Skip to content

kylespink13/Heat_Sink_Shape_Optimization

Repository files navigation

Partial Differential Equations Final Project

Final Project Topic:

Optimal Control of Systems Governed by Partial Differential Equations, Heat Sink Shape Optimization using the Adjoint Method


The purpose of the MATLAB files are to a) solve 2D Laplace Equation with Dirichlet and convective boundary conditions (Math_112C_Project_Main.m) (constant heat emitted from the bottom of the heat sink in ambient temperature), b) calculate the defined objective function (objectiveFunction.m), $$J = \lambda_{1}\frac{area_{k}}{area_{ref}} + \lambda_{2}\frac{temp_{k}}{temp_{ref}}$$ where $\lambda_{1}$ and $\lambda_{2}$ are weights placed on the normalized area and average temperature (to be minimized), and c) generate the updated geometry using the adjoint method with a gradient descent algorithm (yGradient.m).


Heat Sink Shape Evolution gif

Background:

Final project for UC Irvine's Math 112C—Intro to Partial Differential Equations III course.

Material covered in the course:

  1. Numerical Methods for ODEs and PDEs
    • Euler's Method
    • Error and Stability
    • Finite Differences for Heat Equation and 2D Poisson Equation
  2. Reaction-Diffusion Equations
    • Fisher-KPP Equation
    • Wave Speed
    • Turing Pattern Formation
  3. Random Walks
    • Brownian Motion
    • Stochastic Differential Equations
    • Fokker-Planck Equation
    • Ornstein-Uhlenbeck Process
  4. Black-Scholes Equation
    • Ito's Lemma
  5. Fourier Transforms for PDEs

About

Final Project for UC Irvine Math 112C—Intro to PDEs course

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages