Ansys fluent

jackss · November 30, 2020, 18:50

I am trying to line up an experimentally determined wall pressure spectrum due to turbulent fluid flow to my simulation results in ANSYS FLUENT. However, I have tried nearly all solvers ( LES etc.) and the pressure spectrum that i am obtaining in FLUENT is wayyy larger in magnitude. Can anyone help?

DJF · December 2, 2020, 03:39

I've used LES in Fluent in the past to analyze pressure loads and spectra on an inclined plate. Need some more details though.

jackss · December 2, 2020, 10:50

In terms of the boundary conditions, the only thing that i am inputting is the inlet velocity and with the LES model, i am using "no perturbations" . The outlet pressure is (by default) set to 0 pascals. With the LES model, I am using WALES

In terms of the mesh, I am using only an inflation mesh ( in some simulations) which generate millions of elements. Some simulations i use both the body sizing and inflation mesh.

In terms of the time step, i am using 200, but i dont think that should be an issue.

Also, I forgot to mention ( my apologies) that it is fluid flow through a pipe that i am simulating.

jackss · December 2, 2020, 10:51

n terms of the boundary conditions, the only thing that i am inputting is the inlet velocity and with the LES model, i am using "no perturbations" . The outlet pressure is (by default) set to 0 pascals. With the LES model, I am using WALES

In terms of the mesh, I am using only an inflation mesh ( in some simulations) which generate millions of elements. Some simulations i use both the body sizing and inflation mesh.

In terms of the time step, i am using 200, but i dont think that should be an issue.

Also, I forgot to mention ( my apologies) that it is fluid flow through a pipe that i am simulating.

DnyanMiri · December 12, 2020, 01:20

I can help you with your problem but I have not understood your problem clearly.

May I ask why are you using LES for simple problem of flow through a pipe? What are the values of y plus for simulation you performed? more information is needed about turbulence as well as wall functions used if any.

jackss · December 21, 2020, 19:04

Hello, I am using LES because it showed the most potential amongst other solvers. The values for the y+ is around 0.001 . For the turbulence, I used no purtibations and well the solver was LES (WALE). Based on reading, I am thinking that when doing the LES, my y+ value should be 1 ?

DnyanMiri · December 21, 2020, 22:08

I would suggest you to do a quick RANS simulation for your setup and see if the values of pressure are coming as expected. If they are in the proximity then you can move to LES if you want, otherwise check the setup Boundary conditions, reference values etc. Y+ <1 values are recommended when your resolving viscous sublayer near the wall. And for LES you have to use very fina mesh to get its advantages.

jackss · December 22, 2020, 10:03

Thanks for your reply. If my y+ values is around 1e-7 , is that too small and unpractical ? because i noticed when my y+ value was around 0.001, the results looked okayish but when i increased the inflation mesh to a large number of layers ( say 50), the results were not good at all. I am uncertain if the extremely small values of y+ is causing an error or is it actually giving me the 'correct' values ?

mmntmflx · December 25, 2020, 07:00

Quote:

Originally Posted by jackss

Thanks for your reply. If my y+ values is around 1e-7 , is that too small and unpractical ? because i noticed when my y+ value was around 0.001, the results looked okayish but when i increased the inflation mesh to a large number of layers ( say 50), the results were not good at all. I am uncertain if the extremely small values of y+ is causing an error or is it actually giving me the 'correct' values ?

1. How much is your flow velocity?

2. So much inflation layer can reduce your mesh quality thats why you can reach poor results. For ex., aspect ratio is directly effected from inflation layers.

3. Just be sure your mesh quality first such as element quality, aspect ratio and skewness values.

jackss · December 30, 2020, 19:57

Thanks for your response ! My flow velocity is around 95 m/s. When I check my mesh quality, it isnt that great butI have tried other types of mesh but the results are not good. The only good ( not great) results I am getting is with the inflation layers but the mesh quality isnt that great.

A next issue I was having is that when I reduce the element size using the 'body sizing' option, the results that I were getting were very erroneous. Can you suggest anything again? Thanks for your replies, I really appreciate it.

November 30, 2020, 18:50	Ansys fluent	#1
jackss New Member Jack SSIlver Join Date: Nov 2020 Posts: 16 Rep Power: 5	I am trying to line up an experimentally determined wall pressure spectrum due to turbulent fluid flow to my simulation results in ANSYS FLUENT. However, I have tried nearly all solvers ( LES etc.) and the pressure spectrum that i am obtaining in FLUENT is wayyy larger in magnitude. Can anyone help?

December 2, 2020, 03:39	I can help you out	#2
DJF New Member Join Date: Jan 2019 Posts: 12 Rep Power: 7	I've used LES in Fluent in the past to analyze pressure loads and spectra on an inclined plate. Need some more details though.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
GPU acceleration in Ansys Fluent	flotus1	Hardware	63	May 12, 2023 02:48
How to open Icem mesh in Ansys Fluent?	emmkell	FLUENT	27	February 6, 2018 03:34
Using ANSYS Fluent 18.0 on workstations with multiple network cards and PBS Pro	cheezum	Hardware	0	February 28, 2017 10:23
How to combine ANSYS Fluent and Structural analysis?	diwakar	ANSYS	2	June 18, 2015 12:07
error in opening fluent in ansys workbench	tmeysam92	ANSYS	3	March 12, 2013 06:10

December 2, 2020, 10:50		#3
jackss New Member Jack SSIlver Join Date: Nov 2020 Posts: 16 Rep Power: 5	In terms of the boundary conditions, the only thing that i am inputting is the inlet velocity and with the LES model, i am using "no perturbations" . The outlet pressure is (by default) set to 0 pascals. With the LES model, I am using WALES In terms of the mesh, I am using only an inflation mesh ( in some simulations) which generate millions of elements. Some simulations i use both the body sizing and inflation mesh. In terms of the time step, i am using 200, but i dont think that should be an issue. Also, I forgot to mention ( my apologies) that it is fluid flow through a pipe that i am simulating.

December 2, 2020, 10:51		#4
jackss New Member Jack SSIlver Join Date: Nov 2020 Posts: 16 Rep Power: 5	n terms of the boundary conditions, the only thing that i am inputting is the inlet velocity and with the LES model, i am using "no perturbations" . The outlet pressure is (by default) set to 0 pascals. With the LES model, I am using WALES In terms of the mesh, I am using only an inflation mesh ( in some simulations) which generate millions of elements. Some simulations i use both the body sizing and inflation mesh. In terms of the time step, i am using 200, but i dont think that should be an issue. Also, I forgot to mention ( my apologies) that it is fluid flow through a pipe that i am simulating.

December 12, 2020, 01:20		#5
DnyanMiri Member Dnyanesh Mirikar Join Date: Jul 2019 Posts: 35 Rep Power: 6	I can help you with your problem but I have not understood your problem clearly. May I ask why are you using LES for simple problem of flow through a pipe? What are the values of y plus for simulation you performed? more information is needed about turbulence as well as wall functions used if any.

December 21, 2020, 19:04		#6
jackss New Member Jack SSIlver Join Date: Nov 2020 Posts: 16 Rep Power: 5	Hello, I am using LES because it showed the most potential amongst other solvers. The values for the y+ is around 0.001 . For the turbulence, I used no purtibations and well the solver was LES (WALE). Based on reading, I am thinking that when doing the LES, my y+ value should be 1 ?

December 21, 2020, 22:08		#7
DnyanMiri Member Dnyanesh Mirikar Join Date: Jul 2019 Posts: 35 Rep Power: 6	I would suggest you to do a quick RANS simulation for your setup and see if the values of pressure are coming as expected. If they are in the proximity then you can move to LES if you want, otherwise check the setup Boundary conditions, reference values etc. Y+ <1 values are recommended when your resolving viscous sublayer near the wall. And for LES you have to use very fina mesh to get its advantages.

December 22, 2020, 10:03		#8
jackss New Member Jack SSIlver Join Date: Nov 2020 Posts: 16 Rep Power: 5	Thanks for your reply. If my y+ values is around 1e-7 , is that too small and unpractical ? because i noticed when my y+ value was around 0.001, the results looked okayish but when i increased the inflation mesh to a large number of layers ( say 50), the results were not good at all. I am uncertain if the extremely small values of y+ is causing an error or is it actually giving me the 'correct' values ?

December 30, 2020, 19:57		#10
jackss New Member Jack SSIlver Join Date: Nov 2020 Posts: 16 Rep Power: 5	Thanks for your response ! My flow velocity is around 95 m/s. When I check my mesh quality, it isnt that great butI have tried other types of mesh but the results are not good. The only good ( not great) results I am getting is with the inflation layers but the mesh quality isnt that great. A next issue I was having is that when I reduce the element size using the 'body sizing' option, the results that I were getting were very erroneous. Can you suggest anything again? Thanks for your replies, I really appreciate it.