DTIC ADA534233: Silicone Polymer Composites for Thermal Protection System: Fiber Reinforcements and Microstructures

2.1 MB·English

by  Defense Technical Information Center

JOURNAL OF
COMPOSITE
Article MATERIALS
JournalofCompositeMaterials
0(0)1–18
Silicone polymer composites for thermal !TheAuthor(s)2010
Reprintsandpermissions:
sagepub.co.uk/journalsPermissions.nav
protection system: fiber reinforcements
DOI:10.1177/0021998310381536
jcm.sagepub.com
and microstructures
J.H. Koo1, M.J. Miller2, J. Weispfenning3 and C. Blackmon4
Abstract
A new class of silicone polymer matrix composites was evaluated using a simulated solid rocket motor test apparatus.
Conversionofthisorganicsiliconepolymertoaceramic(i.e.silica)structureonexposuretoflameimpingementorhigh
temperature, accountsforitsoutstandingthermalstability.Aresearchprogramwas aimedtodevelopandevaluatethis
newclassofthermalprotectionmaterialsformilitaryapplications.Thisarticlepresentstheeffectsofthetypeandform
of reinforcements on the thermal performance of a novel class of silicone polymer matrix composites. Reinforcement
typessuchasglass,silica,quartz,NextelTM,andNicalonTMwereused.Reinforcementformssuchasrandomcontinuous-
fibermat,chopped-fibermat,2-Dfabric,3-Dfabric,choppedroving,andbroadgoodtapeswithdifferentplyangleswere
tested. Detailed microstructural, mass loss, and peak erosion analyses were conducted on the phenolic-based matrix
composite (control) and silicone-based matrix composites to understand their protective mechanisms.
Keywords
silicone polymer matrix composites, phenolic polymer matrix composites, thermal protection materials, ablatives,
thermal protection systems
Introduction properties relevant to the performance of thermal pro-
tection materials was reviewed. The theoretical and
Background
experimental techniques used to determine thermophy-
Thermal protection materials are required to protect sicalpropertiessuchasdensity,specificheat,decompo-
structural components of space vehicles during the sition kinetic parameters, melt temperature, and
re-entry stage, missile launching systems, and solid thermochemical expansion of polymeric composites
rocket motors (SRMs). A thorough literature survey were summarized. Additionally, thermophysical prop-
was conducted in a series of three papers by Koo erties of selected virgin and charred ablative materials
et al.1–3 for different military and aerospace applica- at elevated temperature were presented. Experimental
tions.Theliteraturesurveywasgroupedinto:(a)numer- testing literature3 as well as erosion and heat transfer
ical modeling,1 (b) material thermophysical properties
characterization,2and(c)experimentaltesting.3
Numerical modeling literature1 was further subdi- 1Department of Mechanical Engineering, The University of Texas at
Austin,TexasMaterialsInstitute,Austin,TX78712-0292,USA
vided into groups based on how the ablation models
2BAESystems,Minneapolis,MN55421-1498,USA
and the computational fluid dynamic analyses were 3CytecEngineeredMaterials,Winona,MN55987-2854,USA
used for different applications. The different applica- 4NavalSurfaceWarfareCenter,DahlgrenDivision,Dahlgren,VA22448-
tion groups were described as thermal protection 5100,USA
system (TPS) for spacecraft, ablative material for mis-
Correspondingauthor:
sile launching systems, internal insulation for SRMs,
J.H.Koo,DepartmentofMechanicalEngineering,TheUniversityofTexas
and nozzle assembly for SRMs. Material properties
atAustin,TexasMaterialsInstitute,Austin,TX78712-0292,USA.
characterization literature2 of various thermophysical Email:[email protected]
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
2010 2. REPORT TYPE 00-00-2010 to 00-00-2010
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Silicone polymer composites for thermal protection system: fiber
5b. GRANT NUMBER
reinforcements and microstructures
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
Naval Surface Warfare Center,Dahlgren REPORT NUMBER
Division,Dahlgren,VA,22448-5100
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
14. ABSTRACT
A new class of silicone polymer matrix composites was evaluated using a simulated solid rocket motor test
apparatus. Conversion of this organic silicone polymer to a ceramic (i.e. silica) structure on exposure to
flame impingement or high temperature, accounts for its outstanding thermal stability. A research
program was aimed to develop and evaluate this new class of thermal protection materials for military
applications. This article presents the effects of the type and form of reinforcements on the thermal
performance of a novel class of silicone polymer matrix composites. Reinforcement types such as glass,
silica, quartz, NextelTM, and NicalonTM were used. Reinforcement forms such as random continuousfiber
mat, chopped-fiber mat, 2-D fabric, 3-D fabric, chopped roving, and broadgood tapes with different ply
angles were tested. Detailed microstructural, mass loss, and peak erosion analyses were conducted on the
phenolic-based matrix composite (control) and silicone-based matrix composites to understand their
protective mechanisms.
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE Same as 18
unclassified unclassified unclassified Report (SAR)
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
2 Journal of Composite Materials 0(0)
datarelevant to theperformance ofthermal protection transient recession on the surface of the ablative mate-
materials for different applications were reviewed. rial when other erosion models are described by the
Laboratory scale apparatus, small-scale, and subscale correlation of the experiment data.
SRMswereusedtotestthesethermalprotectionmate- Studies of Yeh4 have been focused primarily on the
rials. In this article, we include only the literature rele- thermochemicalablation.Theablativematerialusedin
vant to thermal protection materials used for missile theVLSenvironmentisnotonlysubjectedtoathermal
launching systems. attackbutalsosubjectedtoaparticleimpact. Boththe
Theverticallaunchingsystem(VLS)isaspecialkind mechanical erosion due to the particle impact and the
of missile launching system used on surface combat thermochemical ablation caused by the thermal attack
shipsbytheU.S.Navy.Itemploysindividualcanisters can be significant on the surface of the ablative mate-
to store and maintain the missiles in a ready-to-fire rial. To assure the integrity and prolong the lifetime of
condition. The present design of the VLS, originally launcher components, suitable information about the
used for STANDARD missiles and later modified for performance of the ablatives that are exposed to this
TOMAHAWKmissiles,consistsofeightcanisterswith high-temperature, particle-laden flow requires further
missiles as well as a plenum chamber and a central study.
uptake duct. Eight missile modules in the VLS, stored A transient, 2-D material erosion model was devel-
beneath the deck of the ship, have separate cell doors oped by Yang et al.5 to describe the simultaneous pro-
over each canister and can be momentarily opened at cesses of thermochemical ablation and mechanical
the time of missile firing. Following the launching of a erosion of ablative materials. The model predicts the
missile, a two-phase, high temperature and high veloc- rate of surface recession and the erosion pattern.
ityrocketexhaustplumeisproducedduetoburningof Numerical calculations of the surface erosion of
the aluminized propellant in the SRM. The exhaust H41N (phenolic-chopped glass fiber composite) were
plume of the SRM travels downward into the plenum performed. It was found that particle impact substan-
chamber and is then directed out of the VLS through tially alters the thermochemical ablation and thermal
the central uptake duct. effect also changes the mechanical erosion. The pre-
The challenge for the integrity, survivability, and dicted results for overall material erosion of H41N
maintenance of the VLS is met through its innovative arefoundtoagreereasonablywellwithavailableexper-
mechanical design and the extensive use of ablative imental data.
materials. The gas management of the rocket exhaust In the preceding studies, the eroded material pro-
plume, which is of particular concern in the develop- ducedbyeitherthermalablationorparticleimpactero-
mentofVLS,involvesthefollowingissues:strongpres- sion was assumed to be effectively removed by the
sure waves traveling through the canister, and plenum ambient gas stream. As such, no melt layer would
and uptake duct govern the mechanical strength of the formonthesurfaceofthematerial.Inreality,however,
design. Turbulent convective heat transfer from the a melt layer can form on the surface of the material as
rocket plume provides the largest thermal input on observed in the experimental studies of Chaboki et al.6
theinteriorsurfaceoftheVLS.Inaddition,theprocess and Koo et al.7 The composition of the melt layer
ofcompression andexpansionwavesdecays asaresult might include the deposit of the incoming particles
of the interaction of shock, turbulence, and particle- and the eroded high-temperature ablative material.
induced viscous dissipation in the rocket plume. Yang, Cheung, and Koo5,8,9 developed a model for
Yeh4 investigated theoretically and experimentally the ablative wall material which couples the effects of
the effect of internal gas flows on material erosion mechanical erosion due to particle impact and thermo-
and transient internal ballistic processes in a scaled chemical ablation. Their model assumes that particles
VLS. As the two-phase plume impinges the ablative impact and reflect off of the ablative material surface.
material, MXB-360 (phenolic glass mat composite), Duringtheimpact,thekineticenergyoftheparticlesis
located at the bottom of the VLS plenum where ther- transferred as thermal energy to the surface. They
malerosionoftheablativematerialoccurs.Meanwhile, observed that mechanical erosion due to particle
a debris layer, consisting of the liquid-phase aluminum impact is comparable in magnitude to thermochemical
oxidesintheplume,formsonthesurfaceoftheablative ablation. Cheung et al.10 extended the Yang et al.
material, which significantly changes the heat flux to model to include the formation of a layer of molten
the ablative surface. The theoretical part of this study aluminum oxide on the ablative surface. Cheung et al.
aims at mathematically formulating the equations gov- calculated the melt layer thickness near a stagnation
erning the behavior of the plume in the VLS plenum, point by numerically solving integral continuity and
the debris layer between the plume and the ablative momentum equations of the melt layer. The layer acts
material, and the ablative material itself. This coupling to protect the ablative surface by slowing down parti-
model is the first analytical effort to predict the cles before they strike the surface, thus decreasing
Koo et al. 3
mechanical erosion and reducing thermal ablation by Themodelpredictsthedebrislayerthicknessandheat-
increasing the heat transfer resistance. ing to the launcher walls. Integral conservation equa-
Theuseofaluminumtoincreasetheperformanceof tionsaresolvedtopredictthedebrislayerthicknessand
solidpropellantsyieldsatwo-phaseexhaustplumecon- temperature distribution. An implicit finite-difference
sisting of solid and/or liquid aluminum oxide particles equation is solved to predict solidification of the
(Al O ) and gaseous combustion products. Some cur- debris layer on the launcher walls. The model was val-
2 3
rent U.S. Navy rocket motors contain substantial idatedbypredictingerosionofH41Nablativematerial
amounts of aluminum to the extent that 40wt.% of exposed to the plume of an Mk-36 Sidewinder rocket
the exhaust products may be composed of Al O . motor at a standoff distance of 0.91m. The molten
2 3
These aluminum oxides carry a significant amount of debrislayerattenuatesheatingduetoparticleimpinge-
thermal and kinetic energy that is transferred to the ment and convective heating. The model underpre-
launcher when these particles impinge on the launcher dicted erosion by 25%.
walls.Highlylocalizederosionoccurswherealuminized Shih, Cheung, Koo, and Yang13–15 developed a
plumes impinge directly on the launcher wall. A thin physical model to describe the transient ablation phe-
layer of Al O has been observed to form on the nomenon of high-temperature ablative materials for
2 3
launcher walls. Exactly how this layer develops and with and without the formation of a melt layer on the
affects surface heat transfer and ablation is still unre- material surface. The model hasbeen applied to H41N
solved. In the early stages of a missile launch, liquid (larger heat of ablation) and to MXBE-350 (smaller
aluminum particles transport a significant amount of heat of ablation). When exposed to the same external
thermal energy to the walls. Later, after the melt layer heat flux,the growth rate of the melt layer for MXBE-
thickens, the layer of Al O may actually shield the 350 (phenolic rubber-modified glass composite) is
2 3
wallsfromtheexhaustplume.Thislayermustbemod- found to be considerably higher than that of H41N.
eledaccuratelytoestimateheattransferfromtheplume Also, much larger differences in the ablation rates for
to the launcher walls and predict ablation of the thecaseswithandwithoutameltlayerareobservedfor
launcher walls. MXBE-350whencomparedtoH41N.Foragivenabla-
A typical flow inside a VLS is highly complex and tive material the difference in the predicted ablation
contains mixed supersonic and subsonic regions, vari- rates between the cases with and without a melt layer
ous shock structures, turbulent shear layers, and recir- becomes appreciably larger as the imposed external
culationregions.TheAl O particlespossessinggreater heat flux is increased. When a high-temperature abla-
2 3
momentum and thermal inertia than the gaseous flow, tivematerialwithalowheatofablationisexposedtoa
generallylagthegaseousflowinregionsofvelocityand high external heat flux, the melt layer tends to grow
temperaturegradients.Therehavebeenattemptsatthe quickly, and as such, a large portion of energy is
Naval Surface Warfare Center Dahlgren Division to likely to be trapped within the melt layer under such
model the launcher flow using a time-dependent, fully circumstances. The protective effect of the melt layer
coupled two-phase Navier-Stokes solver, CRAFT.11 cannot be neglected at high external heat fluxes for
York et al.12 computed particle heat transfer and materials such as MXBE-350 which has small heat of
surface ablation on a flat plate mounted perpendicu- ablation. Additional experimental data are needed to
larly to a plume from an Mk-36 Sidewinder rocket further validate the Shih et al. model.15
motor. The fully coupled multi-phase Navier-Stokes Hendersonandco-workers16–27publishedaseriesof
flow solver CRAFT was used to model the plume papersfrom1980to1991whichdescribedtheirthermo-
impingement. The authors considered both kinetic physical properties measurement techniques and
energyandthermalenergytransferfromtheAl O par- numerical modeling as well as thermal data of thermal
2 3
ticlestotheablativematerial.Totalerosionpredictions protection materials formissile launching systems such
were significantly higher than experimental data. For as H41N and MXBE-350. Henderson and Tant20
their calculations, the authors assumed a wall temper- reportedmethodsindeterminingthespecificheat,ther-
atureequaltothefailtemperatureoftheablativemate- mochemical expansion, heat of decomposition, and
rial (1600(cid:2)C). They reasoned that the presence of a kineticparameters forboth thelow-temperaturepyrol-
layer of molten Al O would attenuate the particle ysis reactions and the high-temperature carbon–silica
2 3
heating. Subsequently calculations made assuming a reactions for MXBE-350.
wall temperature of 2300(cid:2)C (approximately the melt Henderson and Emmerich22 found experimentally
temperature of Al O ) resulted in better agreement that alteration of the thermophysical properties
2 3
with the experimental data. induced large changes in the thermally induced
A model has been developed by Lewis and response of a certain glass-filled polymer composite.
Anderson11 to predict the formation of a molten The authors reported how to determine the tempera-
Al O debris layer in ducted missile launchers. ture-dependent specific heat of the virgin, char, and
2 3
4 Journal of Composite Materials 0(0)
decomposingmaterialandtheheatofdecompositionof electronmicroscope(SEM)wereconductedonthephe-
the pyrolysis reactions. Henderson et al.22,24,25 used a nolic-based and silicone-based matrix composites to
simultaneous thermal analyzer (STA) to obtain the understandtheirprotectivemechanismsandtheircom-
experimental data and computed the properties using parative ablation performance using SSRM. Peak ero-
ratio method and numerical integration. Powder abla- sion and mass loss of post-test composite specimens
tive samples were machined from larger blocks, then were used as the major criteria to determine the mate-
filtered through a sieve, and stored overnight in rial performance. Selective materials were tested using
vacuum at 35(cid:2)C to remove traces of moisture. To the scaled ducted launcher (SDL) test apparatus under
reduce temperature gradient in the material, 10–15mg solid propellant rocket exhaust plume.37 The overall
of the samples were used and maintained in an argon development and evaluation of this novel class of
atmosphere at purge rate of 100mL/min to prevent silicone polymer matrix composites can be found in
thermo-oxidativedegradationduringtesting.Allexper- Koo et al.37
iments were conducted at heating rates of 20(cid:2)C/min
from 30(cid:2)C to 1050(cid:2)C. An empty sample pan was run
to establish a baseline for correction. Char samples Materials: constituents and fabrication
were obtained by precharring a sample of the virgin
Polymer system
material in the STA to 1050(cid:2)C under the same condi-
tions as the other tests. Silicone polymers can be converted into ceramics
Boyer28developedanimplicitfinite-differencemodel through high thermal heating applications.38 This
tosimulatetheheattransferand expansionincharring method has been used to prepare ceramic parts and
ablators.Variablethermophysicalproperties,endother- obtain ceramic fibers. Limited work has been reported
mic decomposition, convective cooling by pyrolysis on the development of fiber-reinforced silicone
gases, and thermochemical expansion were accounted composites.
in this model. The ablative test sample was instru- A new class of thermal protection materials based
mented with in-depth thermocouples which recorded on silicone polymers was developed. This family of
temperature. Expansion of the sample was also mea- materials is coded as ‘SM8000.’ The silicone polymers
sured. The radiant heat flux history at the ablator sur- used are generically categorized as polysiloxanes.39,40
facewasmonitoredbytworadiometers.Thedatawere Figure 1 shows the chemical structures of this silicone
used to determine the thermophysical properties of polymer system and consists of a homopolymer and a
MXBE-350andtovalidatetheaccuracyoftheseprop- copolymer of polysiloxanes. The curing of these poly-
erties. The thermophysical properties of MXBE-350 mers is facilitated by a catalyst. Curing involves the
were evaluated by combining the above model and ‘crosslinking’ of polymer chains to form a rigid 3-D
experimentdatawithanonlinearestimationalgorithm. polymer network. With exposure to high temperatures
Thespecificheat,thermalconductivity,andcoefficients or flame impingement, the organic polymer decom-
of expansion of the ablator and the specific heat of the poses with the evolution of volatile compounds (i.e.
pyrolysis gas were accurately determined. Both in- water, methanol, and carbon dioxide). Silica (SiO ) is
2
depth temperatures and thermochemical expansion formed during this process.
were accurately predicted with Boyer’s model.
Previous studies by Koo and co-workers29–36 have
been instrumental in undertaking to study a new class
of high-temperature silicone polymer composites and Homopolymer
compare them with the more conventional phenolic
composites. This article summarizes the portion of the R R R
1 1 1
n=10,000
research in which the effects of the type and form of R Si O Si O Si R R =R =meth yl, ethyl
reinforcementsaswellasthetypeandamountoffillers 3 3 1 2or vinyl
on the thermal performance of the silicone polymer R2 R2 nR2
matrix composite. Reinforcement types such as glass,
silica, quartz, NextelTM, and NicalonTM were used.
Copolymer
Reinforcement forms such as random continuous-
fiber mat, chopped-fiber mat, 2-D fabric, 3-D fabric,
R R R
4 5 6 n=10,000
chopped roving, and broadgood tapes with different
ply angles were tested using the high-temperature sim- HO Si O Si O Si O H R4=R5=R6=methy l,
ethyl, vinyl or phenyl
ulatedsolidrocketmotor(SSRM)testapparatus.Silica R R R
4 5 6 n
andaluminafillersofdifferentamountswereevaluated.
Detailed microstructural analyses using scanning Figure 1. Chemicalstructuresof ‘SM8000’silicone polymer.
Koo et al. 5
A resin matrix formulation based on silicone poly- curing, either by compression or autoclave molding.
mers that provided good processability and excellent Typical cure cycles for the SM8000 laminates are in
thermal protectioncharacteristics hasbeen successfully the following steps: (a) apply vacuum and 1.03MPa
developed. The silicone resin matrix is prepared by pressure; (b) ramp for room temperature (RT) to
blending the silicone polymers, catalyst, and filler 130(cid:2)C and hold for 1h to debulk and gel panel; (c)
(silica or alumina). The blend is then used to impreg- ramp from 135(cid:2)C to 191(cid:2)C and hold for 2h to cross-
nate the reinforcing fibers. The impregnated fibers are link the resin; and (d) postcure (freestanding) at 260(cid:2)C
referredtoas‘prepreg.’Atthisstage,nochemicalreac- is optional.
tion of the polymers in the resin blend has occurred. Fabrication of 0.91(cid:3)0.91m2 panels was demon-
Hitco Technology is the original proprietor of the sili- strated for both 0(cid:2) ply orientation of SM8024 and 30(cid:2)
cone polymer composition. This silicone polymer shingle angle SM8027 and SM8029. All of the initial
matrix technology was subsequently licensed to Cytec developmentworkwasevaluatedonpanelsof0.138m2
Engineered Materials, Inc. maximum. The selected scale-up material, SM8029 30(cid:2)
shingle angle laminate fabrication was scaled up in
three stages to a surface area of 2.98m2 as follows:
Fabrication of silicone polymer matrix composites
(a) Stage 1: 0.91(cid:3)0.91m2 laminates (0.83m2); (c)
The silicone matrix resin is prepared by blending the Stage 2: 1.52(cid:3)1.22m2 laminates (1.85m2); and (c)
silicone polymers,catalyst, and filler.The blendisthen Stage 3: 2.44(cid:3)1.22m2 laminates (2.98m2).37,41
used to impregnate reinforcing fibers. These fibers can
beeitherglass,silica,quartz,Kevlar(cid:2),carbon,orother
types of fiber. The fibers can be configured in a variety Thermal performance testing
of forms or ‘architectures,’ i.e., fabrics, mats, unidirec-
Simulated solid rocket motor
tional arrays, braids, or 3-D structures. The impreg-
nated fiber is known as ‘prepreg.’ Impregnation of the An SSRM was used to evaluate the thermal perfor-
various fibers is carried out by conventional hot-melt mance of the candidate materials. The SSRM is an
prepregging and solution techniques. Table 1 summa- established test apparatus developed earlier by Koo
rizesthevariousformulationsinvestigatedinthestudy. et al.7 and has been used extensively in our previous
The molding compound (MC) versions were made by studies.29–31,34–37,41–44 The SSRM is a small-scale,
chopping impregnated fabric into squares with dimen- liquid-fueled rocket motor burning kerosene oxygen
sions of 1.27(cid:3)1.27cm2. as shown in Figure 2. Alumina (Al O ) particles are
2 3
Composite laminates are prepared by stacking a injected into the plume to simulate the particle-laden
number of plies of prepreg in a predetermined flowofsolidrocketexhaust.TheSSRMisacontrolled
sequence. The prepreg stack is then subjected to laboratorydevicecapableofproducingaparticle-laden
exhaust environment with measured heat fluxes from
2840 to 13,728kW/m2. The flame temperature is
Table 1. SM8000variants approximately 2200(cid:2)C and the velocity of the particle-
laden exhaust is approximately 2000m/s.
Constituent Example
Matrix resin Silicone polymers
Filler Silica, alumina
Fiber type Glass, silica,quartz, NextelTM,and
NicalonTM
Fiber architecture Mat,fabric,angleinterlock, roving,
and MC
Laminateply angle 0(cid:2),30(cid:2),60(cid:2),and90(cid:2)
SM8000material code SM8024– glassmat, SM8027– silica
fabric,SM8027A– silica roving,
SM8027MC – silica chopped
fabric,SM8029– quartzfabric,
SM8029A – quartzroving,
SM8029MC – quartzchopped
fabric,SM8029C– 3-Dquartz
fabric,SM8032– NextelTM
312fabric,andSM8033–
NicalonTMfabric
Figure 2. SSRM testing ablative.
6 Journal of Composite Materials 0(0)
The standard sample size is 10.2(cid:3)10.2cm2 and compare their thermal performance. The SSRM test
1.3cm thick. The composite materials are bonded to conditions were 7094kW/m2, 4.5kg/h alumina parti-
10.2(cid:3)10.2(cid:3)0.32cm3 steel substrate. A narrow slot is cles, and 90(cid:2) impingement angle for test duration of
machined into the bondline side of the steel substrate. 12s. Analyses of post-test MXBE-350 sample at mag-
Two thermocouples are embedded into this slot. The nifications of 50(cid:3) and 350(cid:3) showing glass fibers were
beads of the thermocouples are placed at the center exposed after the phenolic resin was eroded when
pointoftheplate.Thetemperaturehistoryofthebond- exposed to the extreme temperatures generated by the
line is recorded with these thermocouples. SSRM. The surface views of the post-test SM8024
The test samples were placed in a fixture down- sample at 50(cid:3) and 350(cid:3) magnifications show some
stream from the SSRM nozzle. The fixture has adjust- meltingofthesiliconeresinandnoglassfibersexposing
ments in axial distance and impingement angle with on the surface as compared to MXBE-350. The cross-
reference to the plume centerline. A normal (90(cid:2)), 60(cid:2), sectionalviewsat350(cid:3)oftheMXBE-350andSM8024
20(cid:2), and 0(cid:2) impingements were used for this study. materialsalsoshowmoreexposureofglassfibersonthe
Three axial distances were used to correspond with surface of the phenolic-based than the silicone-based
heatfluxesof2838,7094,and11,345kW/m2.Testdura- matrix composites. It was noticed that the thermal
tionof12swasused.Startingpriortothemotorburn, wave penetrated deeper into the MXBE-350 than the
the bondline temperature history was recorded for SM8024 indicating SM8024 was less affected by the
10min. Peak erosion was determined by pre- and extreme heat. From the microstructural analyses of
post-test measurements using a pencil-point dial indi- the post-test specimens, it was observed that the
cator. Mass loss was determined by pre- and post-test protective mechanisms of the phenolic-based and
weight loss measurements. Consistent data sets were silicone-based ablatives under these extreme heating
averaged and are presented in this study. environments are very different. The silicone resin is
more thermally stable than the phenolic resin when
exposedtoextremeheat.Asaresult,theSM8024com-
Microstructural char characterization
posite is more ablation resistant than the MXBE-350
In this research program, seven SSRM and three SDL composite both on post-test mass loss and average
motor firings were conducted.37,41 In the first three erosion results.37,41
SSRM firing rounds (Rounds 1–3), they were set to
compare the phenolic-based composite (MXBE-350)
Round 2 SSRM firing
and different variants of silicone-based composites
(SM8000) at the same test conditions. Detailed mor- InRound2,cross-sectionalviewsat350(cid:3)ofthevirgin
phological characterization using SEM was performed MXBE-350 (phenolic glass mat) [Figure 3(a)], SM8024
to understand the material pre-test (virgin) and post- (silicone glass mat) [Figure 3(b)], SM8027 (silicone
test (char) microstructures and their protective mecha- silica fabric) [Figure 3(c)], SM8027A (silicone silica
nisms. SEM is a JOEL 5601 and SEM specimens were roving) [Figure 3(d)], SM8027MC (silicone silica
coated with gold–palladium to reduce electron charg- MC) [Figure 3(e)], SM8029 (silicone quartz fabric)
ing. This article summarizes the detailed microstruc- [Figure 3(f)], and SM8029MC (silicone quartz MC)
tural analyses of the first three SSRM firing rounds [Figure 3(g)] were examined using SEM. The different
(Rounds 1–3). A different objective was set for each fiber architectures can be clearly seen from this set of
of these seven SSRM rounds. For the last four SSRM SEM micrographs. This set of SM8000 materials were
firing rounds (Rounds 4, 5, 7, and 9), 69 SM8000 sam- tested using SSRM under the same test conditions.
plesweretestedalongwith24MXBE-350controlsam- They were exposed to heat fluxes of 6240 and
ples. Peak erosion and mass loss material data of the 9643kW/m2 with 4.5kg/h of alumina particles at an
SSRM firings are included in this article. In addition, impingement angle of 90(cid:2) for test duration of 12s.
threeSDLfiringrounds(Rounds6,8,and10)consisted A series of SEM micrographs at 50(cid:3), 350(cid:3), and
of 15 test firings were conducted. Fifty-three SM8000 1000(cid:3) of the surface views of MXBE-350 tested at
samplesalongwith79MXBE-350controlsampleswere 6240kW/m2 show the phenolic resin was pyrolyzed
tested in this research program. A different objective exposing the glass mat on the surface. A similar series
was also set for each of the three SDL rounds. SDL of SEM micrographs of MXBE-350 tested at
test results will be discussed elsewhere.37,41 9643kW/m2 [Figure 4(a) and (b)] show more glass mat
exposed and some melting of phenolic resin at higher
heat flux level. Cross-sectional views of MXBE-350 at
Round 1 SSRM firing
350(cid:3)exposedto6240and9643kW/m2(Figure5)were
In Round 1, MXBE-350 and SM8024 materials were analyzed. No melting of the phenolic resin underneath
tested using the SSRM at the same conditions to the surface was observed at 6240kW/m2 test case and
Koo et al. 7
Figure 3. Cross-sectionalviewof(a)virginMXBE-350sampleat350(cid:3);(b)virginSM8024sampleat350(cid:3);(c)virginSM8027sample
at350(cid:3);(d)virginSM8027Asampleat350(cid:3);(e)virginSM8027MCsampleat350(cid:3);(f)virginSM8029sampleat350(cid:3);and(g)virgin
SM8029MC sampleat350(cid:3).
Note: Scalebar is100mm.
8 Journal of Composite Materials 0(0)
morephenolicmeltingwasobservedunderneaththesur- micrographs of SM8024 tested at 9643kW/m2
faceatthemoresevere9643kW/m2testcase(Figure5). [Figure 6(a) and (b)] show the similar phenomenon as
Melting of phenolic resin occurred as deep as 0.77mm observedat6240kW/m2testcase.Cross-sectionalviews
fromthesurface(Figure5). of SM8024 at 350(cid:3) exposed to 6240 and 9643kW/m2
A series of SEM micrographs at 50(cid:3), 350(cid:3), and (Figure 7) were analyzed. More melting of the silicone
1000(cid:3) were analyzed showing the surface views of resin and voids was observed underneath the surface
SM8024 (silicone glass mat) tested at 6240kW/m2. at the 9643kW/m2 test case (Figure 7) than the
Silicone resin melted and a lot of alumina particles 6240kW/m2 test case.
were trapped with the melted silicone with no glass AseriesofSEMmicrographsatdifferentmagnifica-
fibers exposing on the surface. Another series of SEM tions of the surface views of SM8027 (silicone silica
Figure 4. SurfaceviewofMXBE-350sampleat(a)50(cid:3)tested Figure 6. SurfaceviewofSM8024sampleat(a)50(cid:3)testedat
at9643kW/m2;(b) 350(cid:3)tested at9643kW/m2. 9643kW/m2;(b) 350(cid:3)tested at9643kW/m2.
Note: Scalebar is100mm. Note: Scalebar is100mm.
Figure 5. Cross-sectionalviewofMXBE-350sampleat350(cid:3)testedat9643kW/m2showingthermalwavemovingfromrighttoleft
across thespecimen (scale bar is100mm).
Koo et al. 9
Figure 7. Cross-sectionalviewofSM8024sampleat350(cid:3)testedat9643kW/m2showingthermalwavemovingfromrighttoleft
across thespecimen (scale bar is100mm).
the silica fabrics were not totally wetted out in these
SM8027 specimens. More melting of the silicone resin
and voids were observed at 6240kW/m2 test case than
the9643kW/m2testcase(Figure9).
AseriesofSEMmicrographsatdifferentmagnifica-
tionsofthesurfaceviewsofSM8027MC(siliconesilica
MC) tested at 6240kW/m2 show the silicone resin was
melting and aluminaparticles were trapped on thesur-
face, some silica fibers were exposed, and large cracks
are visible on the surface. A similar series of SEM
micrographs of SM8027MC at 9643kW/m2 [Figure
10(a) and (b)] show large holes (about 10mm) were
observed. Cross-sectional view of SM8027MC at
350(cid:3) exposed to 9643kW/m2 is shown in Figure 11.
More melting of the silicone resin underneath the sur-
face was observed at 9643kW/m2 test case (Figure 11)
thanatthe6240kW/m2testcase.Largevoidswerealso
observed underneath the surface for the 6240kW/m2
test case.
AseriesofSEMmicrographsatdifferentmagnifica-
tions of the surface views of SM8029 (silicone quartz
fabric) tested at 6240kW/m2 show that the silicone
resin was eroded and a small amount of quartz fibers
were exposed on the surface. Large holes (about
100mm)wereobserved.AsimilarseriesofSEMmicro-
graphsofSM8029at9643kW/m2[Figure12(a)and(b)]
show larger holes and cracks were observed. A cross-
sectional view of SM8029 at 350(cid:3) exposed to
9643kW/m2 is shown in Figure 13. More melting of
Figure 8. SurfaceviewofSM8027sampleat(a)50(cid:3)testedat
9643kW/m2and(b) 350(cid:3)testedat9643kW/m2. the silicone resin and voids was observed underneath
Note: Scalebar is100mm. thesurfaceat9643kW/m2testcase(Figure13)thanthe
6240kW/m2 test case. The thermal wave penetrates to
about 0.8mm from the front surface of the ablative
fabric)testedat6240kW/m2showthesiliconeresinwas (Figure 13).
meltingandaluminaparticlesweretrappedandnosilica AseriesofSEMmicrographsatdifferentmagnifica-
fibers were exposed on the surface. A series of SEM tions of the surface views of SM8029MC (silicone
micrographs of SM8027 at 9643kW/m2 [Figure 8(a) quartz MC) tested at 6240kW/m2 show that the sili-
and (b)] show the similar phenomenon as observed at cone resin was eroded. Large holes (10mm) and long
6240kW/m2. Cross-sectional view of post-test SM8027 cracks were observed. A similar series of SEM micro-
sample at 350(cid:3) exposed to 9643kW/m2 is shown in graphsofSM8029MCat9643kW/m2[Figure14(a)and
Figure 9. A fair amount of voids appear in between (b)] show even larger holes (10–20mm) on the surface.
the silica fibers and the silicone resin. It indicated that Across-sectionalviewofSM8029MCat350(cid:3)exposed