Convergence problem

star · August 28, 2013, 09:36

Hi friends. I am simulating flow over 2d airfoil at Reynold number of 1000. Flow is laminar and steady state. I run with two different pressure schemes. First I used PRESTO Pressure scheme and 2nd order upwind and the residual became stable (very small oscillations) at 10e-6 and were not decreasing any more. I was monitoring CL and was constantly decreasing with a very minor change.
Then I restarted the whole process with 2nd order Pressure scheme and the solution converged but the result is over predicted from experimental. I know this is steady state case and not transient and my mesh quality, skewness, and Angle is also fine.
Anyone know where is the problem? Thanks in advance

flotus1 · August 28, 2013, 10:34

Why do you thing that this is a problem with convergence? You say that both simulations converge

How high is the error between experiment and simulation?

star · August 28, 2013, 10:51

Thanks Alex for your response. The correct CL is about 0.1634 (at 3 degree angle of attack) but i got 0.167 with 2nd order pressure and momentum schemes. My value was lower when i used coarser mesh but with refining it become over predected. I think it should be equal or less than correct value. When I changed pressure scheme to PRESTO, i got about 0.1620 but not totally stable, it still decreasing (very minor change 0.1% decrease in 1000 iterations) but residuals not decreasing. I run through about 32000 iterations. What do you say about this?

flotus1 · August 28, 2013, 13:41

I say that a mismatch of 2.2% between experiment and simulation is very good.
Unless you know for certain that the error of the measurement is lower than those 2.2%, there is no reason in pushing the simulation towards the "exact" measurement value.

star · August 28, 2013, 23:20

Actually I am doing grid validation at angle of attack 3 degree. one of my friend did the same case and he got almost the same result to the actual (He is not here). I wonder why with two different pressure schemes result is different. In my opinion, using PRESTO pressure scheme the result looks ok but the residuals become stable as i told before so i don't think this is correct.

star · August 29, 2013, 04:06

Quote:

Originally Posted by star

Thanks Alex for your response. The correct CL is about 0.1634 (at 3 degree angle of attack) but i got 0.167 with 2nd order pressure and momentum schemes. My value was lower when i used coarser mesh but with refining it become over predected. I think it should be equal or less than correct value. When I changed pressure scheme to PRESTO, i got about 0.1620 but not totally stable, it still decreasing (very minor change 0.1% decrease in 1000 iterations) but residuals not decreasing. I run through about 32000 iterations. What do you say about this?

Oh i am sorry i came to know that i had made a small mistake with using 2nd order pressure scheme that's why i got a little over predected CL, I corrected it and now it's in process..

.. but the problem with using PRESTO scheme is still there.

Centurion2011 · September 4, 2013, 05:34

try here

August 28, 2013, 09:36	Convergence problem	#1
star Member Harry Join Date: May 2013 Posts: 68 Rep Power: 12	Hi friends. I am simulating flow over 2d airfoil at Reynold number of 1000. Flow is laminar and steady state. I run with two different pressure schemes. First I used PRESTO Pressure scheme and 2nd order upwind and the residual became stable (very small oscillations) at 10e-6 and were not decreasing any more. I was monitoring CL and was constantly decreasing with a very minor change. Then I restarted the whole process with 2nd order Pressure scheme and the solution converged but the result is over predicted from experimental. I know this is steady state case and not transient and my mesh quality, skewness, and Angle is also fine. Anyone know where is the problem? Thanks in advance

September 4, 2013, 05:34		#7
Centurion2011 Member Join Date: Jul 2011 Posts: 78 Blog Entries: 1 Rep Power: 18	try here __________________ I'M NOT A GYNECOLOGIST BUT I'LL TAKE A LOOK.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
convergence problem when use pisoFoam, LES for wind tunnel case	Forrest_Lei	OpenFOAM	3	July 19, 2011 06:00
convergence problem	commonyue	Main CFD Forum	1	December 1, 2009 03:54
Convergence of CFX field in FSI analysis	nasdak	CFX	2	June 29, 2009 01:17
3D Fluid Flow Convergence problem	Emily	FLUENT	2	March 21, 2007 22:18
Non Convergence of 3D Heat transfer cfd problem	Balraj	Main CFD Forum	3	December 9, 2004 00:24

August 28, 2013, 10:34		#2
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,399 Rep Power: 46	Why do you thing that this is a problem with convergence? You say that both simulations converge How high is the error between experiment and simulation?

August 28, 2013, 10:51		#3
star Member Harry Join Date: May 2013 Posts: 68 Rep Power: 12	Thanks Alex for your response. The correct CL is about 0.1634 (at 3 degree angle of attack) but i got 0.167 with 2nd order pressure and momentum schemes. My value was lower when i used coarser mesh but with refining it become over predected. I think it should be equal or less than correct value. When I changed pressure scheme to PRESTO, i got about 0.1620 but not totally stable, it still decreasing (very minor change 0.1% decrease in 1000 iterations) but residuals not decreasing. I run through about 32000 iterations. What do you say about this?

August 28, 2013, 13:41		#4
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,399 Rep Power: 46	I say that a mismatch of 2.2% between experiment and simulation is very good. Unless you know for certain that the error of the measurement is lower than those 2.2%, there is no reason in pushing the simulation towards the "exact" measurement value.

August 28, 2013, 23:20		#5
star Member Harry Join Date: May 2013 Posts: 68 Rep Power: 12	Actually I am doing grid validation at angle of attack 3 degree. one of my friend did the same case and he got almost the same result to the actual (He is not here). I wonder why with two different pressure schemes result is different. In my opinion, using PRESTO pressure scheme the result looks ok but the residuals become stable as i told before so i don't think this is correct.