How to generate acoustic waves in Fluent

Haochey · April 3, 2021, 00:26

Hi,

I've been trying to simulate the viscous effects on acoustic propagation through a small pipe with infinite length. The challenge is to create acoustic plane waves in Fluent. The user menu suggests that not to use velocity inlet and outflow for compressible flow so I use the pressure inlet and use the expression : P_gauge = sin(2*π*20*t) [Pa] and still not sure how to set the outlet boundary. Any suggestion is appreciated!

Thanks!

LuckyTran · April 3, 2021, 09:26

You can use a velocity inlet or pressure inlet. I use velocity inlets a lot for acoustic forcing.

For the outlet, use a pressure outlet with the non-reflecting BC option so the waves don't get reflected back.

Haochey · April 3, 2021, 16:42

Hi Tran,

Thanks for your response. I found the solution of non-reflective pressure outlet does not converge and the residuals goes to infinite after 7 or 8 iterations.

Since the geometry is very simple, I use the automatically generated mesh as attached. Any tips on what way I should go?

I am also not quite sure how to set up the solution method (I tried both implicit and explicit. Neither of them don't work).

Thanks!

LuckyTran · April 4, 2021, 06:02

Looks like you are running into the typical issue of troubleshooting your very first CFD =)

There is no image that I can see but... It's a pipe... that's one of the simplest geometries to mesh after a box. You don't need anything fancy. A uniform mesh will do for the purpose of propagating acoustic waves. If your mesh is crap of course the simulation will diverge. Now assuming your mesh is decent...

You likely have an initialization problem or something else very trivial wrong in your setup if it diverges that fast. Check your initial conditions and so on. Try running a steady case with a simple constant pressure inlet and pressure outlet and go from there.

Haochey · April 5, 2021, 22:48

Hi Tran,

I think the problem is due to compressibility. I tried the simplest case where the pressure inlet gauge pressure is 0 [Pa] and the non-reflective pressure outlet is 0 [Pa] with steady state. The geometry is a 2D rectangular with uniform mesh and its top and bottom bounded by walls. The solution still diverges quickly. Same is for a 3D axisymmetric pipe.

However, it will converge if I use the pressure-based solver. Do you have any simple case files that I can learn from to make the density-based solver converge?

Thanks!

April 3, 2021, 00:26	How to generate acoustic waves in Fluent	#1
Haochey New Member Haocheng Yu Join Date: Apr 2021 Posts: 5 Rep Power: 5	Hi, I've been trying to simulate the viscous effects on acoustic propagation through a small pipe with infinite length. The challenge is to create acoustic plane waves in Fluent. The user menu suggests that not to use velocity inlet and outflow for compressible flow so I use the pressure inlet and use the expression : P_gauge = sin(2π20*t) [Pa] and still not sure how to set the outlet boundary. Any suggestion is appreciated! Thanks!

April 4, 2021, 06:02		#4
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,674 Rep Power: 66	Looks like you are running into the typical issue of troubleshooting your very first CFD =) There is no image that I can see but... It's a pipe... that's one of the simplest geometries to mesh after a box. You don't need anything fancy. A uniform mesh will do for the purpose of propagating acoustic waves. If your mesh is crap of course the simulation will diverge. Now assuming your mesh is decent... You likely have an initialization problem or something else very trivial wrong in your setup if it diverges that fast. Check your initial conditions and so on. Try running a steady case with a simple constant pressure inlet and pressure outlet and go from there. aero_head and Haochey like this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Acoustic Fluent Analysis	aslam9229	FLUENT	6	May 14, 2020 07:07
Acoustic Simulation in Fluent (Pre-defined sound source)	bassam693	FLUENT	1	January 2, 2016 07:06
The fluent stopped and errors with "Emergency: received SIGHUP signal"	yuyuxuan	FLUENT	0	December 3, 2013 22:56
Can we carry out Acoustic problems in Fluent	san	FLUENT	5	April 10, 2006 12:27
Setting up linear waves using fluent	Claire	FLUENT	0	November 16, 2005 03:37

April 3, 2021, 09:26		#2
LuckyTran Senior Member Lucky Join Date: Apr 2011 Location: Orlando, FL USA Posts: 5,674 Rep Power: 66	You can use a velocity inlet or pressure inlet. I use velocity inlets a lot for acoustic forcing. For the outlet, use a pressure outlet with the non-reflecting BC option so the waves don't get reflected back.

April 3, 2021, 16:42		#3
Haochey New Member Haocheng Yu Join Date: Apr 2021 Posts: 5 Rep Power: 5	Hi Tran, Thanks for your response. I found the solution of non-reflective pressure outlet does not converge and the residuals goes to infinite after 7 or 8 iterations. Since the geometry is very simple, I use the automatically generated mesh as attached. Any tips on what way I should go? I am also not quite sure how to set up the solution method (I tried both implicit and explicit. Neither of them don't work). Thanks!

April 5, 2021, 22:48		#5
Haochey New Member Haocheng Yu Join Date: Apr 2021 Posts: 5 Rep Power: 5	Hi Tran, I think the problem is due to compressibility. I tried the simplest case where the pressure inlet gauge pressure is 0 [Pa] and the non-reflective pressure outlet is 0 [Pa] with steady state. The geometry is a 2D rectangular with uniform mesh and its top and bottom bounded by walls. The solution still diverges quickly. Same is for a 3D axisymmetric pipe. However, it will converge if I use the pressure-based solver. Do you have any simple case files that I can learn from to make the density-based solver converge? Thanks!