Decision Analysis for Selecting FDM Process Parameters using Bayesian Network Approach: Abstract ID: 783962

摘要

Additive manufacturing technologies overcome the complex design restrictions imposed by traditional manufacturing processes. Fused deposition modeling (FDM) is one of the most popular additive manufacturing processes. However, the industrial-scale applications for the FDM process are still limited due to inconsistent properties of the printed parts (e.g., compressive strength). Although the features of an FDM printed part can be improved by selecting a proper combination of process parameters, it is typically a tedious process as there are numerous possible combinations of process parameters that can be accounted for. In addition, each combination may have different impacts on the part's properties. Selecting the optimal combination of process parameters is complicated further by the existence of uncertainties in different stages of the additive manufacturing processes. To account for uncertainties in the FDM process, a framework based on the Bayesian network (BN) approach was explored to enhance the decision-making process in determining the optimum combination of FDM process parameters. The unique feature of the proposed method is that it can be applied to both discrete and continuous process parameters. This framework could be leveraged for process parameters selection of other manufacturing processes.