Having problems creating a hole in a geometry of a meshed case

zaidhadi · February 13, 2015, 12:46

Hello everybody
I have the problem in the picture
I have a problem with the appearance of the hole in the upstream side of pier and not do not show in the downstream side
Please help
How can I solve this problem ?? in the picture

JBurnham · February 18, 2015, 17:06

I recommend you use v11's scour module, it's very much improved. In particular, it keeps the movable bed surface sharply-defined, instead of fuzzy like shown in your image.

You need to improve the resolution so that there are at least 10 - 20 cells across the depth of flow. It looks like you only have 3 or 4. Try to make your cells cubic, so they will be finer in x and y as well. If that is too many cells, use a nested block or a conforming block around the pier and some ways downstream, and locally refine the mesh in the smaller block. That's to start.

The next problem is that k-e and RNG turbulence damp out eddies, and that's what you need to form the downstream scour hole. You can definitely get downstream scour with RNG turbulence, but you may need to use LES for maximum accuracy.

Finally, once you have a mesh resolution and a turbulence model you like, you may need to calibrate the scour model Critical Shields Parameter. That is because even with very small cells, the shear stress at the bed may not be exactly correct due to the near-wall turbulence approximation. The solution is to calibrate CSP until the results are correct. Mmodel a known case first with similar flow velocity and depth, then apply the same mesh and CSP to the unknown case.

But like I said, use v11 or later. The earlier scour models are unlikely to give you good results.

SJTU · September 1, 2015, 14:01

Hello JBurnham
1.I noticed he seted a component in the left side of the bed（also the right side of the boundary） to avoid from being impacted by the current or wave directly .But I have saw someone set the component in the left side of the wave boundary（not included in the mesh block），so the wave or current would not formed in the special area.I want to know is this way right？
2.when I define the fluid region,should the fluid cover the sediment？namely ， Z min of fluid is Z min of sediment or Z max of sediment？？
thx for 有人 reply

thunde47 · April 4, 2016, 17:49

In relation to setting the correct parameters for Scour modeling like Critical Shields number, etc., we have recently partnered with CAESES - an optimization software. You can do parametric study to arrive at the correct parameter values. We did some internal study on Scour modeling using CAESES and got amazing results.

February 18, 2015, 17:06		#2
JBurnham Senior Member Jeff Burnham Join Date: Apr 2010 Posts: 204 Rep Power: 17	I recommend you use v11's scour module, it's very much improved. In particular, it keeps the movable bed surface sharply-defined, instead of fuzzy like shown in your image. You need to improve the resolution so that there are at least 10 - 20 cells across the depth of flow. It looks like you only have 3 or 4. Try to make your cells cubic, so they will be finer in x and y as well. If that is too many cells, use a nested block or a conforming block around the pier and some ways downstream, and locally refine the mesh in the smaller block. That's to start. The next problem is that k-e and RNG turbulence damp out eddies, and that's what you need to form the downstream scour hole. You can definitely get downstream scour with RNG turbulence, but you may need to use LES for maximum accuracy. Finally, once you have a mesh resolution and a turbulence model you like, you may need to calibrate the scour model Critical Shields Parameter. That is because even with very small cells, the shear stress at the bed may not be exactly correct due to the near-wall turbulence approximation. The solution is to calibrate CSP until the results are correct. Mmodel a known case first with similar flow velocity and depth, then apply the same mesh and CSP to the unknown case. But like I said, use v11 or later. The earlier scour models are unlikely to give you good results. zaidhadi, SJTU and PICK like this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problems using createBaffles to create a 3D thermal baffle	zfaraday	OpenFOAM Running, Solving & CFD	10	September 6, 2017 18:07
[ICEM] Geometry Problems - Cone nozzle	leo_NM	ANSYS Meshing & Geometry	3	November 1, 2013 09:18
Cloning a case or creating a new case for 16 processes	mellanoxuser	OpenFOAM Running, Solving & CFD	0	February 8, 2008 17:52
Problems with Meshing: Collapsed Cells	Emmanuel Resch	Siemens	1	July 30, 2007 04:02
Serious problems with geometry	Vidya Raja	FLUENT	4	May 28, 2006 12:50

September 1, 2015, 14:01		#3
SJTU New Member xu feng Join Date: Aug 2015 Location: China Shanghai Posts: 4 Rep Power: 11	Hello JBurnham 1.I noticed he seted a component in the left side of the bed（also the right side of the boundary） to avoid from being impacted by the current or wave directly .But I have saw someone set the component in the left side of the wave boundary（not included in the mesh block），so the wave or current would not formed in the special area.I want to know is this way right？ 2.when I define the fluid region,should the fluid cover the sediment？namely ， Z min of fluid is Z min of sediment or Z max of sediment？？ thx for 有人 reply

April 4, 2016, 17:49		#4
thunde47 Senior Member thunde47 Join Date: Dec 2015 Location: India, USA Posts: 133 Blog Entries: 1 Rep Power: 10	In relation to setting the correct parameters for Scour modeling like Critical Shields number, etc., we have recently partnered with CAESES - an optimization software. You can do parametric study to arrive at the correct parameter values. We did some internal study on Scour modeling using CAESES and got amazing results.