The interaction of starting flow fields of an axisymmetric supersonic under-expanded jet with the cross section of a cylinder is simulated. To achieve this goal, an implicit pressure correction algorithm based on a finite volume method on a collocated grid for simulation of compressible flows is presented. A two dimensional test problem, the supersonic flow through a wind tunnel with a step, is simulated and the numerical outcomes agree very well with previous results. Then, the method is extended to axisymmetric flows and validated for a supersonic circular jet flow. Finally, the proposed method is used to analyze the concerned problem. To describe the evolution of the starting flow fields, the Mach number, pressure, density and temperature contours evolutions through time are investigated. Also, the evolution of total force and temperature distribution on the cylinder cross section surface are studied during the collision of mushroom head of supersonic jet with the surface of cylinder.
Keywordswhere γ is the ratio of specific heats, and by using this equation in energy equation (eq. 3), the number of variables are reduced.In the aforementioned governing equations, the conservation forms are used that the fluxes of mass, momentum and energy into and out of the control volume, themselves become important dependent variables rather than just the primitive variables. This is due to the fact that there are discontinuities in primitive variables across the shock; however, the fluxes are constant across the shock wave.
9To non-dimensionalization, the reference variables are used as follow:Where , r r u ρ and L are reference velocity, density and length, respectively. In this way, the non-dimensional governing equations are:
AlgorithmThe variables are collocated in space at the center of control volumes. Moreover, the numerical method is implicit in time. The governing equations can be discretized by integrating over a control volume and using the Gauss theorem as follows:. . . .where the superscripts denote the time steps. Also, the subscripts c,v and face represent the values at the center of control volumes and the values that located on the control volume surfaces, respectively.Here, V and face A determine the volume and surface area of the control volumes. Furthermore, i g is the momentum in the direction of i and N u is the face normal velocity, which is acquired by projection and nor that interpolation.
10To find the values on the control volume surfaces, which there are in the convection term of the governing equations, the following scheme is used:Where Q can be each of the terms on the control volume surfaces. At the vicinity of discontinuity w i takes the value of +1 and -1 depends on the flow direction, but in the rest of the domain it takes the value of zero.Adapted from the researches of Hou & Mahesh, 10,11 the algorithm is based on a pressure-correction method, which the face normal velocities are projected to satisfy a constraint on the divergence that is determined by the energy equation. Thi...