Simulating a flute
 

I want to accurately simulate the sound of a flute. I mean: the musical instrument.
Is CFD the right field?

How would I start? :-)

The sound of a flute has two aspects. One aspect is the air flowing through the flute. That is an atonal sound, which I want to call "noise". The other aspect is the creation of a standing wave. That is the "musical" part, where the pitch is generated.
In order to really capture the characteristics of a flute, you need to simulate both noise and the pitch generation.
Also transients are important, i.e. the sound is different in the very beginning, when you start blowing, compared to "steady state" solutions, when you have sustained the flow for a while.

I should probably add, that I have a solid background in physics, but I am new to CFD.

Interesting project!
In general, this might work, though there might be more suiting, specialized methods around for acoustics. In your case I would start by applying an oscillating boundary condition for the velocity, and convert the pressure oscillations outside of the flute to a sound file.

But we have a constant flow into the flute, which somehow gets modulated there. When we start with a oscillating flow boundary condition we miss the most interesting part.

I agree with JBeilke. I do not want to use an oscillatory boundary condition. I want to start with a constant flow and the osicllations should (hopefully) arise by themselves.

I guess I would need to use DNS for that. But at the moment, I am not quite sure, where/how to start. Anyone got a good resource for DNS, or maybe a ready made solver, which is open source?