Numerical simulation of a convergent divergent supersonic nozzle flow

Ekanayake, S 2013, Numerical simulation of a convergent divergent supersonic nozzle flow, Masters by Research, Mathematical and Geospatial Sciences, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Ekanayake.pdf Thesis Click to show the corresponding preview/stream application/pdf;... 15.17MB
Title Numerical simulation of a convergent divergent supersonic nozzle flow
Author(s) Ekanayake, S
Year 2013
Abstract Supersonic fluid flow through a two dimensional Convergent Divergent (CD) nozzle with angle variation added to the divergent section of the nozzle using ANSYS CFX. . Flow undergoes many forms of unique phenomena, including flow separation, unsteadiness, flow mixing, turbulence, Shock Induced Boundary Layer (SIBL) separation and Mach Shock Diamonds, when gas expanded through a CD nozzle. Some of these phenomena lead to energy loss, thereby reducing the overall thrust generated by the nozzle. The thrust loss due to shock waves and boundary layer separations in nozzle flows remains poorly understood, hence failed to reach maximum potential of an engine.

In this research, two nozzle configurations, with two symmetric and two asymmetric geometry shapes are investigated. Asymmetric is introduced by adding contraction angles at the divergent section. Numerical analysis is focused on the influence of the nozzle geometry, the Nozzle Area Ratio (NAR) and Nozzle Pressure Ratio (NPR) on the flow properties downstream (divergent section) and the external (jet plume) region of the nozzle. Capturing the boundary layer flow characteristics under strong adverse pressure gradients is of particular interest of this study. The NAR is varied to investigate the gas flow direction and speed, for the asymmetric nozzle in Underexpanded conditions at high NPRs. The two symmetric model configurations, NAR 1.5 and NAR 1.66, have a divergent angle at the throat of 2.801 and 3.89 degrees respectively. The two asymmetric geometry configurations, NAR 1.14 and 1.21, consist of a divergent angle of 2.801 degrees at the throat with contraction angle variations in the divergent section. ANSYS CFX is used to solve time-dependent RANS equations for supersonic two dimensional (2D) nozzle flow with Shear Stress Transport (SST) turbulence model. The NPRs for both symmetric and asymmetric are varied between 1.27 – 12.0 under the sea-level condition for the steady state solutions.

It is found that -under low NPRs, the SST model has transcended in performance to capture internal shocks, as well as boundary layer separation, re-circulation zones, shear layer stresses caused by strong adverse pressure gradients accurately in axis-symmetric nozzles. Typical Lambda shocks associated to internal flow separation is not observed for the asymmetry type model, NAR 1.21. Asymmetric nozzles have produced higher Mach number values - than that of symmetric - nozzles.

The research has also found that when varying the divergent section of the asymmetric nozzles, the flow path is vectored away from the nozzle axis line at high NPRs. The flow is significantly offset in the desired direction and is considerably different from the traditional Mach Diamond shock pattern observed in symmetric nozzle shapes at the jet plume region. Asymmetric nozzle geometries have a major contribution towards the size of the Mach disks and Diamond shock patterns within the jet plume region. Varying the angle of the top and the bottom walls has a significant effect on the exhaust flow direction. This could be implemented in the future high speed nozzle designs.
Degree Masters by Research
Institution RMIT University
School, Department or Centre Mathematical and Geospatial Sciences
Keyword(s) Supersonic Flow
Convergent Divergent Nozzle
Flow Separation
Underexpanded
Overexpanded
Nozzle Flow
Asymmetric Nozzle
SST Turbulence Model
Versions
Version Filter Type
Access Statistics: 1034 Abstract Views, 7227 File Downloads  -  Detailed Statistics
Created: Thu, 01 Aug 2013, 15:53:20 EST by Brett Fenton
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us