Hello , can we simulate the fluid flow sounds in Fluent? i.e. when we get a recorded transient solution , can we get the related sounds of fluid flow for this solution ?
Sorry if I do not sound very optimistic about the prospects!! Sound pressure level is about 10^{-5}, while hydrodynamic pressure is about 10^{+5} in SI units. It is unbelievable that any open source software has this kind of ability to straddle across ten orders of magnitude. The alternative practice now (not that I agree with this!) is to calculate the near-field fluid flow by one of these readily available solvers, and then calculate the far-field sound by various empirical approaches, such as Lighthill's approach or Hawkins-Ffocs-Williams' model. This is pure model-type engineering calculations and we accept this, in the absence of any thing better.
Of course there are other deeper issues to address: How good is Navier-Stokes equation with Stokes' hypothesis for computational aero-acoustics where bulk viscosity is taken as zero. Hopefully, this type of issues will be (should be!) addressed in near future!
For the above reasons, many researchers in acoustics instead solve Euler's equation for the sound field. On the numerical side, one must adopt correct dispersion relation preserving (DRP) schemes.
The acoustic waves can be simulated if you resolve in a DNS formulation of the full compressible flow equations.
On the other hand, a LES could be tempted but you need an accurate scheme and a proper SGS model that does not artifically dissipate too much.
Hi Alaa Temimy ,
I'm not an expert in this field, but what you intend to do is named computational (aero-) acoustics prediction. Some theory on this is outlined here:
https://en.wikipedia.org/wiki/Computational_aeroacoustics
ANSYS Fluent is generally capable to do such simulations. The basis for it is a very accurately predicted field of pressure fluctuations, usually in the result of either a LES or a hybrid LES-URANS model (like e.g. SAS or SBES). In order predict from these (turbulent) pressure fluctuations the resulting sound frequencies in a defined location in relation to the sound generating object, ANSYS Fluent applies Lighthill's acoustic analogy, where a Ffowcs Williams and Hawkings equation is solved (FWH-equation).
This should provice you with some pointers in which direction you can do a web search for this topic.
In fact. Can not gets the sound through reviwe motion of the simulation for the fluid. But maybe save the move and used anather program to add the sound fit with simulation. For example motio of sea or bubble and then merged with the video by power point programe.
Dear Thomas Frank
yes, also in some Ansys tutorial was suggested to realize first an incompressible steady simulation and then work with the aeroacustic module but I think that if the computational power is sufficient, the realization of the unsteady compressible flow simulation would allow to get a much more detailed sound field.
Sorry if I do not sound very optimistic about the prospects!! Sound pressure level is about 10^{-5}, while hydrodynamic pressure is about 10^{+5} in SI units. It is unbelievable that any open source software has this kind of ability to straddle across ten orders of magnitude. The alternative practice now (not that I agree with this!) is to calculate the near-field fluid flow by one of these readily available solvers, and then calculate the far-field sound by various empirical approaches, such as Lighthill's approach or Hawkins-Ffocs-Williams' model. This is pure model-type engineering calculations and we accept this, in the absence of any thing better.
Of course there are other deeper issues to address: How good is Navier-Stokes equation with Stokes' hypothesis for computational aero-acoustics where bulk viscosity is taken as zero. Hopefully, this type of issues will be (should be!) addressed in near future!
For the above reasons, many researchers in acoustics instead solve Euler's equation for the sound field. On the numerical side, one must adopt correct dispersion relation preserving (DRP) schemes.
Right Tapan, the full model is very challenging! Fluent could be not sufficiently accurate also in its full second order formulation with central scheme, generally academic in-house code are used for aeroacustic problems. But I would be very curious to see the potentiality of Fluent using a very fine grid.
Thanks for your answer Tapan K. Sengupta .But we can can solve & visualize fluid flow case problem rather than hear the related sounds, which it's clear during experimental test for the case steady.
In addition , the range of sounds are not a problem , because we can select the required range just like volume control , or record it as waves and then examine it with audio programs.
I think , we're just need a link between the mesh cells, variation of pressure for each cell , and pressure to hearing waves analyses and mixer.So that each cell will give its sound and the mixer give the final mixed sounds.
Thanks for your answer Thomas Frank .
Can you provide any documents or guided tutorials of Fluent setup for sounds effects.
Fluent, I think, simulates the flow of a fluid. The fluid identification through its properties along with the physics of the flow is read into Fluent to tell it what it has to do. Fluent isn't designed to handle what happens to sound over a specified period of time, I reckon.
- Badges
- Science topic