quick questions about 2D Low Re simulation over thin cambered Airfoils
 

Hello everybody,

I am simulation the Low Re flow (70 000 > Re > 200 000 ... corresponds about 1 > velocity > 3 m/s) over a family of thin 2D airfoils,
camber from 0 to 10
thickness from 0 to 5
AOA from 0 to 12

I started less than two months ago (with your help and support) using ICEM 14.5 and FLUENT 14.5 as a solver, using transition SST model to capture the Laminar Separation Bubble.

I validate well considering Cl, Cd, and Cp compared with the published measurements and Xfoil simulation.

the domain extended from 15 upstream to 25 downstream
wall Y+ always less than unity, except at the leading and trailing edges, about 1.2
Mesh Quality from 0.6 to 0.9 ... 216 630 cells ... 434 180 faces ... 217 550 nodes

I attached the reply script that I used to generate the mesh and some extreme cases coordinates files.

velocity inlet and pressure outlet ... Ti = 0.1 % and Turb. length scale 0.001 m for both.

I have some quick questions listed from the most important for me ...

1- How can I capture the Laminar Separation Bubble? have I use unsteady analysis? multi zones? How?
2- Should I modify the constants for SST model, specially for most thinnest cases (flat plate and circular arcs)?
3- I got low mesh quality for these extremely thin geometries, how can I improve it?
4- I set the Operating Pressure as 101325 Pa at X = -15 (the INLET section) and ... is that true?
5- I set the Refrence Values ... compute from INLET ... and Reference Zone as FLUID (the entire domain boundary) ... any problems?
6- I didn't change the defaults for the solution methods, controls, or the convergence absolute criteria ... accordingly I got convergence after about 300 to 600 iterations for the thick cases and no convergence at all fot thin cases ... What do you see?

thanks a lot for continue reading until this word
ANY KIND OF HELP, EVEN FOR A SINGLE QUESTION, WILL BE GREATLY APPRECIATED

Best Regards and Respect For ALL

Mamdouh

any help for the first two questions ?
thank you ...

Hi Mamdouh,
As far as I understand the bubble is captured rather easily using the 'right' conditions. By right conditions I mean that the turbulence intensity at the leading edge of your airfoil should be similar to the experiment. Transition is heavily dependent on the freestream conditions and it is important to capture the exact conditions.
For me, I tried simulating a thick airfoil section. The comparison between the experimental and numerical pressure coefficients was pretty good except for the leading edge suction pressure peak. The numerical results did not capture that. This lead to a decreased overall lift prediction. I was hoping you could shed a light on how well your pressure predictions were.

Regards.

Hi AUCH,

1- thank you for your reply,
2- actually, I don't have an experimental data for my geometry, my study is totally theoretical,
3- I just need run settings recommendations for my case (geometry and Re range), to capture both LSB and transition, if anyone has a close experience (either experimental or theoretical) for similar or near simulation,
4- I hope you get a valuable reply for your current thread, and GOOD LUCK for your work,
Regards,
Mamdouh