CHT analysis of graphic card
Conjugate heat transfer analysis on graphics card :-
The term conjugate heat transfer is used to describe processes which invlove variations of temperature within solids and fluids, due to thermal interaction between the solids and fluids. Efficiently combining heat transfer in fluids and solids is the key to designing effective cooling and heating system, in this case designing effective cooling system for electronic equipments. In this simulation, CHT analysis is performed on the graphic card to find at which velocity maximum heat transfer takes place and to identify potential hot spots on the graphic card.
Design model of the graphic card :-
The base, processor and the fins of the graphic card is modeled in CAD softwares like catia or solidworks. Small details which are considered not to be affecting the solution like wires, cables and electronic components are not modeled. An enclosure is created since the simulation is done for the external flow around the graphic cards.
Meshing the graphic card :-
Number of nodes = 70805
Number of elements = 378645
The mesh is refined on the fins as they are responsible for dissipating the large amount of heat generated by the processor and to obtain clear htc contours. Body sizing method is used to refine the mesh on the fins with element size equal to 1.8 mm.
Case setup :-
Solver time = steady state
Solver type = pressure based
Viscous model = k-epsilon and energy equation is solved
Adding materials to the components of graphic card
Base = steel
fins = aluminium
processor = copper
power dissipated by the processor is assumed to be 95 W and the volume of the processor is 6.4*10^-8 m^3. Therefore, the heat dissipated by the processor is 14.8497e8 W/m^3.
The simulation is done on the refined mesh by varying the velocity in steps from 1 m/s to 10 m/s to find for which velocity maximum heat transfer coefficeint is obtained :-
Case 1: velocity = 1 m/s
Residual contour
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 2: velocity = 2 m/s
Residual contour
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 3: velocity = 3 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 4: velocity = 4 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 5: velocity = 5 m/s
Residual plot
Avg nusselt number
HTC contour
Temperature contour
Velocity contour
Case 6: velocity = 6 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 7: velocity = 7 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 8: velocity = 8 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 9: velocity = 9 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
Case 10: velocity = 10 m/s
Residual plot
Avg nusselt number plot
HTC contour
Temperature contour
Velocity contour
In a CFD analysis, the residual measures the local imbalance of a conserved variable in each control volume. Therefore, every cell in the model will have its own residual value for the equations being solved. The lower the residual value is, the accurate the numerical solution. From the above residual plots we can see that the residual value drops to 1e-6 for all the cases and thus we can say that the solution is converged.
The convergence of the numerical solution can also be checked by using quantities of interst. In this simulation average nusselt number is used to check for convergence of the solution. From the above plots, the average nusselt number increases initialy and a little fluctuation is observed between 1-30 iterations and there after the value of the average nusselt number is constant after 50 iterations. Thus, we can conclude that the solution is converged when there is no change in the values of average nusselt number.
From the above table it is clear that the maximum heat transfer coefficient is obtained when the velocity is equal to 10 m/s and also the processor temperature is minimum when the velocity is equal to 10 m/s. Therefore, higher the velocity higher is the heat dissipated from the graphic card which helps in decreasing the temperature of the processor.
In all the ten cases, the highest temperature is observed near the processor, fins base near processor and base of the graphic card near the processor. Therefore, the potential hotspot in the graphic card is the processor with high power dissipation. The temperature of the processor has to be reduced in order to increase the life span of the graphic card.