Depends on what the porous region is.

Drag is only a relevant entity next to a wall; hence why FLUENT will not let you do it.

Can you define the region with walls and holes, then just apply the drag to the walls in between the holes?

Are you using the porous region to suck in the boundary layer? or some other reason?

How do you know that FLUENT is predicing low? is there experiment data available?

Regards

Paul

Hi Paul,

Yes, rightly said! I could apply the drag just to the walls and not to any other boundary condition in Fluent. I am trying to model suction on an aerofoil as a boundary layer control.

I am not validating my results against any experimental data at the moment as I am trying different cases allowing me to choose the best one modeling it. I could say that Fluent is predicting it low because I am not able to select the porous region in the geometry (covers quite a bit of the geometry) while selecting the coefficients. I do see some appreciable changes in the wake region visually but, the coefficients plotted are minus the porous region.

My take on this case would be writing a UDF to define the presssure on the specific points which are not being computed and evaluating the drag (pressure drag) just like the Fluent does that. Does this sound any good?

Cheers,

Vinit.

Hi Vinit,

Can you add some solid to the porous area and scale the surface parameters so that the holes are not included in the computed drag?

Are you including the energy needed to extract the boundary flow into your drag figure?

Paul

First ensure that CFD is predicting well for base line model (non porous case). This guarantees the CFD modelling,(presumed that grid independence study is done). Then move on porous case. If it appears that values are under predicted then try the following-

May be it is possible to get quantities(required for drag/lift coefficients) on porous surface. Use these to define drag/lift(give expression) in  custom field function (avoid lift/drag coefficient defined by FLUENT ). Defined custom defined quantity can also be set as a monitor to check and also it is available in report menu. This may work.

If you put wall on the top and bottom and then calculate drag of the domain minus drag of the walls may be you get a better approch.

Hello,

Paul, I assume modifying the surface parameters wouldn't suit my case as I would like to evaluate the drag force created by the porous region as well which is defined by me based on the material properties in the boundary condition. I wouldn't be including energy terms as I would only be concerned with the co-efficient at this stage.

Mahesh, my results are validated for the non porous model wherein the coefficients are in agreement with the data. My concern now is to include the porous region coefficients into my results which is not being monitored. Your suggestion of using custom defined function sounds feasible to me and would surely give it a go!

Khurram, adding wall on top and bottom wouldn't would't allow me to define drag on the porous region, being my big concern. Do you thing this method would help me to isolate the porous region modeled between the walls?

Highly appreciate your responses!

Vinit

Changing boundary condition /type of surface would completely change the physics under study and could misguide the investigator. Like all velocities normal to wall are zero where as a porous surface would have normal velocities at the hole locations.There are other adverse outcomes which need not be detailed

Hi Abdul,

Appreciate your interest. My take to the problem at the moment would be the approach suggested by Mahesh. Hope it gives my some results. I won't be able to choose a laminar flow for the reason my problem demands defining turbulence. But definitely as you said, a test case on a simpler geometry would be a good approach.

If anyone of you guys help me with any source as to how the coefficients (lift and drag) are evaluated in Fluent? Does the coefficient include pressure and friction parameters? Help!

There is an alternative way to determine loss coefficients by integrating the entropy generation in the flow field around the body under consideration. May be that is something you can use in your case.Details of this approach can be found in:

H. Herwig, B. Schmandt: How to Determine Losses in a Flow Field: A Paradigm Shift Towards the Second Law Analysis, Entropy, Vol.16, 2959- 2989 , 2014

Alternatively use the relation between the loss coefficient and Lift or  Drag coefficient, which is available in any aerodynamics books

Hello Heinz! Thank you very much for your input. I did read your paper and found it very interesting. I definitely feel it being appropriate and a practical approach to evaluate the entropy around the flow field instead of the localized forces on the body under consideration. But, my case is an incompressible case wherein I do not use the energy term. Using the alternate drag coefficient equation proposed by you won't suit my case as there is temperature term involved. We also discussed your approach and surely tempted to try some cases with that. I would definitely keep you updated with that and thanks again.

Hi Mahesh! My take to the problem at the moment is to find the localized forces (integral) over the non wall region and using the same in the coefficient equations. If this analytical method turns out to be reasonable and validated, I would write a UDF to define that particular region. Any source would be highly appreciated.

Cheers guys!

Hi everyone.

Both approaches are described in the attached file.

Conference Paper Numerical Airframe Aerodynamic Performance Prediction: An Ex...

Hi Vinit

I would like to know how you solved the above problem. Did you follow the analytical way of calculating local forces around the surface OR you could find out any method in Fluent itself.

My problem is similar to yours.Trying to get Cd,Cl when a clinder is coated with porous materials.

@mahesh Varpe

Using custom field function is a good idea. But the variables we use for CFF is the static pressure option in Fluent, which is a scalar. If we need drag and lift coeff, force vectors are to be taken for calculations. How can we do that.