Microstructure Control and Property Prediction in Inductively Coupled Selective Laser Melting

Abstract

Metal Additive Manufacturing has earned significant industrial and research inclination in the recent years given faster production times and less wastage of material as compared to subtractive or traditional manufacturing. However, issues and concerns regarding quality, and process control, repeatability and consistency with Additive Manufacturing is still under works. With more demand for tailored manufacturing suitable for specific end-applications, controlling physical properties by modifying process parameters or by inclusion of complimentary processes to Additive Manufacturing has been well known. This proposed study aims at delivering an effective way of addressing the thermal distribution control in a Powder Bed Fusion process for Selective Laser Melting of 316 Stainless Steel. Selective heating of the powder bed through a co-axially integrated induction system with a conventional laser setup is proposed based on literature survey, simulation data and baseline experiments. A system suitable for the proposed concept is designed, fabricated, and assembled. Extensive experimental trials are conducted to study the effect of the auxiliary heating source on the microstructure and the variation in physical properties of the built deposits. Effect of controlling the cooling rate of the melt pool on the resulting mechanical properties is reviewed and discussed. Machine Learning to help predict physical properties and control the process flow given the complex nature, is proposed through a Digital Twin. Analytical data generated through the depositions is used to model the training and testing of the Digital Twin and a framework for a more comprehensive study of the same is laid. A proof of concept demonstrating the use of induction in controlling the microstructure is achieved as well as initial efforts towards the Digital Twin are also successfully achieved.