Conjugate Heat Transfer Analysis on a graphics card

OBJECTIVE :

- To find the maximum temperature attained by processor after simulation.

- To find out the heat transfer coefficient at appropriate areas of the model.

- To identify potential hotspots on the model.

SOFTWARE USED:

ANSYS, SpaceClaim - To prepare FV model.

ANSYS, Meshing - To mesh the given model.

ANSYS, FLUNET - To setup the boundary conditions and to solve the given model.

ANSYS, CFD Post - To perform Post Processing.

THEORY :

In this project , A steady state conjugate heat transfer analysis on a Graphic card has become an everyday used object and a very important part of any computer system , laptops etc. This product is mass produced daily in millions and has made computers exceptionally efficient . As they consume electricity to operate and running a heavy software or graphically task on computer takes a payable use on its graphic card , because of which it generates a large amount of heat .

There is a limit to any system that can hold heat to keep its parts under safe operating condition . We cannot go past these limits as it will damage the system ,Cause failure of the product and system etc. But what we can do is measure to cool this system , cool the graphic card so the system keeps running at optimum speed & condition .

There are various methods of cooling electrical components , through fan , thermal pastes and gels , double fanning , using materials that help dissipating the heat in the fan’s direction , arrangements of components etc.

Cooling of graphic card , i.e by conduction in the PCB base & fins convection with the surrounding air by providing cooling fins over heat source (graphic card processor) which in turn try to remove heat through conduction and additional air is allowed to flow through assembled components which act as a phenomenon called forced convection.

In this simulation conjugate heat transfer analysis is done on graphic card model by applying heat on the processor of the graphic card while it contained in an enclosure . The air is passed at a certain velocity through the enclosure and we observe the heat dissipation through the fins by conduction from the processor , then observe how the flowing air cools the fins & observe the heat transfer taking place through it.

PROCEDURE :

STEP 1 : GEOMETRY USED

Geometry Dimensions :

Importing the 3D model:

NOTE : Check for share topology in order to avoid interpolation error and to get better simulation results.

Part Description :

STEP 2 : MESHING

NOTE : Naming should be given at required areas , since we have to assign boundary conditions in Ansys Setup.

1.BASELINE MESHING (Coarse Mesh) :

Mesh Size : Default size

2.REFINED MESH at Different Locations :

Base – 3 mm

Processor – 0.5 mm

Components – 0.6 mm

Fins – 0.5 mm

Outer enclosure – 3 mm

Statistics :

Nodes = 74210

Elements = 420659

STEP 3 : SET UP & SOLUTION

- Solver type : Pressure based

- Steady state solver

- Check the Energy equation

- Turblent Model : K-Omega SST Model

Boundary Condition :

Inlet Velocity = 5m/s

Initial temperature = 300k

Cell zone conditions :

• set Processor / core as source term as heat is generated by it in the graphic card.

Consider ,

Power supplied to Core (P) = 6 w

We have,

Volume of processor (V) = 64 mm³

- Therefore , Heat generated = P/V = 93750000 w/m³

Materials assigned are as follows :

Base : Steel

Components : Titanium

Fins : aluminum

Processor : Gold

Enclosure is filled with Air .

STEP 4 : RESULTS

Project Schematic :

CASE 1 : Comparing Case with constant Velocity=3 m/s (inlet) with varying the Refinement of Mesh.

RESIDUALS :

From above plots , we can say that Simulation has got converged .

CASE 2 : Comparing Case with Constant mesh with varing the different velocity from 1 m/s to 5 m/s namely 1m/s and 5 m/s , 3 m/s results were mentioned above.

Observations :

From this we can conclude that , As Inlet velocity get increased ---> Heat generated get decreased.