Hypersonic aerothermodynamics of satellite demise
File(s)
Author(s)
Rees, Thomas W
Type
Thesis
Abstract
The goal of the work presented in this thesis is to improve the aerothermodynamic heating
models used in modern satellite re-entry analysis tools. The current generation of heating
models are usually based on correlations developed for shapes such as flat plates and
hemispheres. These overly simplified models can lead to inaccurate predictions of the ground casualty risk associated with a re-entry event. In order to derive new heating models for
shapes more representative of satellite geometries, this work uses a combination of CFD and
wind tunnel measurements to study the heat fluxes experienced by a cuboid at two different
orientations in a Mach 5 flow and at a range of Reynolds number conditions.
2D and 3D CFD simulations showed that the hypersonic flow around a cuboid geometry
has a strong dependence on Reynolds number, with a breakaway separation bubble forming
from the windward expansion edge at very high Reynolds numbers. This separation bubble
can significantly lower the heat fluxes underneath it. In addition to this separation bubble,
the 3D CFD simulations showed that the sharp corners on a 3D geometry can generate
streamwise vortical structures along the streamwise edges of a cuboid. The surface heat
fluxes induced by these structures can be as high as the stagnation point heating value and
could therefore play a significant role in satellite fragmentation during re-entry. Finally, the
CFD simulations suggested that the highest heat flux values experienced by the cuboid occur
at the sharp edges and corners of the geometry. However, the heat flux predicted by CFD in
these regions is seemingly non-physically large.
The CFD simulations were experimentally validated with wind tunnel measurements
of the Stanton number distribution over a cube. The Stanton number measurements were
obtained by recording the temperature history of the wind tunnel model using infrared
thermography and then calculating the convective heat flux using a 3D inverse heat conduction
solver. The experimental Stanton number values generally showed very good agreement
with the CFD results, and comparisons of both the stagnation point Stanton number and
the average heating experienced by the cuboid were favourable. However, in contrast to
the CFD results, the experimental measurements did not show any region of significantly
increased heat flux near the sharp edges of the cuboid. We propose that this is likely due to a breakdown of the continuum assumption in these regions, which cannot be captured with
conventional CFD.
The combined CFD and experimental results were then compared to two currentgeneration
satellite re-entry prediction tools, DRAMA and SAM. These comparisons showed
that DRAMA overpredicts the average heat flux to a cuboid by 65-75% depending the flow
conditions and cube orientation. On the other hand, the average heat fluxes predicted by SAM
agreed well with the experimental and CFD values. Despite SAM’s success at predicting
average heat flux values, comparisons of Stanton number distributions over the surface of the
geometry showed that the tool did not predict the regions of increased heat flux associated
with the streamwise vortex structures generated by the 3D expansions around the cube corner,
while at the same time over-predicting the heat flux to other regions of the cube geometry.
Future generations of re-entry prediction tools will need to be able accurately predict the
Stanton number distributions across entire satellite geometries. This is important because
different fragmentation phenomena may occur depending on which satellite components fail
first.
models used in modern satellite re-entry analysis tools. The current generation of heating
models are usually based on correlations developed for shapes such as flat plates and
hemispheres. These overly simplified models can lead to inaccurate predictions of the ground casualty risk associated with a re-entry event. In order to derive new heating models for
shapes more representative of satellite geometries, this work uses a combination of CFD and
wind tunnel measurements to study the heat fluxes experienced by a cuboid at two different
orientations in a Mach 5 flow and at a range of Reynolds number conditions.
2D and 3D CFD simulations showed that the hypersonic flow around a cuboid geometry
has a strong dependence on Reynolds number, with a breakaway separation bubble forming
from the windward expansion edge at very high Reynolds numbers. This separation bubble
can significantly lower the heat fluxes underneath it. In addition to this separation bubble,
the 3D CFD simulations showed that the sharp corners on a 3D geometry can generate
streamwise vortical structures along the streamwise edges of a cuboid. The surface heat
fluxes induced by these structures can be as high as the stagnation point heating value and
could therefore play a significant role in satellite fragmentation during re-entry. Finally, the
CFD simulations suggested that the highest heat flux values experienced by the cuboid occur
at the sharp edges and corners of the geometry. However, the heat flux predicted by CFD in
these regions is seemingly non-physically large.
The CFD simulations were experimentally validated with wind tunnel measurements
of the Stanton number distribution over a cube. The Stanton number measurements were
obtained by recording the temperature history of the wind tunnel model using infrared
thermography and then calculating the convective heat flux using a 3D inverse heat conduction
solver. The experimental Stanton number values generally showed very good agreement
with the CFD results, and comparisons of both the stagnation point Stanton number and
the average heating experienced by the cuboid were favourable. However, in contrast to
the CFD results, the experimental measurements did not show any region of significantly
increased heat flux near the sharp edges of the cuboid. We propose that this is likely due to a breakdown of the continuum assumption in these regions, which cannot be captured with
conventional CFD.
The combined CFD and experimental results were then compared to two currentgeneration
satellite re-entry prediction tools, DRAMA and SAM. These comparisons showed
that DRAMA overpredicts the average heat flux to a cuboid by 65-75% depending the flow
conditions and cube orientation. On the other hand, the average heat fluxes predicted by SAM
agreed well with the experimental and CFD values. Despite SAM’s success at predicting
average heat flux values, comparisons of Stanton number distributions over the surface of the
geometry showed that the tool did not predict the regions of increased heat flux associated
with the streamwise vortex structures generated by the 3D expansions around the cube corner,
while at the same time over-predicting the heat flux to other regions of the cube geometry.
Future generations of re-entry prediction tools will need to be able accurately predict the
Stanton number distributions across entire satellite geometries. This is important because
different fragmentation phenomena may occur depending on which satellite components fail
first.
Version
Open Access
Date Issued
2020-08
Date Awarded
2021-01
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Bruce, Paul
Sponsor
European Space Agency
Fluid Gravity Engineering
Engineering and Physical Sciences Research Council
Grant Number
NPI 480-2015
EP/L016230/1
Publisher Department
Aeronautics
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)