[woodmark75]

Hi,
I'm in a really time sensitive situation & need to have a good quality mesh generated to solve , lift & drag on a curved wing.

I'll post the geometry files, they contain; the wing, fluid domain i.e. representation of a 'low-speed' wind tunnel, & fixings.
There’s one set, scaled 1:1 with the actual wind tunnel, & the second set are scaled up 35.48:1.

I need to be able to solve in ANSYS/Fluent (14 OR 14.5). I've tried to resolve 'inflation' layers & 'y+' etc but to no avail. ICEM CFD seems to be a very steep learning curve, so that's not an option.

The wing has to be rotated from circa -10 to 20 deg AoA, & the mesh must be able to resolve boundary layer, wake etc.

Any information on turbulence model & solver settings i.e. anything that helps to get this CFD finished, will be properly compensated for.
I understand that this isn’t a particularly difficult task for most posters on this site, but I've tried for weeks & still don’t have the necessary skill-set, or the time to improve.

I’m willing to compensate fairly for any work done, price open to negotiation.
Geometry files posted below….

• Link: https://www.dropbox.com/sh/75lionq3u...p91nOTi5a?dl=0

[strongrj13]

Hello,

I downloaded your geometry files. Can you shoot me an email so we can discuss further?

robert.strong@strongsimulationconsulting.com

Thanks,

Robert Strong

[strongrj13]

If you want to give me an email we can discuss your needs.

robert.strong@strongsimulationconsulting.com

[woodmark75]

Hi Robert, Just sent you an email. Thanks for getting back to me so quickly...

[SourabhJain]

Kite.jpgThis is the initial mesh, no refinement, no preparation for boundary layer, do you need hex? If you can give me some comments, I can work on it.

[woodmark75]

Hi Sourabhjain,

From the link to the files, you'll have seen that there's essentially 2 models/assemblies. The smaller of the 2 is a 1:1 scale with a low-speed wind tunnel test I'm performing. The larger of the 2 is scaled up at 35.48:1.

What I'm trying to do, is 'match' the small CFD model & the wind tunnel results. In order to do this, the volume mesh, between the wing surface & the farfield boundary, has to be of high enough quality to resolve the boundary layer, turbulence, wake etc. For example, the wingtip vortices.

I think it needs some kind of 'structured' mesh, refined to the appropriate 'y+' etc to capture these effects. How this is done, I really don't know, I've been trying for weeks/months!

[SourabhJain]

Hi,

ok, I will make structured mesh on it, main problem will be, it extends upto boundary (and waste cells, also bad convergence) so we need correct topology. I need to spend some time to work on it. Can you tell me freestream velocity/density/dy. viscosity/Re (for defining 1st cell width and clustering ) etc.
For turbulence, just use SA model.

Regards

Sourabh

[woodmark75]

For Model 1.0 i.e. the smaller of the two:

• Velocity = 10 m/s
• Density = 1.225 kg/m^3
• Dynamic Viscosity = 1.7894 x 10^-5 kg/(m.s)
• Reynold's number = 5.6 x 10^4

For Model 2.0 i.e. the larger of the two:
To maintain geometric, dynamic, & kinematic similarity between both models, both have to have equivalent Reynold's numbers. Therefore Re_1.0 = Re_2.0:


As such, one of these three parameters can be scaled, while the other two remain the same.

• Scaling factor = 35.48
• If Velocity = (10 x 35.48)=354.8 m/s
• Density=1.225 kg/m^3
• Dynamic viscosity=1.7894 x 10^-5 kg/(m.s)

Or velocity can remain at 10 m/s, & density can be scaled to 43.463 kg/m^3. The final option would be to keep velocity at 10 m/s, density at 1.225 kg/m^3, and to scale dynamic viscosity to 6.3488 x 10^-4 kg/(m.s)

• With velocity scaled, Re = 7.05 x 10^7
• With density scaled, Re = 7.05 x 10^7
• With viscosity scaled, Re = 5.60 x 10^4

Hope this helps, & thanks...

[Totalsim]

Hi WoodMark,
Have you managed to find someone to sort this for you?

If not I'm sure TotalSim could help.

Website link is www.totalsimulation.co.uk which has our contact details on.

[woodmark75]

Totalsim,

Thanks for your responce. I did manage to find someone who says that they can produce a mesh/simulation that will give results which include an accurate resolution of; Lift forces, Drag forces, Boundary Layer, Wake etc...

Most importantly I must produce accurate Lift & Dag forces over multiple AoA (-5 deg to 15 deg). Also be able to plot accurate Pressure & Velocity contours, and finally be able to show the Wingtip Vortices with Streamlines.

I've tried to do this via an 'unstructured' Tet mesh, including an inflation layer over the wing. But no matter the turbulence model I use, my results are very poor.

I'm being told that I really need to be using; LES or DES, at the very least RSM. But I have a paper that shows accurate resolution of all the aspects I'm interested in, using a single equation Spalart-Allmaras model. The major difference being the mesh.

The mesh used in the paper using the Spalart-Allmaras model, uses a structured mesh. This is my reasoning behind thinking that I need a structured mesh.

If you think that you could provide the mesh & simulation, which provides accurate resolution of; Lift/Drag forces, and wingtip vortices/streamlines. Please get back to me.

My gometry is attached in the link below. The smaller model represents a model inside a wind tunnel. Freestream velocity 10 m/s, AoA -5 deg to 20 deg.

The larger model is the same but scaled up at 35.48:1. Therefore evreything is the same appart from velocity which now is 345.8 m/s.

• Link to geometry: https://www.dropbox.com/sh/75lionq3u...p91nOTi5a?dl=0
• Link to paper: https://www.dropbox.com/sh/r6htec9w8...OlVHL7Iia?dl=0

[SourabhJain]

can you share the paper?

[woodmark75]

Sourabhjain,

There's a link to the 'CFD analysis of a Kite' in post #10. But I'll put another link below...

• CFD analysis of a Kite: https://www.dropbox.com/sh/r6htec9w8...OlVHL7Iia?dl=0

[Totalsim]

Hi WoodMark,

Like you suggest, I think your mesh is likely to be the problem as tet meshes are far from ideal.

If the wing stays attached you shouldn't need higher order models for this. An SST model would be my recomendation, unles you see significant separation and the foil stalls.

[SourabhJain]

This is octree, do you have structured mesh paper?

[SourabhJain]

Hi,
I can do the (structured) meshing work. I will make C-topology along the flow direction, while around kite (perpendicular to flow) I will either make pure O-topology or since its almost 180 degree, I will try to put an inner cylinder and then an outer around it (with in between half ring to complete kite circle).
If you are interested, please let me know about terms and budget of the project.
Sourabh

[SourabhJain]

Hi,
I have not heard from you, are you interested?
Sourabh