Hello Jie,
If I want modify the parameters file (code) --"NsNsToHMNS.par" of the binary neutron stars in the* "gallery example"* to simulate the non vacuum binary neutron stars (and the gas surrounding could be in different density and velocity distribution, should I add some thorns into the parameters file(NsNsToHMNS.par )? if yes, what should I add? )how could I do ?
I do not think that such thorns already exist. You would have to write them on your own.
I do not have such a thorn, but have a attached a very simple skeleton code that demonstrates how one would set values on the grid. The skeleton code was written as a template on how to read spherically symmetric data onto the grid (it is not functional though) but you can just take out the fopen and fscanf calls and compute the density and velocity variables on your own.
Note that doing do will almost certainly violate the Hamiltonian and Momentum constraints of GR so you should keep your densities and velocities small to reduce the violation and may also consider evolving using the ccZ4 formulation of BSSN that the McLachlan code offers in the ML_ccZ4 thorns.
Yours, Roland