Learning Forum

[YH]

Hello

I have been calculated acoustcs created by the water flow in the 3D pipe like the figure. Since the vortex is occur at the pipe connection, acoustics is created there. 

I calculated k-e model to initializel and then calculate LES with Ffowcs- Howkings Williams model to calcualte acoustics. Inlet velocity is 2.0 ft/s, outlet pressure is 6,000psi. Time step size is 5e-5 and Number of time steps is 20,000.The result is the next figure. It is showing a significantly high pressure in some places, the calculations do not appear to be working. Also, I could not understand why ot osn't centerd on 0. Would you tell me why this result is calculated? And what should I do to get the correct result? 

Let me know if you need additional information. Thanks!!

[Federico]

Hello, 

regarding LES, scale resolving simulations are really sensitive to grid sizing and hence the mesh must be finely tuned to what you are looking to resolve. Questions you should be asking yourself:

• What it you grid size in the region of interest? How does it relate to the acoustics that you expect to find?
• Related: given the time step size, what is your acoustic Courant number? You should aim for a Courant number <= 1. Your timestep should also be fine enough to resolve the expected pressure signals with highest frequency. A good rule of thumb is to have at least 10 points for a half-period of the pressure signal with highest frequency.

Regarding FWH, what is your acoustic source and where is your acoustic receiver? Typically, a scale resolving simulation (such as LES) is run to establish the flow with high accuracy and then followed with FWH model. I would need more information on how exactly you set this up you give you more assistance.

I hope his helps!

[YH]

Hi Federico,

Thank you for answering my question. I have not consider about 'acoustic courant number'. Honestly I haven't understand how Fleunt calculate the acoustic pressure. Does the acoustic model use the explicit method? 

This is the additional information and questions of my calculation. I would be happy if you could answer my questions.

• The narrow pipe has 0.01 [in] mesh, it is generated by face mesh. The wider pipe has 0.1 [in] mesh. Since I am thinking the acoustic source is at the narrow pipe. 
• The time step size is 5e-5 [sec]. And acoustic far field speed is 1115.486 [ft/sec]. I'm not familiar with acoustic courant number, but it is C = Vdt/dx = 1115.486 [ft/sec] x 5e-5 [sec]/0.01 [in] = 80. This means that I should increase the mesh size or decrease the time step size.  Is this calculation correct?
• Because I wanna see the frequency lower than 10,000 Hz, 5e-5 [sec] is enough in terms of frequency. 
• I'm sorry I couldn't get 'A good rule of thumb is to have at least 10 points for a half-period of the pressure signal with highest frequency' means. What is the half-period of the pressure signal?

Please let me know if you need more information.