Performing a conjugate heat transfer analysis(CHT) on an Exhaust Manifold using ANSYS FLUENT

OBJECTIVE:

To perform a conjugate heat transfer analysis on an exhaust manifold and postprocess the velocity,temperature,wall heat transfer coefficient.

CONJUGATE HEAT TRANSFER:

It refers to combination of conduction in solid with the convective and radiative heat transfer in the surrounding fluids.some of the application includes

• Cooling of electronic devices
• Heat exchangers
• HVAC
• Engine cooling.

Advantages:

• Accurate and fast
• Suited for complex scenarios.
• All combination of region type.

METHODOLOGY

1.Geometry-space claim:

fig.Exhaust manifold

• Extract the fluid volume and solid volume and make it as a component named as volume exhaust.
• workbench-> share -share coincident topology to get the proper mesh distribution and smooth heat transfer.
• properties->share topology->share.

2.Mesh:

Case 1:

Baseline mesh:

fig.baseline mesh (150mm)

• Naming the selection such as inlet,outlet,outer wall convection.
• created a baseline mesh (default:150mm).

Case 2:

Refined mesh:

fig.refined mesh

• Naming the selection such as inlet,outlet,outer wall convection,inflation layer(fluid volume).
• created a 5 inflation layer with total thickness as 5mm around fluid volume to get the smooth heat flow distribution.

fig.inflation layer around a fluid volume.

• Body sizing-Element size of the solid volume(outer body) as 15mm to get the better results of wall heat transfer coefficient.
• Fluid volume element size as 150mm.

3.Setup:

• solver :Pressure based solver
• Velocity formulation : Absolute
• Time:Steady
• Energy equation
• viscous model: K-epsilon
• Material: air-fluid, Aluminium-solid.
• Boundary conditions: Inlet- velocity-5m/s ,Temperature -700K. outlet-pressure outlet (gauge pressure=0), outer wall convection- h=20 `W/(m^2K`
• Hybrid initialization is done.
• Run the calculation for 500 iterations.

4.Solution:

Case 1:

i)

fig.Residual plot

• Solution is converged at 166 iteration.

ii)

fig.Temperature contour

• from the figure,we observe that the temperature is maximum near the outlet region.

iii)

fig.velocity streamline

• As we see from the figure,streamline velocity is maximum at the region of the outlet.It is because the air velocity from the four inlets having higher magnitude in the one outlet region.
• In order to see this variation,plane (XY) is created.

fig.Velocity magnitude along plane region.

• Velocity is higher at curved surface of the outlet.

iv)

fig.Wall heat transfer coefficient (plane)

• Wall heat transfer coefficient (h) between solid and fluid is found out to be maximum at the curved surface region is 104.14 `W/(m^2K`
• It is caused by the high velocity in this region and it leads to high temperature at the wall.

CASE 2:

i)

fig.Residual plot

• Solution is converged at 497 iteration.

ii)

fig.Temperature contour

• It is producing similar results to the baseline mesh.

iii)

fig.velocity magnitude contour along plane(XY)

• It is clear to see how the velocity is gradually increasing when the fluid enters near the outlet region.
• Inflation layers around the fluid volume gives a smooth transition of heat transfer from fluid to solid.

iv)

fig.wall heat transfer coefficient contour

• we see that the wall heat transfer coefficient is reaching maximum value at the curved surface of the outlet region as CASE 1.

RESULTS:

CONCLUSION:

• Performed conjugate heat transfer analysis for the two cases(baseline and refined mesh)
• Case 2 gives a higher variable values than case 1.
• Number of iteration for convergence is higher in case 2.It is because number of nodes and elements are larger than case 1.
• contour plots in case 2 gives a better result because of inflation layer and mesh refinement.

How would you verify if the HTC predictions from the simulations are right? On what factors does the accuracy of the prediction depend on?

Cfd results can be verified with experimental or theoretical data,performing grid independency test- when we solve any cfd problem the solution should not be affected by the size of grid. Grid independence is done to ensure that the solution is independent on the grid size.we try to solve same problem with fine grid and see the variation in result .

Factors affecting the accuracy of results as follows:

• Number of mesh elements and element size.
• Type of grid and mesh cell faces.
• inflation layers.

Temperature contour Animation

Link:

https://drive.google.com/open?id=1RpSwrc2M4hhV-8-vFVjz8cHnI3jQE34X