Sustainability-Driven Multi-Objective Evolutionary Orienting in Additive Manufacturing

The build orientation of parts which are produced by Additive Manufacturing processes has a considerable effect on material and energy consumptions. The choice of a specific build orientation from the range available is crucial because it has a direct impact on the economic and environmental aspects of part production. Several methods based on functional, economic and environmental requirements have been proposed to automate the choice of an optimal orientation. Those approaches involve trade-off conditions concerning environmental aspects that emerge from scientific investigations. Impact indicators are generally aggregated for the purpose of comparison. This approach allows a reduction in the number of items to be compared before making a decision, but some information provided from the life cycle assessment analysis is lost or not fully exploited. This paper proposes an innovative procedure to overcome these limitations. Firstly, life-cycle impact assessment indicators of the process are considered as the objectives to model the Pareto front of environmentally non-dominated solutions. This first step of the proposed method uses evolutionary algorithms. Then, all the orientations that passed the first selection step are ranked based on their estimated cost to find the optimal solution, which corresponds to the optimal orientation. This approach considers all the environmental impacts during the multi-objective optimisation, according to the principles of impact assessment. This procedure can be applied to any combination of products and additive manufacturing processes. The proposed method is applied to three additive manufacturing technologies to prove its adaptability. Three popular sample parts are used as benchmarks. The application shows that considerable environmental and economical benefits can be achieved through the proposed approach. All tests are repeated using two different evolutionary algorithms to show the effects of the calculation method on Pareto results.