First let’s talk about how noise is generated. Basically, unsteady signals in the flow field are generally caused by turbulence. This is also the reason why RANS is basically not seen in the field of aeroacoustic sound generation because RANS's prediction of turbulence information is unreliable. The most common ones are DNS and LES. DES also needs to step aside. I guess no one can answer how noise is generated from turbulence. We first assume that the motion of turbulent flow obeys the NS equation. So the NS equation basically contains five kinds of waves. Two sound waves propagating in opposite directions, two vortex waves in rotational directions, and a convective entropy wave. Then the main component of the chaotic motion of turbulence is vortex waves (and actually entropy harmony). Under the influence of nonlinearity, part of the energy will be converted into sound energy (this part is very, very small, Lighthill estimated). There is basically no noise problem for stable flow. Noise is a process of continuous compression and expansion, which is an unstable process. Anyone who tells you that he used a steady-state solver to solve a noise generation problem is probably going to hit the jackpot. In addition, let me mention that the problem of sound generation requires high-precision solvers (high-precision does not mean high-order). Numerical dissipation, artificial viscosity, and filtering are particularly rarely used because sound waves attenuate very slowly in actual situations. When calculating, numerical dissipation can cause sound waves to seeutomorrow instantly. The result of low numerical dissipation is that the program is extremely unstable, which is why there are few commercial software for high-precision pneumatic engineering. (As an aside, the most cutting-edge and high-end topic in the field of sound generation is combustion noise. Basically, the top experts are doing work in this direction or other related work in this direction.) The second question is how noise is propagated. This belongs to the field of sound propagation, which is basically considered to be a linear process. In other words, the propagation of sound can be considered as the propagation process of small disturbances on the background flow. Common solution methods are linear Euler equation (LEE), linear perturbation equation (APE), etc. This part is generally solved using high-precision algorithms for the same reason as above. This section is very popular because it covers many engineering issues (open rotor engine noise, airframe noise, duct noise, etc.).