CFD Flow Simulation for Supersonic Nozzles-Conical and its Bell equivalent (RANS-SST)

Flow simulation of nozzles using SU2 solver using structured mesh with RANS-SST combination. As a test case PSLV stage 1 and stage 3 nozzle sizes were used.

Table of contents

• General info
• Screenshots
• Setup
• How to run
• References
• Updates
• To-do list

General info

• Gmsh geometry file is written using python wrapper and generates the .su2 file required for the SU2 solver.
• SU2 configuration is set with RANS (Governing equations - Reynolds-averaged Navier-Stokes) and SST (Turbulence model - Shear Stress Transport).
• RANS-SST is very sensitive to mesh geometry, size, CFL number, multigrid parameters.

Script features

1. Python script 'trans_nozzle_structured.py' designs both conical along with its bell equivalent nozzle and generates structured mesh in .su2 format directly.
2. Generated files are bell_nozzle_cgrid.su2 and conical_nozzle_cgrid.su2.
3. Included in the configuration file, the following test case scenarios. Commented one case.

Test Cases

1. PSLV 1st stage Nozzle: Conical, Area Ratio : 8.0, Throat Radius : 836
   PS1-Max-thrust conditions: MEOP = 5.88 MPa, Tc = 2900 K, Pa = 74293.6 Pa [around approx. 17th sec of flight @ alt = 5.6 Km]
2. PSLV 3rd stage Nozzle: Bell, Area Ratio : 51.0, Throat Radius : 100.52
   PS3-Max-thrust conditions: MEOP = 6.37 MPa, Tc = 2900 K, Pa = 0.0009964 Pa [around approx. 320th sec of flight @ altiutde 136.0 km]

Screenshots

**PSLV 1st stage Nozzle (conical). Also included Bell equivalent. **

Area Ratio : 8.0, Throat Radius : 836

 
 

PSLV 3rd stage Nozzle (bell). Also included conical equivalent.

Area Ratio : 51.0, Throat Radius : 100.52

Setup

Tested the code on Linux based setup.

1. Gmsh (Version-4.7.1) with python module - Mesh generation
2. SU2 (Version-7.1.0) - CFD solver
3. Paraview (Version-5.7.0) - flow visualization

 
SU2 appears in two flavours. Single core/cpu version or MPI version for parallel computations.
If you have multi-core processor, then MPI version speeds up the computation. You need to install MCICH

How to run

Download the files.

Generate mesh

• Verify and install required modules
• run python trans_nozzle_structured.py.

Single CPU mode:

• Execute SU2_CFD nozzle_rans_sst.cfg on terminal.
• Open flow.vtu file through Paraview.

Parallel computation mode:

• Execute mpirun -n 2 SU2_CFD nozzle_rans_sst.cfg on terminal. (2-cores will be used)
• Open flow.vtu file through Paraview.

References

1. Development of Nozzle for PSLV Booster
   [ https://arc.aiaa.org/doi/pdf/10.2514/6.1991-2588 ]
2. Rocket Nozzle Geometries - Jerry M. Seitzman Professor
   [ http://seitzman.gatech.edu/classes/ae6450/nozzle_geometries.pdf ]
3. Bell Nozzle
   [ https://github.com/ravi4ram/Bell-Nozzle ]

Updates

To-do list

• SSLV