NASA Technical Reports Server (NTRS) 20040014967: Sharp Refractory Composite Leading Edges on Hypersonic Vehicles

9 Pages·0.22 MB·English

by  NASA Technical Reports Server (NTRS)

SHARP REFRACTORY COMPOSITE LEADING EDGES ON
HYPERSONIC VEHICLES

Sandra P. Walker
NASA Langley Research Center, Hampton, VA

Brian J. Sullivan
Materials Research and Design, Rosemont, PA

ABSTRACT

On-going research of advanced sharp
refractory composite leading edges for use on
hypersonic air-breathing vehicles is presented in
this paper. Intense magnitudes of heating and of
heating gradients on the leading edge lead to
thermal stresses that challenge the survivability of
current material systems. A fundamental
understanding of the problem is needed to further
design development. Methodology for furthering
the technology along with the use of advanced Figure 1: Air-breathing hypersonic vehicle
fiber architectures to improve the thermal- concept.
structural response is explored in the current work.
Thermal and structural finite element analyses are Ceramic matrix composites utilizing
conducted for several advanced fiber architectures carbon fibers are candidate material systems for
of interest. A tailored thermal shock parameter for sharp leading edge applications under hypersonic
sharp orthotropic leading edges is identified for flight conditions. Although the use of various
evaluating composite material systems. The use of ceramic matrix materials requires investigation for
the tailored thermal shock parameter has the identification of an optimum material system, the
potential to eliminate the need for detailed focus of this preliminary effort is limited to carbon
thermal-structural finite element analyses for matrix composites utilizing carbon fibers (C/C).
initial screening of material systems being The application as a nose leading edge for
considered for a leading edge component. an air-breathing vehicle demonstrator is used as a
basis for this study. Development of the C/C
INTRODUCTION material system is pursued through investigating
the effect of hybrid fiber architectures on
Advanced air-breathing hypersonic improving the thermal-structural response for a
vehicles require sharp leading edges to meet sharp leading edge. A theoretical, semi-empirical
aerodynamic performance requirements. An micro-mechanics approach is utilized to predict
example of an air-breathing hypersonic vehicle the material properties of the advanced C/C
concept is illustrated in Figure 1. Circled are some material system.
locations on the vehicle requiring sharp leading Detailed three-dimensional thermal and
edge technology. Flight trajectories having rapid structural finite element analyses are conducted on
acceleration to Mach 10 impose significant the sharp leading edge component to predict the
thermal stresses along with operating temperatures critical thermal stresses and thermal buckling
exceeding 3800oF on sharp leading edges. To loads. Factors of safety are determined through
meet advanced vehicle requirements, technology comparison of predicted thermal stresses with
development for the leading edge structure has material strength allowables.
been identified as essential in Reference 1.
1
American Institute of Aeronautics and Astronautics
An effort to identify desirable material geometry of the leading edge being analyzed is
attributes for improved factors of safety led to the illustrated in Figure 3. The leading edge
identification of a thermal shock parameter for component is four inches long in the chordwise, x
orthotropic sharp leading edges. The thermal direction (material 1-direction), 18 inches wide in
shock parameter is evaluated here as a material the spanwise, y-direction (material 2-direction),
screening tool for sharp leading edges. and 0.5 inches thick in the z-direction (material 3-
The test specimen requirements to direction) at the root. The tip of the leading edge
structurally validate sharp leading edges are has a radius on 0.03 inches. Due to symmetry in
considered. Initial material property testing the spanwise direction, only one-half of the
requirements are defined for screening materials. leading edge width was modeled. The leading
Additionally, leading edge specimen width edge attachment region was 0.251 inches thick in
requirements for replicating critical thermal stress the z-direction and 0.864 inches long in the x-
magnitudes are identified. The fabrication of direction. A commercially available finite element
advanced hybrid C/C test panels is described. The analysis code, NASTRAN, was utilized for the
material specifications are presented and two analyses.2 The finite element model contained a
architectures of interest were chosen for total of 20,880 eight-node hexahedral solid
fabrication. elements.

AERODYNAMIC HEATING
symmetry
ENVIRONMENT

0.251 in
The aerodynamic heating conditions from
W = 9. in
a specific Mach 10 flight trajectory are utilized in
2 z
this study. The leading edge under evaluation has
x
a tip radius of 0.03 inches, is unswept and L = 4. in
y
operating at zero angle of attack. Figure 2 shows
the stagnation point heat flux on the tip of the
sharp leading edge. In addition to the magnitude Figure 3. Leading edge analysis model.
of the peak heat flux being intense, the temporal
gradient of the heat flux is also intense. A baseline C/C material system with a 3:1
Consequently, the leading edge will also be ratio of pitch type fibers was first analyzed. The
considered here as being subject to a thermal 3:1 ratio identifies the ratio of chordwise to
shock condition. spanwise fibers in a ply weave. All pitch (K321)
fibers were utilized in the 3:1 weave. The pitch
6 fibers were chosen for their high thermal
conductivity in comparison to other commercially
Q,
s 4 available fibers, and thus were utilized to aid in
Btu/in2-s
reducing the peak temperatures occurring at the
2
stagnation point on the tip of the leading edge.
Accordingly, the material was oriented with the
0
higher density of fibers chordwise (i.e., x-
0 20 40 60 80
direction) .
time, sec

Figure 2. Stagnation heat flux for a Mach 10 Thermal Analysis
flight trajectory. A transient thermal analysis was
conducted on the leading edge model assuming an
BASELINE FINITE ELEMENT ANALYSIS initial temperature of 0oF. The temperature
dependent thermal conductivity of the material
Three-dimensional finite element thermal was modeled. The heat flux consisted of a
and structural analyses were conducted to predict predicted chordwise distribution on the upper and
the response of material systems of interest. The lower surfaces using the transient stagnation heat
2
American Institute of Aeronautics and Astronautics
flux (Figure 2) at the tip and applied uniformly A linear bifurication buckling analysis
across the span of the leading edge. The surfaces was also performed for these loading conditions.
at y = 0 and y = 9. inches, the end surface at x = L The results revealed no concern with thermal
and the attachment region were assumed adiabatic. buckling of the leading edge design.
Radiation to space was modeled with a surface
emissivity of 0.75.
Since a uniform heat flux was applied in σ
ymax
the span direction, there was no spanwise
temperature gradient predicted and the predicted
temperature gradient occurring through the
thickness was also insignificant. The predicted
temperature distribution in the x-direction at the
time, t = 79 sec is displayed in Figure 4. A peak
temperature of T = 3800oF occurred at the time, t Figure 5. Spanwise stress distribution at t = 79
= 79 sec. As can be seen in Figure 4, a severe sec.
temperature gradient occurs in the chordwise
direction at the tip of the leading edge. ALTERNATE FIBER ARCHITECTURE
EVALUATION

4000
Because the baseline composite
3000 architecture was predicted to fail due to
T, oF compressive thermal stresses, advanced fiber
2000
architectures for improved thermal and structural
1000 response were investigated.

0
Fiber Selection Considerations
0. 2.5 5.
0.5 3 5.5
The key issues related to the thermal and
x, in structural response of the composite sharp leading
edge include: 1) peak temperature, and 2) peak
Figure 4. Temperature distribution at t = 79 sec. compressive σ , both occurring at the tip of the
y
leading edge.
Structural Analysis The desire to limit the peak temperature is
Non-linear static structural analyses were driven by both the operational limit of the coatings
conducted for the leading edge using temperature to resist oxidation and the effect of excessive
dependent material properties. The attachment temperature degrading structural performance.
area was allowed to freely expand and only free Consequently, the fiber architecture should
body motion was constrained in the analysis. The optimally employ high conductivity pitch fibers
unconstrained boundary condition was expected to chordwise to transfer heat away from the tip and
closely replicate the actual attachment reduce peak temperatures for improved material
configuration constraint. The critical thermal behavior. The ratio of fibers in the chordwise
stresses were predicted to occur at the same time, t direction to fibers in the spanwise direction should
= 79 sec, as the peak temperatures occurred. The also be the maximum possible with permissible
critical maximum stress was compressive factors of safety for the design. The use of the
occurring in the spanwise direction at the center on high conductivity fiber additionally benefits the
the tip of the leading edge. A contour plot of the design yielding lower spatial temperature
critical σ stress distribution is displayed in Figure gradients and hence thermal stresses at the tip.
y
5. The predicted peak σ was significantly An increase in the spanwise compressive
ymax
greater than the material allowable with a factor of allowable stress will also improve the critical
safety of 0.42, implying the design was expected factor of safety for the leading edge design.
to fail. Consequently, employing T-300 fibers spanwise
would be desirable since the T-300 fibers
3
American Institute of Aeronautics and Astronautics
generally have a higher compressive strength composites. This involves the analysis of unit cell
while at the same time possessing a lower models of the hybrids under multiple
compressive modulus then other available fibers. unidirectional stress states to calculate the stresses
This remains the case even when the T-300 fibers necessary to cause failure of the composite for
are heat-treated, as would be necessary for their each stress component.7 In these calculations,
use in this high temperature application. Note that particular emphasis was placed on the material
as-received T-300 fibers are processed at much characterization in the vicinity of 4000oF.
lower temperatures than they will experience in
this composite application. Evaluation Procedure
Based on the aforementioned issues The procedure for evaluating the
relevant to sharp leading edges, hybrid fiber refractory composite material systems under
architectures, utilizing pitch fibers chordwise, consideration include the following steps:
weaved with T-300 fibers spanwise, are being 1. Identify laminate construction for
evaluated here for improving the thermal- evaluation
structural response of the leading edge. In 2. Generate material properties for analyses
addition to using T-300 fibers spanwise, the use of 3. Conduct transient thermal analysis of
varying fiber orientation requires investigation. leading edge subject to aerodynamic
Although optimal ply angle orientation requires heating environment
future consideration, the current investigation is 4. Conduct structural analyses to predict
limited to incorporating ± 45o plies of T-300 fibers critical thermal stresses
in a 1:1 weave in the fiber architecture design. 5. Compute factors of safety for material
system
Material Property Predictions 6. Repeat 1-5 for each material system being
Material properties predictions are needed evaluated
for the composite C/C material systems to conduct 7. Compare factor of safety and peak
the three-dimensional thermal and structural finite temperature for each material system to
element analyses. For each fiber architecture to be identify best material for application of
evaluated, the predicted temperature-dependent interest
properties include: 1) thermal conductivity, k , k ,
x y
and k , 2) engineering modulus, E , E , E , 3) THERMAL SHOCK PARAMETER
z x y z
Poisson’s ratios, ν , ν , ν , ν , ν , ν , 4) shear
yx zx zy xy xz yz
modulus, G , G , G , 5) coefficient of thermal The attempt has been made to define a
xy xz yz
expansion, α , α , α , 6) tensile strength F , F , material parameter as a screening tool for
x y z tx ty
F , 7) compressive strength, F , F , F , and 8) evaluating fiber architecture designs. Use of a
tz cx cy cz
shear strength, T , T , T , where subscripts x, y, material parameter to rank the designs could
xy xz yz
and z are the three material coordinate directions. possibly eliminate the need for detailed thermal
The method for predicting the material and structural analyses of every architecture being
properties utilizes a database containing test data evaluated and thus improve design efficiency. The
from numerous C/C composites. Test data material parameter could also provide insight on
included data for C/C composites that employ the achieving material systems for optimal thermal-
pitch (K321 or P30X) type fibers3, and separately, structural response of the sharp leading edge.
data on C/C composites that employ the T-300 Considering first the material properties
fibers4. The data was used to back-calculate the significantly affecting the critical factor of safety,
effective fiber and matrix properties. Using
classical micromechanical methods5,6, the fiber F
FS = cy (1)
and matrix properties were then projected to critical
σ
predict thermo-elastic and thermal conductivity ymax
properties for the hybrid two-dimensional laminate
constructions of interest. Composite strength The desire to increase the critical FS would be to
predictions were also performed using the back- maximize the right hand side of Equation (1). The
calculated properties in the candidate hybrid peak σymax can be roughly approximated to be
4
American Institute of Aeronautics and Astronautics
dominated by the term E α ∆T, where ∆T is the the temperature dependent nature of the material
y y
change in temperature from the initial temperature. properties of the material systems. Although the
This rough approximation can be justified by appropriate temperatures to use may require future
reasoning that the tip of the leading edge is being evaluation, for this preliminary investigation, the
constrained from spanwise expansion by the mechanical properties (F , E , and α ) of the
cy y y
cooler mass aft of the tip, which is at both lower material at the baseline peak temperature of
temperatures and has a lower coefficient of 3800oF were used to compute the tailored thermal
thermal expansion. Finally, ∆T is most sensitive shock parameter. Alternatively, since the peak
to k , the thermal conductivity in the chordwise temperature rise is to a greater extent affected by
x
direction. This property is very dependent on fiber the thermal conductivity, k , over the transient
x
architecture and is inversely related to ∆T. temperature history, the initial decision here was
Substituting the material property relations in the to use the value of kx predicted at T = 3530oF, in
factor of safety ratio yields the material parameter, the TSP computation, since that value is closer to
an average conductivity.
F k Results of the fiber architecture
TSP = cy x (2) evaluations are presented in Table 1. The first
E α
y y column in the table gives the weave ratio of
spanwise K321 type to chordwise T-300 type
The parameter is denoted as “TSP” because the fibers. The first hybrid listed shown with a 1:0
parameter has the same material properties as the weave were actually unidirectional layers of K321
traditional thermal shock parameter identified for fibers. The second column gives the percentage of
isotropic flat plates subject to thermal shock fibers contained in the laminate oriented at ±45
conditions.8 However, care should be taken in degrees. The ±45 degree fibers were within layers
interpreting this parameter as a thermal shock weaved with a 1:1 ratio of T-300 fibers. The
parameter. In Reference 9, it was shown that a results of the material property predictions yielded
general thermal shock parameter for composite the TSP presented in the following column. The
laminated plates cannot be explicitly expressed. final two columns are the calculated peak
Also, the thermal shock problem has an temperature results and the critical factor of safety
instantaneous step change in heating conditions on in the spanwise direction for each architecture
a relatively thin flat plate. Although the change in evaluated.
heat flux is not instantaneous here, the steep For all hybrid fiber architectures
temporal gradient (Figure 2) should be further evaluated, the computed margins of safety were all
justification of considering thermal shock as an greater than one, and thus all architectures are
issue and utilizing a tailored thermal shock expected to survive the aerodynamic heating
parameter for evaluating sharp leading edge conditions. As compared to the baseline design,
materials. Consequently, the merit of a tailored which was a 3:1 architecture of only K321 fibers
thermal shock parameter (TSP) for orthotropic with a predicted critical factor of safety of 0.42,
leading edges requires investigation for use of the the advantage of hybrid architectures for leading
parameter in screening material systems edge design are evident.
specifically for sharp leading edges and is The most significant increase in the factor
evaluated subsequently. of safety was observed for the architecture with
the 1:0 layers, i.e., which contained no spanwise
FIBER ARCHITECTURE THEORETICAL fibers. When spanwise fibers are then introduced
EVALUATION RESULTS into the architecture, the factor of safety dropped
significantly. Further, compared to the same
Several fiber architectures were identified weave ratio with no 45-degree layers, the addition
for evaluation based on the considerations of 45-degree layers of T-300 fibers did not lead to
previously presented. For each architecture, the any significantly improvement in the factor of
evaluation procedure outlined previously was safety. More significantly, adding 45-degree
followed and the tailored thermal shock parameter layers to the 3:1 and 4:1 hybrids resulted in an
was computed. However, complicating matters are
5
American Institute of Aeronautics and Astronautics
Table 1. Hybrid Fiber Architecture Evaluation Results
HYBRID FIBER MATERIAL FINITE ELEMENT ANALYSIS
ARCHITECTURE ANALYSIS Thermal Structural
Weave % ± 45o Fcykx  Btu  Tmax FS
T-300 fibers E α in−sec (oF) critical
y y
33.3 5.19
1:0
50 6.39 3800 2.12
1:1 0 4.83 4050
2:1 0 4.04 3820 1.16
0 3.87 3720 1.08
3:1
20 3.74 3810 1.09
0 3.62 3660 1
4:1 28.6 3.59 3810 1.08
44.4 3.53 3900 1.07
5:1 14.3 3.44
increase in the predicted peak temperature, which prediction technique for hybrid fiber architecture
is undesirable for the leading edge designs. C/C material systems and experimentally correlate
From all architectures evaluated, the the predicted TSP. Since the key material
predicted TSP was observed to be maximum for properties affecting the thermal structural response
the 1:0 architecture containing 50% of ± 45o T-300 of the sharp leading edge are reasoned to be k , E ,
x y
fibers. The result of the maximum TSP correlated F , and α , preliminary material property testing
y y
well with the maximum FS predicted from the should be limited to these properties to efficiently
critical
finite element analysis. Although the overall trend screen initial materials under consideration.
showed that TSP decreased with decreasing FS, Having experimental data for the key properties
there were some minor discrepancies observed, can then be further utilized to predict to greater
where the FS showed improvement as TSP accuracy the other properties necessary for the
critical
decreased within a given weave. Concurrent with three-dimensional analysis of the leading edge.
the observation of the slight increase in FS while To verify material properties, fabrication
TSP decreased was that T increased. of a ten inch by ten inch panel that is 0.21 inches
max
Consequently, the undesirable increase in peak thick, as illustrated in Figure 6, should be
temperature was also captured with the TSP sufficient. On the panel the specimens shaded in
decreasing slightly. The conclusion can then be gray are those needed for the initial key property
inferred that material system ranking roughly measurements. The test matrix corresponding to
correlates with the TSP. the key properties is presented in Table 2. The
compression test will provide the E and F
y y
VALIDATION TESTING properties with four replicates at the two
temperatures of 70oF and 3800oF. The expansion
Experimental validation is required to test will provide α between 70oF and 3800oF
y
both ensure the ability of the leading edge to through two replicate tests. The thermal
operate without failure when subject to the conductivity test will provide the instantaneous k
x
aerodynamic flight loads and also to validate the at several temperatures between 70oF and 3000oF,
methodology used to predict behavior. which should be sufficient for this property where
material property predictions should suffice for
Material Property Tests higher temperature variations.
Validation testing on the predicted
material properties is required to refine the

6
American Institute of Aeronautics and Astronautics
Figure 6. Material property test panel (all dimensions are in inches).

Table 2. Test Matrix to be repeated for each panel

Description 70oF 3000oF 3800oF Total
Compression-Fill (Cf) 4 4 8
Thermal Expansion – Fill (Exp-f) 2 2
Thermal Conductivity- Warp (Cond-w) 2 2
12

Leading Edge Test Specimens larger specimens. Addressing this issue, finite
For consistency with the properties, element analyses were conducted on varying
leading edge specimens are also being fabricated specimen width. The results of varying specimen
by the same vendor as the material property panels width on the critical thermal stress are presented in
utilizing the same batch of fibers. The pertinent Figure 7. From Figure 7, the observation is made
issues regarding the test specimen include what a that once a specimen width of about 8 inches is
minimally acceptable spanwise width should be reached, the peak stress levels off without any
for replication of the critical state of thermal further increase with increasing width. Therefore,
stress. The issue arose from the ability of test for the current design, a specimen width of 8
facilities to accommodate large specimens in inches will be required to reach the peak design
addition to the increased costs associated with thermal stresses for evaluation.
7
American Institute of Aeronautics and Astronautics
1.2
1
σ 0.8
2max
σ (W=18 ")
2max 0.6
0.4
0.2
0
0 10 20
W, inches

Figure 7. Effect of specimen width on peak critical stress.

FABRICATION OF TEST PANELS 10 flight vehicle trajectory was used as a basis for
this study. Detailed thermal-structural finite
Currently, test panels for both material element analyses for a baseline material revealed
property and leading edge specimens are being that the use of a C/C material system with a single
fabricated under contract. Material specifications fiber type would not survive the application.
include: Based on the predicted thermal-structural
1) Required heat treatment = 2700oC response, the use of a hybrid fiber architecture was
2) Minimum composite density= 0.065 lb/in3 identified to be worthy of investigation. Material
3) Fiber volume fraction = 0.55 property predictions were subsequently conducted
To insure highest possible thermal conductivity, for several hybrid architectures. A tailored
the C/C composites are to be fabricated using a thermal shock parameter for sharp orthotropic
combination of pitch matrix densification and leading edges was derived in an attempt to identify
chemical vapor infiltration. a material parameter to preliminarily screen fiber
The two fiber architectures chosen for architectures. Although the parameter shows
fabrication include the 1:0 architecture with 50% promise in identifying the optimal material
45o T-300 fibers and the 3:1 hybrid architecture systems, further investigation is underway to
with zero 45o plies. The material property test experimentally verify the predicted parameter and
panel fabrication is complete. Preliminary further evaluate the assumptions used with the
inspection revealed the panels to be of quality computation.
construction and material property testing is Key material property tests and the
underway. requirements for test panels necessary to evaluate
the advanced hybrid fiber architectures were
CONCLUDING REMARKS defined. The leading edge test specimen width
required to produce the critical thermal stress
This paper presented work underway for magnitude was identified. Material property test
the development of sharp refractory composite panels have been successfully fabricated with
leading edges for advanced hypersonic vehicles. hybrid architectures and validation testing is
A vehicle nose leading edge for a specific Mach currently in progress.
8
American Institute of Aeronautics and Astronautics
REFERENCES

1. Glass, D. E., Merski, N. R., Glass, C.E.,
Airframe Research and Technology for
Hypersonic Airbreathing Vehicles, NASA TM-
20020211752, July 2002.

2. MSC/PATRAN/NASTRAN, Version 7, Vol.
1, The MacNeal-Schwendler Corperation, Los
Angeles, CA, July, 1997.

3. B.J. Sullivan and K.W. Buesking, Data
Correlation and Evaluation of Affordable Carbon-
Carbon Composites:Effective Properties of K321
and P-30x Fibers in C-C Composites, Proceedings
of the 21st Annual Conference on Composites,
Materials and Structures, Cocoa Beach, FL,
January 1997.

4. J. Spain and S. Starrett, Mechanical and
Thermal Evaluation of Coated Carbon-Carbon,
Southern Research Institute Final Report to
General Dynamics Fort Worth Division, SRI-ME-
91-271-6913, April 1991.

5. Z. Hashin, Analysis of Properties of Fiber
Composites with Anisotropic Constituents, J. Appl.
Mech., Vol. 46, 1979, pp. 543-550.

6. N.J. Pagano and G.P. Tandon, Elastic Response
of Multidirectional Coated Fiber Composites,
Composites Science and Technology, Vol. 31,
1988, 273-293.

7. B.W. Rosen, et al., An Analysis Model for
Spatially Oriented Fiber Composites, Composite
Materials Testing and Design, ASTM STP 617,
1977, pp. 243-259.

8. Lidman, W.G. and Bobrowsky, A. R.,
Correlation of Physical Properties of Ceramic
Materials with Resistance to Fracture by Thermal
Shock, NACA TN 1918, July 1949.

9. Wang, Y. R. and Chou, T. W., Thermal Shock
Resistance of Laminated Ceramic Matrix
Composites, Journal of Material Science, 26, 1991,
pp. 2961-2966.

9
American Institute of Aeronautics and Astronautics