CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   STAR-CCM+ (https://www.cfd-online.com/Forums/star-ccm/)
-   -   Airfoil meshing (https://www.cfd-online.com/Forums/star-ccm/69635-airfoil-meshing.html)

MadsR October 29, 2009 06:22

Airfoil meshing
 
Hi.

I am considering CCM+ as a tool for simulation of 2D airfoils and full wind turbine blades. I have had CD-Adapco presenting their package in detail. I was wondering if any of you guys had real experience with CCM+ for both 2D airfoils and full 3D blades, though.

I am especially concerned about the meshing in the boundary layer and if the mesher of CCM+ is good enough to make orthogonal cells in this area.

Thanks a lot.
/Mads

seang October 30, 2009 02:19

the polyhedral mesher works very well, good prismatic cell layer etc.

Quote:

Originally Posted by MadsR (Post 234475)
Hi.

I am considering CCM+ as a tool for simulation of 2D airfoils and full wind turbine blades. I have had CD-Adapco presenting their package in detail. I was wondering if any of you guys had real experience with CCM+ for both 2D airfoils and full 3D blades, though.

I am especially concerned about the meshing in the boundary layer and if the mesher of CCM+ is good enough to make orthogonal cells in this area.

Thanks a lot.
/Mads


MadsR October 30, 2009 03:36

Hi Kuan.

Thanks a lot for your answer :-)

Are you doing airfoil simulations? CD-Adapco showed me the hex-mesher with the boundary layer tool. Have you used that?

/Mads

aerospaceman November 20, 2009 21:43

Hey there,

I'm afraid that I cannot say that Star CCM+ is very good for meshing airfoils.

I have spent the last 4 weeks full time trying to get this to work, but with no success. The hardest part is to mesh in CCM+.

My Cd for a Naca0012 is 100% off (i.e. 2x what it has to be).

If you want to do 2D, I suggest you mesh with a structured mesher eg gridgen and then transfer it to Star.

If you have the option of using Fluent, it seems to be much more solid if you can use the Linux style interface.

Hope this helps.

ttl November 21, 2009 18:00

Airfoil Meshing
 
Hello,

I have a fair bit of experience meshing 2D airfoils in STAR-CCM+. I have found the code to be very good, meshing is pretty easy and very controllable. The prismatic layering is simple to adjust and the automated mesher is very robust. When I do have issues it is usually with the prism layering but all in all it is pretty good and near full proof for 2d airfoils.
Now that isn't to say STAR-CCM+ is perfect, just pretty good.

To comment on Aerospaceman's post, if your 2x off something is wrong with the mesh or the inputs.

I've also done some work with 3d propeller analysis and found CCM+ to be excellent and very accurate using the rigid body motion capability. The MRF capability is good but not as accurate, as one would expect.

For pure 2d work I usually use something other than CCM+ since it is ultimately a 3d code trying to do 2d work so it is not as fast and there are several good 2d codes available for little to no $$. However, CCM+ is hard to beat when you look at all the capabilities and the total cost.

Personally if I had to buy and use just one code, CCM+ would be it.

Hope that helps.

aerospaceman November 22, 2009 22:03

Hey there TTL,

Thank you very much for the reply.

The reason that I was posted what I said is that this was my experience. It is great to hear that someone has made it to work.

I have been trying for the past 9 months to get this to work. Would you be kind enough to supply me with the meshing settings that you used to get your very good answer?

I would ideally need all the meshing specs to see what exactly I'm doing wrong. I understand that the entire .sim file would be asking too much.

I am pretty sure that my problem is that there is a a huge jump in cell size from my BL block (prism layer model) and the rest, but I have been trying to control this with no success.

Any help would be absolutely fantastic, as I have tried everything that I possible can, and if you could help I would be very greatful.

Thank you very much for your time.

ttl November 22, 2009 23:08

Airfoil Meshing
 
Aerospaceman,

I'd be glad to help. If you post your email address we can talk off the forum. It might work better if you just send me the .sim file you are working with. Just be sure to delete the volume and surface mesh before sending it.

TTL

aerospaceman November 23, 2009 01:58

Hey TTL,

Yes, certainly. My email is greg2earth@yahoo.com

I can send you my .sim file ready to be meshed so you can have a look.

Just inform me of your email and I will send it to you asap.

Thank you very much. I really appreciate your help.

Thanks.

abdul099 November 24, 2009 19:37

I'm working on my bachelor thesis and i have to do some simulation on airfoils. I can just confirm, what ttl said. I used star-cd for simulation, but ccm+ for meshing. The mesh (hex) is mostly o.k., but sometimes the prism layers are looking a little bit strange, especially on the leading edge (where the solution is not very sensitive to the prism layers).

The comparison of the simulation and an experiment gives a discrepancy of 2 - 15%, depending on the angle of attack. Of course, for great or very small (negative) angles of attack with seperating flow there is a larger discrepancy.
But i made the experience too, the solution is very sensitive to the prism layer mesh, especially the number of prism layers and the thickness of near wall prism layer.

Maybe somebody has some hints how to improve the difference between simulation and experiment? Don't matter, if not. I'm satisfied with that differences, but if i forgott something important...

regards

seang November 24, 2009 20:31

Quote:

Originally Posted by abdul099 (Post 237571)
I'm working on my bachelor thesis and i have to do some simulation on airfoils. I can just confirm, what ttl said. I used star-cd for simulation, but ccm+ for meshing. The mesh (hex) is mostly o.k., but sometimes the prism layers are looking a little bit strange, especially on the leading edge (where the solution is not very sensitive to the prism layers).

The comparison of the simulation and an experiment gives a discrepancy of 2 - 15%, depending on the angle of attack. Of course, for great or very small (negative) angles of attack with seperating flow there is a larger discrepancy.
But i made the experience too, the solution is very sensitive to the prism layer mesh, especially the number of prism layers and the thickness of near wall prism layer.

Maybe somebody has some hints how to improve the difference between simulation and experiment? Don't matter, if not. I'm satisfied with that differences, but if i forgott something important...

regards

the prism layer and mesh are quite sensitive to parameters, especially, too quick an expansion. results are mesh and boundary condition sensitive.

For airfoils, there are two things to check, both of which, can be worked out through thin airfoil theory (that means, by hand!); the first is, the slope of CL vs alpha should be 2 pi. the second is, you can check the angle of attack at which CL is zero. Thin airfoil predicts this very accurately.

nomad November 25, 2009 00:57

Dear TTL,
Could you please explain how you setup the DFBI 6 DOF problem for propellers, as I haven't had much success with this? Does it require an interfaced rotating mesh, because the boat tutorial in the manual does not explain this step? I am unable to generate any rpm with my turbine rotor at all.

Also, Cd values for most airfoils are 100% off, whereas the Cl's are within 3%. I don't think STAR CCM+ is able to calculate accurate drag values for airfoils.

Thanks.

nvtrieu November 25, 2009 21:50

Hello ttl,

I've been working on Star-CCM+ for several months. I would like to do simulation for 2D airfoil but up to now, but still got some problem on meshing. The mesh near leading edge is not good even I used volume control at this region to make the cells as small as possible. So I got the Cd and Cl wrong compare with experiment data from NACA. When I did on NACA 0012, at the attack angle zero, but the Cl is not equal to zero and the Cd is lager than the right value!
So via this forum I would like to get the help for my problem! that would be great if you show me how to solve it!
Thanks in advance! and thanks to all!

Trieu.
ps: Email: trieuckgt@gmail.com
MSN: trieu.dut@live.com
YM: trieu_tme@yahoo.com

abdul099 November 26, 2009 04:09

Quote:

Originally Posted by nomad (Post 237591)
Dear TTL,
Also, Cd values for most airfoils are 100% off, whereas the Cl's are within 3%. I don't think STAR CCM+ is able to calculate accurate drag values for airfoils.

Thanks.

I did the experience with STAR-CD, that the drag is very sensitive to the used turbulence model. Maybe you will get better results with another turbulence model. On my first shot with STAR-CCM+ (unfortunately i had to change) i had a discrepancy on drag of 13%, so STAR-CCM+ is able to calculate right, if you use the right settings on an accurate mesh.

Thanks a lot, seang. I use an asymmetric airfoil, so it is a little bit to hard to check the angle of attack, at which Cl is zero. The slope of Cl vs. Alpha is about 6, which is o.k. if i consider the discrepancy of my Cl (about 5%). I just thought, i could have forget some settings. But compared with other people, my problems seem to be minor...

Good luck @all

nvtrieu January 3, 2010 01:51

How to simulate the flow over an 2-D airfoil in Star-CCM+ ?
 
Quote:

Originally Posted by ttl (Post 237194)
Hello,

I have a fair bit of experience meshing 2D airfoils in STAR-CCM+. I have found the code to be very good, meshing is pretty easy and very controllable. The prismatic layering is simple to adjust and the automated mesher is very robust. When I do have issues it is usually with the prism layering but all in all it is pretty good and near full proof for 2d airfoils.
Now that isn't to say STAR-CCM+ is perfect, just pretty good.

To comment on Aerospaceman's post, if your 2x off something is wrong with the mesh or the inputs.

I've also done some work with 3d propeller analysis and found CCM+ to be excellent and very accurate using the rigid body motion capability. The MRF capability is good but not as accurate, as one would expect.

For pure 2d work I usually use something other than CCM+ since it is ultimately a 3d code trying to do 2d work so it is not as fast and there are several good 2d codes available for little to no $$. However, CCM+ is hard to beat when you look at all the capabilities and the total cost.

Personally if I had to buy and use just one code, CCM+ would be it.

Hope that helps.

Hello ttl,

I'm very happy when I found a person who said "meshing is pretty easy and very controllable. The prismatic layering is simple to adjust and the automated mesher is very robust" in star-CCM+, because I've been doing simulation the flow over an 2-D airfoil in Star-CCM+ but I've not been success until now. The problem is how to generate a good mesh. I've tried many many time on it but still wrong! so could you teach me about setting up the parameter for grid generation? Thanks you so much!

email: trieuckgt@gmail.com
MSN: trieu.dut@livemail.com
YM: trieu_tme@yahoo.com

Dan709 August 15, 2016 22:02

1 Attachment(s)
Hi guys,

Sorry to revive this old thread here, but I am trying to mesh an airfoil in STAR-CCM+ and am having a lot of trouble at the trailing edge in particular. It seems that no matter how much I try and configure the settings and read the user guide, my prism layer thickness always retracts to zero at the trailing edge (see attached image).

I was just wondering if anyone has come up with a work around to solve this problem in STAR-CCM+? A lot of my colleagues that work in my graduate lab with me use Fluent and are able to keep the thickness of their boundary layer constant right up to the trailing edge of the airfoil. So, it appears to be an issue with the way that STAR-CCM+ generates its boundary layer mesh. Very frustrating!

If it would help at all, I'd also be happy to provide the exact settings I am using in the prism layer mesher. Any suggestions would be super appreciated, especially as this is my first post :)

Cheers,
Dan

cc2ley March 4, 2018 19:47

Hi,
Did u try changing the prism layer reduction percentage to 0 percent?


All times are GMT -4. The time now is 10:08.