Model change: k-epsilon to k-omega SST results in divergence

kroetenwurz · November 23, 2011, 11:14

Hey there,

At first, I am still a beginner in CFD with StarCCM+.

I have a simulation of a stirred tank reactor. Using the k-epsilon-model works fine and the solution converges. Now I want to run the same simulation with k-omega SST, but the solution doesn't converge. I didn't change any default values.

Do you have any general advices for me to achieve convergency?

Thank you and best wishes,
kw

kyle · November 23, 2011, 12:18

Did it diverge or is it just oscillating? If your flow has unsteady features, they will show up more readily with K-omega than K-epsilon, even on a steady state simulation. K-epsilon is more diffusive, so it has a tendency to stamp out these oscillations.

kroetenwurz · November 24, 2011, 08:31

At first it looked like that it wanted to diverge. I reduced the under-relaxation factors of Segrgated Flow, which helped. Now especially the Sdr-Plot is oscillating between 1e-4 and 1e-5 (no normilization).

November 23, 2011, 11:14	Model change: k-epsilon to k-omega SST results in divergence	#1
kroetenwurz New Member Join Date: Nov 2011 Posts: 2 Rep Power: 0	Hey there, At first, I am still a beginner in CFD with StarCCM+. I have a simulation of a stirred tank reactor. Using the k-epsilon-model works fine and the solution converges. Now I want to run the same simulation with k-omega SST, but the solution doesn't converge. I didn't change any default values. Do you have any general advices for me to achieve convergency? Thank you and best wishes, kw

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
k-omega SST simulation of turbulent flow around a circular cylinder	DanM	OpenFOAM Running, Solving & CFD	17	October 13, 2016 14:29
expert parameters of SST model?	David	CFX	1	December 2, 2005 07:36
SST model with rough wall?	David	CFX	4	November 30, 2005 12:22
SST model Autometic wall function parameters?	David	CFX	0	November 24, 2005 05:47
about Y+ in SST model	zuozicheng	CFX	2	July 12, 2005 11:05

November 23, 2011, 12:18		#2
kyle Senior Member Join Date: Mar 2009 Location: Austin, TX Posts: 160 Rep Power: 18	Did it diverge or is it just oscillating? If your flow has unsteady features, they will show up more readily with K-omega than K-epsilon, even on a steady state simulation. K-epsilon is more diffusive, so it has a tendency to stamp out these oscillations.

November 24, 2011, 08:31		#3
kroetenwurz New Member Join Date: Nov 2011 Posts: 2 Rep Power: 0	At first it looked like that it wanted to diverge. I reduced the under-relaxation factors of Segrgated Flow, which helped. Now especially the Sdr-Plot is oscillating between 1e-4 and 1e-5 (no normilization).