Simulation of Conjugate Heat Transfer Analysis on a graphics card

OBJECTIVE 

      To Perform a steady-state conjugate heat transfer analysis on a graphics card & Identify the hotspots in the model.

TASK 

      1. Find out the maximum temperature attained by the processor.

      2. Prove that the simulation has achieved convergence.

      3. Find out the heat transfer coefficient in appropriate areas of the model.

Conjugate heat transfer

      Conjugate heat transfer (CHT) is used to describe processes which involve variations of temperature within solids and fluids, due to thermal interaction between the solids and fluids. The exchange of thermal energy between the two physical bodies is called the study of Heat transfer, the rate of transferred heat is directly proportional to the temperature difference between the bodies. Conjugate heat transfer corresponds with the combination of heat transfer in solids and heat transfer in fluids. In solids, conduction often dominates whereas, in fluids, convection usually dominates. Efficiently combining heat transfer in fluids and solids is the key to designing effective coolers, heaters, or heat exchangers. Forced convection is the most common way to achieve a high heat transfer rate. In some applications, the performances are further improved by combining convection with phase change.

Graphics Card

      It is computer hardware that generates computer graphics and allows them to be shown on a display, usually using a graphics card in combination with a device driver to create the images on the screen.

Three parts in a Graphics card

i) Base - thermoplastic materials where all the electronics components are assembled.

ii) Processor - Silicon is used for this manufacturing processing.

iii) Heat sink - The function of this is to improve the heat transfer rate from the processor & the materials like aluminum & copper which have high heat conductivity are used, for our case we use aluminum.

PROCEDURE

       Graphics card model is imported into the Space claim. The model has different parts separate them into fluid & solid components depending on the material. The base, processor & fins are assigned to the solid component & region around is set to the fluid component. The share topology is used to obtain conforming mesh.

        The baseline mesh is created & four named selection to assign the boundary conditions.

i) Front face - Inlet ii) Backface - Outlet iii) Four walls - Wall convection iv) Processor - Processor.

        In the fluent,  Solver - Pressure based, Time - Steady-state, Velocity - Absolute &      k-epsilon model is chosen to capture the turbulence. 

        Two solid materials are created for the silicon & thermoplastic.

        Boundary conditions: i) Inlet v = 10 m/s & T = 300k. ii) Outlet P = 0 gauge. iii)Processor T = 423 k. iv) wall convection h = 20 W/m^2k. The velocity is assumed to be 10 m/s & processor is consider as the only heat source for the graphics card.

        The solver is initialized & create contour & animation plots for the temperature. The solver ran for 500 iterations.

The simulation gets converged at 94 iterations.

Case 2

The outer region is set to hex dominant mesh to capture the physics properly & the mesh size is reduced to 2mm. The processor, base & fins are regions where more heat transfer occurs so using the face sizing option the mesh size of these parts are set to 0.5 mm. The other parameters remain unchanged. The solver ran for 500 iterations.

The simulation gets converged around 400 iterations. 

RESULTS

Temperature plot

Velocity plot

Pressure plot

Heat transfer coefficient

INFERENCE

• The temperature plot for both the cases is almost equal & in the second case, the results are more accurate at inlet temperature.
• The max velocity in the first case is 12.19  m/s & in the second case, the max velocity is 12.2 m/s & variation of the velocity is around 0.08%  whereas the mesh cell count increases by around 400 % but case 2 takes more time to attain convergence.
• From the table is clearly depicted that the value of the heat transfer coefficient is dependent on the cell count & the variation of the HTC increases by 54.1%. HTC is a boundary layer concept & in case 2 we create refine mesh around the fluid layer to capture physics more accurately. The accuracy factor depends upon the element size, the total number of cells, mesh quality & inflation layer at the interface. So we can conclude that finer mesh produces accurate results.   
• The potential hotspots of the model are the heat sink & processor where the temperature is very high or regions near the heat source.
• Simulation has no change on all the convergence criteria, after certain iterations or timesteps, then we can consider that the solution has reached a steady-state (i.e)converged. The role of this residual plot is to monitor the change in each and every convergence criterion for every time step and plot those values for every iteration or timestep. There are two cases while using the residuals plot for convergence i) No Change after some iterations ii) Following the same pattern. In both cases, simulation gets converged by following the second pattern.