (tensile strength, strain, and modulus of elasticity). An artificial neural network model based on the Broyden-Fletcher-Goldfarb-Shanno iterative algorithm was investigated for prediction of thermal conductivity and mechanical properties of the composites showing satisfactory results. This is relevant for applications for optimization of dental materials to produce dentures, which were exposed to variations in temperature during the application.

Keywords

• composite materials
• thermal conductivity
• al2o3 particles