[nuimlabib]

i simulate airfoil using K-W turbulence model. i set the turbulence intensity about 0.05%, and turbulent length scale 0.0001, when i simulate with angle of attack with range -4 to 8 deggre can convergence. but when increasing the angle of attack with 10 up to 16 cannot convergence..iam using second order upwind scheme, chord length 1 m, and with velocity inlet 2.5 m/s, the properties of air is standard. why with larger of angle of attack 10 deggre this is cannot convergence...help me..please..thanks

[gaurav]

which Analysis software are you using? try to run a time accurate mode and see if you get a good result.

[nuimlabib]

i using fluent 6.2.16..
I've tried using coupled solution, and the parameters under relaxation factor was my volatile but still could not converging...whether there should be justified?

[gaurav]

Fluent can run only a second order scheme and from what information I have gathered from my professor u need to run the third and fifth order at this high angle of attack as there are a lot of viscous effects that are not considered in the 2nd order.

[nuimlabib]

simulation at a high angle of attack could be convergent with the first order scheme, but the result is less precise.and I have done the validation of a low angle of attack with a second order scheme and the results are better than the first order scheme.
whether to use segregated or coupled solution, which is better for the simulation Airfoil with low Reynolds numbers, please advice

[jstults]

Originally Posted by nuimlabib

simulation at a high angle of attack could be convergent with the first order scheme

This is because the first order scheme has large numerical dissipation, so the unsteadiness that is preventing the residuals from converging is damped out.

, but the result is less precise.and I have done the validation of a low angle of attack with a second order scheme and the results are better than the first order scheme.

The higher-order scheme should have less numerical dissipation, so you will more accurately capture the unsteadiness, leading to residuals that don't 'converge'.

Converging the residuals with the first order scheme may feel good, but it is probably not a physically relevant solution if the higher order scheme doesn't converge too.

[frmonto]

hi
with big separation bubbles the solution becomes unsteady because physically the bubble will move and your time marching scheme will follow this behaviour
do not play with fluent options to make the bubble stable, run unsteady
the low tu level says to the code that you are looking for almost a laminar solution, it is not surprising a big bubble

Originally Posted by nuimlabib

i simulate airfoil using K-W turbulence model. i set the turbulence intensity about 0.05%, and turbulent length scale 0.0001, when i simulate with angle of attack with range -4 to 8 deggre can convergence. but when increasing the angle of attack with 10 up to 16 cannot convergence..iam using second order upwind scheme, chord length 1 m, and with velocity inlet 2.5 m/s, the properties of air is standard. why with larger of angle of attack 10 deggre this is cannot convergence...help me..please..thanks