CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Can't seem to get the correct cl and cd values (https://www.cfd-online.com/Forums/fluent/223506-cant-seem-get-correct-cl-cd-values.html)

NonStopEagle January 14, 2020 05:31

Can't seem to get the correct cl and cd values
 
Hey guys,
I am trying to simulate a NACA 0015 airfoil in ansys fluent. I just cant seem to get the correct cl and cd values. I think my mesh is fine, the y+ values are equal to or less than 1. I am using the k-omega SST model. Any ideas what might be going wrong?
the airfoil chord is 0.4 m and I have input this value in the reference area section of the reference values. I am running a pressure based simulation, with a velocity inlet( vel = 11 m/s) and a pressure outlet. The simulation is unsteady and I stop the simulation when I start getting the same cl values across 5-6 timesteps.
Please tell me why I am not getting the right cl and cd values.

NonStopEagle

arnie333 January 16, 2020 06:01

Two quick questions...

1. Are you comparing your analysis values to the experimental values at similar velocity (Re number) ? - recall that Cd and Cl are dependent on velocity
2. have you confirmed that your geometry was imported at the correct scale ?

NonStopEagle January 17, 2020 15:05

Quote:

Originally Posted by arnie333 (Post 754808)
Two quick questions...

1. Are you comparing your analysis values to the experimental values at similar velocity (Re number) ? - recall that Cd and Cl are dependent on velocity
2. have you confirmed that your geometry was imported at the correct scale ?

Thanks for the reply arnie333,
yes, the Reynolds number for my simulation is 300000, accordingly, the chord length is 0.4m and the velocity at inlet is about 11 m/s.
I imported the geometry such that the chord was 1m, I then scaled the geometry down in fluent by a scaling factor of 0.4

NonStopEagle

arnie333 January 18, 2020 06:58

1. You don't say how much different your analysis results Cd & Cl are from experiment ... are you a few % out or orders of magnitude ?
2. Are you analysing a 2D or 3D case ?
3. Are you running a steady state or transient analysis ?
4. If steady state, have you tried with "Pseudo Transient" ON and OFF
5. Have you tried reducing your "Timescale Factor' or/and your "relaxation Factors"
6. You say its "unsteady"... perhaps run a transient analysis, if not done so already. Reduce the cell size from the training edge going downstream for at least 5 chord lengths
7. I assume that you have checked cell Skewness and Orthogonality and these are within the recommended values !?
8. Are you using Double Precision ?
9. Have you changed your 'Turbulent Kinetic Energy' and 'Specific Dissipation Rate' to 2nd Order ?

NonStopEagle January 22, 2020 14:10

Quote:

Originally Posted by arnie333 (Post 754963)
1. You don't say how much different your analysis results Cd & Cl are from experiment ... are you a few % out or orders of magnitude ?
2. Are you analysing a 2D or 3D case ?
3. Are you running a steady state or transient analysis ?
4. If steady state, have you tried with "Pseudo Transient" ON and OFF
5. Have you tried reducing your "Timescale Factor' or/and your "relaxation Factors"
6. You say its "unsteady"... perhaps run a transient analysis, if not done so already. Reduce the cell size from the training edge going downstream for at least 5 chord lengths
7. I assume that you have checked cell Skewness and Orthogonality and these are within the recommended values !?
8. Are you using Double Precision ?
9. Have you changed your 'Turbulent Kinetic Energy' and 'Specific Dissipation Rate' to 2nd Order ?

Hi arnie333,
sorry for the late reply.
1. I am having trouble getting the correct values at high AoA(s) where flow separation occurs. While the lift differs from the experimental values by about 34% the drag can be differ by an order of 2 sometimes.
2. I am running a 2D simulation.
3. I run a steady simulation initially and after the residuals have stabilised, I switch over to a transient simulation and wait till the cl and cd values become constant.
4. I don't really have any convergence issues so to say, so I don't think changing the relaxation factors will do me any good.
5. I have kept the pressure outlet sufficiently far so as to eliminate the effects of the boundary condition imposed there.
6. I am indeed running on double precision.
7. The order of discretisation for all the quantities(except pressure) is third order MUSCL.
Do you have any idea what i can do to get accurate values after the flow seperation happens. also, i have noted that at specific AoA(s) there is vortex shedding from the airfoil, due to which the cl and cd oscillate. Under such condition how do i get a single value of cl and cd which i can compare with the experimental data i have?

thanks in advance

NonStopEagle

arnie333 January 22, 2020 14:19

In your Cd calculations, are you considering the change in your frontal area (perpendicular to airflow) due to your change in AoA ?

In your Cl calculations, are you considering that your lift is perpendicular to the airfoil and not vertical ?

NonStopEagle January 22, 2020 14:24

Quote:

Originally Posted by arnie333 (Post 755351)
In your Cd calculations, are you considering the change in your frontal area (perpendicular to airflow) due to your change in AoA ?

In your Cl calculations, are you considering that your lift is perpendicular to the airfoil and not vertical ?

hi arnie333,
i am using the fluent force report to print out the cl and cd values. Since I rotate the entire mesh(domain) to a specific AoA, I think its the right thing to do?
Also the lift, as far as i know is perpendicular to the free stream flow and drag is parallel the the free stream. Perhaps you're confusing the normal and axial forces to lift and drag?
with regards to the change in frontal area, I believe ansys does that when we specify the direction along which the forces and coefficients are to be evaluated.

thanks

NonStopEagle

arnie333 January 22, 2020 14:42

Cd: To my knowledge, the Reference Area is not updated by rotating the fluid/mesh domain. You can go to Reports->Projected Areas and compute in the direction (frontal) you need.
Cl: For the lift (with AoA) I am not 100% sure, but I would think that you would need to enter the Cartesian components of your angle - just enter these in the Cl Monitor report X & Y (2D) for fluent to calculate automatically....try it and let us know.


All times are GMT -4. The time now is 09:43.