Conjugate Heat Transfer Analysis on a graphics card.

Aim: The aim of the challenge is to perform a conjugate heat transfer on graphic card and learn various aspects of CHT analysis.

Objective: Measure the temperatiure field across the processor which is a heat source in our case. Air is the cooling medium and their is a heat transfer coeffcient between processor and air, which is analyzed at varying velocity of incoming air to study the effect of its velocity on heat transfer coeffcient and temperature across the surface of processor.

Geometry: we have been provided with a geometry and we need to prepare it for simulation using spaceclaim.The geometry provided to us is enclosed in a enclosure. the geometry consists of Base, Fins, and processor.

Meshing:

The coarse mesh is first generated over the geometry which is then reined using mesh sizing tool for individual parts of graphic card.

The components of the graphic card are meshed with element size of 0.5 mm.

Number of element:

Setup:

For the solution purpose we use staedy state pressure based solver for analysis.

enery equation is turned on and we use K-epsilon realizable turbulance model.

Material: for fluid-standard material AIR is applied

Fins- Aluminium 

Base- PCB material FR4 (user defined)

Density=1850 kg/m3

specific heat (cp)=950 j/kgk

thermal conductivity=0.29 w/mk

processor-gold

The processor is the heat source with power consumption of 20 watt

The power density is hence 78125000w/m3

Boundary conditions:

Inlet-velocity is 2m/s, 4m/s, 5m/s

Inlet temperature=300k

Outlet=pressure outlet.

hybrid initialization is done with 300 number of iterations.

Case 1) Inlet velocity 2m/s

Residual Plot:

Local range temperature contour:

The local range option uses only the variable values on the current object at the current time step to set the max and min range values. it is useful to obtain full colour range on the object.

The above temperature plot shows that the hot spot lies on the upper surface of the processor, we can see maximum temperature region marked by red colour at the centre of the fin region which is almost 399 k. The region that comes in contact of the inlet air first is the front part of base and the fin hence temperature in that area is very low as compared to over all tempearature of graphic card.The inlet air temperature increases as air moves ahead due to convective heat transfer.

Global range temperature contour:

The global range option uses variable values from the results in all domains regardless of the domain selected for geometry and all time steps to determine the min and max range.

Velocity Contour plot:

The maximum velocity attained by inlet air is 2.4 m/s the velocity is maximum in the upper region where their are no obstacles. the velocity decreases as comes in contact with componenets of graphic card.

Wall Heat Transfer coeffcient:

The maximum Heat Transfer coeffcient attained is 958.9 w/m2k

Temperature streamline:

The streamlines entering the graphic card is at a very low temperature but as it passes over thr fin region and leaves from behind the temperature of streamlines increses from 300 k to almost 371 k.

Temperature distribution on processor:

The temperature distribution shows that the front part of processor receives cooler air than the rare part which leads to higher temperature region in rare end of the processor.

Case 2) Inlet Velocity=4m/s

Residual plot:

Local range temperature contour:

we can see maximum temperature region marked by red colour at the centre of the fin region which is around 377.4 k. The maximum temperature is decreased in this case owing to increase in velocity of air.

Global range temperature contour:

The global range option uses variable values from the results in all domains regardless of the domain selected for geometry and all time steps to determine the min and max range.

Velocity Contour plot:

The maximum velocity is 4.628m/s

Wall Heat Transfer coeffcient:

The maximum Heat Transfer coeffcient attained is 958.9 w/m2k

Temperature streamline:

Due to increase in velocity their is decrease in maximum temperature and also the minimum tempearture is also decreased.

Temperature distribution on processor:

 Case 3) Inlet velocity =5m/s

Residual plot:

Local range temperature contour:

we can see maximum temperature region marked by red colour at the centre of the fin region which is around 371.3k. The maximum temperature is decreased in this case owing to increase in velocity of air.

Global range temperature contour:

The global range option uses variable values from the results in all domains regardless of the domain selected for geometry and all time steps to determine the min and max range.

Velocity Contour plot:

The maximum velocity is 5.814 m/s

Wall Heat Transfer coeffcient:

Temperature streamline:

Temperature distribution on processor:

The overall heat distribution on processor is much evenly spread as compared to earlier cases the reason is that velocity of air is enough to maintain a good heat tramsfer coeffcient over the surface hence the hotspot in this case lies eaxctly at the centre of the processor.

Results:

1.The results confirms that rate of heat transfer increases with increase in fluid flow rate which proves the defination of forced convection.this also cnfirms that our simulation is converged.

2.The potential hotspot is at the centre fin in the middle line of the fins.

Reference:

https://forum.ansys.com/discussion/2863/global-vs-local-ranges

https://forum.ansys.com/discussion/1232/heat-transfer-coefficient

https://en.wikipedia.org/wiki/FR-4

https://www.engineering.com/story/choosing-the-right-turbulence-model-for-your-cfd-simulation

https://www.cfd-online.com/Forums/main/8398-pressure-based-density-based-solver.html#:~:text=and%20momentum%20equations.%22-,The%20pressure%2Dbased%20solver%20traditionally%20has%20been%20used%20for%20incompressible,for%20high%2Dspeed%20compressible%20flows.

https://www.tomshardware.com/reviews/geforce-radeon-power,2122-7.html#:~:text=Standard%20CPUs%20use%20between%2065,run%20simultaneously%20at%20full%20load.