Energy component in the density of selective laser melting fabricated prototype

Due to recent trends in the growing demand for customized and better product quality, inappropriate selection of linear energy density values during the selective laser melting (SLM) process will not only result in a decrease in productivity but will also have negative implications on environmental performance. The SLM process is complex and involves multiple inputs; therefore, the physical-based models are difficult to be formulated. In this context, the present work proposes two variants of the evolutionary approach using genetic programming (GP) in the formulation of a functional expression between the density of fabricated parts and seven inputs of the SLM process. These variants are proposed by introducing two model selection criteria from statistical learning theory to be used as fitness functions in the GP framework. The performance of the proposed energy-based density models is evaluated against the actual experimental data based on five statistical metrics and hypothesis testing. The relationships between the system parameters are unveiled and can be used for effectively monitoring the economic and environmental performance of SLM processes. It was found that the hatching space and linear energy density have the highest impact on the SLM density component. A major contribution of the study is that the optimum values of the inputs can be selected to curtail the energy usage from the SLM process.