Conjugate Heat Transfer Analysis on a graphics card.
Problem Definition
Perform a steady-state conjugate heat transfer analysis on a model of a graphics card. We have to use appropriate materials for the simulation. We have to explain the reason to choose the mesh.
1. Run the simulation by varying the velocity from 1m/sec to 5m/sec for at least 3 velocities and discuss the results.
2. Find out the maximum temperature and heat transfer coefficient attained by the processor.
3. Prove that the simulation has achieved convergence with appropriate images and plots.
4. Identify potential hotspots on the model.
Introduction
Conjugate Heat Transfer (CHT)
The Conjugate Heat Transfer (CHT) analysis allows for the simulation of heat transfer between solid and fluid domains by exchanging thermal energy at the interfaces between them. In CHT we focus on conjugating the boundary condition at the fluid-solid interface through coupled heat transfer analysis. This coupled field is termed as Conjugate Heat Transfer (CHT).
Conjugate Heat Transfer (CHT) solver is used to understand the interface properties and flow properties due to change in the wall heating rates.
It is also useful to examine the heat penetration of solid in various wall materials with different time scales. CHT can also be used where the solution domain consists of more than one sub-domains in which heat transfer governed by different differential equations.
This model describes the heat transfer between a body and a fluid flowing over or inside it as a result of the interaction of two objects.
CHT studies are focused to understand the limit of applicability of the assumptions which are considered to recover the experimental data. It is also considered as a tool for the simultaneous prediction of wall heat flux and temperature.
Conjugate heat transfer problems can be solved by either analytical methods or by numerical methods. The analytical CHT studies require a thorough understanding of the given heat transfer problem and involve complex mathematics. Such studies involve a lot of simplifications based on certain valid assumptions. Hence the correctness of the solution depends largely on the validity of assumptions made.
On the other hand, numerical CHT solution methods are relatively simpler if the numerical methods to be followed are clear. Moreover, a correctly implemented numerical CHT solver is not just limited to one CHT problem. Hence the same CHT algorithm can be applied to several CHT problems with minimum efforts in terms of solution domain modifications.
Case Setup
1. First we are going to open the workbench and going to click on Fluid Flow(Fluent). It is going to open the module in the Project Schematic.
2. Now we are going to click on the geometry. It will take you to the spaceClaim. Here we are going to open the existing design. So now we can see the original design. On which we are going to do our simulation.
3. We want the mesh to be finer in the plates of the fins as compared to the base of the fins. So we are going to split the fins into two components. One is fins-base and the second one is fins-plates. So in order to that, we are going to use the split body command and we are going to create two components.
4. Now we want that information should flow into all body parts of the processor so we are going to select the share option in the workbench.
5. Now we are going to open the mesh. Here first we are going to insert the mesh and we are going to give the enclosure mesh to be hexagonal and we are going to give the element size to be 15 mm.
6. Now we are going to give the body size to every part of the graphic card. So we are going to insert the sizing for every body part. We are going to give the following mesh element size:-
We are going to give the 3 mm element size to be the base of the graphic cards. We don't want to
We are going to give the 0.5 mm element size to be processor. Processor is our source of energy that's why we are going to give the fine mesh to the processor.
We are going to give the 1.5 mm element size to the base of the fins. The work of the fins is to keep processor cool so the fins should be fine.
Now we are going to give the 0.22 mm element size to the plates of the fins. Plates are very thin and the heat is going to the atmosphere from the plates. So it will be better if we make the plates finer as compared to the base of the fins.
We have taken the element size to 0.5 mm to the capacitor. There is also some heat generation in the capacitor so we are going to make it finer as compared to the base of the graphic card.
Now we are going to take the element size 2 mm to the four blocks of the graphic cards.
7. We are going to naming the body parts.
8. Now we are going to update the mesh. Now we are going to Setup the case. We are going to take the model to be turbulent k-`omega` SST model.
9. In the Fluent we will go to Physics. Here we choose the type of solver, we have taken the pressure-based type solver. The Mach number value is less than 0.33 that's why we are using pressure-based solver.
10. Now we have to choose the velocity formulation, we have taken absolute velocity formulation. Now we are going to choose the time, we are using a steady time. We want to calculate the heat transfer coefficient so we are going to consider the energy equation also.
11. Now we are going to give the initial boundary condition. So at the inlet, we are going to give 2.5 m/s inlet velocity and at the outlet, we are going to give the pressure to be zero (static pressure).
12. Now we are going to select the material. We are taking air to be our fluid material and for the solid body, we are going to use different-different materials. We want the processor to be cooled. So we are going to take material for the processor to be copper. Because copper is a good heat conductor so the heat transfer coefficient will be high for the copper. Because of that more heat will go into the atmosphere and the processor will become cooled.
In the graphic card, there are fins present on the processor. The work of the fins is to cool down the processor. So for the same reason, we are going to use copper material for the fins.
For the other parts, we are going to use Aluminium material. Like in the base of the graphic card, Capacitor and blocks. Because we don't want for heat to go to the base of the graphic card and Aluminium is a low conductor as compared to Copper.
13. The heat will generate from the processor. So we are going to make the processor be our heat source. In order to that, we are going to go to the cell zone and where we are to click on the processor and we are going to edit it. Now we are going to enable the source box. Then we are going to take the power to be 128 Watt and the area of the processor is 64 `mm^3`. So the energy we are going to take is 2e9 (W/`m^3`).
In the cell zone, we defined which type of material we are using in which body part of the Graphic card.
14. Now we are going to initialize the simulation and we are going to create the temperature contour. After that, we are going to start the simulation.
Result
Residual
From the residual graph, we can see that after some time it is repeating the pattern so we can say that we get the result gets converge.
Temperature
We are getting a maximum temperature of 1246 k in the graphic card.
We can see from the above pictures that the maximum temperature we are getting at the base below the processor and wall at the end of the fins base. The temperature is high at the base of the processor because the heat is going directly from the processor to the base of the graphic card and it has less heat transfer to the atmosphere.
The wall at the end of the fin base has a high temperature because when the air passes from the fins plates and above the fins base, the temperature of the air increases. So when it hits the wall, the wall absorbs that heat, and temperature increases.
The processor is going to be hotter because it is our energy source.
Heat Transfer Coefficient
We are getting a maximum heat transfer coefficient of 946.9 w/`m^2k`.
We are getting maximum heat transfer at the fins plate and at the capacitor. The plates have a high heat transfer coefficient because the maximum heat generated in the processor goes to the fins so the heat will go to the maximum from the fins to the air by the fins.
Vector Plot
In the above picture, we can see the circulation of the air in the graphic card. The velocity of the air is very less in between the fins plates and after the fins.
Case-2
Coarse Mesh
In the coarse mesh these element sizes I have taken:-
Enclosure = 20 mm
Graphic Card base = 4 mm
Processor = 0.8 mm
Fin Base = 2 mm
Fin Plates = 0.3 mm
Capacitor = 0.8 mm
Blocks = 2.4 mm
Result
Residual
From the residual graph, we can see that after some time it is repeating the pattern so we can say that we get the result gets converge.
Temperature
We are getting a maximum temperature of 1248 k in the graphic card.
Heat Transfer Coefficient
We are getting a maximum heat transfer coefficient of 743.1 w/`m^2k`.
Vector Plot
In the above picture, we can see the circulation of the air in the graphic card. The velocity of the air is very less in between the fins plates and after the fins.
Grid Independence
| Case | Mesh | Number of elements | Maximum Temperature (K) | Maximum Heat Transfer Coefficient (W/`m^2`-K) |
| Case-1 | Fine Mesh | 504954 | 1246 | 946.9 |
| Case-2 | Coarse Mesh | 317828 | 1248 | 743 |
So From the table, we can see that the maximum temperature of the graphic cards we are getting is almost the same but the heat transfer coefficient is different. We got the heat transfer coefficient high for the fine mesh. So we can say from the result that when we make the mesh to be finer the temperature difference between the surface of the graphic card and the fluid decreases. So we get the high heat transfer coefficient. The heat transfer coefficient depends upon the first element location. So we can't say about the grid Independence on the basis of the Heat Transfer Coefficient.
We get the temperature value almost the same. So from the temperature value, we can say that it passes the grid independence test. So the simulation we have done is right. So we are going to use the same method for the next cases.
Case-3
In this case, we are going to take the inlet velocity to be 1 m/s and we are going to simulate for the fine mesh as same as the case-1.
Result
Residual
From the residual graph, we can see that after some time it is repeating the pattern so we can say that we get the result gets converge.
Temperature
We are getting a maximum temperature of 2081 k in the graphic card.
Heat Transfer Coefficient
We are getting a maximum heat transfer coefficient of 946.9 w/`m^2k`.
Vector Plot
In the above picture, we can see the circulation of the air in the graphic card. The velocity of the air is very less in between the fins plates and after the fins.
Case-4
In this case, we are going to take the inlet velocity to be 5 m/s and we are going to simulate for the fine mesh as same as the case-1.
Result
Residual
From the residual graph, we can see that after some time it is repeating the pattern so we can say that we get the result gets converge.
Temperature
We are getting a maximum temperature of 907.3 k in the graphic card.
Heat Transfer Coefficient
We are getting a maximum heat transfer coefficient of 946.9 w/`m^2k`.
Vector Plot
In the above picture, we can see the circulation of the air in the graphic card. The velocity of the air is very less in between the fins plates and after the fins.
Comparision
| Velocity (m/s) | Maximum Temperature (K) | Maximum Heat Transfer Coefficient (W/`m^2`-K) |
| 1 | 2081 | 946.9 |
| 2.5 | 1246 | 946.9 |
| 5 | 907.3 | 946.9 |
From the table, we can see that with the increase of the velocity of the air the temperature value decreases but we got the heat transfer coefficient value same. Because the velocity value doesn't change that much.
Conclusion
When we increase the velocity temperature of the graphic card decreases. Because the air is taking heat from the graphic card and it is going out fast.
In these cases, we are seeing that with the increase in the velocity there is no change in the heat transfer coefficient but I have done one simulation with the 50 m/s inlet velocity than the heat transfer coefficient was 1116 W/`m^2`-K. So we can say with the increase in the velocity, the heat transfer coefficient also increases.
The maximum temperature we are getting at the base below the processor and wall at the end of the fins base. The temperature is high at the base of the processor because the heat is going directly from the processor to the base of the graphic card and it has less heat transfer to the atmosphere.
The wall at the end of the fin base has a high temperature because when the air passes from the fins plates and above the fins base, the temperature of the air increases. So when it hits the wall, the wall absorbs that heat, and temperature increases.
The processor is going to be hotter because it is our energy source.
We are getting maximum heat transfer at the fins plate and at the capacitor. The plates have a high heat transfer coefficient because the maximum heat generated in the processor goes to the fins so the heat will go to the maximum from the fins to the air by the fins.