CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > Siemens > STAR-CCM+

Prism Layers at Slip Wall to Non-Slip Wall Transition

Register Blogs Members List Search Today's Posts Mark Forums Read

Like Tree2Likes
  • 1 Post By me3840
  • 1 Post By me3840

Reply
 
LinkBack Thread Tools Display Modes
Old   November 25, 2017, 10:53
Default Prism Layers at Slip Wall to Non-Slip Wall Transition
  #1
New Member
 
InfernoxCJC's Avatar
 
Conor Crickmore
Join Date: Oct 2017
Location: Loughborough University, UK
Posts: 12
Rep Power: 2
InfernoxCJC is on a distinguished road
Hello everyone,
This is a follow-up to the thread I posted a few days ago.

The format of the simulation is a replication of a University Wind Tunnel with extended Slip Walls upstream and downstream of the Working Section to minimise inaccurate Boundary Layer formation, as used in this paper.

I was having issues with a strange 'phantom' Boundary Layer forming on my Slip Walls which I eventually contributed to the presence of Prism Layers and a very low Y+ value along the Slip Wall. Having removed the Prism Layers, I have been getting crashes caused by divergence. It seems I am getting a couple of cells, either at the Inlet-Working Section transition or the Working Section-Outlet transition that are recording Y+ and Velocity values in the thousands followed by the crash.

I've been playing around with Prism Layer settings. Keeping my standard Prism Layers throughout causes the strange Boundary Layer Behaviour, any form of transition from one type of Prism Layer (Thickness, Number of Layers etc.) to another causes this strange single cell phenomena, whereas using one consistent Prism Layer with higher Wall Y+ values seems to remove the issue, but puts me into the unwanted 1 < Y+ < 30 region.

Investigating my Mesh further, it seems to be caused by this Prism Layer behaviour at the transition. As you can see, it is far from uniform at the point where the floor meets the outer wall but I have no idea how to address this.

As before, I understand that this may not be sufficient information to go on, but any advice you could offer would be appreciated!

Conor
InfernoxCJC is offline   Reply With Quote

Old   November 26, 2017, 18:33
Default
  #2
Senior Member
 
Join Date: Nov 2010
Location: USA
Posts: 1,035
Rep Power: 16
me3840 is on a distinguished road
I don't really understand what geometry I'm looking at from that picture.

It does look to me though that you haven't chosen your prism settings very carefully. The one wall only has 3 prisms and the last prism is about 25% of the size of the first hexahedron, while the bottom wall looks like it has a very large number of prisms and while the last cell to first hexa transition looks better, the growth rate on the prism layers must be very large.

It looks like in the corner it retains the low y+ layers but there's face stuck with the larger prism setting on the bottom for whatever reason (are you doing this by volume control?).
InfernoxCJC likes this.
me3840 is offline   Reply With Quote

Old   November 26, 2017, 18:53
Default
  #3
New Member
 
InfernoxCJC's Avatar
 
Conor Crickmore
Join Date: Oct 2017
Location: Loughborough University, UK
Posts: 12
Rep Power: 2
InfernoxCJC is on a distinguished road
Quote:
Originally Posted by me3840 View Post
I don't really understand what geometry I'm looking at from that picture.

It does look to me though that you haven't chosen your prism settings very carefully. The one wall only has 3 prisms and the last prism is about 25% of the size of the first hexahedron, while the bottom wall looks like it has a very large number of prisms and while the last cell to first hexa transition looks better, the growth rate on the prism layers must be very large.

It looks like in the corner it retains the low y+ layers but there's face stuck with the larger prism setting on the bottom for whatever reason (are you doing this by volume control?).
Thanks for your response, me!
This is actually occurring in a region away from the test geometry, so you are looking at the 'wall' and 'floor' of my 'wind tunnel' while facing into the flow direction.

This seems to be my problem, the 'wall' doesn't seem to be properly applying the Prism Layers while the floor is and I have no idea why.

The problem is seemingly exacerbated by that fact that I am trying to transition from No-Slip Walls and Prism Layers during the 3.4m 'Working Section' of my Wind Tunnel to Slip Walls and no Prism Layers during the extended 'inlet' region, the reason for which is explained in the paper I linked too in the original post.

I control my Prism Layers on the large-scale using the Reference Values and Volume Controls for finer details. In this case I have also been disabling them on the aforementioned Slip Walls.
InfernoxCJC is offline   Reply With Quote

Old   November 26, 2017, 18:58
Default
  #4
Senior Member
 
Join Date: Nov 2010
Location: USA
Posts: 1,035
Rep Power: 16
me3840 is on a distinguished road
If your prism layers are good quality there should be no reason to remove them on a slipping wall.

I would not suggest using volume controls to adjust prism layers, especially with such a simple geometry. Transitioning from one layer to another is very difficult if there's no clear boundary which the surface always provides. It's better to use surface controls.

So I would switch to using surface controls and make sure the inputs you are using aren't too extreme.
InfernoxCJC likes this.
me3840 is offline   Reply With Quote

Old   November 26, 2017, 19:06
Default
  #5
New Member
 
InfernoxCJC's Avatar
 
Conor Crickmore
Join Date: Oct 2017
Location: Loughborough University, UK
Posts: 12
Rep Power: 2
InfernoxCJC is on a distinguished road
Quote:
Originally Posted by me3840 View Post
If your prism layers are good quality there should be no reason to remove them on a slipping wall.

I would not suggest using volume controls to adjust prism layers, especially with such a simple geometry. Transitioning from one layer to another is very difficult if there's no clear boundary which the surface always provides. It's better to use surface controls.

So I would switch to using surface controls and make sure the inputs you are using aren't too extreme.
Again, thanks for this!
The problem is, having the Prism Layers on my Slip Walls is what caused this behaviour which I was seeking help for. It seems the small-scale of the Near Wall Prism Layer was causing some form of 'phantom' Boundary Layer to form which was only resolved by significantly increasing the size of the near wall layer, much to the detriment of my Y+ values.

My initial idea was therefore to have my low Y+ Prism Layers transition to high Y+ Prism Layers (with difference near wall size, number of layers etc) at the boundary between the Non-Slip and Slip Walls, using the Region > Boundary > Custom Prism Layer options but any 'transition' caused the huge velocity values in single-cells described in the original post and these poor quality Prism Layers
InfernoxCJC is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Enhanced Wall Treatment paduchev FLUENT 24 January 8, 2018 11:55
Wall functions vs. boundary layers kandelabr OpenFOAM Running, Solving & CFD 3 June 12, 2017 07:05
Centrifugal fan-reverse flow in outlet lesds to a mass in flow field xiexing CFX 3 March 29, 2017 10:00
Icemcfd: Preventing prism inflation collisions? Joe CFX 1 July 31, 2007 09:13
Multicomponent fluid Andrea CFX 2 October 11, 2004 05:12


All times are GMT -4. The time now is 05:26.