Running Python Simulations in the Browser—The Case of Thermal Metamaterial Optimization

Scientific WebApps are gaining popularity both in the classroom and for research. Notable examples include Apps from the NanoHub ecosystem and the Material Project. Current mainstream platforms, however, require a backend that sends Javascript code to the front end, which in turn presents the WebApp to the end user. In this talk, we will outline a paradigm-shifting approach where Python code is executed in the browser, and the resulting WebApp can even run without an Internet connection. We show an example, based on PyScript [1], where a thermal metamaterial is optimized via a Graphical User Interface (GUI); the GUI allows the user to specify the prescribed full effective thermal conductivity tensor. The code executes topology optimization of a periodic 2D domain in the browser and visualizes the optimized structure. Finally, the corresponding STL file can be downloaded for 3D printing. The WebApp, dubbed HeatOpt, can run locally or be deployed via GitHub. Because of their cost-effective (in most cases, free) deployment and ability to use mainstream languages (e.g. Python), we anticipate this approach will challenge current server-based.<br/>[1] https://pyscript.net/