Machinability and tribological optimization of origami-inspired Almond Shell-PMMA via RSM, ML, and TOPSIS.

An integrated approach combining Response Surface Methodology (RSM), Machine Learning (ML-SVM) and TOPSIS optimization method is applied in this study to analyse the tribological behaviour of 3D printed patterns of almond shell-PMMA (polymethyl methacrylate) origami inspired composites and machinability. The process of taking advantage of fold-based geometrical patterns (such as Miura-ori or triangular tessellation) to enhance load distribution and energy absorption in 3D printed specimen is called origami-inspired. These patterns promote a certain degree of structural rigidity and cause weakened materials to deform under applied stress in a controlled manner in general to improve the mechanical strength and wear-resistance. Besides tribological performance factors such as wear rate and friction coefficient, the factors to be evaluated on the machinability properties of cutting force, surface roughness and material removal rate include spindle speed (3000-9000 rpm), feed rate (0.05-0.15 mm/rev), and depth of cut (0.2-0.6 mm). Although the machine learning algorithms were able to make predictive models concerning wear performance and machinability, RSM was addressed to plan the experiments and conclusion of the parameters interaction. TOPSIS method identified the parameters combination that will serve the best by balancing between tribological efficiency and machinability. The novelty aspect of the current work is the inclusion of agricultural waste (10 percent almond shells) to the polymer matrices and and the use of a hybrid optimization strategy on the ways to optimize its functional properties with respect to being used in wear and machining applications. Notable findings indicate that the most influential variable affecting machinability and tribological results is the feed rate; in the best case scenario there is achievement of surface roughness of 1.2 10-15 m and wear rate aside at 1.5 in 10-4 mm 3/Nm. The proposed model was able to give a workable and industrially friendly composite method with great efficiency and sustainability in terms of optimization performance and predictability (R2 > 0.95).