عنوان مقاله [English]
Study of cooling procedure is one of the main steps in designing space propulsion sub-systems in order to reduce costs and improving the performance. In the present study, the conjugate heat transfer of the methane coolant inside a rectangular channel is numerically simulated at supercritical pressure conditions. The methane flow inside the channel is compressible with low Mach number and enters the cooling channel with supercritical pressure and subcritical temperature. The coolant flow absorbs heat from the heated wall and exits the channel with supercritical temperature. A finite volume scheme is utilized for discretization of the governing equations on a collocated grid in general boundary-fitted coordinates. An iterative solution method based on the SIMPLEC (Semi-Implicit Method for Pressure Linked Equations-Consistent) algorithm is used to solve the equations. Also, central differencing scheme is used for the discretization of the diffusion fluxes and density approximation on the control volume boundaries. Upwind scheme is used for the density correlation approximation, and hybrid scheme is used for the convective fluxes discretization on the control volume surfaces. The solver is developed based on the thermodynamic and transport property relations corresponding to the coolant flow conditions in the transcritical regime. The solver is validated with the experimental data of MTP test, and the thermal behavior of methane inside the rectangular cooling channel is investigated. Also, a relation is derived for calculation of the pseudo-critical temperature of methane according to pressure. The relative error of this relation with NIST data is less than 0.5 percent and it operate in a range of pressure from 4.6 MPa to 30 MPa. Furthermore, the Nusselt relations presented for coolant flows with supercritical pressures are studied and corrected for the methane coolant at supercritical pressure conditions in 3D rectangular cooling channels. The relative error of modified Nusselt relations with numerical data are less than 1 percent.