Modeling and Control for Laser Based Additive Manufacturing Processes

Additive manufacturing, also called 3D printing, is well known for its ability to produce complex shapes, with great reductions in manufacturing time and material over traditional machining. This project focuses on additive manufacturing processes that use a powerful laser to melt a pool of metal powder onto a growing work piece. Inaccurate knowledge of the temperature in the vicinity of the melt pool can lead to use of too much or too little laser energy, and thence to errors in the dimensions of the final part. The complex geometry of 3D-printed components can make accurate knowledge of temperature difficult. This project studies the use of feedforward control to address this problem, by augmenting real-time sensor measurement of the melt pool with z family of pre-computed thermal models representing the part as it is formed. This strategy anticipates necessary changes in input thermal energy, and thereby allows much higher accuracy than responding to errors only after they occur. Results from this project will have potential applications ranging from customized medical implants to aerospace fuel nozzles to turbine blade repair. This project will advance the state of the art in additive manufacturing for metal components. The project also supports recruitment of women and minorities through university-level initiatives, specifically Penn State's Women in Science and Engineering Research (WISER)and Minority Undergraduate Research Experience (MURE) programs, and curriculum development for an additive manufacturing summer camp for high school teachers.

Despite its huge potential, additive manufacturing has not yet achieved widespread adoption in industry due to multiple challenges in the process, including accuracy of part geometry, mechanical and material properties of the processed part, and surface roughness. To resolve these issues, fundamental understanding and advanced technologies for modeling and control are in demand to improve the accuracy and process stability of additive manufacturing. This project investigates novel modeling methodologies and control algorithms for laser-based additive manufacturing processes, specifically for directed energy deposition. Intellectual significance offered by the research to the additive manufacturing scientific community include: 1) A physics-based, control-oriented dynamic model that accounts for 3D part geometry and thermal history. This model will have high fidelity yet still be amenable for real-time multivariable control; and 2) A nonlinear multi-input, multi-output control paradigm. This control paradigm takes advantage of the special nonlinear structure of the additive manufacturing process dynamics without resorting to linearization, and also addresses constraints for input/state constraints, performance optimization and robustness with respect to uncertain model parameters. Experimental tests will be performed by the research team to evaluate the effectiveness of the modeling and control methodologies.