Content uploaded by Georg Herdrich
Author content
All content in this area was uploaded by Georg Herdrich on Jul 12, 2017
Content may be subject to copyright.
1
Experimental Investigation of Material Demisability
in Uncontrolled Earth Re-entries
By Adam S. PAGAN,1) Bartomeu MASSUTI-BALLESTER,1) Georg HERDRICH,1)
James A. MERRIFIELD,2) James C. BECK,3) Volker LIEDTKE,4) and Benoit BONVOISIN5)
1)Institute of Space Systems, University of Stuttgart, Stuttgart, Germany
2)Fluid Gravity Engineering Ltd., Emsworth, United Kingdom
3)Belstead Research Ltd., Ashford, United Kingdom
4)Aerospace and Advanced Composites GmbH, Wiener Neustadt, Austria
5)European Space Agency, European Space Research and Technology Centre (ESA-ESTEC), Noordwijk, The Netherlands
(Received April 24th, 2017)
An overview of the results of an extensive experimental campaign is given, in which eight spaceflight-relevant
materials as well as segments cut from a Composite Overwrapped Pressure Vessel (COPV) were characterised with regards
to their aerothermal demisability, using simulated uncontrolled atmospheric entry conditions generated in the Institute of
Space Systems’ (IRS) plasma wind tunnel facilities. The facilities, measurement techniques and procedures as well as the
various test conditions employed are presented. Data extracted from the experiments include spatially and temporally
resolved heating histories of exposed samples, visible and near-visible range boundary layer spectra,
temperature-dependent spectral and total surface emissivities as well as extensive visual and physical observations and
measurements. Due to the extensive scope of the activity, only an overview containing a limited selection of representative
results is presented. On the basis of these findings, a brief discussion of each specimen type’s specific demise-relevant
behaviour and some observations of note are provided.
Key Words: Design for Demise, Uncontrolled Re-Entry, COPV, Ablation, Material Characterisation
Nomenclature
I
: current
h
: enthalpy
K
: gas-specific constant
ṁ
: mass flow rate
Ma
: Mach number
p
: pressure
: stagnation point heat flux
R
: radius
U
: voltage
x
: distance to generator nozzle
Subscripts
0
: stagnation point
amb
: ambient
Ar
: Argon flow
CuO
: Cold-wall copper oxide calorimeter
cyl80
: cylindrical geometry, Ø 80 mm
eff
: effective
fc
: fully catalytic
flat50
: flat-faced geometry, Ø 50 mm
hemi50
: hemispherical geometry, Ø 50 mm
mag
: magnet
N
: nose
N2
: nitrogen flow
O2
: oxygen flow
1. Introduction
Spacecraft re-entering Earth’s atmosphere in an
uncontrolled manner are subjected to intense aerothermal
heating as their relative kinetic energy is dissipated. Where no
thermal protection system is provided, vehicles and other
structures will usually break up mid-flight, releasing
substructures and components, which may or may not proceed
to succumb to the extreme heating environment.1) With the
safety of life and property on Earth in mind, the emerging
Design for Design (D4D) philosophy aims at reducing the
risks associated with the impact of space debris by
encouraging the break-up and demise of any spacecraft at risk
of being subjected to an uncontrolled entry event through an
according design.2) One important aspect of such designs lies
in the choice of materials, which exhibit widely varying
responses towards re-entry relevant environments. In the
course of the ESA TRP “Characterisation of Demisable
Materials”, a selection of representative aerospace structure
materials was subjected to extensive investigations in order to
determine their respective propensities towards and
mechanisms of demise.
In this context, eight different aerospace materials were
characterised and investigated under experimentally simulated
high enthalpy air flows at the Institute of Space Systems (IRS)
of the University of Stuttgart. The test conditions were
selected to be relevant for uncontrolled re-entries of spacecraft
components from LEO following a re-entry breakup event and
were effected within the IRS plasma wind tunnels PWK1 and
2
PWK4.3-5)
The selection of materials included four metallic alloys, one
high-temperature ceramic as well as three distinct organic and
semi-organic composites, which were subjected to
aerothermal heating at different heating rates until a demise of
the material specimen was observed or a steady state
condition was attained. To this end, radiation-cooled and
well-insulated sample holders were deployed to emulate 1D
heating conditions. Surface temperatures were monitored
through pyrometry at both the exposed front and enclosed
back surfaces of the test specimens. Optical Emission
Spectroscopy (OES) was employed for the identification of
gas species in the boundary layer as well as for the tracking of
material-related species abundance.6) Further, the progression
of material demise was monitored visually, which, together
with an extensive post-test analysis of tested material samples,
provides insight into the specific mechanisms and
phenomenology of aerothermal demise.7,8)
Flanking this test campaign, an investigation of the
temperature-dependent total pyrometer-specific spectral
emissivities was conducted for most of the candidate materials
both in pristine and post-exposure conditions using the
Emissivity Measurement Facility (EMF) at IRS.9,10) This
activity provided a reference for the correction of pyrometric
temperature measurements as well as for modelling purposes.
In conjunction with the plasma wind tunnel experiment data,
these results illustrate the high significance of optical surface
properties and their behaviour when subjected to the harsh
thermochemical environment encountered during re-entry.
Finally, in addition to the eight basic materials, sections cut
from a Composite Overwrapped Pressure Vessel (COPV)
were subjected to similar test conditions to additionally
provide insights into the demise-specific interaction of
different materials in heterogeneous structures.8)
An overview of the results of the test campaign is presented
and general findings are discussed. The observed mechanisms
encouraging and prolonging the aerothermal demise of
specific types of materials are explored.
2. Material Selection and Grouping
A total of nine different materials and structures were
investigated within two phases in the course of this research
project. For the sake of a more structured discussion of the
results, these are grouped into four different categories:
High-temperature ceramics, metallic alloys, laminate
composites and compound structures.
The sample geometry used for the eight single-material
sample types constitutes a flat truncated cone, with a minor
diameter of 26.5 mm and an opening angle of 60°.
Thicknesses varied between 3 mm and 4.2 mm, depending on
the material type. This coin-shaped geometry was selected for
compatibility both with the IRS standard radiation-cooled
50 mm material plasma probe and with the Emissivity
Measurement Facility (EMF) to simplify emissivity
investigations of both virgin and pre-exposed specimens.
As the only representative of the high-temperature ceramic
group, Silicon Carbide (SSiC) was tested, with its
well-reproducible and comparatively durable response to
aerothermal heating illustrating the importance of material
selection for demisable spacecraft and providing an excellent
baseline reference for the remaining materials.
Four metallic aerospace alloys were tested, including AISI
type 316L stainless steel, grade 5 titanium Ti-6Al-4V, the
aluminium alloy 7075 as well as the aluminium-lithium alloy
2099.
Amongst the composites tested were two types of Carbon
Fibre-Reinforced Polymer, one composed of a Torayca M55J
carbon fibre laminate with a TenCate EX-1515 cyanate ester
resin (CFRP EX-1515/M55J), and another featuring a
Polyether Ether Ketone matrix (CFRP/PEEK). A further
variant of CFRP provided by JAXA-ISAS was tested in
cooperation in the context of a thematically related activity,
however these results will be discussed at a later time. The
final composite material type tested in the course of this the
reported activity was Glass Laminate Aluminium Reinforced
Epoxy (GLARE).
The only compound structures investigated were extracted
from a composite-overwrapped He tank, consisting of a 1 mm
Ti-6Al-4V liner overwrapped with approximately 8 mm of
CFRP. Circular segments with a 39.8 mm diameter were cut
out of the tank structure and instrumented with four type K
thermocouples located at depths of 2.5 mm, 4.5 mm
and 6.5 mm from the front surface and at the back face
respectively. This was integrated with a 30 mm thick
aluminium silicate insulator and glued to a concentrically
placed aluminium back plate for compatibility with the IRS
standard 80 mm ablator material sample holder. Due to the
necessary deviation in sample and thus probe geometries,
separate test conditions had to be calibrated.
3. Facilities
An overview of the IRS test facilities employed in the course
of this research activity is provided in the following.
3.1. Plasma Wind Tunnels
Most of the plasma wind tunnel tests were conducted in the
IRS facility PWK4, with the exception of the steady state
response characterisation activity as well as the initial tests at
low heat fluxes, which were conducted in PWK1. With the
exception of the plasma source, both facilities are essentially
identical with regards to their setup (see Fig. 1 for a
schematic), and are connected to the same powerful in-house
vacuum system, capable of providing a maximum suction
power of 6000 m³/h and up to 250 000 m³/h at atmospheric
pressure 10 Pa, respectively, as measured at the 1 m diameter
intake of the system. In effect, a wide variety of conditions
relevant to Earth atmospheric entries at altitudes up to 90 km
can be emulated between the IRS plasma wind tunnels. The
plasma generator installed in each facility is flanged to the lid
of a cylindrical tank, which features a 2 m diameter and a
length of 6 m.4,5)
3
PWK1 uses the magnetoplasmadynamic generator RD5,
which features a nozzle exit diameter of 125 mm and is
capable of emulating the thermochemical boundary conditions
experienced during hyperbolic Earth entries. Bulk enthalpies
lie between 3 and 150 MJ/kg at the nozzle exit and stagnation
pressure levels of up to 50 hPa can be effected for air mass
flows of up to 50 g/s. The resulting cold-wall calorimetric
reference stagnation point heat fluxes can exceed 15 MW/m²
for all relevant plasma probe geometries. Cathode and anode
erosion through oxidation are respectively minimised by
injecting the oxygen component into the air gas mixture
around the arc region only after the nitrogen component has
passed by the cathode and by further injecting a small
enveloping flow of argon around the surface of the nozzle.4,5)
PWK4 deploys the thermal plasma generator RB3 with a
nozzle exit diameter of 75 mm, capable of providing an air
flow with an enthalpy of 6 to 30 MJ/kg at the nozzle exit,
capable of effecting reference heat fluxes of up to 5 MW/m²
and total pressures up to 1 bar for air mass flows. As for RD5,
cathode erosion is minimised by adding the oxygen
component at the downstream end of the anode. A spotty arc
attachment is avoided through the generation of an axial
magnetic field through an external coil, which constantly
circulates the arc.4,5)
PWK1 and PWK4 are fitted with four-axis CNC tables,
enabling a free positioning of a water-cooled plasma probe.
An array of optical windows provides full visual access to the
entirety of the test section.4,5)
3.2. Emissivity Measurement Facility
The IRS Emissivity Measurement Facility (EMF) features a
pressure vessel, which can be evacuated or within which a
protective argon atmosphere can be generated. A long
graphite tube, designed to provide a near-isothermal
environment when heated, is placed at the centre of the
chamber between two water-cooled electrodes which are
attached to an external power supply. The tube can
accordingly be heated to any desired temperature level up to
2200 K and beyond. A pneumatic piston is installed at the rear
end of the chamber, onto which a graphite material sample
holder can be mounted. An interchangeable optical window is
installed in the flange of the other side of the chamber,
through which the sample can be observed throughout the
entire length of the tube. At its backward position, the sample
is heated through radiation and convection from the heated
tube within which it remains immersed until a thermal
near-equilibrium is created (see also Fig. 2). Seen through the
window, its effective surface emissivity is greater than 0.999,
at which point it can be considered a black body for all intents
and purposes. Once the surface temperature has equilibrated at
the desired temperature level, the piston is shot forward
through the graphite tube within 0.2 s, coming to a stop near
the window. Most materials experience no significant cooling
during this short time, whereas they are momentarily removed
from the black body cavity, now displaying grey body
properties. Depending on the selection of pyrometers
deployed, device-specific, spectral and total emissivities can
accordingly be extracted from the relation of measured
temperatures measured before and immediately after the
piston’s movement is completed.9,10)
Upon completion of a test shot, the piston is reset to its
original position, and is heated either to the same temperature
level again, or to a higher level if the heating current is
modified respectively.
4. Measurement Techniques
In the following, the measurement techniques deployed in
the course of this activity are described.
4.1. Plasma Probes
Plasma probes are intrusive measurement devices mounted
onto the moveable platform within the vacuum chamber of a
plasma wind tunnel, to be moved into the plasma plume
during testing. These are used to characterise flow conditions
or to expose material samples to an accordingly characterised
high-enthalpy gas flow in a controlled manner.
Water-cooled, cold-wall copper calorimeter probes were
used to measure a reference heat flux at a given generator
condition and probe position. All plasma wind tunnel test
conditions calibrated for testing the coin-shaped homogeneous
Fig. 1. Schematic of IRS plasma wind tunnel facilities PWK1/4.
Fig. 2. Functional schematic of IRS Emissivity Measurement
Facility EMF.
4
material samples were characterised using a 50 mm diameter
hemispherical enthalpy probe which doubles as a heat flux
probe. The stagnation pressure was measured using a
flat-faced Pitot probe of the same diameter.
The 50 mm material investigation probe constitutes a
flat-faced design previously used e.g. for the investigation of
catalytic material properties, features a radiation-cooled SSiC
casing, and is compatible with the conical material sample
geometry. It is designed towards creating a well-insulated,
near-1D wall heat conduction scenario through the material
sample’s thickness. Fig. 3 depicts the sample support system
for materials in a stagnation point configuration. Radial heat
losses are minimised through a dedicated insulation ring
between the heated sample and radiation-cooled cap, creating
a knife-edge contact, emulating the context of a closed,
thin-walled structure both in a thermal and mechanical sense.
The interior of the probe’s head is fitted with ample insulation
against the water-cooled probe structure. Three ceramic rods
press the material specimen against the SSiC cap via the
insulation ring, with pressure being applied by three springs
manufactured from a high-temperature resistance steel alloy.
The backside temperature of the sample can be monitored
through a miniature linear pyrometer developed at IRS
(Mini-PYREX) by collimating the backside radiation through
an SSiC tube into an optical fibre.
Due to the inherent incompatibility of the COPV segment
with this 50 mm material probe geometry, these samples were
fitted into a cylindrical sample holder of 80 mm diameter,
typically used for testing ablators (see also Fig. 4). This type
of material probe features a water-cooled copper cap
enclosing the material sample at the sides, thus limiting the
material-plasma interactions to the front surface. A small gap
to the concentrically centred sample ensures conductive radial
insulation, with only the aluminium back plate of the
integrated sample providing a conductive interface to the
water-cooled structure, thus also emulating a near-1D-heat
conduction environment. A set of ceramic insulation rings,
against which the back plate is pressed through a fastening
ring, surrounds the sample and ensures that it is flush with the
cap’s front surface.
A combined heat flux / Pitot pressure insert was placed into
the 80 mm sample holder in place of a material sample to
characterise the high-enthalpy air flow environment.
4.2. Temperature Measurements
Front surface temperatures were monitored using an LP3
80/20 linear pyrometer and a LumaSense MCS640
thermographic imaging camera, which are calibrated for
measurement wavelengths around 958.1 nm and 960 nm,
respectively, within which air plasmas are mostly transparent.
These devices are sensitive above approximately 900 K. For
tests on aluminium alloys which exhibit low melting
temperatures, a Maurer TMR95-EA narrowband pyrometer
was deployed, with a measuring wavelength band between 1.5
and 2.6 µm, which is sensitive between black body
temperatures of 420 and 920 K.7)
For the correction of temperature measurements obtained in
the course of the plasma wind tunnel test campaign,
device-specific surface emissivities were determined for each
material both in a pre- and post-oxidised state using the
EMF.9,10) Where such measurements have not been conducted,
assumptions were made based on experiences with similar
materials or literature values.
The design of the 50 mm material probe allowed for
pyrometric measurements at the rear surface of each sample,
using the IRS Mini-PYREX pyrometer. The Mini-PYREX, or
MP3, is sensitive between 0.9 and 1.7 µm. Within reasonable
accuracy, the cavity behind the sample can be assumed to
approximate a black body environment, so that no emissivity
correction needed to be performed once a thermal steady-state
condition was attained. The Mini-PYREX is fitted with a
variety of optical filters to adjust the desired calibration curve,
which was generated using a black body radiation source
available at IRS. Between material tests, the sensitivity of the
setup was verified using a well-characterised radiation source
to identify and account for any potential contamination of the
SiC tube or collimator lens.
As rear surface pyrometry is not possible with the 80 mm
ablator probe setup, the COPV segments were instrumented
with four thermocouples each as described in Chapter 2.
The thermopile pyrometer GIRL CaF2 is sensitive within
bandwidths of 0.4 to 8 µm and 2 to 22 µm, when paired with a
CaF2 optical window in the EMF. As it captures more than
90% of the measurement target’s thermal radiation between
temperatures of 400 to 2100 K in this configuration, it was
deployed to measure total, temperature-dependent emissivities
over the respectively relevant temperature ranges.7)
4.3. Emission Spectroscopy
Optical emission spectroscopy was conducted in the
boundary layer region for the purpose of identifying atomic
and diatomic lines and bands emerging over the course of a
material sample’s exposure to the high-enthalpy air flow.6-8)
Fig. 4. Left: instrumented and assembled PWT sample of COPV
segment. Right: IRS 80 mm ablator probe with combined HF/Pitot
pressure insert.
Fig. 3. Schematic of an IRS 50 mm material probe head.
5
For those tests conducted in PWK1, an Acton SpectraPro
2750 spectrograph was used in conjunction with an Andor
Newton DU920N-OE EMCCD camera. A periscope setup
allows for the monitoring of the spectra along the stagnation
line within approximately 50 mm of the surface at an effective
spectral resolution of 0.115 nm/pixel. Depending on the
configuration, a bandwidth of approximately 120 nm can be
covered at any given time within the range of 200 to 1000 nm.
An Ocean Optics S2000 spectrometer was deployed for
those tests conducted in PWK4 and focused on a measurement
volume approximately 3 mm across at a point along the
stagnation line in immediate vicinity of the sample’s surface.
This setup features a resolution of approximately
0.34 nm/pixel and covers a range of approximately 300 to
880 nm.
Measurements were taken at intervals from the onset of the
test until the termination of the test and exposure times were
adjusted in between exposures as required.
In post-processing, atomic emission lines were identified
using the National Institute of Standards and Technology’s
(NIST) Atomic Spectra Database11) as a primary reference,
whereas diatomic molecular bands were identified based on
relevant available literature.
A more detailed description of the process as well as a
representative selection and discussion of spectra obtained for
SSiC, type 316L stainless steel, grade 5 titanium, aluminium
alloy 7075, and CFRP EX-1515/M55J is presented in Ref. 6).
A computational rebuild of the boundary layer spectra for
each material using the Plasma Radiation Database
(PARADE) is currently ongoing.12)
4.4. Physical Inspection
Before and upon completion of a plasma wind tunnel test,
the thickness and mass of the specimen were measured. The
surfaces of previously exposed specimens were then inspected
using an optical microscope setup. In addition, the Vickers
hardness was tested both near the centre and the rim of the
respective specimens for silicon carbide, type 316L stainless
steel, grade 5 titanium, aluminium alloy 7075, and CFRP
EX-1515/M55J.
5. Test Setup and Conditions
The setup, procedures and conditions for the different types
of experimental investigations conducted in the course of the
Characterisation of Demisable Materials TRP are summarised.
5.1. Emissivity Measurements
Total and device-specific spectral, temperature-dependent
emissivities were measured in the Emissivity Measurement
Facility for materials both in their pristine state and having
previously been subjected to plasma wind tunnel testing.
Emissivity data was measured for temperatures up to a point
near the critical temperature of the respective test candidates,
and obtained for all materials with the exception of CFRP and
CFRP/PEEK.
The emissivity datasets obtained for the materials SSiC,
type 316L stainless steel, grade 5 titanium, and aluminium
alloy 7075, are presented and discussed in Ref. 10).
5.2. Plasma Wind Tunnel Test Conditions
Plasma wind tunnel test conditions were selected based on
analyses of relevant LEO entry trajectories. To account for the
heat-mitigating effects of tumbling, a low heat flux (LHF)
reference condition was defined, assuming a tumble-averaged
heat flux of 260 kW/m² as a representative value for
re-entering spacecraft debris. The corresponding tumble-free
stagnation point reference heat flux was determined at
1400 kW/m² and designated the high heat flux condition
(HHF). A moderate heat flux condition (MHF) was defined at
the approximate logarithmic midpoint between these extremes
at 520 kW/m².3) An additional very low heat flux condition
(VHLF) at 125 kW/m² was added to examine the
decomposition behaviour of particularly quickly melting
materials. An overview of all test conditions used in the
course of the campaign is provided in Table 1.
Steady state thermal response characterisations of the three
material candidates considered least prone to demise were
conducted based on the LHF-1 generator conditions defined
for PWK1, as well initial transient thermal response
investigations on four materials. Due to scheduling issues and
sample pre-heating considerations, the remaining
plasma-wind-tunnel-based testing activities were carried out
Table 1. Plasma wind tunnel test conditions for Characterisation of Demisable Materials TRP.7,8)
Condition
VLHF
LHF-1
LHF-4
LHF-80
MHF
MHF-80
HHF
HHF-80
Facility/Generator
PWK1/RD5
PWK1/RD5
PWK4/RB3
PWK4/RB3
PWK4/RB3
PWK4/RB3
PWK4/RB3
PWK4/RB3
I, A
1420
1420
460
460
460
368
605
800
IMag, A
-
-
40
40
40
40
40
40
U, V
68
68
88
88
88
83
88
93
pamb, Pa
10
10
41
20
41
20
343
188
ṁN2, g s-1
1.53
1.53
5.00
5.00
5.00
5.00
5.00
5.00
ṁO2, g s-1
0.47
0.47
1.52
1.52
1.52
1.52
1.52
1.52
ṁAr, g s-1
0.3
0.3
-
-
-
-
-
-
x, mm
920
550
325
285
205
170
350
173
, kW m-2
125
263
264
-
520
-
1400
-
, kW m-2
125
263
174
-
342
-
923
-
, kW m-2
-
-
-
262
-
362
-
1405
p0, Pa
15
17
192
250
415
503
1900
1824
h, MJ kg-1
17.4
36
10.1
19.2
13.7
18.7
17.3
38.2
Ma
0.73
0.87
1.93
3.32
2.92
4.64
2.10
2.83
6
in PWK4. Accordingly, an additional LHF-4 condition was
calibrated in addition to the MHF and HHF conditions.
The test conditions were calibrated using the hemispherical
50 mm copper calorimeter and Pitot probes to determine the
reference heat flux and stagnation pressure, respectively.
For the tests conducted in PWK4, the corresponding
effective reference heat flux for flat-faced probe geometries is
subject to viscous effects, which can be accounted for via the
correction shown in Eq. (1), assuming an effective radius of
Reff = 2.3 RN for the flat-faced geometry based on Ref. 13),
with
. (1)
Due to the high Knudsen number regime existing in the
low-pressure environment of the VLHF and LHF-1 conditions,
a correction for viscous effects is not conducted here.
The fundamentally different sample and plasma probe
geometry employed for the characterisation of the COPV
segment’s demise behaviour prompted the use of an
equivalent combined calorimeter / Pitot pressure probe.
Accordingly, three additional conditions, labelled LHF-80,
MHF-80 and HHF-80 were calibrated, whereas the MHF-80
test condition is calibrated at a lower reference heat flux as
compared to the original MHF condition.
The local mass-specific enthalpy is approximate using the
semi-empirical relation derived by Marvin and Pope,14) as
shown in Eq. (2):
, (2)
with the fully-catalytic heat flux approximated with
, and the gas-specific constant K =
3.67 ∙ 10-4 W kg Pa-0.5 J-1 m-1.5. The effective radius is Reff = RN,
when based on 50 mm hemispherical probe measurements and
Reff = 2.9 RN for the 80 mm cylindrical geometry.13)
The Mach number for each condition is calculated in
approximation from the respective total and ambient pressures
using either the Bernoulli equation for compressible flows, or
iteratively via the Rayleigh-Pitot equation, depending on the
Mach number regime.15)
5.2. Steady State Thermal Response Investigation
As the materials SSiC, type 316L stainless steel and grade 5
titanium were identified as the most resilient materials, a
steady-state thermal response investigation was conducted for
each in PWK1 based on the VLHF/LHF-1 generator
conditions.16)
Starting at the furthest possible position from the plasma
generator, the sample holder was successively shifted closer to
the generator. After each shift, it would be allowed sufficient
time for a thermal steady state to be attained, after which the
front surface temperature would be recorded. This step was
repeated until the demise of each sample. Assuming
comparatively limited heat losses through the material probe,
the effective net heat flux was calculated from the measured
temperature profiles, employing the Stefan-Boltzmann law
and using emissivities measured for pre- and post-exposure
samples.
An according copper-based calorimetric heat flux profile
was generated for the LHF-1 generator condition, covering the
accessible axial length of the facility.
This data allows for an initial assessment of the basic
respective survivability of the tested materials and provides a
baseline from which to extract quantitative data for effective
relative surface catalycities. The results of the steady state
thermal response investigation are presented and discussed in
Ref. 16).
5.3. Transient Thermal Response Investigation
The transient thermal response investigation comprises the
primary activity of the test campaign reported here. Pristine
samples were typically subjected to three of the test conditions
presented in Table 1 to characterise and quantify their thermal
and auto-destructive response, as well as identifying species
emanating from the respective surface into the boundary layer
through optical emission spectroscopy.
Tests were open-ended, and were typically terminated once
either the exposed sample had visibly begun to demise or
when a steady-state condition had set in, as signified by
constant temperature measurements. In cases where demise
occurred gradually over extended durations, the respective test
was terminated arbitrarily.
First results of the transient thermal response investigation
activities are reported in Refs. 6-8,17).
6. Results and Discussion
Due to the limited scope of this paper, only an overview of
the findings regarding the aerothermal demise behaviour the
respective material types and specific materials can be given.
Where available, related publications providing more detailed
information on relevant activities are referenced.
6.1. High-Temperature Ceramics
High-temperature ceramics fulfil many functions in
spaceflight, forming the basis for radiation-cooled heat shield
materials due to their often extreme thermal durability and
good heat dissipation properties. Another notable application
is also found in space-bound telescope mirror optics due to
their thermomechanical properties, in which case large,
monolithic components are often required.18) Due to the
circumstances of these applications as well as the materials’
thermal properties, it is obvious that such ceramics generally
have an inherently low demisability. As many investigations
of such materials have been conducted in the past concerning
their behaviour in high-enthalpy air flows due to their
relevance for thermal protection systems,18) only one
candidate, silicon carbide, often employed for mirrors and
supporting structures for optical instruments, was selected for
the Characterisation of Demisable Materials TRP as a
representative of this material group.
SSiC was the only ceramic tested in the course of this
activity, and was selected as a candidate for the steady-state
thermal response characterisation activity conducted using the
LHF-1 generator conditions. It was further subjected to
thermal response tests using the MHF and HHF conditions.
Due to its generally heat-resistant characteristics, including
a comparatively low catalycity,16) very high surface
emissivities10) and a high critical temperature, subject to the
7
onset of active oxidation processes, which in turn is
pressure-dependent,19) none of the samples tested for their
transient thermal response demised. It was confirmed that the
material exhibits a very straightforward, easily reproducible
heating behaviour. No species which could be associated with
the material’s surface or related boundary-layer interactions
could be identified at any time for either test.6)
The steady state material response investigation resulted in
a full demise of the specimen for a cold-wall calorimetric
reference heat flux of approximately 1500 kW/m² at a
stagnation pressure around 60 Pa, succumbing to active
oxidation. 16) It must be noted that the low pressure
encouraging the onset of active oxidation is not representative
for the entirety of a destructive entry trajectory.
As was expected, SSiC displayed next to not tendency to
demise at any practical rate when subjected to conditions
relevant for uncontrolled entries from LEO.
6.2. Metallic Alloys
As in space applications metallic alloys are primarily
employed for structures with no primary use as heat shield
materials, few experimental investigations concerning their
specific behaviour in re-entry environments have been
conducted in the past. Accordingly, exploring the demise
behaviour of metallic alloys constitutes one of the prime
interests of this activity.
It was found that many predictive tools may have
underestimated the enhancing effect of surface degradation on
the alloys’ respective surface emissivities and thus their
ability to dissipate the incoming heat flux. Comparatively high
emissivities were measured for metallic specimens which had
previously subjected to high-enthalpy air flows in a plasma
wind tunnel.10)
Whereas different alloys vary in their respective melting
temperatures, this was initially projected to constitute the
critical temperature above which aerothermal demise would
occur without delay. However, the experimental activities
reported here show that the formation and retention of oxide
and/or nitride layers, while naturally subject to shear forces,
may effectively prolong an object’s demise. Such layers allow
the surface to be superheated to a temperature level
significantly beyond the melting temperature of the pristine
material, while simultaneously providing some limited
additional thermal insulation for the bulk material and/or
mechanically upholding the specimen’s structural framework.
This effectively delays or even prevents a molten phase from
physically dissipating within a realistic timeframe for an
uncontrolled atmospheric entry.7)
As the examination of the results obtained for the
aluminium alloy 7075 and aluminium-lithium alloy 2099 has
shown, the occurrence of demise-prolonging oxidation and/or
nitration is apparent, yet hardly relevant for materials with
very low melting temperatures. While a thin solid oxide layer
typically serves to retain a molten phase for some limited time,
the delay is typically in the order of seconds and ultimately
suffers mechanical failure (see also Fig. 5).7,8)
However, the formation of oxide/nitride layers is more
significant for stainless steel and titanium. In the case of type
316L stainless steel, demise was only observed for the HHF
condition at front surface temperatures exceeding the liquidus
of the alloy and with a delay of approximately 180 s following
the onset of the test. Visual and thermographic recordings
depict the formation of a comparatively thick and sturdy
oxide/nitride layer, which is eventually penetrated by the
formation of an expanding liquid hot spot, before the general
structural integrity of the sample is eventually broken (see
Fig. 6).
None of the tested grade 5 titanium specimens suffered
demise, despite of those tested at the HHF condition being
superheated, with back surface temperatures remaining below
the solidus at a comparatively high temperature difference to
the front surface. This indicates some thermal blockage
mechanism that is possibly related to the formation and
persistence of the oxide and/or nitride layer at the front
surface. In particular, vanadium(V)-oxide was observed,
covering the front surface of the samples exposed to high
heating rates and manifesting itself as a thin liquid film (see
Fig. 6).6,7)
Temperature histories measured for stainless steel and
titanium show that heating of the front surface is delayed once
an elevated temperature regime is attained, whereas at the
same time the thermal response of the specimen’s back face
appears erratic in nature. This is particularly the case for the
titanium alloy, where the back surface temperature goes
through multiple cycles of increase and decline before finally
Fig. 5. Left: aluminium alloy 7075 sample melting under MHF
condition in PWK4. Right: aluminium-lithium alloy 2099 melting
under HHF test condition in PWK4.
Fig. 6. Left: type 316L stainless steel melting under HHF test
condition in PWK4. Right: grade 5 titanium Ti-6Al-4V sample with
liquid Vanadium(V) oxide film under HHF test condition in PWK4.
8
stabilising. For stainless steel, merely one such cycle is
observed.
It is believed that the internal thermal dynamics indicated
by this behaviour are connected to the latent heats and
changes in effective thermophysical and optical properties
associated with phase transitions within the bulk material,
such as the beta phase transition for titanium. Whereas the
observable onset of such a transition occurs with a delay
especially at higher heating rates, an examination of the
heating history of the test of titanium conducted at the LHF-1
and MHF conditions reveals a correlating effect on
respectively the front and back surface temperature as the beta
transus at around 1253 K is locally surpassed. Further, results
from the emissivity measurement activity provide strong
evidence of a sudden and significant drop in surface
emissivity attributable to the beta phase transformation.10)
A further, likely complementary cause for the erratic
temperature histories may be found in the formation process
of oxide/nitride layers on the front surface, which is believed
to occur mainly during the primary heating phase. The results
of the optical emission spectroscopy measurements
accordingly indicate a definite increase in boundary layer
activity related to material-specific atomic and diatomic
species during the phase at which front surface heating is
delayed and a steady-state has not yet set in. Naturally, this
activity increases and changes for samples undergoing melt,
however in cases where a steady state or quasi-steady state is
attained, the occurrence and intensity of atomic and molecular
emission lines is heavily reduced.6)
Both type 316L stainless steel and grade 5 titanium were
subjected to steady-state thermal response investigations
under the generator conditions corresponding to VLHF and
LHF-1. Due to the limited availability of specimens, the
responses at different positions were mostly tested with a very
limited number of samples. The resulting variation in surface
morphologies and thermal responses between repeat
measurements indicate a strong dependency of the material’s
behaviour to the preceding heating history. However, the
results were sufficiently consistent to illustrate relative
differences with regards to the respective catalytic efficiencies
at different conditions. It was determined that while stainless
exhibits a consistently higher catalycity than both SSiC and
titanium under the given test conditions, net resulting heat
fluxes extracted for the titanium alloy were, at higher heating
rates effected at positions in relatively close proximity to the
plasma generator, significantly lower than those measured for
SSiC.16)
6.3. Laminate Composites
As an alternative to metal, laminate composites are of
primary interest as structural materials for many spaceflight
applications, providing high mechanical strength at a
comparatively low mass.
The results of the experimental campaign depict a
behaviour which is roughly similar to that found in typical
charring ablators when exposed to aerothermal heating, in that
volume ablation effects precede and/or coincide with surface
ablation phenomena, depending on the heating conditions (see
also Ref. 19)), as would be expected for any material based on
organic compounds. The 2D laminate reinforcement also
generally features a very low heat conduction through any
exposed specimen’s thickness, which further adds to the
overall impression that these materials essentially behave like
ablative heat shields for all intents and purposes.
Due to the lack of isotropic reinforcement, the relevance of
this durability may be subject to macroscopic material failure.
Delamination may occur as an effect of shear forces
associated with high stagnation pressures, essentially breaking
down the parent object into more demisable items or into
fragments with a sufficiently low terminal kinetic energy as to
no longer present an on-ground casualty risk. However, the
structural context of components made of such materials is
often closed (see also Section 6.4), meaning that few or no
exposed edges exist to provide vulnerabilities to
aeromechanical shearing. By suppressing lateral
plasma-material interactions and providing mechanical
constriction via the insulation ring, the 50 mm material probe
emulates such a closed structural context and limits the
occurrence of delamination events,
Experiments on CFRP EX-1515/M55J at the LHF-1
condition have resulted in a near-steady state being attained.
Whereas swelling and outgassing of the specimen were
observed, no onset of surface ablation was registered, as the
surface temperature apparently remained below the threshold
of combustion. At the MHF and HHF conditions, however, a
low number of isolated delamination events were observed as
well as surface ablation, gradually reducing the thickness of
the exposed specimens, which was further reflected by a
bilateral convergence of the front and back surface
temperatures up to the point of perforation. Surface ablation
rates were directly correlated with the respectively effected
heating rate.7)
Tested at the LHF-4, MHF and HHF conditions, the
CFRP/PEEK specimens were subject to severe initial swelling
of up to 400% of their initial thickness, combined with
thermoplastic deformation on account of the PEEK matrix.
Because of the swelling well beyond the sample’s front
surface, the flow field was likely deformed, effecting locally
increased heat fluxes. Accordingly, surface ablation was
observed for all samples following their initial swelling,
Fig. 7. Left: CFRP EX-1515/M55J specimen ablating under HHF
condition. Right: CFRP/PEEK specimen ablating under HHF
condition.
9
resulting recession rates correlating with the reference heat
flux (see also Fig. 7).8)
Due to the hybrid nature of its composition, the demise
behaviour of tested Glass Laminate Aluminium-Reinforced
Epoxy (GLARE) displayed a combination of different demise
behaviours such as outgassing, delamination and melt,
respectively characteristic for laminate composites and metals.
Whereas the temperature increase was comparatively smooth
during the test conducted at the LHF-4 condition, the
dynamics of destruction became more violent at higher
heating rates, for which both the visual recordings as well as
periodically alternating front surface temperatures indicate a
successive layer-by-layer heating and subsequent removal of
material (see also Fig. 8).8)
Emissivity measurements could not be conducted to
completion for any of the three investigated composites. An
extremely fast cooling rate of pre-charred CFRP
EX-1515/M55J during the positional shift in the EMF resulted
in correspondingly large inaccuracies, which increased
successively at higher temperatures. Because of thermoplastic
deformation, no emissivities could be extracted for
CFRP/PEEK. While GLARE could not be subjected to
relevant temperature levels beyond the melting point of its
aluminium component for extended durations, its front surface
further exhibited modifications from charring within the
protective argon atmosphere of the EMF which did not
correlate to the rapidly changing surface morphology
observed during exposure to an air plasma plume. Based on
the available data, previous experience, as well as various
literature sources, the surface emissivity of CFRP
EX-1515/M55J and CFRP/PEEK were assumed as 85%,
whereas that of GLARE is tentatively estimated at 30%.8)
The evaluated optical emission spectra measured for the
laminate materials were generally dominated by organic
diatomic bands, including the C2 Swan bands, CH, CN and
NH. In some cases, a sodium line was observed around 589
nm as the only atomic line not stemming from the plasma
plume.6)
6.4. He COPV segment
The COPV segment samples were tested in PWK4, using
the 80 mm diameter ablator material probe under the
accordingly calibrated conditions LHF-80, MHF-80, and
HHF-80.8)
Whereas the tests at the LHF-80 and MHF-80 conditions
were restricted to a duration of 600 s each, the final test was
terminated only after all four thermocouples had failed and the
8 mm layer of CFRP had been ablated to reveal the underlying
Ti-6Al-4V tank liner.
The initially exposed layers of CFRP ablated in much the
same manner as was seen for CFRP EX-1515/M55J at the
MHF and HHF conditions, with the surface recession rate
correlating primarily with the effected heat flux and resulting
surface temperature. However, significantly more
delamination was observed in the initial phases of each test,
possibly as a result of the slight curvature of the tank segment
not being flush with the front surface of the sample holder cap,
revealing exposed cutting edges, where shear forces would
take effect. Fewer delamination events were observed as the
surface respectively retreated into the sample holder cap.
Unlike for the respective tests conducted for the
homogeneous Ti-6Al-4V samples at the HHF condition, the
titanium liner melted once the enclosing CFRP layers were
ablated during the test conducted at the HHF-80 condition,
which is a result of the different geometric reference heat flux
probe geometries with which the conditions were respectively
characterised. The penetration of the CFRP layers was
reflected by a visibly altered boundary layer activity (see also
Fig. 9. Left to right: COPV segment sample subjected top HHF-80 test condition after 20s, 240s, 650s and shortly after the generator shutdown.
Fig. 8. Clockwise from top left: virgin GLARE specimen, GLARE
decomposing under HHF condition in PWK, post-test specimen of
GLARE.
10
Fig. 9). Until that point, OES measurements show largely
identical organic species emission bands as for the
homogeneous CFRP variants.6,8)
7. Conclusion
An overview of the experimental activities conducted at the
Institute of Space Systems (IRS) at the University of Stuttgart
in the context of the ESA TRP Characterisation of Demisable
Materials was presented.
The sample types examined were briefly described and
grouped into four categories, which encompass
high-temperature ceramics, metallic alloys, laminate
composites and Composite Overwrapped Pressure Vessel
(COPV) segments.
The experimental activities encompassed measurements of
the steady state and transient thermal responses of the
candidate materials subjected to different heating rates in
simulated uncontrolled Earth entries from Low Earth Orbit
(LEO).
The plasma wind tunnels in which the experiments were
conducted were described, as well as the IRS Emissivity
Measurement Facility (EMF), which was used to extract
temperature-dependent spectral and total emissivities of
pristine and pre-oxidised samples to perform emissivity
corrections of pyrometric measurements and for modelling
purposes.
This overview further describes the contact-less and in-situ
measurement techniques and procedures deployed, as well as
the respectively effected test conditions.
Due to the limited scope of this paper, only the tentative
conclusions of the result analyses could be presented for the
different sample types. While some of the activities have
previously been reported upon in detail, the dissemination of
other detailed results is as of yet pending.6-8,10,16,17)
Acknowledgments
The experiments reported and discussed were performed at
the Institute of Space Systems at the University of Stuttgart in
the state of Baden-Württemberg, Germany. The authors would
like to gratefully acknowledge funding of these research
efforts by the European Space Agency (ESA) under contract
no. 4000109981/13/NL/CP.
References
1)
Kelley, R.L.: Using the Design for Demise Philosophy to Reduce
Casualty Risk due to Reentering Spacecraft. 63rd International
Astronautical Congress (2012).
2)
Lemmens, S., Funke, Q. and & Krag, H.: On-ground Casualty Risk
Reduction by Structural Design for Demise. Adv Space Res, 55
(2015), 2592-2606.
3)
Smith, A.J. and Merrifield, J.A.: TP2.1 Test Objectives and
Matrices for the Identification of Key Mechanisms for Material
Demise (2014), Fluid Gravity Engineering Ltd., CR085/14,
Emsworth, United Kingdom.
4)
Herdrich, G., Fertig, M. and Löhle, S.: Experimental Simulation of
High-Enthalpy Planetary Entries, TOPPJ, 2 (2009), 150-164.
5)
Löhle, S., Fasoulas, S., Herdrich, G., Hermann, T.,
Massuti-Ballester, B., Meindl, A., Pagan, A. S. and Zander, F.: The
Plasma Wind Tunnels at the Institute of Space Systems: Current
Status and Challenges, 46th AIAA Thermophysics Conference
(2016), Washington D.C., USA.
6)
Pagan, A.S., Massuti-Ballester, B., Herdrich, G., Merrifield, J.A.,
Beck, J.C., Liedtke, V. and Bonvoisin, B.: Investigation of the
Surface and Boundary Layer Composition for Demising Aerospace
Materials. 7th International Workshop on Radiation of High
Temperature Gases in Atmospheric Entry (2016), Stuttgart,
Germany.
7)
Pagan, A.S, Massuti-Ballester, B., & Herdrich, G.:
Characterisation of Demisable Materials TR2.1: Demisability
Testing to Characterise Material Behaviour due to Plasma
Exposure. Technical Report IRS-15-P10 (2015). Institute of Space
Systems, University of Stuttgart, Germany.
8)
Pagan, A.S, Massuti-Ballester, B., & Herdrich, G.:
Characterisation of Demisable Materials TR3.1: Demisability
Testing to Characterise Material Behaviour due to Plasma
Exposure – Phase 2. Technical Report IRS-17-P01 (2017). Institute
of Space Systems, University of Stuttgart, Germany.
9)
Schüßler M., Auweter-Kurtz M., Herdrich G. and Lein S.: Surface
Characterization of Metallic and Ceramic TPS-materials for
Reusable Space Vehicles. ACTA ASTRONAUT, 65 (2009),
676-686.
10)
Pagan, A.S., Massuti-Ballester, B. and Herdrich, G.: Total and
Spectral Emissivities of Demising Aerospace Materials.
FRONTIER OF APPLIED PLASMA TECHNOLOGY, 9 (2016),
7-12.
11)
Kramida, A., Ralchenko, Y., Reader, J. and NIST ASD Team:
NIST Atomic Spectra Database (version 5.4). Online at
http://physics.nist.gov/asd, (accessed March 2017). National
Institute of Standards and Technology, Gaithersburg, MD, USA.
12)
Liebhart, H., Wernitz, R., Herdrich, G., Fasoulas, S., Röser, H.-P.,
Merrifield, J. and Beck, J.: Advances for Radiation Modelling for
Earth Re-entry in PARADE: Application to the Stardust
Atmospheric Entry, 43rd AIAA Thermophysics Conference (2012),
New Orleans, LA, USA.
13)
ASTM E637-05: Standard Test Method for Calculation of
Stagnation Enthalpy and Experimental Measurements of
Stagnation-Point Heat Transfer and Pressure, (2005).
14)
Marvin, J. G. and Pope, R. B.: Laminar Convective Heating and
Ablation in the Mars atmosphere, AIAA J., 5 (1967), 240-248.
15)
Kolesnikov, A.F.: Combined Measurements and Computations of
High Enthalpy and Plasma Flows for Determination of TPM
Surface Catalycity, RTO AVT/VKI Special Course on
“Measurement Techniques for High Enthalpy Plasma Flows”, von
Karman Institute for Fluid Dynamics, RTO EN-1 (1999),
Rhode-Saint-Genèse, Belgium.
16)
Pagan, A.S., Massuti-Ballester, B., Herdrich, G., Merrifield, J.A.,
Beck, J.C., Liedtke, V. and Bonvoisin, B.: Experimental
Demisability Investigation of Common Spaceflight Materials, 30th
International Symposium on Space Technology and Science (2015),
Kobe, Japan.
17)
Pagan, A.S., Massuti-Ballester, B. and Herdrich, G.: Experimental
Thermal Response and Demisability Investigations on five
Aerospace Structure Materials under Simulated Destructive
Re-Entry Conditions, 46th AIAA Thermophysics Conference (2016),
Washington D.C., USA.
18)
Toulemont, Y., Passvogel, T., Pibratt, G.L., de Chambure, D.,
Pierot, D. and Castel, D.: The 3.5-m all-SiC telescope for
HERSCHEL. Proc. of SPIE, 5487 (2004), 1119-1128.
19)
Herdrich, G., Auweter-Kurtz, M., Fertig, M., Löhle, S., Pidan, S.
and Laux, T.: Oxidation Behaviour of SiC-based Thermal
Protection System Materials using newly developed Probe
Techniques, J SPACECRAFT ROCKETS, 42 (2005), 817-824.
20)
Pagan, A.S., Zuber, C., Massuti-Ballester, B., Herdrich, G., Hald,
H. and Fasoulas, S.: The Ablation Performance and Dynamics of
the Heat Shield Material ZURAM®, 31st International Symposium
on Space Technology and Science (2017), Matsuyama, Japan.
