A Methodology for Developing a Model for Energy Prediction in Additive Manufacturing Exemplified by High-Speed Laser Directed Energy Deposition

Abstract

AbstractThe need for energy-efficient manufacturing technologies is growing due to the increasing pressure from climate change, consumers, and regulations. Additive manufacturing is claimed to be a sustainable manufacturing technology, especially for individualized products and small batches. To include the energy demand in the decision-making process on whether a part should be manufactured by additive or rather by subtractive or formative manufacturing, the energy demand which arises during manufacturing of a part must be predicted before the manufacturing process. For this, individual energy prognosis models are needed for each individual AM system. This paper, therefore, presents a methodology that enables users to develop a customized model to predict the energy demand of their AM System.Four steps are necessary to create a model for energy prediction. First, the structure of the investigated system has to be captured. Here the subsystems and their corresponding process parameters are identified. Then the build cycle is analyzed and divided into several process steps in which the power consumption of the subsystems repeatedly follows the same pattern. Afterwards, those process parameters, that have a significant influence on the energy demand of each subsystem are identified within full factorial design of experiments and subsequently analyzed in detail. In the final step, individual models are developed for the energy demand of each subsystem for each process step. These individual models are then aggregated to create an overall model. The application of the methodology is also demonstrated and validated by the example of high-speed laser directed energy deposition.