Welcome to my personal website, built with Franklin.jl and powered by Julia, my preferred language for scientific computing and engineering simulation.
I work on the modeling, simulation, control, and visualization of thermal and energy systems, with a strong focus on thermal management. My work covers data-center cooling, buildings, HVAC&R systems, heat pumps, electronics cooling, heat recovery, and integrated energy systems. My goal is to build practical simulation tools that connect physical models with real engineering decisions.
My work combines system modeling, numerical simulation, performance analysis, and control-oriented methods such as PI control, extremum-seeking control, model predictive control, and reinforcement learning for physical-system control.
I also enjoy turning simulation results into interactive visual tools. I am interested in digital-twin-style workflows, where geometry, simulation data, and real-time interaction are brought together. Recent examples include Julia/GLMakie-based visualization of CFD-driven airflow around electronics cooling hardware, and EnergyPlus-to-3D-mesh workflows for building inspection, rendering, sun-path studies, and shadow analysis.
I do not believe in a single “perfect” programming language. I have worked with MATLAB, C/C++, Python, and Modelica, but Julia is the language I most often reach for today. It offers a strong balance between performance, clarity, and scientific expressiveness, especially for the type of simulation, optimization, visualization, and modeling work I do.
More recently, I have been integrating uncertainty-aware and machine-learning-assisted methods into engineering simulations, including uncertainty quantification, global sensitivity analysis, Bayesian inference, model calibration, and physics-informed neural networks for surrogate modeling.
This website is where I share selected projects, research notes, visual experiments, and small software demonstrations related to thermal management, scientific machine learning, Julia-based modeling, digital-twin-oriented simulation, and engineering visualization.
Real-time GLMakie visualization of streamline particles flowing through a CPU tower cooler over an ATX motherboard, driven by Ansys Fluent CFD data with reactive switching across fan-RPM operating points.
This GLMakie animation was created for non-commercial academic visualization. The geometry was adapted from the GrabCAD model "iwongou 12cm CPU Cooler": original GrabCAD model page. The CAD model was converted to OBJ, rescaled, and rendered only for visualization purposes. No CAD files, OBJ files, or modified geometry files are redistributed here. All rights to the original CAD model remain with its creator. This visualization is not affiliated with or endorsed by GrabCAD, Stratasys, the original creator, or any real-world company/product represented by or resembling the model.
This GLMakie animation was created for non-commercial academic visualization. The motherboard geometry was adapted from the Sketchfab model "Asus Strix b-550-f Gaming Motherboard Realistic" by MUSHROOM_BUILDS, licensed under Creative Commons Attribution: original Sketchfab model page. The model was converted, rescaled, and rendered only for visualization purposes. No original model files, converted OBJ files, or modified geometry files are redistributed here. All rights to the original model remain with its creator. This visualization is not affiliated with or endorsed by Sketchfab, Epic Games, ASUS, the original creator, or any real-world company/product represented by or resembling the model. GLMakie animation of a building generated from an EnergyPlus IDF file via my IDFMesh.jl converter — extruded walls, fenestrations, baseboards, trims, and roof eaves rendered for visualization.
This GLMakie animation was created for non-commercial academic visualization. The building geometry was generated from an EnergyPlus example input file and converted to OBJ only for rendering and animation purposes. No EnergyPlus input files, generated OBJ files, DXF files, or modified geometry files are redistributed here. The visualization is shown only as a rendered video. EnergyPlus is developed by the U.S. Department of Energy and its contributors, and all rights, trademarks, and license terms associated with EnergyPlus remain with their respective owners. This visualization is not affiliated with or endorsed by the U.S. Department of Energy, NREL, LBNL, or the EnergyPlus development team.