How to keep y+ below 1

Mohamed911 · September 20, 2021, 16:05

Hello everyone.

I'm working on an axial pump. in literature, they recommend that y+ should be kept below 1.

how do I keep y+ below 1?

note that I'm using TurboGrid for meshing, and I'm using K-epsilon as a Turbulence model

Opaque · September 20, 2021, 17:32

Confused..

Why would you want to keep y+ ~ 1, and then use k-epsilon? The point of using k-epsilon is to be able to use coarser meshes and rely on the wall function to cover for the mesh deficiencies.

k-epsilon on wall resolved meshes is a waste of computer resources with no increase in accuracy

karachun · September 21, 2021, 03:43

When using k-e model with small y+ values you can even broke your solution so yeu gen nonphysical solution.

Mohamed911 · September 21, 2021, 07:37

Quote:

Originally Posted by Opaque

Confused..

Why would you want to keep y+ ~ 1, and then use k-epsilon? The point of using k-epsilon is to be able to use coarser meshes and rely on the wall function to cover for the mesh deficiencies.

k-epsilon on wall resolved meshes is a waste of computer resources with no increase in accuracy

so in your opinion. which turbulence model would be better to use?

Note: I am studying the induced shear stress

siw · September 21, 2021, 11:22

I recommend you read these CFD Leap Australia blogs articles hyperlinked in my post (2nd in thread).

CFX SST model with wall function and yplus

Better yet is someone sticks theses useful resources at the top of the CFX forum.

ghorrocks · September 21, 2021, 19:02

Yes, the SST turbulence model is the best model for most applications for CFX. It is definitely the most widely applicable across a large range of flow types, as it has sub-models for integration to the wall, rotating flow corrections, laminar to turbulent transitions, high mach number corrections, wall roughness models, ability to link to LES style models and more.

Gert-Jan · September 21, 2021, 20:33

SST is now the default turbulence model in Fluent.
I don't know why it isn't in CFX yet.
For reasons Glenn just mentioned, it should be upgraded to the default in my opinion.

Mohamed911 · September 22, 2021, 05:14

I tried to do simulations with what you suggested (SST).
for the mesh, I used TurboGrid as the mesh tool. a wall function Y + = 1 with 30 layers. the Reynolds numbers were set at 1E-05 for the stationary parts and 1E-06 for the rotor.
however, the results show that Y + is greater than 10

Gert-Jan · September 22, 2021, 06:52

Then reduce the first layer thickness such that your Y+ becomes below 1.

Mohamed911 · September 23, 2021, 13:21

Quote:

Originally Posted by Gert-Jan

Then reduce the first layer thickness such that your Y+ becomes below 1.

I tried to do this. but I couldn't find how to modify the layer thickness in TurboGrid.
all I could get is "Layer offset" and "wall Y+".

September 20, 2021, 16:05	How to keep y+ below 1	#1
Mohamed911 Member Join Date: Dec 2019 Posts: 31 Rep Power: 6	Hello everyone. I'm working on an axial pump. in literature, they recommend that y+ should be kept below 1. how do I keep y+ below 1? note that I'm using TurboGrid for meshing, and I'm using K-epsilon as a Turbulence model

September 20, 2021, 17:32		#2
Opaque Senior Member Join Date: Jun 2009 Posts: 1,880 Rep Power: 33	Confused.. Why would you want to keep y+ ~ 1, and then use k-epsilon? The point of using k-epsilon is to be able to use coarser meshes and rely on the wall function to cover for the mesh deficiencies. k-epsilon on wall resolved meshes is a waste of computer resources with no increase in accuracy __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

September 21, 2021, 19:02		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,870 Rep Power: 144	Yes, the SST turbulence model is the best model for most applications for CFX. It is definitely the most widely applicable across a large range of flow types, as it has sub-models for integration to the wall, rotating flow corrections, laminar to turbulent transitions, high mach number corrections, wall roughness models, ability to link to LES style models and more. __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

September 21, 2021, 03:43		#3
karachun Senior Member Join Date: Nov 2015 Posts: 246 Rep Power: 12	When using k-e model with small y+ values you can even broke your solution so yeu gen nonphysical solution.

September 21, 2021, 11:22		#5
siw Senior Member Stuart Join Date: Jul 2009 Location: Portsmouth, England Posts: 742 Rep Power: 26	I recommend you read these CFD Leap Australia blogs articles hyperlinked in my post (2nd in thread). CFX SST model with wall function and yplus Better yet is someone sticks theses useful resources at the top of the CFX forum.

September 21, 2021, 20:33		#7
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,928 Rep Power: 28	SST is now the default turbulence model in Fluent. I don't know why it isn't in CFX yet. For reasons Glenn just mentioned, it should be upgraded to the default in my opinion.

September 22, 2021, 05:14		#8
Mohamed911 Member Join Date: Dec 2019 Posts: 31 Rep Power: 6	I tried to do simulations with what you suggested (SST). for the mesh, I used TurboGrid as the mesh tool. a wall function Y + = 1 with 30 layers. the Reynolds numbers were set at 1E-05 for the stationary parts and 1E-06 for the rotor. however, the results show that Y + is greater than 10

September 22, 2021, 06:52		#9
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,928 Rep Power: 28	Then reduce the first layer thickness such that your Y+ becomes below 1.