Week 6 - CHT Analysis on a Graphics card
AIM: Perform steady state CHT analysis of heat flow over Graphic card.
OBJECTIVES:
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.
The given problem is completed using Ansys tool, the necessary steps taken are as follows:
Import the geometry.
Perform necessary geometry clean ups, like identify shared regions and fix the intersecting geometries.
Create shared geometry to have a conformal mesh for different components in later stages.
After this geometry is ready for meshing, the meshing is done by using different sizes for the different components.
Each component is meshed with different sizes.
The mesh is conformal at the nearby regions.
SETUP:
The analysis setup for the baseline model and the other subsequent models is same. The setup involves assigning the inlet velocity and initial temperature of the flowing air. The processor is generating heat of 10 W.
The heat released by processor is computed manually which came out to be 156250000 W/m3.
The components are assigned with there respective zones and the appropriate materials, as shown in the figures below:
The desired outputs are the heat transfer coefficient and the temperature of the processor. With this the setup is complete and the model is ready for the analysis.
Mesh Independence Test:
The mesh sizes were varied for some components in order to have optimum number of elements. The optimum mesh will be the one having converged results and less run time with good accuracy.
The element size for the following components was changed to generate three different mesh models for the same operating conditions just to get a best possible model.
The mesh sizes used are mentioned in table below:
|
|
Model 1 |
Model 2 |
Model 3 |
|
PROCESSOR |
2 |
1 |
0.5 |
|
FINS |
2 |
1 |
0.5 |
|
BASE |
7 |
6 |
4 |
|
ENCLOSURE |
7 |
6 |
4 |
|
TOTAL ELEMENTS |
119300 |
150365 |
233609 |
|
HEAT TRANSFER COEFFICIENT (W/m2K) |
246.1 |
252.8 |
258.9 |
|
TEMPERATURE (K) |
542.3 |
535.6 |
529.7 |
The analysis of the above-mentioned models gives following results:
MODEL 1:
MODEL 2:
MODEL 3:
The results are for temperature are decreasing and that of heat transfer coefficient are increasing as the mesh size is reduced.
The results are seen to be converging fastest for the MODEL 2.
The results variation may seem similar in terms of magnitude but moving from model 1 to model around 33000 elements are more but moving from model 2 to model 3 there is increment of 83000 elements in the model. This much increment will result in high run time, which is not justified for the changes obtained.
The results for the MODEL 2 are seen to be stable for most of the iterations, therefore it is obvious to use the MODEL 2 to perform the velocity variation analysis.
The MODEL 2 will be used for the analysis with varying velocity. The velocity is varied from 1 m/sec to 5 m/sec in step of 2m/sec.
RESULTS
Velocity = 3 m/sec:
Velocity = 5 m/sec:
|
|
1 M/SEC |
3 M/SEC |
5 M/SEC |
|
HEAT TRANSFER COEFFICIENT (W/m2K) |
258.9 |
472.55 |
628.88 |
|
TEMPERATURE (K) |
529.7 |
420.6 |
387.7 |
CONCLUSIONS
From the figures and table above the following conclusions can be drawn:
With increase in the velocity the heat transferred from the graphic card is increased. The maximum is seen for 5 m/sec and least for 1 m/sec.
The average temperature of the components is reduced as flow velocity is increased. This implies proper mixing and transport of the heat.
The fins component is seen to be a global hotspot in the given graphic card assembly, since most of the heat is taken by fins only. So, it’s necessary that fin material should have high thermal conductivity values as well as high melting point.