aspect ratio

keeper · March 26, 2012, 05:40

Hi all.

Ii is commonly known that big aspect ratio significantly affects convergence. But can simulation be affected if there is the mesh which conctains 2.5M elements and only one elements has aspect ratio of 30000?

Thanks for reply.

Best Regards

ghorrocks · March 26, 2012, 17:57

It depends.

An element with an aspect ratio that high is bound to cause problems. But if it is in a region of no significance then it is fine. If it is an area where accuracy is requierd then you have a problem. It is also highly likely that this element will have the highest residuals in the simulation and therefore slow convergence for the entire simulation.

It would be worth your while to improve the mesh quality to remove this element..

keeper · March 27, 2012, 04:40

thanks for the answer. I found that aspect ratio should be up to 50 away from the boundary. Is there any limitation concerning inflates? That is the region where high aspect ratio elements ussually appear.

ghorrocks · March 27, 2012, 05:42

There is no hard and fast rule for what the aspect ratio should be. The main problem with high aspect ratio elements is that they are highly diffusive, however, if the flow is aligned with the faces of the element the diffusion can be limited. This is the case in inflation layers as the flow is roughly parallel with the wall. But it can also be the case away from the wall as well.

If you are doing a simple model you can run high aspect ratio elements. If you are doing free surface modelling with surface tension an aspect ratio of 1.5 is going to cause problems. Your rule of thumb is nothing more than a guess and the real situation is very complex.

March 26, 2012, 05:40	aspect ratio	#1
keeper Senior Member Join Date: Jul 2010 Posts: 152 Rep Power: 15	Hi all. Ii is commonly known that big aspect ratio significantly affects convergence. But can simulation be affected if there is the mesh which conctains 2.5M elements and only one elements has aspect ratio of 30000? Thanks for reply. Best Regards

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ICEM] check aspect ratio	camoesas	ANSYS Meshing & Geometry	2	November 14, 2011 09:38
pressure eq. "converges" after few time steps	maddalena	OpenFOAM Running, Solving & CFD	69	July 21, 2011 07:42
aspect ratio vs. y+	hammam	CFX	3	August 6, 2007 10:41
High Aspect Ratio elements	Flavio	CFX	2	November 24, 2006 12:01
does aspect ratio affect analysis	mahesh	Main CFD Forum	2	October 10, 2005 09:12

March 26, 2012, 17:57		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	It depends. An element with an aspect ratio that high is bound to cause problems. But if it is in a region of no significance then it is fine. If it is an area where accuracy is requierd then you have a problem. It is also highly likely that this element will have the highest residuals in the simulation and therefore slow convergence for the entire simulation. It would be worth your while to improve the mesh quality to remove this element..

March 27, 2012, 04:40		#3
keeper Senior Member Join Date: Jul 2010 Posts: 152 Rep Power: 15	thanks for the answer. I found that aspect ratio should be up to 50 away from the boundary. Is there any limitation concerning inflates? That is the region where high aspect ratio elements ussually appear.

March 27, 2012, 05:42		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	There is no hard and fast rule for what the aspect ratio should be. The main problem with high aspect ratio elements is that they are highly diffusive, however, if the flow is aligned with the faces of the element the diffusion can be limited. This is the case in inflation layers as the flow is roughly parallel with the wall. But it can also be the case away from the wall as well. If you are doing a simple model you can run high aspect ratio elements. If you are doing free surface modelling with surface tension an aspect ratio of 1.5 is going to cause problems. Your rule of thumb is nothing more than a guess and the real situation is very complex.