Radiative heat transfer in porous media

Programming language : MATLAB

Tags: simulation, heat transfer, machine learning, deep learning, etc

Objective

I conducted research during my master’s program to investigate the characteristics of radiative heat transfer in porous materials with surface or internal voids. A machine learning model was essential for predicting the radiative properties of porous media, varying with different porous structure, material composition, and types of radiative heat. To accomplish this task, I applied thermodynamic domain knowledge, machine learning optimization techniques, and simulation tools. The research findings can be applied in various engineering fields, such as controlling the production of complex structures in metal 3D printing processes by sintering metal powders with lasers or designing radiation shielding materials.

• Performed Monte Carlo Ray Tracing simulation to approximate radiative properties of porous media
• Prepared training sets for Neural Networks through simulations and data cleaning processes
• Developed Neural Networks surrogate model to predict radiative properties of porous media accurately and efficiently
• Remodeled open-source Fortran code in Linux environment to obtain complex geometry data
• Utilized parallel computing in simulations to improve space and time efficiency

Monte Carlo ray tracing simulation

Publications & Conferences

• Hyun Hee Kang, Mine Kaya and Shima Hajimirza, “A data driven artificial neural network model for predicting radiative properties of metallic packed beds”, Journal of Quantitative Spectroscopy and Radiative Transfer, March 2019
• Hyun Hee Kang, Shima Hajimirza,” Modeling of Multi-Parameters Radiation in Porous Metal Via Machine Learning”, Thermal & Fluids Analysis Workshop (TFAWS), NASA Johnson Space Center,Houston, TX, 2018
• M. Kaya, H. H. Kang, and S.Hajimirza, “ A Machine Learning Approach for Modeling Radiation in Packed Bed Systems’, Proc. Of the the 6th International Conference on Computational Thermal Radiation in Participating Media, Cascais, Portugal, April 11-13, 2018