Thank you Ian Hinder.At the first step, I just want to add a low density fluid to the BBH simulation (mimic the atmosphere around the BH) to see how it evolution (I remember I had read a paper which describe this kind simulation, and I want to repeat it.). I put the fluid static in the initial data.I think I will give some constraints on the fluid static in the initial data in the future. Maybe it need to write a new thorn, I think.Yours,Zhi-Chao.--2017-08-18 22:43 GMT+08:00 Ian Hinder <ian.hinder@aei.mpg.de>:On 18 Aug 2017, at 10:05, 赵志超 <yanyuechuixue@gmail.com> wrote:Thank you Ian Hinder.Yes I want to add a low density fluid to a BBH simulation.I review the source code of GRHydro_InitialData Thorn's GRHydro_OnlyAtmo.F90 . I found it give any position ( I mean every nx,ny,nz) a rho value which is a constant equal to "rho_abs_min" or "initial_rho_abs_min".so I think it can give a low density fluid to a BBH simulation.But it report "NAN" or "Con2Prim: count > GRHydro_countmax".OK, thanks for the clarification!I'm not an expert on GRHydro, but maybe someone else can suggest something?Are you planning to solve the constraints?