Article

Experimental Characterization of C–C Composite Destruction

Under Impact of High Thermal Flux in Atmosphere and
Hypersonic Airflow
Ryan Bencivengo 1 , Alin Ilie Stoica 1 , Sergey B. Leonov 1, * and Richard Gulotty 2

1 Department of Aerospace and Mechanical Engineering, University of Notre Dame, Notre Dame, IN 46556,
USA; rbencive@nd.edu (R.B.); astoica@nd.edu (A.I.S.)
2 Honeywell International Inc., 50 E Algonquin Rd, Des Plaines, IL 60017, USA; richard.gulotty@honeywell.com
* Correspondence: sleonov@nd.edu

Abstract: Hypersonic flight in the atmosphere is associated with high thermal flux impact-
ing the vehicle surface. The nose, leading edges, and some elements of the engine typically
require the implementation of highly refractory materials or an active thermal protection
system to maintain structural stability during the vehicle mission. Carbon–carbon (C–C)
composites are commonly considered for the application thanks to their unique thermal
and mechanical properties. However, C–C composites’ ablation and oxidation under long
cruise flights at high speeds (Mach number > 5) are the limiting factors for their application.
In this paper, the results of an experimental study of C–C composite thermal ablation and
oxidation with test article surface temperatures up to 2000 K are presented. The tests were
performed under atmospheric conditions and hypersonic flow in the ND_ArcJet facility at
the University of Notre Dame. The test articles were preheated with CW laser radiation
and then exposed to M = 6 flow at stagnation pressures up to 14 bar. It was found that C–C
composite oxidation and mechanical erosion rates are significantly increased in hypersonic
airflow compared to those at ambient conditions and nitrogen M = 6 flow. Compared to
atmospheric air, mass loss occurred at a rate of 1.5 orders of magnitude faster for M = 6
airflow. During high-speed flow conditions, rapid chemical oxidation and the mechanical
Academic Editors: Ikhyun Kim and destruction of weakened C-fibers likely cause the accelerated degradation of C–C composite
Yosheph Yang material. In this study, a post-mortem microscopic analysis of the morphology of the C–C
Received: 30 November 2024 surface is used to explain the physical processes of the material destruction.
Revised: 10 January 2025
Accepted: 10 January 2025 Keywords: carbon–carbon (C–C) composites; hypersonic airflow; mechanical destruction;
Published: 11 January 2025 thermal ablation; oxidation
Citation: Bencivengo, R.; Stoica, A.I.;
Leonov, S.B.; Gulotty, R. Experimental
Characterization of C–C Composite
Destruction Under Impact of High 1. Introduction
Thermal Flux in Atmosphere and
Hypersonic Airflow. Aerospace 2025,
Vehicles that re-enter Earth’s atmosphere or vehicles that fly at hypersonic velocities in
12, 43. https://doi.org/10.3390/ the atmosphere experience high thermal and mechanical loads as a result of the air compres-
aerospace12010043 sion (stagnation temperature T0 > 2000 K and stagnation pressure P0 > 3 MPa). To overcome
Copyright: © 2025 by the authors.
this problem, highly refractory materials or active thermal protection systems are utilized
Licensee MDPI, Basel, Switzerland. for the construction of hypersonic vehicles. Thermal protection systems are typically heavy,
This article is an open access article expensive, and time-consuming to design. In many cases, the application of refractory
distributed under the terms and materials is more practical, but is challenging because the properties of the materials dictate
conditions of the Creative Commons
the operating parameters of the vehicle. C–C composites are a class of prospective materials
Attribution (CC BY) license
for hypersonic vehicle construction. The current study provides insight into the physical
(https://creativecommons.org/
licenses/by/4.0/).

Aerospace 2025, 12, 43 https://doi.org/10.3390/aerospace12010043

Aerospace 2025, 12, 43 2 of 20

mechanisms of C–C composites destruction under harsh environmental conditions relevant

to hypersonic flight.
The design of vehicles capable of flying in the hypersonic regime has been under
extensive development for many decades [1]. A key problem in the advance of hypersonic
vehicles is managing the intense thermal heat flux and temperature spikes that occur
because of air compression over the vehicle surface [2,3]. Additionally, for vehicles utilizing
air breathing propulsion systems, strict limitations are applied to components within the
engine as high thermal loads are encountered. This requires the implementation of systems
designed for extreme heat flux protection. Most affected by high heat fluxes are the nose
tip and the leading edges, which oftentimes have quite small surface areas compared to
the vehicle total surface. Due to these constraints, the ability for active cooling of the
leading edges becomes challenging [4]. A potential solution to this problem is the use of
refractory materials possessing high temperature melting/ablation points and high thermal
conductivities, which would allow the material to spread and dissipate heat efficiently [5,6].
The latter design challenge requires that certain components of the vehicle, such as the
leading edge or heat shield, be constructed of materials that can withstand extremely
high temperatures of 2000 K or greater [7,8] without melting while also maintaining high
mechanical strength for the given flight conditions.
Conventional aerospace materials, such as most metal alloys or their derivatives,
have lower melting points than what is required for hypersonic flight or are prohibitively
heavy. Ultra-high temperature ceramics may be an alternative thanks to their refractory
properties [6,9], but they are typically brittle and difficult to machine efficiently. Also,
surface cracks (caused by thermal shocks) may occur after installation or during flight.
Therefore, though the most conventional materials may provide ready-made technology for
this application, they may not fulfill all the requirements, such as efficient withstanding of
high heat flux going into the vehicle, which may cause serious problems for extended flights.
Other types of materials, such as C–C composites, have been in research and devel-
opment for many years for various applications in the aeronautics industry, including
hypersonic flight [10]. However, limited research has been performed to study the be-
havior of these materials at high time-variable thermal fluxes, when surface temperatures
exceed 2000 K, not to mention 2000 K [11] under high-speed flow. Several computational
efforts have been performed to study the oxidation rate and mechanisms of carbon–carbon
composite destruction under hypersonic flight conditions [12]. A series of studies were
published related to C–C composite nozzle material recession [13–15]. A few experimental
studies have been completed and published regarding the C–C composite behavior at high
temperature, including oxidation, which can be seen in references [16–18]. The use of C–C
composites for hypersonic flight becomes of high interest as this material had showed
highly promising thermal properties combined with relatively strong mechanical charac-
teristics [19] such as: low density (1.60–1.98 g/cm3 ), low coefficient of thermal expansion
(−0.85 to 1.1 × 10−6 1/K), high modulus of elasticity (200 GPa), relatively high thermal con-
ductivity (4–35 W/(m-K)), and the retaining of mechanical properties at temperatures up
to 2300 K in inert environments. The most commonly considered engineering applications
include the implementation of C–C composites in the construction of hypersonic vehicles
leading edges [7,19], air inlet cowl leaps, nozzles, fins, and other surfaces experiencing high
thermal loads.
In the hypersonic flow regime, the main sources of heat transfer/exchange for materi-
als at high thermal loads are axial and radial thermal conduction, gas convection, radiation,
ablation, and chemical reactions. The emissivity of the C–C composite is generally high,
close to 0.86 at 1990 K, which makes it efficient at removing heat when considering radi-
Aerospace 2025, 12, 43 3 of 20

ation [20]. The C–C composite materials have multiple ways of removing heat from the
system at the cost of replacement after each flight.
Ablation, which is to be determined in this work, is also a significant contributor
to the heat absorbance at Tsurface > 2500 K [21], or even lower temperatures as it is hy-
pothesized in this work. The most significant issue facing the implementation of C–C
composites in thermal protection systems is oxidation in a chemically active environment,
such as atmospheric air at high temperatures. For cruise flights, oxidation is the major
limiting factor. With increasingly more extreme hypersonic environments, two protection
approaches are being developed: (1) deposition of high-temperature protective coatings,
and (2) modification of the carbon–carbon matrix. An example of C–C matrix modification
would be the addition of HfB2 to improve oxidation resistance in air. Ablation processes
may also be reduced at low heat fluxes such as 2.4 MW/m2 or lower. However, cracking of
the protective layer is a limiting factor at heat fluxes greater than 4.2 MW/m2 [22].
Since flight testing is highly time- and funds-consuming, and simulations of complex
composition and shapes are challenging and inaccurate without extensive validation,
intensive and robust experimental ground testing is of high demand for the collection of
reliable data and to properly validate the results of computational modeling. In the case of
ground testing, there is still limited availability of measured data for the C–C composite
in hypersonic environments, which may be reproduced to obtain consistent databases.
Note, that in general, there is no one single method of ground testing which can properly
simulate the hypersonic environment in all relevant aspects [23].
In environments such as air with high heat flux and temperature, mass ablation rates
rapidly increase with increasing exposure time while the linear erosion rate demonstrates a
bilinear response decrease after an initial rise [24]. This is due to the machining process
of the articles causing fiber fragmentation, which weakens the fiber and matrix interfaces,
resulting in an initial surge in mass loss due to ablation for the first test run of a new
article. Furthermore, long fibers have better resistance to the shear forces generated by
flows like plasma streams [24]. The oxidation process of the carbon fibers leads to a pointy
morphology. The interface between the fiber and the matrix oxidizes at a much higher
rate when compared to the core of the fiber. For long exposure times, at temperatures
lower than 1100 K the chemical reaction rate is the governing factor since the oxidation is
uniform throughout the articles. However, for higher temperatures, greater than 1400 K,
the oxidation of the C–C composite occurs prevalently along the edges. This shows that gas
phase diffusion significantly contributes to the oxidation process. Under a microscopic view,
there are oxidation voids in the articles exposed to high temperatures either in a furnace
or arc jet flow. However, these oxidation voids are distinguishable from imperfections
due to machining or processing because oxidation voids are only present on pointy fibers
with no surrounding matrix, and on delineated fibers along the edge [25]. These results
were observed under electron microscopy scanning for raw C–C articles. Adjacent to
the fibers, there are areas of increased local stress in the matrix, which crack gradually
under oxidation, allowing oxygen diffusion into the matrix/fibers [26]. Furthermore, for
the C–C composites made via the Chemical Vapor Infiltration (CVI) process, the material
structure consists of areas with increased porosity and noticeable density gradients in
the C–C material volume causing the recession rate to be dependent on the porosity and
density of the raw material [27].
The objective of this study is to characterize and quantify the process of C–C composite
destruction under high-temperature conditions, including the mass loss as the result of
thermal ablation and oxidation in atmosphere and at hypersonic flow exposure. The paper
poses the detailed experimental results relevant to the destruction of a C–C composite under
various conditions. Destruction is assessed by the mass loss and the surface morphology
Aerospace 2025, 12, x FOR PEER REVIEW 4 of 21

Aerospace 2025, 12, 43 4 of 20

morphology modification. Most of the analysis was focused on the material destruction
under M = 6 hypersonic flow at a controlled temperature of the test article surface above
1000 K. Both the
modification. physical
Most of themechanisms
analysis wasoffocused
the test-article material destruction
on the material recession and a quantifi-
under M=6
cation of theflow
hypersonic ratesatofa recession
controlledwere analyzed.ofAthe
temperature post-mortem microscopic
test article surface aboveanalysis
1000 K.ofBoth
the
morphology
the of the C-C surface
physical mechanisms is used to explain
of the test-article materialthe physicaland
recession processes of the material
a quantification of the
destruction.
rates of recession were analyzed. A post-mortem microscopic analysis of the morphology
of the C–C surface is used to explain the physical processes of the material destruction.
2. Materials and Methods
2.
2.1.Materials and
Heat Transfer Methods
Mechanisms
2.1. Heat Transfer Mechanisms
The ablation and oxidation processes of the test article exposed at hypersonic flight
The ablation
conditions and oxidation
are of major interest. Toprocesses
provideof the test article
consistent exposed
and reliable at hypersonic
testing, flight
the test article,
conditions are of major interest. To provide consistent and reliable testing,
in the form of a blunt body, is placed in three environmental conditions: (1) atmospheric the test article,
in the formwith
conditions of a blunt body, isstill
surrounding placed in three
air and laserenvironmental
exposure centeredconditions: (1) atmospheric
to the front face of the
conditions with surrounding still air and laser exposure centered to the
test article, (2) vacuum conditions with laser exposure, and (3) M = 6 flow conditions front face of thewith
test
article, (2) vacuum
laser exposure. conditions
Laser exposure withwas laser
set exposure,
at different and (3) Mratings
power = 6 flowtoconditions
generate up withto laser
10.6
exposure. Laser exposure was set at different power ratings to generate up to 10.6 MW/m 2
MW/m for 360 s of exposure time and was consistent within each test. The beam diameter
2

for 360high-power
of the s of exposure timeisand
laser 8 mm wasand consistent within each
can be expanded test. needed.
when The beam diameter
The of the
heat transfer
high-power laser is 8 mm and can be expanded when needed. The heat
problem is displayed in Figure 1 for each test type. This procedure is mainly focused on transfer problem is
displayed
providing in Figureon
insights 1 for each
mass losstest type. This
dynamics withprocedure
respect toismeasured
mainly focused
surfaceon andproviding
internal
insights on mass loss dynamics with respect to
temperature and determining ablation and oxidation rates. measured surface and internal temperature
and determining ablation and oxidation rates.

Figure 1. Schematics of the test article thermal balance under hypersonic M = 6 air flow. For
atmospheric testing conditions,
Figure 1. Schematics convection
of the test article would
thermal coolunder
balance downhypersonic
the article,Mwhereas, under
= 6 air flow. Forvacuum
atmos-
conditions, convection and oxidation do not occur.
pheric testing conditions, convection would cool down the article, whereas, under vacuum condi-
tions, convection and oxidation do not occur.
The means of heat transfer shown in Figure 1 illustrate how the article is heated and
coolsThe
down. A control
means volume shown
of heat transfer drawn in around
Figurethe article describes
1 illustrate how the the general
article thermal
is heated and
balance equation for the test article in airflow, as shown in Figure 1 (the sign
cools down. A control volume drawn around the article describes the general thermal of the heat
flux q depends
balance on for
equation thethe
gastest
andarticle
surfaceintemperatures, see below)
airflow, as shown is provided
in Figure in Equation
1 (the sign (1).
of the heat
flux q depends on the gas and surface temperatures, see below) is provided in Equation ∂T
(1). q projected laser + qconvection + qoxidation − qradiation − q ablation − qconduction = ρc p ∂t (1)

𝜕𝑇
𝑞 +In𝑞Equation+(1), 𝑞 the positive
− 𝑞 terms − 𝑞 to heat
refer − 𝑞entering the = 𝜌𝑐
control volume, while (1)
𝜕𝑡
the negative terms provide heat exiting the control volume. The heating of the article is
In Equation
provided (1), the positive terms
by the continuous-wave refer projected
laser beam to heat entering the control
on the front face of volume, while
the article, the
the negative terms provide heat exiting the control volume. The heating of
convective heat transfer is realized with the hypersonic flow within the compression and the article is
provided by
boundary the and
layer, continuous-wave
conduction is laser beam
through theprojected on the
article with front face
anisotropic of the article,
thermal the
properties
convective heat transfer is realized with the hypersonic flow within the compression
(varying thermal conductivity axially and radially). The hot surface of the article will and
boundary
radiate layer,
heat and
out of theconduction is through
control volume. the article
The ablation ofwith anisotropic
the C–C materialthermal
providesproperties
a means
(varying thermal conductivity axially and radially). The hot surface of the
of cooling, while oxidation adds more heat to the article. The conductive, convective, article willand
ra-
diate heatheat
radiative out of the control
transfers can bevolume. The in
described ablation of the C-C
a conventional material
way, provides
see below a means
for the basic
of cooling, while oxidation adds more heat to the article. The conductive,
equations; however, the ablation and oxidation require further attention. convective, and
The projected heat flux coming from the laser and going into the control volume is
simply calculated by using the power rating of the laser equipment used for the run divided
by the surface area of the beam projection on the article front side. The total heat flux that is
Aerospace 2025, 12, 43 5 of 20

projected onto the article would be at its maximum 10.6 MW/m2 . Taking into account 8%
light reflection from the optical window and about 96% of the light absorption by the test
article surface, actual total heat flux is about 9.4 MW/m2 at the laser maximum power. Note
that for a real cruise flight, the calculated convective heating would be up to 50 MW/m2
and a temperature of 2400 K at the surface of the leading edge of the vehicle travelling at
M = 6 cruise speeds. While the test article is being exposed to the laser radiation, thermal
conduction occurs throughout the article in the radial, angular, and axial directions, (r,ϕ,z).
C–C materials have anisotropic thermal properties, and thermal conductivity can then be
modeled as dependent on the direction, k(r,ϕ,z), and an equation for thermal conductivity
can be written in cylindrical coordinates as shown in Equation (2a), which, in the one-
dimensional case, could be simplified to Equation (2b). Convective heat transfer from high
enthalpy hypersonic airflow can be estimated by Equation (3a), where for hypersonic flow
the free stream enthalpy is h∞ >> hw and therefore the enthalpy ratio, hw /h∞ , approaches
0, ρ is the density of gas at the stagnation point, and Rn = d/2 is the leading-edge radius
and d is the diameter of the blunt body. This shows the need of either high-enthalpy flow
(arcjet, for example) or laser preheating of the article to replicate the thermal conditions of
the hypersonic flight. Under the current conditions with laser preheating, the processes of
oxidation, ablation, and thermal shocks are simulated by a cold hypersonic flow impinging
the test article for a short duration when the surface temperature is high and relatively
stable. The forced convective heat transfer between the preheated article and the hypersonic
flow can be estimated by Equation (3b), where h is the convective heat transfer coefficient,
which is a function of Reynolds, Re, and Prandtl, Pr, numbers (used to calculate Nusselt
number, Nu). For the current test conditions (estimated values Red ≈ 106 ; Pr = 0.715 for
air; Nud ≈ 600; h ≈ 2.4 × 103 W/m2 K), the negative convective heat flux is estimated
in a range of qconvection = −2 MW/m2 to −4 MW/m2 that is less in magnitude than the
laser heating thermal flux. Radiative cooling of the article is a main source for heat transfer
under high temperature conditions. For any given point on the surface of the article and
neglecting the roughness and the shrinking of the article, the radiative cooling can be
described by Equation (4), where σ is the Stefan–Boltzmann constant and ε is the emissivity
of the test article surface.
     
1 ∂T ∂T 1 ∂ ∂T ∂ ∂T ∂T
rk(r ) + 2 k(ϕ) + k(z) = ρc p , (2a)
r ∂t ∂r r ∂ϕ ∂ϕ ∂z ∂z ∂t

dT
qconduction = −k (2b)
dz
 1  
ρ 2 hw
qconvection = 1.63 × 10−4 1− (3a)
Rn h∞
  k
qconvection = −h· Tsur f − T f low , h = Nud · , Nud = 2 + 0.4· Re1/2
d · Pr
1/3
(3b)
d
4
qradiation at point(r,ϕ,z) = εσTsur f (r,ϕ,z) (4)

dm ablation
q ablation = Ab · (5)
dt
The ablation of the material of the test article can be described using Equation (5),
where Ab represents ablation heat factor, similar to latent heat, while dmablation /dt is the
mass loss due to ablation. One of the sources of heat addition is the oxidation process. The
carbon in the C–C composite chemically reacts with the O2 in the working gas. The main
Aerospace 2025, 12, 43 6 of 20

products of carbon oxidation are CO and CO2 [16] and the heterogenous global chemical
reactions are described in Equation (6) through Equation (8):

C(s) + O2 → CO2 with ∆G ◦ = −399.13 + 2.48 × 10−3 T (kJ/mol) at T < 500 ◦ C (6)
 
1
C(s) + O2 → CO with ∆G ◦ = −220.9 + 0.179 T (kJ/mol) at T > 500 ◦ C (7)
2
C(s) + CO2 ↔ 2CO with ∆G ◦ = −172.2 + 0.177 T (kJ/mol) (8)

The volumetric oxidation process is defined by the following global reaction:

C2( g) + O2 → 2CO with ∆G ◦ = −937.0 + 0.1912 T (kJ/mol) (9)

CO + (1/2)O2 → CO2 with ∆G ◦ = −565.4 − 0.174 T (kJ/mol) at T > 500 ◦ C (10)

where ∆G◦ is the Gibbs free energy change. The heat flux generated by the oxidation
reaction may then be described as shown in Equation (11):

dmoxidation
qoxidation = F × , (11)
dt
where F represents the factor that accounts for a portion of the heat flux produced by
oxidation going into the control volume, and dmoxidation /dt is the mass loss rate due to
oxidation.
After simplifications, the analysis of Equations (1)–(5) allows for the estimation of a
few numbers important for the test planning, such as a characteristic time of the test model
heating by laser irradiation, theat = 30–50 s; maximum achievable temperature at predefined
laser power when Tsurface is lower than 2000 K; and the cooling rate in the high-speed cold
flow, which is greater than 200 K/s. The result of a simplified analysis of the contribution
of the chemical reactions, Equations (6)–(10), in the thermal balance is briefly discussed in
Section 3.3.

2.2. Experimental Test Setup and Conditions

The ground test facility ND_ArcJet schematics are shown in Figure 2. The ND_ArcJet
is a short-duration arc-heated hypersonic wind tunnel at the University of Notre Dame
capable of simulating the high temperature conditions of hypersonic flight. The facility be-
came operational in 2012 and offers a unique test platform for experimental studies of high
enthalpy phenomena in hypersonic flows, non-equilibrium plasma-structure interactions,
materials testing, and scramjet turbulent combustion problems. The flow parameters are as
follows: flow Mach numbers M = 4.5, 6 and 9; stagnation temperatures T0 = 300–6000 K;
total gas pressures P0 = 1–40 bar; flow run time up to 1 s. A significant drawback of the
testing at the ND_ArcJet is a relatively short time of the hypersonic flow exposure. To
overcome this, the test article was preheated by a near-IR CW laser up to a predefined
surface temperature from Tsurface = 300 to 2000 K. For the article preheating, a multi-mode
ytterbium fiber laser YLR-500-MM-AC-Y14 (IPG Photononics, Oxford, MA, USA), 500 W
CW power was utilized. For the purpose of the described testing, the facility is equipped
with high-speed filtered imaging cameras, an IR camera (surface temperature monitoring),
fast pressure sensors, and a high definition schlieren system.
Aerospace 2025,12,
Aerospace2025, 12,43
x FOR PEER REVIEW 77 of 20
21

Figure2.
Figure 2. ND_ArcJet
ND_ArcJet test
test facility
facilitywith
withcapabilities
capabilitiesof
ofMM== 66 high
high enthalpy
enthalpy flow
flow of
of various
various gases.
gases.

Ground
Ground testing of ofthethematerials
materialsatathypersonic
hypersonic conditions
conditions cancan be performed
be performed using using
var-
various shapes of the test article. The basic shapes include
ious shapes of the test article. The basic shapes include a conic tip, leading edge a conic tip, leading edge (wedge),
blunt
blunt body,
body,acreage,
acreage,and andtheir
theircombinations.
combinations.In In terms
terms of of the
the reproducibility
reproducibility of of results
results and and
the validation of simulations, the most representative shape
the validation of simulations, the most representative shape relevant to ground testing relevant to ground testing and
which
and whichprovides more accurate
provides more accurateresults of the capability
results of the material
of the capability of theismaterial
the bluntisbody.the blunt The
data could be then reduced to other shapes once a database
body. The data could be then reduced to other shapes once a database for the specific C- for the specific C–C shape is
available.
C shape isThree-directional needled felt PAN-based
available. Three-directional needled feltcarbon PAN-basedfiber reinforced
carbon fiber carbon matrix
reinforced
commercial composites were provided by Honeywell Aircraft
carbon matrix commercial composites were provided by Honeywell Aircraft Landing Sys- Landing Systems, similar
to those
tems, provided
similar in previous
to those providedinvestigations [28,29]. The C–C
in previous investigations material
[28,29]. The C-C is graphitized
material is
via
graphitized via heat treatment after Chemical Vapor Infiltration (CVI) densification. and
heat treatment after Chemical Vapor Infiltration (CVI) densification. The shape The
dimensions of the test of
shape and dimensions article
the testare article
based are primarily on the capabilities
based primarily and efficiency
on the capabilities and effi- of
measurements. The article is shaped in a blunt body geometry,
ciency of measurements. The article is shaped in a blunt body geometry, see Figures 3 and see Figures 3 and 5, and
is
5, hollow with 7.6with
and is hollow mm7.6 thickness
mm thicknessof solidofC–C solidmaterial
C-C materialalong alongand on and theonfront
the end
frontwith end
the following dimensions: outer diameter of 25.4 mm and
with the following dimensions: outer diameter of 25.4 mm and inner diameter of 9.7 mm inner diameter of 9.7 mm to
provide
to provide access
accessforfor
a K-type
a K-type thermocouple.
thermocouple. TheThe front
frontend endis 9.5 mm
is 9.5 mm thick with
thick with solid
solid C–CC-
material. The thermocouple is initially located on the interior
C material. The thermocouple is initially located on the interior surface of the front end, surface of the front end,
9.5
9.5 mm
mm awayaway from
from thethe flow
flow stagnation
stagnation point point where
where the the laser
laser beam
beam projects
projects the the heat
heat flux.
flux.
The entire test article is expected to shrink as it is exposed to
The entire test article is expected to shrink as it is exposed to heat flux and consequently heat flux and consequently
ablation
ablation and oxidation occur.
and oxidation occur. Two Twoorifices
orificesofof1.51.5 mm mm diameter
diameter areare drilled
drilled throughout
throughout the
the body to provide access for Alumina-based pins, which
body to provide access for Alumina-based pins, which hold the article in place in the test hold the article in place in
the testHeat
stand. stand. Heat transfer
transfer via the pins via the pins interface
interface would bewould neglectedbe neglected
due to the due
smallto the small
footprint
footprint compared to the entire article surface and very low
compared to the entire article surface and very low thermal conductivity of the pins. Mul- thermal conductivity of the
pins.
tiple Multiple
test articlestestofarticles
the same of thesizesame
and size and features
features were machinedwere machined and used andforused for the
the testing.
testing. Machining of the test articles was made such that
Machining of the test articles was made such that the main fiber orientation is either par-the main fiber orientation is
either
allel orparallel or perpendicular
perpendicular to the front to face
the front faceand
surface, surface,
thus and thus parallel/perpendicular
parallel/perpendicular to the hy-
to
personic flow. This paper is specifically focused on the uncoated “parallel” articlearticle
the hypersonic flow. This paper is specifically focused on the uncoated “parallel” mass
mass loss dynamics.
loss dynamics. Material Material
properties properties
considered considered
for thisfor thiswere
study studysimilar
were similar
to [19] suchto [19] as
such as low density (1.60–1.98 g/cm 3 ), low coefficient of thermal expansion (−0.85 to
low density (1.60–1.98 g/cm ), low coefficient of thermal expansion (−0.85 to 1.1 × 10 1/K),
3 −6

1.1 × 10 −6 1/K), high modulus of elasticity (200 GPa), relatively high thermal conductivity
high modulus of elasticity (200 GPa), relatively high thermal conductivity (4–35 W/(m-
(4–35 W/(m-K)),
K)), and the retaining and the of retaining
mechanical of properties
mechanicalatproperties
temperatures at temperatures
up to 2300 K upintoinert
2300en- K
in inert
vironments.environments.
The
The experimental
experimental procedure
procedure of of any
any test
test starts
starts withwith measuring
measuring the the mass
mass of of the
the test
test
article beforehand. The test article is then placed inside the
article beforehand. The test article is then placed inside the wind tunnel, as shown in Fig- wind tunnel, as shown in
Figure
ure 3, 1003, 100
mm mm away
away fromthe
from theexit
exitplane
planeofofthe theM M==66nozzle.
nozzle.The Thecontinuous
continuous wave wave laser,
laser,
IPG Photonics YLR-500-AC S/N PLMP322-1728 (IPG Photonics,
IPG Photonics YLR-500-AC S/N PLMP322-1728, is then set at the predefined power, rang- Oxford, MA, USA), is
then set at0 the
ing from predefined
to 534 W, where power, ranging
the latter from 0the
represents to 534 W, where
maximum the latter
power ratingrepresents
for the equip- the
maximum power rating for the equipment, with an 8 mm
ment, with an 8 mm beam, and the emission in the infrared spectrum λ = 1.06 µm. The beam, and the emission in the
infrared
durationspectrum λ = 1.06 µm.
of laser exposure Thetoduration
is set 360 s whereof laser forexposure
any given is set
heatto flux,
360 s the
where for any
maximum
given heat flux, the maximum
surface temperature is reached at 330 s. surface temperature is reached at 330 s.
Aerospace
Aerospace2025,
2025,12,
12,x43
FOR PEER REVIEW 8 of
8 of2120

(a) (b)

Figure
Figure3.3.(a)
(a)Top
Topview
viewschematics
schematicsofoftest
testsetup
setupfor
forMM==6 6flow
flowtest
testwith
withthe
thedrawing
drawingnot
nottotoreal
realscale;
scale;
(b)schlieren
(b) schlierenimage
imageof
ofthe
thetest
testarticle
articlein
inMM==66flow.
flow.

AnAnOptris™
Optris™XiXi1M 1MInfraRed
InfraRedcameracamera(Optris
(OptrisGmH GmH&&Co., Co.,Berlin,
Berlin,Germany),
Germany),spectralspectral
rangearound
range around λ λ= = 0.8 µm,isisthen
0.8 µm, thenadjusted
adjustedand and triggered
triggered to record
to recordsurface
surfacetemperature
temperature with
a sampling frequency of 30 Hz and 100 × 180 pixel resolution.
with a sampling frequency of 30 Hz and 100 × 180 pixel resolution. The internal tempera- The internal temperature
of the
ture article
of the is measured
article is measured withwith a aK-type
K-typethermocouple,
thermocouple, using using room temperature as
room temperature asaa
reference point. An oscilloscope set at 4 kHz sampling
reference point. An oscilloscope set at 4 kHz sampling frequency records the internal tem-frequency records the internal
temperature
perature signalsignal
from from the thermocouple.
the thermocouple. Both theBothIRthe IR camera
camera and thermocouple
and thermocouple are trig- are
triggered synchronously when laser emission is started.
gered synchronously when laser emission is started. Laser emission, set at different power Laser emission, set at different
poweras
ratings ratings
describedas described above, isonto
above, is projected projected
the testonto thewith
article, test the
article,
pointwith the point of
of impingement
impingement being centered on the front face. The laser
being centered on the front face. The laser power rating remains constant throughout power rating remains constant
the
throughout the entire run. When the vacuum test is
entire run. When the vacuum test is performed, the wind tunnel test section is first performed, the wind tunnel test
section is first pumped out to a total pressure of less than
pumped out to a total pressure of less than 1 mbar. Once vacuum conditions are reached, 1 mbar. Once vacuum conditions
are reached,
laser emissionlaser emission
is started, withisthe
started,
longest with the longest
exposure time exposure
being 360time s. Whenbeing 360
the M s.= 6When
test
is performed, the wind tunnel is first pumped out to vacuum, then the article is the
the M = 6 test is performed, the wind tunnel is first pumped out to vacuum, then article
exposed
toislaser
exposed to laser
emission andemission
allowedand allowed
to reach to reach
a steady a steady
state state temperature
temperature at the surface at the
andsurface
inte-
and interior. With the laser still on, M = 6 flow of either air or
rior. With the laser still on, M = 6 flow of either air or nitrogen is released with a steady- nitrogen is released with a
steady-state
state flow duration
flow duration of 1atmospheric
of 1 s. For s. For atmospheric room temperature
room temperature and pressure
and pressure testing,
testing, the
the article is allowed to cool down back to room temperature
article is allowed to cool down back to room temperature for 930 s after 360 s of laser for 930 s after 360 s of laser
exposuretime.
exposure time.For Forvacuum
vacuumand andMM==66flow flowtests,
tests,the thetest
testarticle
articleisisalso
alsocooled
cooleddown
downfor for
930 s after 360 s of laser exposure time to allow the temperature
930 s after 360 s of laser exposure time to allow the temperature of the article to come back of the article to come
toback
the to the reference
reference point and point and reduce
reduce any oxidation
any oxidation from bringingfrom bringing
the tunnel theback
tunnel back to
to atmos-
pheric pressure. Mass flow rate of the M = 6 flow is controlled by changing the driverdriver
atmospheric pressure. Mass flow rate of the M = 6 flow is controlled by changing the tank
tank pressure (higher pressure for
pressure (higher pressure for larger mass flow). larger mass flow).
Thetest
The testarrangement
arrangementfor forthetheMM = 6= flow
6 flow is shown
is shown in Figure
in Figure 3a. 3a.
The The working
working gas forgas
forhypersonic
the the hypersonic flow flow is provided
is provided by industrial
by industrial grade gradegas stored gas stored in compressed
in compressed gas cylin- gas
cylinders. The same components, such as the laser, test
ders. The same components, such as the laser, test article shape and features, and meas- article shape and features, and
measurement techniques, are used for atmospheric testing and vacuum conditions with
urement techniques, are used for atmospheric testing and vacuum conditions with no
no flow. The test article mass is measured before each test. Then, after the test article is
flow. The test article mass is measured before each test. Then, after the test article is natu-
naturally cooled, the article is weighed again, and the mass measurement is recorded. For
rally cooled, the article is weighed again, and the mass measurement is recorded. For M =
M = 6 flow testing, a schlieren visualization was also taken for a few runs at cold flow (laser
6 flow testing, a schlieren visualization was also taken for a few runs at cold flow (laser
off) and hot conditions (laser on) at 20 kHz framerate. This system is triggered to record
off) and hot conditions (laser on) at 20 kHz framerate. This system is triggered to record
images only when the M = 6 wind tunnel is triggered. The schlieren images were also
images only when the M = 6 wind tunnel is triggered. The schlieren images were also
utilized to confirm that the flow is steady for 1 s, and that it is ideally expanded at the exit
utilized to confirm that the flow is steady for 1 s, and that it is ideally expanded at the exit
of the converging–diverging nozzle and for a distance of, at least, 1250 mm. An example
of the converging–diverging nozzle and for a distance of, at least, 1250 mm. An example
of the schlieren images is shown in Figure 3b. An example of mass loss data processing
of the schlieren images is shown in Figure 3b. An example of mass loss data processing
according to the characterization in Figure 4 for hypersonic flow runs is shown in Table 1.
according to the characterization in Figure 4 for hypersonic flow runs is shown in Table
1.
 Aerospace
Aerospace 2025,
2025, 12,12,
43x FOR PEER REVIEW 99ofof2021

(a) (b) (c) (d)

Figure4.4.Article
Figure Articleimages
imagesofofboth
bothmachining
machiningorientations:
orientations:(a)
(a)parallel
parallelnew;
new;(b)
(b)parallel
paralleldecommis-
decommis-
sioned(after
sioned (afteratatleast
least3600
3600s of
s of cumulativelaser
cumulative laserexposure);
exposure);(c)(c) perpendiculardecommissioned;
perpendicular decommissioned;and
and
(d)
(d)perpendicular
perpendicularnew. new.

Table
Table Relevant
1. 1. data
Relevant from
data MM
from = =6 6flow
flowexperiments.
experiments.

ParameterParameter Run 1 Run 1Run 2 Run 2 RunRun3 3

Flow Gas
Flow Gas Employed Employed Air Air Air Air Nitrogen
Nitrogen
Stagnation
Stagnation PressurePressure
(bar) (bar) 6.9 6.9 14 14 6.96.9
Flow Rate through
Flow nozzle (kg/s)
Rate through nozzle (kg/s) 0.073 0.073 0.146 0.146 0.073
0.073
Total Mass Loss
Total RateLoss
Mass (mg/s)
Rate (mg/s) 88.5 88.5 93.5 93.5 31.2
31.2
MassMass
LossLoss
Rate Rate
due to Oxidation
due to Oxidation (mg/s) 58.2 61.4 61.4 0.00
58.2 0.00
(mg/s)
Mass Loss Rate due to Ablation 车
Mass Loss Rate due to Ablation 30.3 32.1 31.2
and Mechanical Erosion (mg/s) 30.3 32.1 31.2
and Mechanical Erosion (mg/s)

Live images of the article undergoing testing were achieved using a Phantom v2512
camera, images
Live as shown of the article 3a.
in Figure undergoing
A 550 nmtesting werefilter
shortpass achieved
and ausing a Phantom
hot glass IR filter v2512
were
camera, as shown in Figure 3a. A 550 nm shortpass filter and a
placed in front of the camera lens to block the article’s blackbody radiation and laserhot glass IR filter werein-
placed in front of the camera lens to block the article’s blackbody
frared radiation reflection. A high-resolution optical lens with a 200 mm focal length was radiation and laser
infrared
used onradiation reflection.
the camera A high-resolution
for zoomed imaging at the optical lens with
required a 200distance.
working mm focalAdditionally,
length was
used
two on the cameralenses
teleconverter for zoomed imaging
were used, one at2× the
andrequired
one 1.4× working
to provide distance.
enhanced Additionally,
zooming on
two teleconverter lenses were used, one 2 × and one 1.4 ×
the face of the article. Due to the required short exposure time of the camera, the zooming
to provide enhanced test article
on the face of the article. Due to the required short exposure time
was illuminated in 532 nm green light using a SW laser, synced with the camera sampling of the camera, the test
article was illuminated in 532 nm green light using a SW laser,
frequency, with a maximum power of 2 W, positioned outside of the tunnel. The green synced with the camera
sampling
laser wasfrequency,
positioned with a maximum
to illuminate thepower
portion of of
2 W,
thepositioned
article beingoutside of the tunnel. The
imaged.
green laser was positioned to illuminate the portion of the article being
The specific characteristics of material destruction studied include an overall quanti- imaged.
fication specific
The characteristics
of material mass loss and of material
a specificdestruction
qualitativestudied
analysisinclude an overall
of carbon fiber and quantifi-
carbon
cation
matrix of morphological
material mass loss and a specific
alteration. qualitative
The testing underanalysis
different of environments
carbon fiber and and carbon
corre-
matrix morphological alteration. The testing under different environments
sponding analyses using high-framerate footage, mass loss quantification, and scanning and correspond-
ing analyses
electron using high-framerate
microscopy allows for the footage, mass lossofquantification,
determination and scanning
various morphological electron
changes, such
microscopy allows for the determination of various morphological
as the inherent structure and the shape of the internal carbon fibers, to be attributed changes, such as theto
inherent
specific structure
mechanisms, and such
the shape of the internal
as mechanical erosion,carbon
rapidfibers, to be etc.
oxidation, attributed
It should to specific
be noted
mechanisms, such as mechanical erosion, rapid oxidation, etc.
that fracture toughness testing of the bulk material is out of scope of this paper. It should be noted that
fracture toughness testing of the bulk material is out of scope of this paper.
2.3. Calibration of Instrumentation
2.3. Calibration of Instrumentation
Calibration of the test instrumentation includes calibrating the lab grade scale accord-
Calibration of the test instrumentation includes calibrating the lab grade scale ac-
ing to specs for every day of testing with a precise 1000 mg weight provided by the man-
cording to specs for every day of testing with a precise 1000 mg weight provided by the
ufacturer, and assuring the scale sits on a flat, level surface. Continuous use of the scale
manufacturer, and assuring the scale sits on a flat, level surface. Continuous use of the
proved that there is a standard deviation of mass measurements of ±1.0 mg, and for the
scale proved that there is a standard deviation of mass measurements of ±1.0 mg, and for
mass loss rates it is ±0.1 mg/s. The thermocouple is calibrated to measure internal temper-
the mass loss rates it is ±0.1 mg/s. The thermocouple is calibrated to measure internal
ature of the test article with room temperature as a reference point and was set at 4000
temperature of the test article with room temperature as a reference point and was set
samples/s recording using the oscilloscope. The IR camera is calibrated using software
at 4000 samples/s recording using the oscilloscope. The IR camera is calibrated using
provided by the manufacturer and by installing a spectral filter to block any reflection
Aerospace 2025, 12, 43 10 of 20

software provided by the manufacturer and by installing a spectral filter to block any
reflection from the lasers. The IR camera temperature data was initially compared to optical
spectrometer data obtained for the same laser exposure conditions and the temperatures
were confirmed to be consistent. A formal error on the temperature readout indicated by
IR video is ±1% for a surface temperature lower than 1800 K and ±2% for temperatures
greater than 1800 K, but that does not take into account the temperature variation over the
surface on a microscale.
Mass loss measurements for the hypersonic testing conditions imply several calibration
tests need to be performed and considered for any mass changes. The calibration tests
helped to quantify mass changes due to equipment/experimental procedure rather than the
thermal ablation and oxidation mechanisms. The following calibration tests, although with
minimal mass losses, were accounted for: handling of article during installation/removal
(friction), vacuum conditions with no heat flux exposure (fracture of cold fibers caused
by bringing the test section from atmospheric pressure to vacuum, vacuum duration, and
repressurizing after testing completion), any oxidation due to residual air in the tunnel after
bringing to vacuum, and any losses caused by cold M = 6 flow (breaking of fibers when the
article is not exposed to the laser but only to hypersonic airflow). The latter would be used
for comparison to a hot article under M = 6 airflow to demonstrate that the ablation and
oxidation is driven by the high heat flux and consequently the high temperature. The mass
loss caused by ablation is measured by performing the following tests: vacuum and laser
exposure to the article (thermal stress causing weakening of fibers), and M = 6 N2 flow
causing mechanical erosion. The ablation–oxidation measurement is performed with the
article being exposed to heat flux generated by the laser, and the highly compressed airflow
released when the article is held at constant maximum temperature, causing oxidation.
The assumptions made were as follows: reference temperature to be room temperature,
and articles being considered blackbodies with emissivity and absorptivity close to 1. For
hypersonic conditions, mass loss dynamics were categorized in three classes: calibration
(mass loss due to the testing procedure and not caused by ablation/oxidation), ablation
mass loss, and ablation–oxidation mass loss. These calibration and measurement tests offer
the possibility of characterizing the mass loss dynamics quantitatively. For instance, one
performs an M = 6 test with air as the working gas and a mass loss number is obtained.
Performing a test with N2 or Ar should result in a different mass loss quantity due to the
lack of the oxidation process, but ablation due to very high heat flux provided by the laser
and the hypersonic flow conditions would still provide the ablation mechanism at the
surface level.

3. Results
3.1. Ablation/Oxidation Measurements at Atmospheric Conditions
Heat flux played a substantial role in the mass loss rate of the test articles under
atmospheric conditions. The total energy deposited into the test article is a function of
the output power of the laser and the total run time of the test. To vary the total energy
deposited into the test article, either the laser output power, or the run time can be varied.
Tests were performed to achieve an identical total energy deposition, but with differing
run times and differing laser output powers. The cases with higher laser output powers
had a notably higher mass loss rate, indicating that the rate of energy delivery, namely
the heat flux on the test article surface, has a significantly more impactful effect on the
thermal response and the mass loss behavior of the material when compared to simple
run time. The reason for this is likely due to a combination of a few factors. First, a higher
surface temperature on the test article leads to faster rates of oxidation of the test article.
As indicated in Equations (7) and (8), as the temperature on the test article increases, ∆G◦
Aerospace 2025, 12, 43 11 of 20

becomes increasingly negative, which makes the direct oxidation reaction, Equation (6),
and volumetric oxidation more thermodynamically favorable. Additionally, rapid temper-
ature increases because of higher heat fluxes can cause intense thermal gradients within
the material, potentially leading to points of structural weaknesses induced by thermal
stress. The temperature gradient may achieve significant magnitude, such as when the
maximal outer surface temperature of 1770 K and internal temperature at the location of
the thermocouple of 1300 K were measured. However, both methods converged to a mass
loss rate of around 3.0 ± 0.1 mg/s as total energy deposition increased.
According to Figure 4, the machining orientation (parallel/perpendicular of main
fibers’ layers to the flow) yield different shapes of the composite article after testing. Parallel
articles, Figure 4b, ablate cylindrically and more uniformly, whereas perpendicular articles
ablate in a rather conical shape as shown in Figure 4c. These patterns are due to the
machining orientation as they change axial and radial thermal conductivity between the
two orientations. The parallel main fibers will allow heat to transport through the article
streamwise, thus creating a more uniform temperature distribution and causing oxidation
and ablation to occur uniformly throughout the entire article. This will dissipate the heat
more efficiently streamwise down the test article. Perpendicular orientation of the main
fibers will reduce radial thermal conductivity through the article, therefore causing the
fibers hidden under the surface layers to be cooler and not ablate and oxidize at the same
rate as the fibers at the surface. Additionally, heat needs to move through stacked layers
of varying carbon fiber layers. Therefore, when new, the initial mass is 17,000 ± 10 mg
and outer diameter at the thermocouple probe plane is 24.9 ± 0.1 mm for any orientation.
However, that diameter shrinks, reducing to an average of 19.9 ± 0.1 mm for the parallel
orientation, and 20.3 ± 0.1 mm for the perpendicular test article. This average was based
off three articles of each type, after 3600 s of cumulative heat flux exposure and atmospheric
air with temperatures beyond 800 K. Therefore, the axial recession rate of material was
1.7 × 10−3 mm/s for the parallel articles and 1.4 × 10−3 mm/s for perpendicular articles.
For vacuum and hypersonic flow testing, the focus was primarily on the parallel article
given the timeframe of the project.
The IR camera observations of the test article surface temperature allow the correlation
of laser power, surface temperature and the rate of the test article mass loss, as is shown
in Figures 5 and 6 for a single day of experimentation. Some data scattering is due to the
test articles replacement (results differ slightly for newer vs. older test articles). Figure 5a
demonstrates the test article steady state surface temperature acquired by the thermal
camera versus the projected heat flux. The camera’s proper readout starts from the surface
temperature of 800 K. Two distinct ranges are considered: one for temperatures lower than
1400 K where the thermal radiation could be neglected, and one for temperatures greater
than 1400 K where the thermal radiation is a major mechanism of the test article cooling.
The data in Figure 5b excludes an initial preheating and indicates the mass loss rate at a
given steady state surface temperature. Depending on the laser power, a noticeable test
article recession is observed at temperatures greater than 1000 K. It gradually increases
with the surface temperature magnitude within the range of the temperatures explored
in this test series. Figure 6 plots mass loss rate against laser exposure time for maximum
projected heat flux. This demonstrates that a steady state mass loss rate, dm/dt, greater
than 3 mg/s is achieved for the maximum laser output power. For highly used test articles,
this steady state mass loss rate of 3.0 mg/s was reached even earlier, at around 60 s of laser
exposure time. A small uncertainty in the weighing of the test articles and in the startup
and shutdown times of the laser is also accounted for in Figure 5b.
Aerospace 2025,
Aerospace 12, 12,
2025, x FOR PEER
x FOR REVIEW
PEER REVIEW 12 12
of of
21 21
Aerospace 2025, 12, 43 12 of 20

(a)(a) (b)(b)

Figure
Figure
Figure 5.5.(a) Test
5.(a)(a)
Test article
Testarticlesurface
article
surfacetemperature
surface
temperatureversus
temperature
versuslaser
versus power
laser
laser and
power
power (b)
and
and mass
(b)
(b) loss
mass
mass rate
loss
loss versus
rate
rate test
versus
versus test
test
article surface
article surface temperature.
temperature.
article surface temperature.

Figure
Figure
Figure 6.Mass
6.6.Mass
Massloss
loss rate
rate
loss versus
versus
rate laser
laser
versus exposure
exposure
laser time
time
exposure atatmaximal
time maximal
at power.
power.
maximal power.

3.2. Mass Loss at Vacuum Conditions

3.2.3.2.
MassMassLoss at Vacuum
Loss at Vacuum Conditions
Conditions
Tests were also performed for the parallel articles under vacuum conditions at elevated
Tests were
Tests were also performed
also performed forfor
thethe
parallel articles
parallel under
articles under vacuum
vacuum conditions
conditions at at
ele-ele-
temperatures. In these experiments, the parallel test articles were exposed to a heat flux
vated
vatedtemperatures.
temperatures. In In
these experiments,
these experiments, thethe
parallel testtest
parallel articles were
articles were exposed
exposed to to
a heat
a heat
provided by the laser for 6 min of exposure time at 9.4 MW/m2 . The2 test article experienced
flux provided
flux provided byby thethelaser forfor
laser 6 min
6 minof of
exposure
exposuretime
timeat 9.4 MW/m
at 9.4 MW/m . The testtest
2. The article expe-
article expe-
about 28 ± 0.1 mg of total mass loss. This corresponds to a negligibly low mass loss rate
rienced
rienced about
about28 28
± 0.1 mgmg
± 0.1 of of
total mass
total massloss. This
loss. Thiscorresponds
corresponds to to
a negligibly
a negligibly lowlowmass
mass
of around 0.08 mg/s compared to atmospheric and flow conditions. Regardless, it can be
loss rate
loss of of
rate around
around 0.08 mg/s
0.08 mg/scompared
compared to to
atmospheric
atmospheric and andflow conditions.
flow conditions. Regardless,
Regardless,
assumed that some percentage of mass loss under cold flow comes from thermal ablation.
it can be be
it can assumed
assumed that some
that somepercentage
percentage of of
mass
massloss under
loss under cold flow
cold flowcomes
comes fromfromthermal
thermal
Likely, this mass loss is due to the thermal ablation of superheated fibers at the front of the
ablation.
ablation.Likely,
Likely,this mass
this massloss is due
loss to to
is due thethe
thermal
thermal ablation
ablation of of
superheated
superheated fibers at at
fibers thethe
test article or the vaporization of residual resin material from the laying out of carbon fibers.
front
frontof of
thethe
testtest
article or or
article thethe
vaporization
vaporization of of
residual
residualresin material
resin material from
fromthethe laying
layingoutout
of3.3.
carbon
of carbonfibers.
fibers.
Mass Loss Under Hypersonic Flow Conditions
The mass loss of the test article was experimentally determined under M = 6 flows of
3.3.3.3.
Mass Loss
Mass LossUnder
UnderHypersonic
Hypersonic Flow Conditions
Flow Conditions
both air and nitrogen at flow stagnation temperature T0 = 300 K. The hypersonic testing
was The Themass
mass
performed loss of of
atloss
one the testtest
the
single articleofwas
article
value was
the experimentally
experimentally
laser power since determinedwereunder
determined
there under
several Mother
M= 6=flows of of
6variable
flows
both
bothairair
parametersandandnitrogen
to nitrogen
consider.at Each
flow
at flow stagnation
stagnation
test temperature
temperature
run is time-consuming, T0 T=which
0 300 K.leads
= 300 The
K. hypersonic
The
to ahypersonic testing
testing
limited number
waswasperformed
performed
of tests. at one
at single
one single value of
value the
of laser
the power
laser since
power sincethere were
there wereseveral
severalother vari-
other vari-
able
ableparameters
parameters
Shown to to
in Table consider.
are theEach
1consider. Eachtest
results run
test
of is time-consuming,
run
multiple is time-consuming, which
experiments performedwhichleads to cold
leads
under a limited
to aMlimited
=6
number
number of tests.
of tests.
flow with an article preheated by laser exposure at the radiative heat transfer of 9.4 MW/m2 .
Aerospace 2025, 12, x FOR PEER REVIEW 13 of 21

Aerospace 2025, 12, 43

Shown in Table 1 are the results of multiple experiments performed under cold 13 Mof=20
6 flow with an article preheated by laser exposure at the radiative heat transfer of 9.4
MW/m2. As it was measured, the mass loss rate under cold hypersonic flow is significantly
As it was measured, the mass loss rate under cold hypersonic flow is significantly higher,
higher, approximately 40 times, than that under atmospheric conditions. As a note, in-
approximately 40 times, than that under atmospheric conditions. As a note, increasing the
creasing the stagnation pressure from 6.9 bar to 14 bar did not cause a substantial increase
stagnation pressure from 6.9 bar to 14 bar did not cause a substantial increase in the mass
in the mass loss rate of the article. Further, when compared to a flow of nitrogen under
loss rate of the article. Further, when compared to a flow of nitrogen under similar surface
similar surface temperature and mass flow rate conditions, the test article experienced a
temperature and mass flow rate conditions, the test article experienced a higher mass
higher mass loss rate being exposed at airflow. This difference is suggested to be due to
loss rate being exposed at airflow. This difference is suggested to be due to the oxidation
the oxidation reaction on the surface of the test article.
reaction on the surface of the test article.
For the parallel articles, a comparison was made to observe the mass loss dynamics
For the parallel articles, a comparison was made to observe the mass loss dynamics
between the three testing conditions as shown by the diagram in Figure 7. In each test the
between the three testing conditions as shown by the diagram in Figure 7. In each test
article was exposed to a heat flux of 9.4 MW/m2 for 3602s. The maximum surface temper-
the article was exposed to a heat flux of 9.4 MW/m for 360 s. The maximum surface
ature exceeded 2000 K. Note that the atmospheric testing with still air provides a reference
temperature exceeded 2000 K. Note that the atmospheric testing with still air provides a
point when compared to the vacuum and hypersonic flow at M = 6 of either air or nitrogen.
reference point when compared to the vacuum and hypersonic flow at M = 6 of either air or
Mass loss rate at atmospheric conditions was 3.28 ± 0.1 mg/s. For the vacuum, the mass
nitrogen. Mass loss rate at atmospheric conditions was 3.28 ± 0.1 mg/s. For the vacuum,
loss was 0.08 ± 0.1 mg/s which is below the error of mass measurement and therefore
the mass loss was 0.08 ± 0.1 mg/s which is below the error of mass measurement and
negligible. This shows that there is no oxidation occurring in vacuum conditions. To as-
therefore negligible. This shows that there is no oxidation occurring in vacuum conditions.
certain a value for mass loss due purely to the flows of air or nitrogen, a mass loss of 28
To ascertain a value for mass loss due purely to the flows of air or nitrogen, a mass loss
mg due to laser erosion and thermal stress observed under vacuum conditions was as-
of 28 mg due to laser erosion and thermal stress observed under vacuum conditions was
sumed to be present in all tests and is compensated for in the relevant mass loss rate cal-
assumed to be present in all tests and is compensated for in the relevant mass loss rate
culations. This leads to a mass loss rate under nitrogen flow of 31.2 mg/s. However, for
calculations. This leads to a mass loss rate under nitrogen flow of 31.2 mg/s. However, for
newer test articles the total mass loss due to nitrogen flow testing was slightly lower, likely
newer test articles the total mass loss due to nitrogen flow testing was slightly lower, likely
due to a smaller number of structural weaknesses. Given that for nitrogen flow there is no
due to a smaller number of structural weaknesses. Given that for nitrogen flow there is no
oxidation process, the ablation process dominates the mass loss dynamics, quantifying the
oxidation process, the ablation process dominates the mass loss dynamics, quantifying the
mass
massloss
lossrate
ratedue
duetotopure
puremechanical
mechanicalflow
flowerosion
erosionand andablation
ablationat
at31.2
31.2 ±±0.1
0.1mg/s.
mg/s.The
The
mass
mass loss
lossrate
rateunder
underairairflow
flowwas
wascalculated
calculatedtotobebe91.0
91.0mg/s.
mg/s.

Figure 7. Mass loss rate comparison between the three testing conditions. For each test, the article
Figure 7. Mass loss rate comparison between the three testing conditions. For each test, the article
was exposed to 9.4 MW/m2 of heat flux for 360 s. For M = 6 flow, P0 = 6.9 bar. The errors for each
was exposed to 9.4 MW/m2 of heat flux for 360 s. For M = 6 flow, P0 = 6.9 bar. The errors for each
mass loss rate measurement were ±0.1 mg/s.
mass loss rate measurement were ±0.1 mg/s.
The dynamics of the test article surface temperature are shown in Figure 8 for air flow,
andThe dynamics
in Figure of the
8b for test article
N2 flow. surface
The time temperature
sequence are shown
of the flow run is inas Figure
follows: 8 for
the air
first
flow, and in Figure 8b for N 2 flow. The time sequence of the flow run is as follows: the first
1.2 s is M = 6 flow (indicated in Figure 8a by a red bar), then about 0.5 s of supersonic
1.2unsteady
s is M =flow,
6 flowsee(indicated
Figure 12in forFigure 8a by and
the details, a redthen
bar), then0.6
about about
s of 0.5
weaks ofsubsonic
supersonicflow
unsteady flow, see Figure 12 for the details, and then about 0.6 s of weak subsonic
of a remnant gas from the plenum section. There is a significant difference in the surface flow of
a temperature
remnant gas dynamics
from the plenum section. There is a significant difference
between these two cases. The cooling rate observed under the in the surface
temperature dynamics between
flow of nitrogen was higher than these two
that cases.
under Theofcooling
flow air. Therate observedisunder
hypothesis the flow
that, likely, this
ofisnitrogen
because the oxidation reaction is exothermic and releases additional heat on this
was higher than that under flow of air. The hypothesis is that, likely, is
its own
during the phases of both steady-state M = 6 flow and subsonic weak flow. Despite many
uncertainties in chemical reactions details, a rough estimate of the chemical heat flux at
 becausethe
because theoxidation
oxidationreaction
reactionisisexothermic
exothermicand andreleases
releasesadditional
additionalheat heatononitsitsown
owndur-dur-
ing the phases of both steady-state M = 6 flow and subsonic
ing the phases of both steady-state M = 6 flow and subsonic weak flow. Despite manyweak flow. Despite many
Aerospace 2025, 12, 43 14 of 20
uncertaintiesininchemical
uncertainties chemicalreactions
reactionsdetails,
details,a arough
roughestimate
estimateofofthe
thechemical
chemicalheat
heatfluxfluxatat5050
± ±0.1
0.1mg/s
mg/sofofcarbon
carbonoxidation
oxidationreturns
returnsa avalue
valuewell-above
well-abovethethe1010MW/m
MW/m . Duringthe
2. 2During thephase
phase
ofoscillating
of50 oscillating
± 0.1 mg/s flow,
flow, thecooling
the
of carbon cooling rateis
rate
oxidation ishigher
higher
returns thanthe
a than
value thechemical
chemical
well-above reaction’s
reaction’s
the 10 MW/m contribution
contribution
2 . During to to
the
the
the
phase heat flux
heatofflux balance.
balance.flow, the cooling rate is higher than the chemical reaction’s contribution
oscillating
to the heat flux balance.

(a)(a) (b)
(b)
Figure
Figure 8.
8.8. Surface
Surface
Surface temperature
temperature
temperature dynamics
dynamics
dynamics during
during
during (a)
(a)
(a)MMM
==6=6airflow
6airflow
airflow and
and
and (b)
(b)
(b)MMM
==6= 6N
6N
N 2 flow.
2 flow. The
The red
red bar
bar
Figure 2 flow. The red bar
indicates
indicates
indicates a duration
duration
a aduration of
ofof
MMM
= =6= 6flow.
6 flow.
flow.

4.4. Physical
4.Physical Mechanisms
Mechanismsofof
PhysicalMechanisms C–C
ofC-C CompositeAblation
C-CComposite
Composite Ablation
Ablation
Live camera
Livecamera
Live cameraimagesimages
imagesofof the
ofthe heated
theheated article
heatedarticle under
articleunder atmospheric
underatmospheric conditions
atmosphericconditions reveal
reveala aarapid
conditionsreveal rapid
rapid
steady
steady decrease
steadydecrease
decreasein ininthe
the size
sizeof
thesize superheated
ofof superheated
superheated fibers atatthe
fibers
fibers atthe front
the ofof
front
front the article,
ofthe
the indicating
article,
article, intense
indicating
indicating in-in-
oxidation.
tenseoxidation.
tense A comparison
oxidation.AAcomparison of the
comparisonofofthe test article
thetest images
testarticle is
articleimages shown in
imagesisisshown Figure 9.
shownininFigure Note the decrease
Figure9.9.Note
Notethe the
in size of
decrease
decrease superheated
inin sizeofofsuperheated
size fibers (lighter
superheated gray
fibers(lighter
fibers and
(lighter white
gray
gray andand in color)
white
white ininmoving from Figure
color)moving
color) moving 9a
fromFigure
from to d.
Figure
9aThe white
9atotod.
d.TheTheand
white gray
white and fibers
and gray
gray shrink
fibersin
fibers size as
shrink
shrink ininexposure time increases.
sizeasasexposure
size exposure This is highlighted
timeincreases.
time increases. Thisisishigh-
This high-in
the red circled
lightedininthe
lighted theredregions
redcircled in Figure
circledregions 9b,c. The
regionsininFigure shrinkage
Figure9b,c.
9b,c.The of fibers
Theshrinkage forms a
shrinkageofoffibers pit in
fibersformsthe center
formsa apit of
pitininthethe
the
front face,
centerofofthe
center as evident
thefront
frontface, in Figure
face,asasevident 9a,d.
evidentininFigure
Figure9a,d.
9a,d.

(a)(a) (b)
(b) (c)(c) (d)
(d)
Figure9.9.
Figure
Figure 9.(a)(a)
(a)Article
Articlefront
Article frontsurface
front surfacebefore
surface beforeheat
before heatflux
heat fluxexposure;
flux exposure;(b)
exposure; (b)zoomed-in
(b) zoomed-inimage
zoomed-in imageshowing
image showinginitial
showing initial
initial
materialrecession
material
material recessionareas;
recession areas;(c)
areas; (c)zoomed-in
(c) zoomed-inimage
zoomed-in image showing
showing
showing material
material
material recessed
recessed
recessed during
during
during the
the
the formation
formation
formation of of of
the
pit
thein the
pit in laser
the impingement
laser impingement spot; and
spot; (d)
and 360
(d) s of
360 laser
s of exposure
laser at
exposure atmospheric
at conditions.
atmospheric
the pit in the laser impingement spot; and (d) 360 s of laser exposure at atmospheric conditions. conditions.

Evident
Evidentinin
Evident high-framerate
inhigh-framerate
high-frameratevideo video footage
footageofof
videofootage the
ofthe article
thearticle undergoing
articleundergoing
undergoingM MM===6 66flow
flow
flowofof air
airisis
ofair is
an
an increased
increased rate
rateof flaking
of of
flaking fibers.
of The
fibers. process
The of
process flaking
of is shown
flaking is in
shown
an increased rate of flaking of fibers. The process of flaking is shown in Figure 10 and Figure
in 10
Figure and occurs
10 and
in the in
occurs
occurs time
inthespan
the time
time ofspan
360 of
span µs. The
of360 flaking
360µs.
µs.The isflaking
especially
Theflaking intense intense
isisespecially
especially during
intensethe flowthe
during
during phase.
theflow Whereas
flow phase.
phase.
the article under atmospheric conditions lost material due to the steady oxidation of fibers
which led to the dissipation of the material on the microscale, M = 6 air flow causes rapid
oxidation on the microscale with the added effect of mechanical shearing and breakage on
the macroscale, which occurs on the scale of microseconds throughout the duration of flow.
Additionally, flaking occurs sporadically throughout the entire time span of the flow. This
 Whereas the article under atmospheric conditions lost material due to the steady oxida-
tion of fibers which led to the dissipation of the material on the microscale, M = 6 air flow
Aerospace 2025, 12, 43 causes rapid oxidation on the microscale with the added effect of mechanical shearing and
15 of 20
breakage on the macroscale, which occurs on the scale of microseconds throughout the
duration of flow. Additionally, flaking occurs sporadically throughout the entire time
span of thethat
indicates flow. Thismechanical
simple indicates that simpleismechanical
breakage not the onlybreakage
mechanism is not the only mecha-
of destruction. As the
nism
first of
airdestruction.
molecules reach As the thefirst air molecules
article, fibers thatreach
werethe article, fibers
structurally weakthat duewere structurally
to manufacturing
weak due to manufacturing
imperfections or prior testing imperfections
are blown out or prior testing are
immediately. blownfibers
Stronger out immediately.
will undergo
Stronger
oxidation, fibers will undergo
forming oxidationoxidation,
pits, which forming
lead tooxidation pits, which
more structural lead to more
weaknesses. As thestruc-
wave
tural
of airweaknesses. As the wave
continues blowing, these of air continues
created structuralblowing,
weaknesses theseleadcreated
to thestructural
breakage weak-of even
nesses
more leadfibers,toasthe breakagebyofthe
indicated even more fibers,
continued flakingas of
indicated by the
fibers after thecontinued
initial wave flaking
of air.ofTo
fibers
confirmafterthis,
the the
initial wavetest
heated of air. To confirm
article was testedthis,under
the heateda flow test
of article was tested
N2 , where a lower under
mass
a loss
flowrateof N 2, where
was observed,a lower mass loss
as shown rate was
in Figure observed, as shown
6. High-framerate videoinfootage
Figure of 6. the
High-test
framerate
article under video Nfootage of the atest
2 flow reveals article
rapid underout
blowing N2 offlow reveals
small a rapid
filaments blowing
at the out ofof
beginning
small filaments
the flow. This at the beginning
is shown in Figure of the
11. flow. This is have
The photos shown beenin Figure
edited11. The photos
to help visualize havethe
been
small edited to help white
superheated visualize theleaving
fibers small superheated white fibers
the surface directly in frontleaving
of the thetest surface di-
article. This
rectly
processin front
occurs of during
the testthearticle.
veryThis
first process occursofduring
milliseconds flow and the finishes
very first milliseconds
about a quarterofof
flow and finishes
a second into flow. about a quarter of
Throughout the a second
rest of theintoflow,
flow.occasional
Throughout the rest
flaking of of the flow,
superheated
occasional
fibers occurs, flakingbutof at superheated
a substantially fibers
loweroccurs, but at compared
rate when a substantially to videolower rate when
footage of the
compared
test articletoundervideoairflow.
footage Additionally,
of the test article under airflow.
the observed size ofAdditionally,
flakes breaking the off
observed
the test
size of flakes
article were breaking
smaller when off the test article
compared towere smaller
the test article when
under compared
air flow.toThis the leads
test article
to the
under air flow.
conclusion thatThis
for aleads to the conclusion
moderately that for
used test article, a moderately
about one third usedof thetestmass article, about
loss rate can
one third of the
be attributed tomass loss rate can
the mechanical be attributed
blowing to the mechanical
out of weakened fibers andblowing
two thirds outofoftheweak-
mass
ened fibers
loss rate canandbe two thirdsto
attributed ofthe
therapid
massoxidation
loss rate and
can destruction
be attributed of to the rapid
strong fibers.oxidation
For a new,
and destruction
unused of strong
test article, it wasfibers.
observedFor athatnew,theunused
mechanicaltest article,
breakage it was observed
of fibers that theto
contributes
mechanical breakage
a slightly smaller of fibers
portion contributes
of the overall massto a slightly
loss rate, smaller
likely portion
due to the of the
loweroverall
numbermassof
loss rate, likely
structural due to the lower number of structural weaknesses.
weaknesses.

(a) (b) (c) (d)

Figure
Figure10.
10.Flaking
Flakingmechanism
mechanismofofsuperheated superheatedfibers fiberson onthe
thesurface
surfaceof
ofthe
theparallel
paralleltest
testarticle,
article,ar-
arrows
rows indicating
indicating
Aerospace 2025, 12, x FOR PEER REVIEW flakes
flakes moving moving
during during
flow. flow.
Each Each
imageimage
showsshowsflakesflakes removed
removed from
from the the surface
surface at 16 ofat21
different
time instances,
different (a) t1 , (b)(a)t2t1>, (b)
time instances, t1 , (c)
t2 >t3t1,>(c)
t2 ,tand
3 > t2,(d)
andt4 (d)
> t3t.4 > t3.

(a) (b) (c) (d)

Figure
Figure11.
11.Blowout
Blowoutmechanism
mechanismof ofsmall,
small,weakened
weakenedfibers fiberson onthe
thesurface
surfaceof ofthe
theparallel
paralleltest
test article
article
under
underaaMM==66flow
flowof
ofnitrogen.
nitrogen.(a)
(a)t1t,1 (b)
, (b)t2t>2 >
t1,t(c) t3 >t3t2>, and
1 , (c) (d) (d)
t2 , and t4 > tt43.> t3 .

The mechanical destruction of the overheated individual carbon filaments is likely to

be caused by a high-speed flow blowout. However, hypothetically, it might be signifi-
cantly intensified as the bow shock wave instability resulted in pressure oscillations in the
compression area of the blunt body. A comparison of instant schlieren images of a steady
 (a) (b) (c) (d)
Aerospace 2025, 12, 43
Figure 11. Blowout mechanism of small, weakened fibers on the surface of the parallel 16 of 20
test article
under a M = 6 flow of nitrogen. (a) t1, (b) t2 > t1, (c) t3 > t2, and (d) t4 > t3.

The mechanical destruction of the overheated individual carbon filaments is likely to

The mechanical destruction of the overheated individual carbon filaments is likely to
be caused by a high-speed flow blowout. However, hypothetically, it might be significantly
be caused by a high-speed flow blowout. However, hypothetically, it might be signifi-
intensified as the bow shock wave instability resulted in pressure oscillations in the com-
cantly intensified as the bow shock wave instability resulted in pressure oscillations in the
pression area of the blunt body. A comparison of instant schlieren images of a steady flow
compression area of the blunt body. A comparison of instant schlieren images of a steady
over the heated test article with a strongly oscillating flow is shown in Figure 12.
flow over the heated test article with a strongly oscillating flow is shown in Figure 12.

(a) (b)

Figure
Figure 12.Schlieren
12. (a) (a) Schlieren images
images of steady
of steady bowbow shock
shock wavewave
withwith
arrowarrow showing
showing direction
direction of theofflow
the
and flow andunder
(b) one (b) one under oscillating
oscillating mode. mode.

To explore
To explore the the
physical
physical details
detailsof of
fiber destruction,
fiber destruction,images imageswereweretaken
takento to examine the
the microstructure
microstructure of of the
the test
test article
articlesurface
surfaceafter
afterablation
ablationand andoxidation
oxidationusing usingscanning
scanningelec-
electron microscopy. Figure 13 shows a group of fibers on the test
tron microscopy. Figure 13 shows a group of fibers on the test article surface for article surface for various
various
testing conditions.
testing conditions. As seen
As seen in Figure
in Figure13, many
13, manysmall pitspits
small are are
visible on both
visible on boththe the
internal
internal
solid carbon
solid carbonfiber andand
fiber sheathing
sheathing CVICVIcarbon
carbon material.
material.There
Thereare arelikely
likelytwo two reasons
reasons for the
the existence
existence of of these
these pits.
pits. Firstly,
Firstly, the
the smalls
smallspits
pitson on the
the CVI
CVI carbon
carbon are are evident
evident in in all
all test
test articles.
articles.InInquantity,
quantity,they
theyare areleast
leastprevalent
prevalentininthe thetest
testarticle
articletested
testedunderundervacuum
vacuumcon-
conditions
ditions andand are
aresmaller
smallerininsizesizeasaswell.
well.Likely,
Likely, imperfections
imperfections in the process
in the processof chemical
of chemical
vapor infiltration leads to the initial creation of these pits. The pits could
vapor infiltration leads to the initial creation of these pits. The pits could be enlarged be enlarged by by
the the
vaporization
vaporization of leftover
of leftoverresin material
resin from
material fromthe the
laying out out
laying of carbon
of carbon fibers, or from
fibers, or from
thermal stresses caused by the laser preheating. These pits are
thermal stresses caused by the laser preheating. These pits are also evident on the also evident on the test test
article that that
article underwent
underwentN2 flow,
N2 indicating that at that
flow, indicating elevated temperatures
at elevated this material
temperatures this likely
material
weakens
likely weakens and is mechanically blown out in sections by the flow. As evident in13c,
and is mechanically blown out in sections by the flow. As evident in Figure Figure
much 13c,ofmuch
this CVI carbon
of this CVImatrix
carbon is missing,
matrix iswith sharp-shaped
missing, fibers left behind.
with sharp-shaped fibers left Likely,
behind.
thisLikely,
is the material
this is thethat is weakened
material that is and mechanically
weakened blown out blown
and mechanically by the outflowby first.
the Lastly,
flow first.
the Lastly,
oxidation thepits on the pits
oxidation CVIon carbon
the CVImatrix are matrix
carbon large and areplentiful
large andon both theon
plentiful test article
both the test
thatarticle
was heated under atmospheric conditions and under M = 6 air
that was heated under atmospheric conditions and under M = 6 air flow. Therefore, flow. Therefore, it
is noted that these pits are local regions where oxidation occurs and causes structural
weaknesses. The structural failures of the CVI carbon matrix are seen where there are gaps
in the structure. This structural failure occurs even without flow, as seen in the gaps in the
structure in Figure 13a.
 itit is
is noted
noted that
that these
these pits
pits are
are local
local regions
regions where
where oxidation
oxidation occurs
occurs and
and causes
causes structural
structural
weaknesses. The
weaknesses. The structural
structural failures
failures of
of the
the CVI
CVI carbon
carbon matrix
matrix are
are seen
seen where
where there
there are
are gaps
gaps
Aerospace 2025, 12, 43
in the structure. This structural failure occurs even without flow, as seen in the gaps inofthe
in the structure. This structural failure occurs even without flow, as seen in the gaps 17
in 20
the
structure in
structure in Figure
Figure 13a.
13a.

(a)
(a) (b)
(b) (c)
(c) (d)
(d)
Figure 13.
Figure 13. Fiber
Fiberbundle
Fiber bundleon
bundle onarticle
on articlesurface
article surfaceafter
surface after
after laser
laser
laser exposure
exposure during:
during:
exposure (a)
(a)(a)
during: atmospheric
atmospheric conditions;
conditions;
atmospheric (b)
conditions;
(b)
M =M6 = 6 airflow;
airflow; (c) (c)
M =M6 = 6 nitrogen
nitrogen flow;
flow; andand
(d) (d) vacuum
vacuum conditions.
conditions.
(b) M = 6 airflow; (c) M = 6 nitrogen flow; and (d) vacuum conditions.

Figure 14
Figure 14shows
14 showsaaasingle
shows single
single fiber
fiber
fiber onon
on thethe
the surface
surface
surface for each
for each
for each testedtested
tested article.
article. article.seenAs
As seen
As seen
in Figures
in Figures in
Figures
13a and13a
13a and 14a,and
14a, the14a,
the testthe
test test heated
article
article article heated
heated under
under atmospheric
under atmospheric
atmospheric conditions
conditions
conditions has carbon
has carbon
has carbon fibersfibers
fibers with
with
with
many many
many oxidation oxidation
oxidation pits. pits.
pits. These
These are These
are also are
also evident also
evident by evident
by observing by
observing the observing
the images
images of the images
of the
the test of
test articlethe test
article that
that
article
underwent
underwent that underwent
the M
the theof
M == 66 flow
flow Mair,
of = 6seen
air, flowin
seen ofFigures
in air, seen13b
Figures in and
13b Figures
and 14b.14b.13bNote
Note and the
the 14b. Notesmoothness
relative
relative the relative
smoothness
smoothness
of the
of the test of
test articles the
articles that test articles
that were
were heated that
heated in were
in the heated
the presence in
presence of the presence
of oxygen,
oxygen, Figure of oxygen,
Figure 14a,b, Figure
14a,b, whenwhen14a,b,com-
com-
when
pared compared
to the test to the
articlestest articles
heated heated
without without
oxygen oxygen
present,
pared to the test articles heated without oxygen present, Figure 14c,d. Likely, the oxygen present,
Figure Figure
14c,d. 14c,d.
Likely, Likely,
the oxygenthe
oxygen
present present
present around
around the around
the test the
test articletest
article willarticle will
will oxidize
oxidize the oxidize
the small the small
small imperfections imperfections
imperfections around around thearound
the edge the
edge of edge
of the
the
of the
internal internal
fiber fiber
first, first,
leading leading
to a to
smoother a smoother
surface, surface,
but still but
a
internal fiber first, leading to a smoother surface, but still a pointy morphology of the fiber, still
pointy a pointy
morphology morphology
of the of
fiber,
the fiber,
which agrees
which which
agrees with agrees
with [24]. with
[24]. Then, [24].
Then, small Then,
small pits small
pits will pits
will form will
form on on theform
the fiber on the
fiber where fiber
where more where
more oxidation more
oxidation takes oxida-
takes
tion
place,takes place,
creating creating
voids as voids
described as described
by [25], by
either [25],
due either
place, creating voids as described by [25], either due to elevated temperature, structuralto due
elevated to elevated
temperature, temperature,
structural
structural
weakness,weakness,
weakness, or as
or as [26] or as [26]
[26] states,
states, thatstates,
that oxygen
oxygen that oxygen
diffuses
diffuses atdiffuses
at that specific
that at that
specific specificAlso
location.
location. location.
Also of note
of note Also of
is the
is the
note is the
sharpness of
sharpness sharpness
of the
the fibers. of the
fibers. The fibers.
The fibers
fibers onThe
on thefibers
the test on the
test article test
article that article
that underwent that underwent
underwent nitrogen nitrogen flow nitrogen
flow areare
flow are
sharper than
sharper sharper
than the than
the fibers the
fibers on fibers
on the
the other on the
other test other
test articles. test
articles. The articles.
The flow The
flow itself flow
itself appears itself
appears to appears
to have to
have decreased have
decreased
the radial size of the fiber through mechanical abrasion and sheared away the away
decreased
the radial the
size radial
of the size
fiber of the
through fiber through
mechanical mechanical
abrasion abrasion
and sheared and sheared
away the edgesthe
edges on
on
edges on
the fibers
fibers of the
of the fibers
the test of the
test article test
article that article
that underwent that
underwent nitrogenunderwent
nitrogen flow, nitrogen
flow, rather flow,
rather than rather
than corrode
corrode andthan corrode
and smooth
smooth
the
and
the smoothlike
surface the the surface like of
process theoxidation
process of oxidation
does. This does. of
process This process ofabrasion
mechanical mechanical and
the surface like the process of oxidation does. This process of mechanical abrasion and
abrasion and
edge-shearing is edge-shearing
is also
also present
present on is also
on the present
the test
test articleon the
article that test article
that underwent
underwent the that underwent
the flowflow ofof air, the
air, given flow
given that of
that
edge-shearing
air, given
the fibers that
fibers form the
form sharper fibers form
sharper points sharper
points than than the points
the test than
test article the test
article heated article
heated under heated
under atmospheric under atmospheric
atmospheric conditions.
conditions.
the
conditions.
However, these However,
these processes these
processes are processes
are occurring are occurring
occurring simultaneously simultaneously
simultaneously with with the with the
the smoothing-out smoothing-out
smoothing-out mecha- mecha-
However,
mechanism
nism of of oxidation,
oxidation, leading leading
to points to points
not as not
sharpas sharp
as those as under
those undernitrogen nitrogen
flow. flow.
nism of oxidation, leading to points not as sharp as those under nitrogen flow.

(a)
(a) (b)
(b) (c)
(c) (d)
(d)
Figure 14.
Figure
Figure 14. Individual
14.Individual
Individual fiber
fiber onon
on
fiber thethe
the article
article surface
surface
article afterafter
after
surface laserlaser
laser exposure
exposure during:
during:
exposure (a) atmospheric
(a)
during: atmospheric con-
con-
(a) atmospheric
ditions;
ditions; (b)
(b) M
conditions; M =
(b)=M6 airflow;
6 airflow; (c)
(c) M
= 6 airflow; M =
(c)=M6 nitrogen
6 nitrogen flow; and
flow;flow;
= 6 nitrogen and and (d) vacuum
(d) vacuum conditions.
conditions.
(d) vacuum conditions.

Figures 13d and 14d shows the surface structure after the test article is exposed to
laser heating under vacuum conditions and no flow: the surface structure is not different
when compared to the initial surface structure after the machining of the test article. Still,
Aerospace 2025, 12, 43 18 of 20

note the initial imperfections in the material caused by machining and imperfections in the
CVI process.

5. Discussion and Conclusions

Designing vehicles capable of withstanding the harsh conditions exhibited during
hypersonic flight is a rather difficult problem to tackle. Specifically, dealing with the high
heat flux generated at the leading edges of the vehicle and in the engine is key for the
survival of the vehicle. According to available publications [5,23] related to existing and
prospective thermal protection systems, conventional materials may prove to be ready
for applications [9], but they have limited capabilities in terms of withstanding their
maximum surface temperature and their survivability in the harsh environment during
hypersonic flight. For thermal protection systems of hypersonic vehicles, materials such
as C–C composites may be a reasonable selection thanks to the capacity of maintaining
high mechanical strength at temperatures greater than 2000 K [2,26]. However, lower cost
of the materials fabrication, reduction in the physical parameters’ uncertainties, and a
detailed physical model of the material’s interaction with high-enthalpy flows are still of
high demand.
A combination of ground testing to study the phenomena, computational modeling,
and flight testing is a proper pathway to routine practical implementation. The ground
testing can simulate the harsh environment relevant to hypersonic flight and allows the
collection of a basic dataset to characterize material properties and the details of behavior
on macro and micro scales. In this work, the commercial C–C composite material was
tested under high thermal flux, causing a steady state surface temperature up to 2000 K at
variable conditions, including open atmosphere, vacuum, and high-speed flows of both air
and nitrogen.
This testing allowed for the accurate and consistent measurement of material mass loss
rates due to thermal ablation and oxidation. These processes are explained qualitatively
through the analysis of scanning electron microscopy images and high-framerate live
videos. Quantitatively, these processes were characterized through mass tracking and
surface and internal temperatures measurements. High-framerate images taken during
M = 6 flow testing provided insights on how superheated small flakes mechanically broke
from the article, showing evidence of ablation thus leading to rapid destruction of the
material. Scanning electron microscopy images revealed that, under quiescent atmospheric
air at high temperature, the sheathing material surrounding the individual carbon fibers
exhibited a rough surface, with numerous pits created by the oxidation process. Under
hypersonic air flow, in which oxidation and ablation are present, the morphology of the
material was observed to be pointy and that the carbon fiber was oxidizing slower than the
sheathing material, which qualitatively agrees with the results found by other authors [24].
However, the microscopic images for the article that was tested under hypersonic nitrogen
flow exhibited a different behavior. The exposed fibers are shaped into pointy forms,
sharper compared to the ones exposed to air flow, indicating the effect of accelerated
oxidation. This shows that, under nitrogen flow, the mechanical erosion in combination
with thermal stresses induced by the high thermal gradient weaken the structure of the
carbon fiber, creating an abrasive effect eroding the fibers.
The main contributor to the material mass loss process was determined to be oxidation,
followed by the combined thermal ablation and mechanical breakage of the fibers due to
flow mechanical stress. The results of the ground testing of the test article with dimensions
d × x = 25.4 × 25.4 mm show that the mass loss rate of C–C composites in M = 6 flow
was up to 91.0 ± 0.1 mg/s for air flow at temperatures exceeding 1300 K. Keeping in
mind the front surface area S of the test article, these numbers correspond to the mass loss
Aerospace 2025, 12, 43 19 of 20

flux (1/S)(∂m/∂t > 0.2 kg/m2 s), or the material linear decrease ∂x/∂t > 0.1 mm/s. Under
a M = 6 flow of nitrogen, the mass loss rate was measured as high as 31.2 ± 0.1 mg/s.
Under air flow, the intense oxidation process increases the mass loss rate compared to
that under nitrogen flow. Under a M = 6 flow of air, about two thirds of the mass loss
rate can be attributed to the effects of oxidation. Under nitrogen flow the recession occurs
due to thermal/mechanical ablation and additional structural weakening of the fibers.
For reference, in atmospheric quiescent air, the mass loss was around 3.0 ± 0.1 mg/s for
the same heat flux exposure. Under vacuum conditions and identical heat flux exposure,
testing proved that a very small to negligible thermal ablation occurs (<0.1 mg/s) with
no oxidation.
The study concluded that the destruction mechanism of C–C composite materials
under high-speed airflow differs significantly from one under stagnant conditions. As
a result, there was an increase of about 1.5 orders of magnitude in the rate of material
destruction. As part of the ongoing study, results from the uncoated C–C composites will
be compared to those from coated articles, including their oxidation rate and thermal shock
resistance. Future work will include data processing refinement to improve the graphing
visualization, and more vacuum and hypersonic flow testing. Thermal and mechanical
stresses will also be examined in the degradation of mechanical properties and the recession
of composite materials.

Author Contributions: Each author contributed to the work on this research in multiple ways as
follows. Conceptualization, S.B.L. and R.G.; methodology, A.I.S. and R.B.; formal analysis A.I.S. and
R.B.; investigation, A.I.S., R.B. and S.B.L.; resources, S.B.L.; data curation S.B.L.; writing—original
draft preparation, R.B.; writing—review and editing, A.I.S., S.B.L. and R.G.; visualization, R.B.;
supervision S.B.L. and R.G.; funding acquisition, S.B.L. All authors have read and agreed to the
published version of the manuscript.

Funding: This research was funded by the US Air Force Office of Scientific Research (PM Dr. Fariba
Fahroo), grant number FA9550-22-1-0065.

Data Availability Statement: Due to privacy restrictions, data used for the evaluation and creation of
the results presented in this manuscript remain unavailable for public access. However, some data
related to specific measurements could be shared on demand.

Acknowledgments: Three-directional needled felt PAN-based carbon fiber reinforced carbon matrix
commercial composites were kindly provided by Honeywell Aircraft Landing Systems. The authors
are grateful to Philip Lax for his help in schlieren visualization. The authors would also like to thank
Tatyana Orlova at the Notre Dame Integrated Imaging Facility for guiding the electron microscopy
visualization of the exposed articles.

Conflicts of Interest: The author Richard Gulotty is employed by the company Honeywell Inter-
national Inc. The remaining authors declare that the research was conducted in the absence of any
commercial or financial relationships that could be construed as a potential conflict of interest. The
material presented in this work was reviewed by Honeywell Aircraft Landing Systems and approved
for publication.

References
1. Van Wie, D.M. Hypersonics: Past Present, and Potential Future. Johns Hopkins APL Tech. Dig. 2021, 35, 335–341.
2. Glass, D.E. Thermal Protection Systems and Hot Structures for Hypersonic Vehicles. In Aerospace Materials and Applications;
American Institute of Aeronautics and Astronautics: Reston, VA, USA, 2018; pp. 531–578.
3. Tenney, D.R.; Lisagor, W.B.; Dixon, S.C. Materials and structures for hypersonic vehicles. J. Aircr. 1989, 26, 953–970. [CrossRef]
4. Zhang, S.; Li, X.; Zuo, J.; Qin, J.; Cheng, K.; Feng, Y.; Bao, W. Research progress on active thermal protection for hypersonic
vehicles. Prog. Aerosp. Sci. 2020, 119, 100646. [CrossRef]
5. Peters, A.B.; Zhang, D.; Chen, S.; Ott, C.; Oses, C.; Curtarolo, S.; McCue, I.; Pollock, T.M.; Prameela, S.E. Materials design for
hypersonics. Nat. Commun. 2024, 15, 3328. [CrossRef] [PubMed]
Aerospace 2025, 12, 43 20 of 20

6. Zhu, T.; Wang, Z. Advances in processing and ablation properties of carbon fiber reinforced ultra-high temperature ceramic
composites. Rev. Adv. Mater. Sci. 2024, 63, 20240029. [CrossRef]
7. Ohlhorst, C.W.; Glass, D.E.; Bruce, W.E.; Lindell, M.C.; Vaughn, W.L.; Dirling, R.B.; Hogenson, P.A.; Nichols, J.M.; Risner, N.W.;
Thompson, D.R. Development of X-43-A Mach 10 Leading Edge. In Proceedings of the 56th International Astronautical Congress
in 2005, Fukuoka, Japan, 16–21 October 2005.
8. Glass, D. Physical Challenges and Limitations Confronting the Use of UHTCs on Hypersonic Vehicles. In Proceedings of the 17th
AIAA International Space Planes and Hypersonic Systems and Technologies Conference, Santa Clara, CA, USA, 11–14 April 2011;
pp. 2011–2304.
9. Fahrenholtz, W.G.; Hilmas, G.E. Ultra-high temperature ceramics: Materials for extreme environment. Scr. Mater. 2017, 129, 94–99.
[CrossRef]
10. Windhorst, T.; Blount, G. Carbon-carbon composites: A summary of recent developments and applications. Mater. Des. 1997, 18,
11–15. [CrossRef]
11. Ohlhorst, C.W.; Vaughn, W.L.; Ransone, P.O.; Tsou, H.-T. Thermal Conductivity Database of Various Structural Carbon-Carbon
Composite Materials; NASA: Linthicum Heights, MD, USA; NTIS: Springfield, VA, USA, 1997.
12. Barrios-Lobelle, A.; Davuluri, R.; Fu, R.; Martin, A.; Poovathingal, S.J. Surface Oxidation of Carbon/Carbon Composites in
Hypersonic Environments. In Proceedings of the AIAA Scitech 2021 Forum, Online, 11–15,19–21 January 2021.
13. Keswani, S.T.; Kuo, K.K. Validation of an Aerothermochemical Model for Graphite Nozzle Recession and Heat-Transfer Processes.
Combust. Sci. Technol. 1986, 47, 177–192. [CrossRef]
14. Borie, V.; Brulard, J.; Lengelle, G. Aerothermochemical Analysis of Carbon-Carbon Nozzle Regression in Solid-Propellant Rocket
Motors. J. Propuls. Power 1989, 5, 665–673. [CrossRef]
15. Thakre, P.; Yang, V. Chemical Erosion of Carbon–Carbon/Graphite Nozzles in Solid-Propellant Rocket Motors. J. Propuls. Power
2008, 24, 822–833. [CrossRef]
16. Panerai, F.; Cochell, T.; Martin, A.; White, J.D. Experimental measurements of the high-temperature oxidation of carbon fibers.
Int. J. Heat Mass Transf. 2019, 136, 972–986. [CrossRef]
17. Ge, H.; Zhang, L.; Zhang, H.; Wang, F.; Gao, X.; Song, Y. Cyclic Ablation Properties of C/SiC-ZrC Composites. Aerospace 2024, 11,
432. [CrossRef]
18. Zhou, W.; Xiao, P.; Luo, W.; Li, Y. Microstructural evolution of SiC coating on C/C composites exposed to 1500 ◦ C in ambient air.
Ceram. Int. 2019, 45, 854–860. [CrossRef]
19. Marshall, L.A.; Bahm, C.; Corpening, G.; Sherrill, R. Overview With Results and Lessons Learned of the X-43A Mach 10 Flight. In
Proceedings of the AIAA Scitech Forum, Capua, Italy, 16–20 May 2005; pp. 1–23.
20. Slobodkin, L.S.; Flyaks, M.Y. Emissivity measurement for a carbon-carbon composite and a phenolic carbon plastic. J. Eng. Phys.
1989, 57, 972–975. [CrossRef]
21. Sun, X.; Yang, H.; Mi, T. Heat Transfer and Ablation Prediction of Carbon/Carbon Composites in a Hypersonic Environment
Using Fluid-Thermal-Ablation Multiphysical Coupling. Int. J. Aerosp. Eng. 2020, 2020, 9232684. [CrossRef]
22. Zhang, J. Oxidation Protection of Carbon/Carbon Composites and Non-Destructive Characterization Methodology Development.
Ph.D. Thesis, Sorbonne University, Paris, French, 2019.
23. Bertin, J.J.; Cummings, R.M. Fifty years of hypersonics: Where we’ve been, where we’re going. Prog. Aerosp. Sci. 2003, 39, 511–536.
[CrossRef]
24. Farhan, S.; UI-Haq, N.; Kuo, W.-S. Degradation behavior of 4D carbon/carbon composites under supersonic oxidative air plasma.
Ceram. Int. 2013, 39, 7135–7142. [CrossRef]
25. Jacobson, N.; Curry, D. Oxidation Microstructure Studies of Reinforced Carbon/Carbon. Carbon 2006, 44, 1142–1150. [CrossRef]
26. Han, M.; Zhou, C.; Sillberschmidt, V.; Bi, Q. Oxidation behaviour and residual mechanical properties of carbon/carbon composites.
Carbon Lett. 2023, 33, 1241–1252. [CrossRef]
27. Fradin, M.; Vignolis, G.; Rebillat, F.; Haras, K.; Gregis, C. Oxidation behavior of nozzle throat carbon/carbon composites featuring
variable densities. In Proceedings of the 9th European Conference for Aeronautics and Space Sciences, Lille, France, 27 June–1
July 2022; pp. 1–10.
28. Ozcan, S.; Filip, P. Wear of carbon fiber reinforced carbon matrix composites: Study of abrasive, oxidative wear and influence of
humidity. Carbon 2013, 62, 240–247. [CrossRef]
29. Ozcan, S.; Tezcan, J.; Filip, P. Microstructure and elastic properties of individual components of C/C composites. Carbon 2009, 47,
3403–3514. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.