You need to have an variable area duct from your tank exit...I suggest to have a look to the books of Zucrow

The most important part seems to be the sonic throat. Other issues include condensation and reflection of pressure waves at the duct walls. It is a well established twchnology to build such a wind tunnel. There should be lots of information in the public domain.

https://en.wikipedia.org/wiki/Supersonic_wind_tunnel

https://en.wikipedia.org/wiki/Ludwieg_tube

You need a convergent / divergent nozzle.

At the minimum area, there will be choked flow (at Mach 1) that limits mass flow. the divergent section will then increase velocity to super-sonic at a speed determined by the input pressure.

Have a look at these references:

https://www.grc.nasa.gov/www/k-12/airplane/nozzled.html

http://www.ivorbittle.co.uk/Books/Fluids%20book/Chapter%2013%20%20web%20docs/Chapter%2013%20Part%203%20Complete%20doc.htm

https://en.wikibooks.org/wiki/Fluid_Mechanics_Applications/B25:Supersonic_Flow_In_Convergent-Divergent_Type_of_Nozzles

Please mark up this answer if you find it useful

Paul

Hello  Parth.

I'm agree with  Paul comments, I add:

Firstly , the  convergent- divergent nozzle is necessary to obtain  supersonic flow

The ratio of the section of exit and the section of the throat, determines the Mach number  of the exit.

The flow depends mainly on the pressure in the tank and the section of the throat , but also of the effectiveness and the quality of the divergent nozzle

 During the relaxation the pressure in the tank will decrease  quickly and to limit the duration of the supersonic mode.

To thus think of a sufficient  flow of compressor to fill the tank during your experiment. 

a good design of your convergent divergent nozzle will be able to achieve the mach no