[lami_nar]

As a part of our project we are trying to simulate the results from Theory of wing sections by Ira Abott for NACA 2412. Our Cl values are matching the graph values for 5.7e6 reynolds number with standard surface roughness, but we are not able to do the same for smooth wall. Even though the default settings of ansys fluent are that for a smooth wall, we are only able to match the standard roughness values. We have tried improving the mesh several times to no avail.

Any insights into why this is happening will be appreciated.

Model- K omega sst
Reynolds number- 5.7e6

[vinerm]

Are you getting lower or higher than expected value?

[lami_nar]

Originally Posted by vinerm

Are you getting lower or higher than expected value?

I am getting lower values. For eg. Max Cl from simulation is 1.21 and from the experimental value from theory of wing sections is 1.7

[vinerm]

First of all, experimental values are not possible for smooth walls; there is always some roughness. However, if the roughness is smaller than viscous layer, then the effect of roughness is negligible. However, the difference you are observing is significant. Do check if you are using correct reference values in Fluent for determining . If those are correct, then you should check your material properties and boundary conditions. If everything is good, then it would be very difficult to debug since the difference is too large to be numerical error.

[lami_nar]

Thank you for your inputs

[lami_nar]

Originally Posted by vinerm

First of all, experimental values are not possible for smooth walls; there is always some roughness. However, if the roughness is smaller than viscous layer, then the effect of roughness is negligible. However, the difference you are observing is significant. Do check if you are using correct reference values in Fluent for determining . If those are correct, then you should check your material properties and boundary conditions. If everything is good, then it would be very difficult to debug since the difference is too large to be numerical error.

Just to give you a reference, our values are matching the line marked with triangles ie reynolds number 5.7e6 having standard roughness.
But the settings in Ansys fluent correspond to a smooth wall (as roughness height is zero). How is it possible that the values of simulation of smooth wall match with value of standard roughness condition? Are these simulations acceptable?
Kindly find the attachment image

[duri]

Originally Posted by lami_nar

Just to give you a reference, our values are matching the line marked with triangles ie reynolds number 5.7e6 having standard roughness.
But the settings in Ansys fluent correspond to a smooth wall (as roughness height is zero). How is it possible that the values of simulation of smooth wall match with value of standard roughness condition? Are these simulations acceptable?
Kindly find the attachment image

Looking into the data suggest standard roughness has LE trip and probably smooth wall doesn't have trip. If this is true, try using transition model with laminar flow till some percent of chord.

[vinerm]

The effect of roughness on the flow is significant only if the roughness is larger than the viscous sublayer. Any roughness smaller than that implies smooth wall. So, you have to look at the thickness of boundary layer in your case and the roughness value. If the case is more or less smooth, then the results should match. If roughness is much more than viscous sublayer, then result match could also be due to the reason that the boundary layer is very thin and, hence, viscosity does not play much role and whole of the stress is due to turbulence.