Revisiting the optimal shape of cooling fins: A one-dimensional analytical study using optimality criteria

Joe Alexandersen,Ole Sigmund

DOI: https://doi.org/10.1109/ITherm51669.2021.9503196

2021-01-11

Abstract:This paper revisits the optimal shape problem of a single cooling fin using a one-dimensional heat conduction equation with convection boundary conditions. Firstly, in contrast to previous works, we apply an approach using optimality conditions based on requiring stationarity of the Lagrangian functional of the optimisation problem. This yields an optimality condition basis for the commonly touted constant temperature gradient condition. Secondly, we seek to minimise the root temperature for a prescribed thermal power, rather than maximising the heat transfer rate for a constant root temperature as previous works. The optimal solution is shown to be fully equivalent for the two, which may seem obvious but to our knowledge has not been shown directly before. Lastly, it is shown that optimal cooling fins have a Biot number of 1, exhibiting perfect balance between conductive and convective resistances.

Optimization and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is about the optimal shape design of a single heat sink. Specifically, the author uses the one - dimensional heat conduction equation and convective boundary conditions to re - examine the optimal shape problem of the heat sink. Different from previous studies, this paper mainly makes the following improvements:

1. **Optimization method**: Different from previous work, this paper uses the optimization conditions based on the Lagrange functional to solve the problem. This method obtains the optimization conditions by requiring the stationarity of the Lagrange functional, thus providing a simpler basis for the common constant - temperature - gradient condition.

2. **Objective function**: The objective of this paper is to minimize the root temperature under a given heat - power input, rather than to maximize the heat - transfer rate under a constant root temperature. Although these two problems are different, this paper proves that their final solutions are completely equivalent.

3. **Optimal - solution characteristics**: This paper also proves that the Biot number of the optimal heat sink is 1, which indicates that a perfect balance between the conduction and convection resistances has been achieved.

### Main contributions of the paper

- **Optimization conditions**: By introducing the optimization conditions of the Lagrange functional, this paper provides a more concise physical explanation for the constant - temperature - gradient condition.
- **Equivalence proof**: It proves the equivalence of the two optimization objectives of minimizing the root temperature under a given heat - power input and maximizing the heat - transfer rate under a constant root temperature.
- **Biot - number analysis**: It points out that the Biot number of the optimal heat sink is 1, which means that the balance between the conduction and convection resistances has reached the best state.

### Mathematical model and derivation

This paper uses the one - dimensional heat conduction equation and convective boundary conditions to describe the temperature distribution of the heat sink. The specific equations and boundary conditions are as follows:

\[
-\frac{d}{dx}\left(k t(x) \frac{d\theta}{dx}\right)+2h\theta(x) = 0\quad\text{for}\quad0 < x < L\tag{1a}
\]

\[
-k t(0) \frac{d\theta}{dx}\bigg|_{x = 0}=\tilde{Q}_0\tag{1b}
\]

\[
-k t(L) \frac{d\theta}{dx}\bigg|_{x = L}=0\tag{1c}
\]

where:
- \(k\) is the thermal conductivity of the material,
- \(t(x)\) is the thickness function,
- \(\theta(x)=T(x)-T_{\infty}\) is the temperature difference,
- \(\tilde{Q}_0=\frac{Q_0}{b}\) is the heat - power input per unit width.

### Optimal solution

By solving the above equations and boundary conditions, the optimal solution is obtained:

- **Optimal length**:

\[
L=\sqrt[3]{\frac{3kA}{h}}\tag{6}
\]

- **Optimal thickness function**:

\[
t(x)=\frac{h}{k}(L - x)^2\tag{7}
\]

- **Optimal temperature function**:

\[
\theta(x)=\frac{\tilde{Q}_0}{hL}\left(1-\frac{x}{L}\right)\tag{8}
\]

- **Optimal thermal resistance**:

\[
\tilde{C}=\frac{\tilde{Q}_0^2}{hL}\tag{9}
\]

These results indicate that the optimal design of the heat sink can not only effectively reduce the root temperature, but also achieve the best balance between conduction and convection.