Modeling of the Process Parameters Influencing Cold Metal Transfer (CMT): Development of an Approach Based in Causal Networks

Abstract

The new Additive Manufacturing technologies combined with other transformations such as increasing digitalization and automation are creating new opportunities and associated challenges. For all the AM technologies, both functional and topological design of parts need to be completely reconsidered, for example, it becomes possible to integrate multiple functions into a single part. From the manufacturing process outlook, the links between process parameters and design requirements have to be unveiled. Discovering those links is a challenging process because of the existence of cross-impacts. The nature of the relationships is also probably highly nonlinear in some cases. Using a traditional design of experiment approach to discover those links might be time-consuming and the number of parameters to test might be enormous. This article applies an approach based on SI metrics combined with the functional representation of the manufacturing process to form causal-graphs. Those graphs are used in a preliminary phase to generate models of the interrelations between manufacturing and design parameters. Those models are also used to guide the experimental process and to minimize the amount of experiments to be conducted to validate the model. The approach is applied to the Cold Metal Transfer (CMT) technology currently in test in our laboratory