Process characterisation of 3D-printed FDM components using improved evolutionary computational approach

Fused deposition modelling (FDM) is an additive manufacturing technique deployed to fabricate the functional components leading to shorter product development times with less human intervention. Typical characteristics such as surface roughness, mechanical strength and dimensional accuracy are found to influence the wear strength of FDM fabricated components. It would be useful to determine an explicit numerical model to describe the correlation between various output process parameters and input parameters. In this paper, we have proposed an improved approach of multi-gene genetic programming (Im-MGGP) to formulate the functional relationship between wear strength and input process variables of the FDM process. It was found that the improved approach performs better than MGGP, SVR and ANN models and is able to generalise wear strength of the FDM prototype satisfactorily. Further, sensitivity and parametric analysis is conducted to study the influence of each input variable on the wear strength of the FDM fabricated components. It was found that the input parameter, air gap, has the maximum influence on the wear strength of the FDM fabricated component.