Conjugate heat transfer analysis of exhaust port by simulation through ANSYS

Conjugate heat transfer refers to the coupling of heat transfer in solids and heat transfer in fluids. Conjugate heat transfer analysis is used to study the heat transfer from a solid to fluid or vice-versa. It is also used to study the heat transfer in interface of solid and fluid. Conjugate heat transfer analysis is quite important as majority of heat transfer applications involve more than one phase. Conjugate heat transfer analysis is employed in designing and simulating of heat transfer appliances like shell and tube heat exchanger, air conditioners, radiators etc., to increase effective heat transfer in them and improve their energy saving. Only through conjugate heat transfer analysis parameters like Nusselt number can be identified.

For the simulation of exhaust port a pressure based steadty state solver setup is considered with energy equations included and choosing the standard k-epsilon model to simulate turbulence. For wall material aluminium is considered and the fluid material flowing inside is considered to be air in this simulation. The inlet velocity is taken as 5m/s with inlet temperature at 700k. A convective boundary condition of 20 W/(m2-k) is considered for outer wall and the outlet is taken as pressure outlet. The simulation is carried out first for a baseline mesh and then carried out for a refined mesh. For the latter case the mesh is refined until the point allowed by Ansys Academic licence.

Geometry view 1:

Geometry view 2:

Unrefined mesh:

The unrefined mesh has 33604 nodes and 137707 elements.

Unrefined mesh metrics:

Convergence plot:

The simulation is run until the residual started varying with constancy.

Wall Heat transfer coefficient at XY plane at z=0:

Velocity streamlines of fluid(air) inside:

Outer wall temperature:

Velocity of fluid(air) using XY plane at z=0:

Temperature of fluid(air) and solid(aluminium) at XY plane at z=0:

Refined mesh with inflation layers on fluid volume:

The mesh is refined and inflation layers are added to have 137664 nodes and 495014 elements which is close to 0.5 million elements limit in Ansys academic licence.

Refined mesh metrics:

Convergence plot:

Wall heat transfer coefficient at XY plane at z=0:

Close up of wall heat transfer coefficient:

Velocity streamlines of fluid(air) inside:

Outer wall temperature:

Velocity of fluid(air) at XY plane at z=0:

Temperature of fluid(air) and solid(aluminium) at XY plane at z=0:

From the simulation results we can see at that the part of the exhaust port where the air flow culminates the wall temperature is higher compared to the wall temperature of the part of exhaust port where the air flows enters i.e., inlets, this is because the part of the exhaust port where the air flow culminates has higher velocity as shown by velocity streamline pictures due to mass flow conservation law. The higher velocity results in higher reynold's number in that part thereby higher heat tranfer coefficient which can be seen from the zoomed picture of refined wall heat transfer coefficient, that is why the wall temperature is higher at the part of the exhaust port where the flow culminates.

As the mesh is getting more and more refined the elements get smaller and the results are captured more precisely as seen in the case of wall heat transfer coefficient where we can see that including inflation layers produced body fitted mesh that captured the wall heat transfer closely and precisely near to the wall. Also as the mesh is refined the number of elements increase so more data is caputured compared to unrefined mesh which will have adverse effect on the results obtained from the simulation. e.g., here in this case max heat transfer coefficient is 105.82 W/(m2-k) in case of baseline mesh and it is 218.67 W/(m2-k) in case of the refined mesh.

Heat transfer coefficient (HTC) values obtained from the simulation can be verified by comparing it with the values obtained from doing the project experimentally or by comparing the values obtained with the data obatined from trusted journals. Another method of verification is by checking whether the CFD setup is accurate by using known standard reference models and comparing the values obtained with the experimental data available for that reference model, doing so would verify that the CFD setup considered is valid therefore implying that the simulation carried out using the same CFD setup will give valid results. 

The accuracy of results obtained from the simulation depends upon the refinement of mesh. The mesh should be refined until the solution becomes mesh independent i.e., performing grid independence test. The grid independence is performed here but since Ansys Academic licence allows only upto 0.5million elements in the mesh the test is stopped at nearly 0.5 million cells. Further to get accurate HTC values inflation layers were added with total thickness of 5mm so as to get body fitted mesh which accurately captures the HTC values near the wall.