HeatOpt is a "computational paper" showcasing a heat transport topology optimization solver accompanied by technical details.
The main goal is to optimize a material with a prescribed effective thermal conductivity tensor of the form
The tensor is reconstructed by running three separate simulations, one for each direction of the applied temperature gradient. The thermal conductivity along a given direction
We solve the heat conduction for the directions
The engine is written in Python and is based on PyScript
Author: Giuseppe Romano (romanog@mit.edu)
Computational Paper: Link
To run HeatOpt as a stand-alone program, run the following code
python -m venv env
source env/bin/activate
pip install -r requirements.txtOptimizing a material with tensor
can be achieved with the following code
from heatopt import get_optimizer,write_stl
import matplotlib.pylab as plt
optimizer,grid = get_optimizer()
kappa,fig,x = optimizer(0.1,0.2,-0.04)
#Always check if the final kappa is close to the prescribed one
print(kappa)
with open('structure.stl','w') as f:
f.write(write_stl(x))
plt.savefig('structure.png')
plt.show()Note that the thermal conductivity for the base material is 1 W/m/K.