[ICEM] Tetra/Prism - Small Y+

Jim0101 · September 2, 2014, 09:07

Hi Guys,

I'm looking at trying to accurately predict drag on aircraft fuselage and wing at transonic flight conditions (Mach=0.7).

I've been trying with ICEM to produce a hybrid Teta/prism mesh that will resolve the boundary layer down to the wall, giving me a Y+<=1. Effectively this requires a first cell height of roughly 0.005 mm (small!).

Despite producing a mesh that seems to pass all the ICEM mesh checks and has reasonable quality (>0.1) when I run it in fluent the case converges for the first 600 iterations then blows up! it appears that temperature and pressure limits are reached in a couple of the boundary layer cells, the first from the wall. But they are in areas where I can't see any issue with the mesh........

I've produced a mesh with a larger y value of 0.15 mm ish and the solution uses wall functions and converges quite happily.... So I'm happy with my case setup etc. It is definitely mesh related!

I've been using ICEM for a couple of years and follow the following steps to get my mesh:

1.0 Octree Volume mesh, to give me a watertight volume.
2.0 Delete the volume mesh.
3.0 Smooth & manually improve the surface mesh (Quality >0.5).
4.0 Then do a Delauney volume fill with advancing front option.
5.0 Smooth that volume mesh to obtain a quality above 0.3.
6.0 Grow 8 prism layers with an expansion ratio of 1.25, but no initial height set to allow the prisms to float.
7.0 Split the prisms in 3 to produce 24 layers, redistribute to give desired initial wall height.
8.0 Smooth tetras freezing pentas & any pyramids produced.
9.0 smooth everything up to around 0.05.

Something bad is happening in the first cell from the wall, but I don't know what.

Any advice is greatly appreciated!

Many Thanks

Jim

jjfai · September 3, 2014, 05:39

I have the same question!!!!

Jim0101 · September 3, 2014, 09:50

It's literally driving me nuts!

I've just tried reducing the triangulation tolerance in ICEM for the underlying surfaces to see if that helped, but getting the same behavior in fluent!

Something else a little strange is that the range of wall prism heights varies from 0.0055 mm to 0.0088 mm in ICEM but then in fluent they're 0.001 mm to 0.003 mm ish. I'm now concerned it may be some sort of write precision issues when converting mesh to .msh in ICEM........

Help!!!

September 2, 2014, 09:07	Tetra/Prism - Small Y+	#1
Jim0101 New Member Jim Join Date: Sep 2014 Location: UK Posts: 2 Rep Power: 0	Hi Guys, I'm looking at trying to accurately predict drag on aircraft fuselage and wing at transonic flight conditions (Mach=0.7). I've been trying with ICEM to produce a hybrid Teta/prism mesh that will resolve the boundary layer down to the wall, giving me a Y+<=1. Effectively this requires a first cell height of roughly 0.005 mm (small!). Despite producing a mesh that seems to pass all the ICEM mesh checks and has reasonable quality (>0.1) when I run it in fluent the case converges for the first 600 iterations then blows up! it appears that temperature and pressure limits are reached in a couple of the boundary layer cells, the first from the wall. But they are in areas where I can't see any issue with the mesh........ I've produced a mesh with a larger y value of 0.15 mm ish and the solution uses wall functions and converges quite happily.... So I'm happy with my case setup etc. It is definitely mesh related! I've been using ICEM for a couple of years and follow the following steps to get my mesh: 1.0 Octree Volume mesh, to give me a watertight volume. 2.0 Delete the volume mesh. 3.0 Smooth & manually improve the surface mesh (Quality >0.5). 4.0 Then do a Delauney volume fill with advancing front option. 5.0 Smooth that volume mesh to obtain a quality above 0.3. 6.0 Grow 8 prism layers with an expansion ratio of 1.25, but no initial height set to allow the prisms to float. 7.0 Split the prisms in 3 to produce 24 layers, redistribute to give desired initial wall height. 8.0 Smooth tetras freezing pentas & any pyramids produced. 9.0 smooth everything up to around 0.05. Something bad is happening in the first cell from the wall, but I don't know what. Any advice is greatly appreciated! Many Thanks Jim

September 3, 2014, 05:39		#2
jjfai New Member jjfai Join Date: Dec 2013 Posts: 1 Rep Power: 0	I have the same question!!!! Jim0101 likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[snappyHexMesh] .STL: non-closed manifold surface	giulio.topazio	OpenFOAM Meshing & Mesh Conversion	32	November 25, 2016 03:15
why did Time step is too small in stiffbs?	satum	FLUENT	1	December 10, 2012 09:15
Simulating small features on large models	siw	CFX	1	February 16, 2012 17:10
Small influence of convection in PISO if Courant number is small	sebi	OpenFOAM	0	February 22, 2011 14:55
Time step too small?	derz	CFX	2	September 8, 2010 00:03

September 3, 2014, 09:50		#3
Jim0101 New Member Jim Join Date: Sep 2014 Location: UK Posts: 2 Rep Power: 0	It's literally driving me nuts! I've just tried reducing the triangulation tolerance in ICEM for the underlying surfaces to see if that helped, but getting the same behavior in fluent! Something else a little strange is that the range of wall prism heights varies from 0.0055 mm to 0.0088 mm in ICEM but then in fluent they're 0.001 mm to 0.003 mm ish. I'm now concerned it may be some sort of write precision issues when converting mesh to .msh in ICEM........ Help!!!