Hi Mehdi,

I guess that Comsol software can be a good way to simulate the experiment.

For measuring, in this case you mentioned high power levels, so I suggest that the best option is a 1/8 in. pressure field microphone. These are microphone that can support large levels without saturation although they feature smaller SNR than bigger size models. You can see G.R.A.S. or Brûel&Kjær web site for more info about these microphones and the signal conditioners they require.

I suppose your signal is going to be a sinusoidal of 20 kHz. In such a case, the size of the chamber is much larger than the wavelength and interference due to reflections may happen. Therefore, the result of the measurement may be affected by the microphone position inside the chamber. I recommend to measure several positions and average the resulting RMS values between positions. On the other hand, if you are only interested in the pressure at one specific position then just measure there.

I hope this helps :)

If your pulse duration is much greater than the sound transit time in the gas chamber (at least many milliseconds in air), then the signal is quasi-steady, and an analytic calculation of the steady-state field pattern, using Kinsler and Frey et al., will probably be simpler and more reliable than a simulation.  Your microphone will need to be calibrated so you can convert your voltage signal at 20 kHz to a pressure signal.  

You can use Piezoelectric probe for measuring sound pressure

At or more than 20kHz, you're dealing with wavelengths about 1.2cm and below; so the size of your microphone can greatly influence the field patterns in your measurements. Go with as small a microphone as possible. Also, working at ultrasonic frequencies require adequate microphones. 

Fiber optic probes will be ideal but aren't commercial. I suggest an ultrasonic MEMS microphone such as the Zero-Height SiSonic Microphone. I've used it with great results. Part # SPU0410LR5H-QB