Time step too small?

derz · September 6, 2010, 02:28

Hi guys,

I'm running a transient simulation of a little truck model we have in the wind tunnel. Width Re is around 640,000 for this and I'm using DES.

They have been getting a Cd value of around 0.29. I was always getting around 0.33 but ever since I reduced the size of the time steps I have gotten it to around 0.30, which is great.

However, I noticed that when I reduce the size of the time step to around 0.002 s, the vortex shedding I had observed with time step 0.005 s disappears. So my drag coefficient has improved, but I feel there should be some shedding at the rear, as that's what is happening in the wind tunnel.

Is my time step somehow too small to resolve the shedding somehow? I always thought going as small as possible was the goal.

* It should be noted I made a small geometric change before lowering the time step...it is possible that small change has somehow eradicated the shedding, but I highly doubt it.

ghorrocks · September 6, 2010, 07:39

The smaller timestep will usually mean increased temporal resolution and therefore more shedding, not less. If it results in less shedding then you may be approaching numerical round-off and that is limiting convergence. Have you tried running double precision?

derz · September 8, 2010, 00:03

Thanks, Glen. I'll see what happens when I run it in double precision mode...

September 6, 2010, 02:28	Time step too small?	#1
derz Member Derwin Parkin Join Date: Feb 2010 Posts: 35 Rep Power: 16	Hi guys, I'm running a transient simulation of a little truck model we have in the wind tunnel. Width Re is around 640,000 for this and I'm using DES. They have been getting a Cd value of around 0.29. I was always getting around 0.33 but ever since I reduced the size of the time steps I have gotten it to around 0.30, which is great. However, I noticed that when I reduce the size of the time step to around 0.002 s, the vortex shedding I had observed with time step 0.005 s disappears. So my drag coefficient has improved, but I feel there should be some shedding at the rear, as that's what is happening in the wind tunnel. Is my time step somehow too small to resolve the shedding somehow? I always thought going as small as possible was the goal. * It should be noted I made a small geometric change before lowering the time step...it is possible that small change has somehow eradicated the shedding, but I highly doubt it.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Superlinear speedup in OpenFOAM 13	msrinath80	OpenFOAM Running, Solving & CFD	18	March 3, 2015 05:36
How to write k and epsilon before the abnormal end	xiuying	OpenFOAM Running, Solving & CFD	8	August 27, 2013 15:33
Differences between serial and parallel runs	carsten	OpenFOAM Bugs	11	September 12, 2008 11:16
AMG versus ICCG	msrinath80	OpenFOAM Running, Solving & CFD	2	November 7, 2006 15:15
VOF	özgür	FLUENT	8	January 6, 2004 08:23

September 6, 2010, 07:39		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	The smaller timestep will usually mean increased temporal resolution and therefore more shedding, not less. If it results in less shedding then you may be approaching numerical round-off and that is limiting convergence. Have you tried running double precision?

September 8, 2010, 00:03		#3
derz Member Derwin Parkin Join Date: Feb 2010 Posts: 35 Rep Power: 16	Thanks, Glen. I'll see what happens when I run it in double precision mode...