TMS Home Page Presenting a Web-Enhanced
Article from JOM

View Current Issue




Aluminum: Cast Shop and Alloys: Overview Vol. 61, No.1 pp. 23-31

The Development of a Multifunctional Composite Material
for Use in Human Space Exploration Beyond Low-Earth Orbit

S. Sen, E. Schofield , J. S. O’Dell, L. Deka, and S. Pillay

About this Issue


JOM in Print
The print and/or PDF versions of the article can be acquired.



Figure 1
Dose equivalent as a function of areal density (or material thickness) for PE fiber-reinforced epoxy matrix Composites 1 and 2; pure PE and aluminum results are from calculations using the 1986–87 solar-minimum GCR environment. Similar results were obtained when the 1989 solar maximum environment was used. Note that the material thickness can be obtained by dividing the areal density by density of the material.



Figure 2
(a) A schematic representation of a plain-weave pattern showing interlaced warp and fill fibers. 16
Figure 2
(b) A woven UHMWPE Spectra 1000 fabric used for the present investigation.
Figure 2
(c) A typical product form, Composite 1, after hand lay up of fabric plies and autoclaving.



Figure 3
(a) The microstructure of the carbon foam used in this investigation showing an open cell structure and fiber-like ligaments.
Figure 3
(b) VPS deposited B4C coating on carbon foam.
Figure 3
(c) A transverse section showing the complete architecture of composite 2.



Figure 4
The absolute fragment fluence per incident ion as a function of fragment charge from 800 MeV/u 28Si beam exposure. All samples had an areal density of 4.4 gm/cm2. The fragmentation efficiency of both composites was comparable to the PE benchmark material.



Figure 5
A schematic showing the dimensions (in centimeters) of the tensile test sample as per ASTM D638-03. Also shown is an actual test sample that was machined using a waterjet.



Figure 6
(a) Typical stress-strain characteristics of Composite 1.
Figure 6
(b) A scanning electron microscope image of a typical fractured surface of Composite 1 showing the PE fiber bundles are still bound.
Figure 6
(c) A higher magnification view of the fiber bundles at the fractured surface showing good bonding between PE and epoxy matrix.



Figure 7
The operational details of a two-stage micro-light gas gun.



Figure 8
(a) The ballistic impact damage on Composite 1. Note the damage area, delamination, and crater diameter on the front side and dimpling on the back side (insert) of the composite in comparison to.
Figure 8
(b) Impact response of VPS-deposited B4C on Composite 2 which shows only a crater diameter with no corresponding damage area. No damage was observed in the underlying carbon foam and PE composite.



Figure 9
Finite element modeling results (a) showing the effect of hypervelocity impact as a function of material property, namely failure strain and (b) quantifying the magnitude of compressive stress propagating through the B4C coating, (c) illustrating the reflection of compressive wave from the carbide-foam interface.



Figure 10
(a) 12.7 cm thick open cell carbon foam with VPS-deposited B4C that was used for thermal testing.
Figure 10
(b) The temperature profile obtained from B4C surface adjacent to plasma torch impingement location (pyrometer), immediately below the B4C surface (TC1), and at the bottom of the carbon foam block (TC6). Only the B4C face was exposed to the plasma. The other sides of the carbon foam were protected with a high-temperature alumina paste..












Questions? Contact
© 2009 The Minerals, Metals & Materials Society

Long-duration human exploration beyond the low Earth orbit (LEO) mandates development of materials to minimize crew and equipment exposure to the interplanetary radiation environment. The potential for biological damage by the relatively low percentage of high-energy heavy ions in the galactic cosmic ray spectrum far outweigh that due to lighter particles because of their ionizing power and the quality of the resulting biological damage. To avoid paying a penalty due to additional weight, it would be beneficial to develop a multifunctional material as an integral part of a spacecraft structure to provide shielding effectiveness and structural integrity. This paper discusses the development of polyethylene fiber reinforced epoxy matrix structural composites that effectively satisfy both primary requirements.


NASA’s current vision for space exploration includes long-duration human travel beyond lower Earth orbit (LEO) and sustained human presence on other planetary surfaces. For this vision to be a reality, one of the major challenges that need to be overcome is to minimize the radiation exposure to crew and equipment from the interplanetary radiation environment. Life on Earth is well protected from this radiation environment by a combination of the geomagnetic field and the atmospheric overburden. In LEO, the radiation exposure of astronauts is kept below the National Council on Radiation Protection (NCRP) limits by limiting their exposure time and by taking advantage of the shielding still offered by the geomagnetic field.1 However, exposure to free space cosmic radiation during an approximately two-year round trip to Mars or during extended stays of up to a few months on the surface of the moon or Mars could result in significant biological damage. This paper discusses the development of a multifunctional composite material that will provide shielding from cosmic radiation while also providing structural integrity, thermal management, and protection from micro-meteoroid impact. This emphasis on multifunctionality is to avoid paying a significant cant penalty in weight and cost due to addition of material solely for radiation shielding. The design for shielding solutions is dictated by the nature of interaction between the cosmic radiation environment and the proposed shielding material. It is therefore worthwhile to first describe the cosmic radiation environment and the physics of its interaction with materials.

The galactic radiation environment consists primarily of a continuous flux of galactic cosmic rays (GCRs)2,3 and transient but intense fluxes of solar energetic particles (SEPs).3,4 The primary constituents of the GCR spectrum are 85% protons, 14% alpha particles, and 1% heavy nuclei with energies ranging from 10 MeV/nucleon to 10 GeV/nucleon.5 The intensity of these particles depends on the solar cycle as well as on the location in the inner solar system.2,5 Despite their low flux, the heavy ions in the GCR pose a serious health risk because they are highly ionizing (energy loss is proportional to the square of the atomic number for charged particles with same velocity) and because the quality of the resulting biological damage is high.6,7 The SEPs, on the other hand, consist primarily of protons and alpha particles with energies ranging from a few MeV/nucleon to few hundreds of MeV/nucleon.2,4,5,7 The likelihood of occurrence for these events is highest during solar maxima and their occurrences can be predicted with some degree of confidence. Although the SEP spectrum does not contain heavy ions and their energy range is much lower than that of GCRs, they still pose serious risks to crew and equipment, particularly in the event of a severe solar storm.

Interaction of the charged particles present in the interplanetary radiation environment with a material takes place through several specific atomic and nuclear processes. Of these processes, two are of particular relevance here, namely, energy loss and fragmentation. The energy loss of charged particles per unit length of material traversed (also know as the stopping power or the linear energy transfer, LET)6,8 is directly proportional to the square of their atomic number and inversely proportional to their energy. As discussed earlier, the radiation risk from GCR exposure is dominated by the small but highly ionizing flux of heavy ions. Fragmentation of the incident heavy ion projectile leads to the formation of smaller fragments moving at the same velocity as the incident particle, and are less ionizing due to their lower atomic number. Breaking up the heavy ions in the GCR flux into smaller fragments with lower ionizing power is the only realistic solution for passive radiation shielding design. It is also important in this process to minimize the production of secondaries from target fragmentation that can otherwise add to the radiation risk.9,10 Therefore, any proposed radiation shielding material for use in outer space must be composed of nuclei that maximize the likelihood of projectile fragmentation while producing the minimum number of target fragments. In this respect, polyethylene (PE) has been found to be one of the best-suited materials for radiation shielding11,12 since it has a very high density of hydrogen atoms (see Table I). As hydrogen has the smallest atomic diameter, it provides a large number of interaction points or high cross section for projectile fragmentation. Moreover, the absence of elements heavier than carbon minimizes the production of target fragments, since the hydrogen nuclei consist of a single proton. It is therefore not surprising that in some quarters of the International Space Station passive radiation shielding in the form of polymeric materials is currently being used.13

As is evident from the above discussion, the superior radiation shielding effectiveness of PE has already been established. The emphasis of this paper is to develop a composite architecture based on PE that would not only be an effective radiation shield but also would have sufficient structural integrity to be considered as structural elements of a crew vehicle for long-duration exploration beyond LEO. In addition to radiation shielding and structural integrity, a truly multifunctional material for a crew vehicle should address thermal management required for mitigating the effects of temperature fluctuations in outer space, and severe re-entry temperatures, and offer ballistic protection against micrometeoroid impacts.

…describe the overall significance of this paper?
Long-duration human exploration beyond the low Earth orbit mandates development of materials to minimize crew and equipment exposure to the interplanetary radiation environment. A polyethylene fiber reinforced epoxy matrix composite with an open cell carbon foam and vacuum plasma deposited boron carbide coating was developed to potentially satisfy the primary requirements for radiation shielding, structural integrity, micrometeoroid impact, and atmospheric re-entry temperature resistance.

…describe this work to a materials science and engineering professional with no experience in your technical specialty?

Interaction of the charged particles in the interplanetary radiation environment with a shielding material takes place through several specific atomic and nuclear processes. Using a shielding material to break the heavy ions in the galactic cosmic ray flux into smaller fragments with lower ionizing power is the only realistic solution for passive radiation shielding design. The emphasis of this work was to develop a multifunctional composite architecture that will satisfy the requirements for deep space radiation shielding and also for structural integrity, micro-meteoroid impact, and re-entry temperatures.

…describe this work to a layperson?
A challenge to NASA’s vision for long-duration human space exploration is to minimize the radiation exposure to the interplanetary radiation environment. This paper discusses a multifunctional composite material that will provide shielding from cosmic radiation while also providing structural integrity, thermal management, and protection from micro-meteoroid impact.

Radiation Transport Calculations

To determine the optimum compositions for composite fabrication, radiation transport calculations were performed. The transport code essentially solves the one-dimensional Boltzman equation where the flux of particles of a given atomic number, energy, and spatial location are determined. Detailed discussion on the transport code is beyond the scope of this paper, and readers are directed elsewhere.14 Transport calculations were performed using both the 1986–1987 solar-minimum and the 1989 solar-maximum GCR environment.15 Two observables were used to evaluate the shielding effectiveness: the absorbed dose and the dose equivalent. Absorbed dose is defined as the energy absorbed by a target per unit mass from any kind of ionizing radiation. The international unit for absorbed dose is Gray (Gy) or 1 J/kg. However, it has been established that the absorbed dose required to obtain the same level of biological damage can be different for different kinds of radiation. To account for this difference in absorbed dose, the concept of dose equivalent was introduced. Dose equivalent is expressed in the units of Sievert (Sv) and defined as the product of the absorbed dose and a dimensionless quality factor, Q. This quality factor is dependent on the LET of the radiation and is prescribed by organizations such as NCRP and the International Commission of Radiological Protection. Further discussion on some of these fundamental metrics for radiation shielding can be found elsewhere.8 For the present analysis, to minimize systematic uncertainties in the calculated results, dose and dose equivalent relative to PE were analyzed. Two composite architectures were evaluated for radiation shielding effectiveness.

Composite 1, an epoxy matrix reinforced with ultra-high molecular weight (UHMW) PE fabric, formed the baseline composite for structural and radiation shielding requirements. The nominal composition of this composite was 68.0% PE and 32.0% epoxy matrix by weight. Composite 2 consisted of Composite 1 with the addition of an open cell carbon foam and plasma-deposited B4C coating to address thermal management and ballistic protection requirements. The nominal composition for this composite was 57.2% PE, 22.0% epoxy matrix, 8.5% boron, and 12.3% carbon by weight.

Figure 1 shows transport calculation results for the two described composites. Calculation results for aluminum and PE are also included for comparison. Results presented in Figure 1 are based on the assumption of a solar-minimum condition when the GCR flux is at a maximum. Similar results were obtained when the 1989 solar maximum environment was used. These calculations indicate that over thicknesses ranging from 1 g/cm2 to 20 g/cm2 the shielding effectiveness of Composite 1 is only about 1% to 5% less than that for pure PE. The entire composite structure, including thermal and ballistic protection, is only 2% to 8% less effective than pure PE over the same thickness range. Terrestrial and LEO requirements mandate keeping the exposure below 50 cSv/ year for blood forming organs.1 Using this requirement as an example, it can be seen from Figure 1 that compared to PE only a marginal increase in composite areal density (13.86 g/cm2 for Composite 1 and 14.13 g/cm2 for Composite 2 in comparison to 13.18 g/cm2 for pure PE) will be required to achieve the same level of shielding provided by PE. However, as will be illustrated later, this increase in areal density compared to PE is offset by gains in the composites’ multifunctional properties.

Development of the Composite Architecture

Composite 1: Structure and Radiation Shielding
The material product form used to fabricate the composite consisted of a high-strength fabric that was plain woven from UHMW PE Spectra 1000 fibers. Individual fibers were 30 µm in diameter with ultimate tensile strength (UTS) and modulus in the range of 3 GPa and 103 GPa, respectively.16 A typical plain weave consists of fibers laid in the 0º warp and 90º fill direction as schematically shown in Figure 2a. A single-ply UHMWPE fabric is shown in Figure 2b. Several such plies were hand laid using a thermoset resin typically employed for aerospace applications. The resin system was selected based on its long pot life desirable for hand lay-up operations and its reasonably high glass transition temperature of 90ºC. To enhance adhesion between the PE fabric and resin, the fabric was gas plasma treated prior to the composite lay up. For the current application, the fabric was oxidized during the plasma treatment. Gas plasma treatment promotes superior adhesion through surface roughening and increasing the surface area of the fiber, and by generating oxygen-containing functional groups on the fiber surface.17 After lay up, the composite was autoclaved at 600 kPa for 24 hours during final curing. The natural exotherm produced by the resin system during curing was monitored and kept below 90ºC. The finished composite was in the form of a plate with desired dimensions for either radiation or mechanical testing. A typical 30.5 cm3 × 15.2 cm3 × 1.27 cm3 Composite 1 plate is shown in Figure 2c. Composite 1 was used for characterization of mechanical and radiation shielding properties.

Composite 2: Architecture with Thermal and Ballistic Protection Multifunctionality
A second composite sample, designated Composite 2, was fabricated to further enhance the multifunctional nature of the composite by addressing thermal management and ballistic impact resistance properties. A thermal protection system (TPS) is typically employed to withstand the space environment and extremely high heat flux encountered during re-entry. For example, the black high-temperature reusable surface insulation (HRSI) tiles typically seen on the belly of the space shuttle are exposed to approximately 1,300ºC during re-entry.18 It is anticipated that re-entry from a lunar mission will expose the vehicle to even higher temperatures. A combination of an open-cell carbon foam and plasma deposited B4C coating on the exterior surface of the carbon foam was used to address these requirements (see Figure 3a). A coal-based carbon foam was selected primarily because of its low density (0.268 g/cm3), low thermal conductivity (0.25–5 W/mK depending on the cell structure), and ability to withstand temperatures up to 3,000ºC in a nonoxidizing atmosphere or with suitable surface protection. Thermal conductivity of the carbon foam is comparable to that of the HRSI tiles used on the space shuttle.

B4C was deposited on the surface of the carbon foam via vacuum plasma spraying (VPS). An as-deposited B4C coating is shown in Figure 3b. The texture of the B4C coating reflects the texture of the carbon foam substrate surface. The 1 mm thick coating was deposited using a B4C powder feedstock fed through an argon plasma with H2 assist gas. The VPS technique was used for this application because it provides a durable mechanical and metallurgical bond between the carbon and B4C without the use of low-temperature materials or bonding agents. Durability is essential for high-temperature applications such as re-entry vehicles. In addition, the VPS process has the ability to rapidly produce an adherent deposit on curved surfaces and for acreage applications. A 1.25 cm thick carbon foam brick with the B4C coating was bonded to the PE composite using the matrix epoxy. The complete architecture of Composite 2 is shown in Figure 3c. This composite was characterized for radiation shielding effectiveness, hypervelocity ballistic impact resistance, and thermal exposure characteristics.

Table I. Density of Hydrogen
Atoms in Different Materials

Material # Atoms/cm3 × 1022
Hydrogen 5.7
Water 6.7
Polyethylene 7.9
Polystyrene 4.7
Polyimide 2.2
Polyamide 3.0

Heavy Ion Exposure of Composites

Since it is currently not feasible to test the radiation-shielding effectiveness of new materials in the free space GCR environment, the only realistic way to experimentally assess the shielding effectiveness of these materials is to expose them to heavy ion beams at an accelerator facility. Availability of beam time and appropriate beam type for such experiments can be limited since only two such facilities exist worldwide: the NASA Space Radiation Laboratory (NSRL) at the Brookhaven National Laboratory, and the Heavy Ion Medical Accelerator (HIMAC) in Chiba, Japan. An 800 MeV/u 28Si (incident LET 45.96 keV/µm) beam, a reasonably good GCR-proxy beam for shielding effectiveness evaluation,11,12 was made available for exposure of the composite samples at HIMAC. Exposures were carried out with the assistance of personnel at NSRL. During the exposure several silicon detectors were placed upstream and downstream of the sample. Upstream detectors were used to verify the monochromatic beam source while the downstream detectors were used to determine the charges of the particles emitting from the sample material. Further details related to detector stacks, data acquisition, and shielding analysis have been presented elsewhere.11

Figure 4 shows the absolute fragment fluence per incident ion as a function of fragment charge for the composite samples as well as for PE. Fragment fluence, one of the metrics used for measuring radiation-shielding effectiveness, is defined as the number of particles with charge Z emerging from the downstream side of the sample within a 1 cm2 circle centered on the beam axis. This metric is essentially a measure of the capability of a material to fragment heavy charged particles into lighter charged particles. On this basis, Figure 4 clearly illustrates that both Composites 1 and 2 are capable of fragmenting the incoming silicon beam (Z = 14), and their performance is comparable to that of the benchmark PE material. It should be noted that fragments with Z < 6 were not resolved by the detectors and as such are not shown in the figure.

It is customary for accelerator exposure measurements not to estimate the systematic uncertainties (which are typically of the order of 2% to 5%) for the sake of time and resource optimization. As a result only values that are relative to a benchmark material, PE in this case, should be evaluated. Table II lists some of these relative metrics to evaluate the shielding effectiveness of the two composites. In terms of surviving fractions (or attenuation) of the primary silicon beam, the composites are a maximum of 7.2% inferior to PE. In addition, the two composites are marginally (3.4–4.2%) inferior to PE both in terms of dose and dose-equivalent. It is equally important to note from the data that the measured shielding effectiveness of the composite samples (areal density = 4.4 g/cm2) are in agreement with the transport calculation results presented in Figure 1. Hence, based on both experimental measurements and transport calculations it is reasonable to conclude that the developed composites will have shielding effectiveness comparable to PE and definitely much superior to currently used aluminum alloys.

Table II. Absolute and Relative (to PE) Values of Radiation Shielding Effectiveness Metrics for the Composite Samples from the 800 MeV/u28Si Beam Exposure*
  Surviving Fraction
of the Primaries
Absorbed Dose (nGy) AVG. Q After
Dose Equivalent
Abs. Rel. Abs. Rel. Target Abs. Rel.
C #1 0.675 1.060 57.6 1.034 12.64 727.9 1.037
C #2 0.683 1.072 57.8 1.038 12.65 731.5 1.042
PE 0.637 1.000 55.7 1.000 12.61 702.0 1.000

All samples were approximately 4.4 gm/cm2 thick.; C #1 = Composite 1, C #2 = Composite 2.

Composite 1: Characterization of Mechanical Properties

Panels of Composite 1 were machined using a waterjet to obtain dogbone tensile testing samples according to ASTM D 638-03 specifications (see Figure 5). During testing, titanium tabs were used to reduce stress concentration and to prevent grip damage at the sample ends. Constant displacement tensile testing was performed using a load cell of 100 kN. A total of seven Composite 1 samples were tested. A representative stress-strain curve for Composite 1 is shown in Figure 6. Unlike metallic systems, a typical yield point is not obvious for such composites. As expected, the epoxy matrix failed first, and this is indicated by the change in slope between 1.5% and 2.4% strain. Beyond this point the load was completely transferred to the PE fabric until failure occurred at about 4.4% strain. Figure 6b shows the failure of the epoxy matrix while the PE fiber bundles appear to be bound and compact. Even at final fracture of the composite (see Figure 6c) there was hardly any indication of fibrillation at the fractured surface since the presence of epoxy is clearly evident between the fiber bundles. This desirable behavior of the composite structure can be attributed to superior surface adhesion between the fabric and the epoxy matrix as a result of the fabric surface treatment using gas plasma.

The average measured UTS and elastic modulus of Composite 1 along with properties of aluminum alloys such as Al 5052, Al 2219, and Al 2024 are listed19,20 in Table III. These alloys are typically used for space applications such as the International Space Station modules and space shuttle fuselage. It is evident that the PE fabric composite has UTS and specific modulus values comparable to the aluminum alloys. However, their main advantage is highlighted in the two density and UTS columns within Table III. Composite 1 is about 2.8 times lighter (lower density) than the aluminum alloys, and consequently its specific strength (strength/ weight ratio) is 2.5 to 4 times greater than typical aerospace aluminum alloys. The preliminary radiation shielding and mechanical testing data presented in this paper for Composite 1 clearly illustrate that the proposed composite deserves further attention as a viable multifunctional material for replacing traditional aluminum alloys.

Ballistic Properties of Composites 1 and 2

For any composite architecture to be considered for a crew vehicle application its ballistic properties, for example, micrometeoroid impact response, have to be addressed. Terrestrially, the micrometeoroid environment can be simulated by impacting the test material with soda lime glass bead projectiles 0.4 mm in diameter at velocities in the range of 6–7 km/s. The experimental system used for such testing is a two stage micro-light gas gun (MLGG) as shown in Figure 7. In the first stage, an explosion at the breach end moves a piston forward, which compresses H2 gas behind a diaphragm. Once the diaphragm is ruptured, the second stage initiates where the compressed gas enters the barrel and moves the projectile at hyper velocities. The projectile finally enters the target chamber where it impacts the target. The low molecular weight of H2 provides the ultra-highvelocity flows needed to achieve hyper velocities.

The effectiveness of micrometeoroid impact resistance was quantified by measuring the crater diameter and the damage diameter of the samples after ballistic impact testing. The initial impact of the projectile results in a straight and narrow track, and its diameter is defined as the crater diameter. The subsequent shock wave generated within the target material is absorbed either through straining of the matrix or by creating new surfaces. The new surfaces are manifested as a damaged area and can be quantified by a corresponding damage diameter. As can be seen from Figure 8a, the projectile impact created a 3.2 mm diameter crater on the impact surface of Composite 1. In addition, the extremely high strain rate was manifested in the form of delamination and dimpling on the back surface of the composite. In comparison, the VPS-deposited B4C coating only had a comparable crater diameter with no associated damage area (Figure 8b). There was no indication of spalling or erosion on the underlying carbon foam.

To obtain a better insight into the significant improvement in ballistic resistance of Composite 2, a numerical modeling scheme was undertaken. Relevant properties, such as density, modulus of elasticity, shear modulus and Poisson’s ratio, for both B4C and carbon foam were obtained from the literature.21 A finite element model builder was used to mesh the B4C coating with 162,500 elements and the glass bead projectile with 2,275 elements. To match the experimental conditions, the imposed projectile impact velocity was 6 km/s. Based on the material’s properties, failure strain (FS) of 0.12 and 0.01 were used for the B4C and projectile, respectively. As part of the modeling scheme, a failure strain erosion criterion was used such that the elements within the B4C coating and glass bead projectile were eroded once the plastic strain within those elements exceed 0.12 and 0.01, respectively. Figure 9a shows an example of the effect of hypervelocity ballistic impact as a function of failure strain. For FS = 0.12, the damage is more localized, and plastic strain is sufficient to cause erosion (spalling) of elements at the distal side of the B4C coating. This is exactly what was observed for Composite 2 (Figure 8b). If, for example, B4C was replaced by a material with FS = 0.5, no spalling would be observed at the distal side. However, the damage would no longer be localized, and would instead be manifested as bulking or volume change. The FS value for PE is much higher than 0.5, and therefore the more widespread damage observed for Composite 1 (Figure 8a) is not surprising. The projectile was completely eroded at the very early stages of impact. The model predicted that on impact a compressive stress wave of 4 GPa would be generated in the composite ahead of the projectile. Figure 9b shows the propagation of this compressive wave as a function of time. The effect of this compressive wave on the underlying carbon foam needs further analysis.

The impact response of the two component system consisting of the B4C deposit and the underlying carbon foam was simulated to investigate whether the compressive stress wave would traverse through the carbon foam. Figure 9c shows the mesh profile at the interface of the two component system. The longitudinal and transverse wave velocity, VE and VG respectively, in a medium can be expressed as where K1 is bulk modulus, G is the shear modulus, and o is the density. Since both G and K for carbon foam are almost three orders of magnitude less than the values of B4C, VE, and VG in carbon foam are almost 18 times less than in B4C. The model predicts that the compressive wave is not transmitted through the carbon foam but is instead reflected back into the B4C top layer. This reflection is shown in Figure 9c at time steps 2.92 × 10–2 µs and 3.92 × 10–2 μs. This reflected wave potentially caused spalling on the distal side of the B4C layer. Previous experimental and modeling studies conducted on multilayer composites have also shown that a low modulus material such as carbon foam is preferable for prevention or minimization of longitudinal wave transmission.22 The experimental and modeling results presented here demonstrate that the proposed composite architecture is capable of protecting the underlying PE structural layer from hypervelocity micrometeoroid impact.

Table III. Measured Mechanical Properties of Composite 1 in Comparison to Some Traditional Al Alloys used in Spacecraft Structures
Material Type UTS (MPa) Density (kg/m3) Specific UTS (N-m/kg) Specific Modulus (N-m/kg)
Composite 1 447 980 45 x 105 21.5
Al 2024/T81 481 2,768 1.7 x 105 25.5
Al 2219/T81 455 2,851 1.5 x 105 26.8
Al 5052/H38 290 2,680 1.08 x 105 25.2

Thermal Characteristics of Composite 2

A preliminary assessment of the thermal characteristics of Composite 2 was performed to determine the ability of B4C coating and carbon foam to protect the underlying structural PE composite when subjected to the high temperatures experienced during spacecraft re-entry in Earth’s atmosphere. Several pieces of carbon foam were adhered together using a high-temperature carbon-based adhesive to fabricate a 10.1 cm3 × 10.1 cm3 × 12.7 cm3 test block (Figure 10a). The test block was dried at room temperature for 4–8 hours, then cured at 130ºC for 4 hours and 260ºC for 2 hours to remove adhesive volatiles. Sample thickness was 12.7 cm to closely duplicate the thickness of the HRSI space shuttle tiles. The top surface of the carbon foam was coated with a 1 mm thick B4C layer using the VPS technique as described earlier. The sides of the carbon foam were coated with an Al2O3 paste to prevent excessive oxidation during the high-temperature test. Six thermocouples located along the centerline of the test block were used to monitor the temperature profile through the thickness.

During testing the B4C face of the test block was exposed to a high-temperature plasma for a duration of 7 minutes to produce the temperature profile shown in Figure 10b. A maximum surface temperature of 1,200ºC was measured for the B4C coating using a two-color pyrometer. However, it should be emphasized that the pyrometer was placed at an approximate angle of 45º to the B4C face, and it was aimed a few centimeters away from the point of plasma impingement. Therefore, the temperature at the point of impingement was greater than 1,200ºC. The distance between the plasma and the test block was reduced twice to obtain higher test temperatures. Visible light glow signifying extreme heating at the point of plasma impingement was observed only when the distance was reduced to 3 cm. The test duration of 7 minutes was reasonably close to the re-entry time of 8–12 minutes recorded during the Apollo lunar missions.23 Two thermocouple profiles (TC1 and TC6) are shown in Figure 10b. TC1 was placed 6.7 mm below the B4C and the carbon foam interface while TC6 was positioned 12.1 cm below the coating interface of the 12.7 cm thick test block. TC1 recorded a maximum temperature of about 800ºC during the test. This indicates that the B4C coating provided significant thermal protection by dropping the temperature by more than 400ºC over a thickness of 6.7 mm. More importantly, the temperature profile recorded by TC6 indicates a maximum temperature of 39ºC for the duration of the test. Therefore, a structural PE composite placed below the 12.7 cm thick thermal protection system would be adequately protected. The thermal protection capability presented here is preliminary. Further refinement of the surface temperature measurement system is required to accurately estimate the high-temperature capabilities of the proposed system. However, this preliminary testing demonstrates a path toward development of a complementary thermal protection system without significant sacrifice of the radiation shielding effectiveness.


Simulation results indicated that over thicknesses ranging from 1 g/cm2 to 20 g/cm2, Composite 1 was only about 1% to 5% less effective than pure PE, while the entire composite structure, including thermal and ballistic protection, was only 2% to 8% less effective compared to pure PE. However, transport code analysis indicated that the composites were much superior to traditional aluminum aerospace alloys. These predictions were validated by subsequent sample construction and 800 MeV/µ 28Si beam testing. The measured mechanical, ballistic, and thermal properties indicate that a multifunctional composite approach can significantly compensate for the marginal increase in areal density required for the composites to be comparable to pure PE in terms of radiation shielding effectiveness. Tensile strength measurement demonstrated that the developed composites have specific strength 2.5 to 4 times greater than typical aluminum aerospace alloys. Numerical evaluation and experimental validation demonstrated that selection of an open cell carbon foam material with a VPS deposited B4C coating can be instrumental in providing protection from micro-meteoroid impact and re-entry temperatures to an underlying PE structural composite.

During ballistic testing, the B4C coating was responsible for complete erosion of the projectile almost on impact. Extremely low bulk and shear modulus of the carbon foam prevented transmission of the compressive stress wave generated during impact, thereby protecting the underlying structural composite material. Also of note is the versatility of the VPS technique in depositing a well adhered B4C coating even on an irregular surface such as the open cell C foam, circumventing the problems typically associated with depositing coatings on curved and uneven surfaces for high-temperature use. Thermal testing indicated that on exposing the B4C layer to temperatures 1,200ºC and above, the maximum temperature recorded at the bottom of the 12.7 cm thick carbon foam sample did not exceed 40ºC. This thermal capability is comparable to that of the high temperature ceramic tiles that are currently used on the space shuttle.

It should be noted that expected reentry temperatures from a lunar mission will be higher, and hence much more elaborate and accurate characterization of thermal management schemes are mandated. Nevertheless, this paper has demonstrated the validity and importance of a multifunctional composite architecture that could potentially circumvent the problem of paying a penalty in additional weight to implement passive radiation shielding. Such a multifunctional composite holds the key for safe long duration missions beyond LEO. As a first step, the developed composites were selected by NASA’s Materials International Space Station Experiment program for further evaluation. As part of this evaluation these samples have recently been attached to the aft side of the International Space Station and will be exposed to the LEO environment for approximately 6 months.


The authors are grateful to the NASA Small Business Innovative Research program for funding this effort. The experimental nuclear physics group at Brookhaven National Laboratory, New York, is acknowledged for support and expertise during exposure of the samples to heavy ion beams and data reduction. The authors are also grateful to Dr. Sheila Thibeault of NASA’s Langley Research Center for technical discussions and encouragement during the course of the project. The authors wish to express their gratitude to Ms. Ernestine Cothran of BAE Systems for her patience and dedication through resolving administrative issues toward the project.


1. “Guidance on Radiation Received in Space Activities,” National Council on Radiation Protection and Measurements Report No. 98 (1989).
2. G.D. Badhwar and P. O’Neill, Adv. Space Res., 17 (2) (1996), pp. 7–17.
3. J.W. Wilson, Health Phys., 79 (5) (2000), p. 470.
4. M.A. Shea and D.F. Smart, Solar Phys., 127 (June 1990), pp. 297–320.
5. T.A. Parnell, J.W. Watts, and T.W. Armstrong, “Radiation Effects and Protection for Moon and Mars Mission,” Proceedings of the Sixth International Conference and Exposition on Engineering, Construction, and Operations in Space, Space 98 (Reston, VA: American Society of Civil Engineers, 1998), pp. 232–234.
6. H. Smith, editor, 1990 Recommendations of the International Commission on Radiological Protection, Annals of the ICRP, vol. 21/1-3 (New York: Pergamon Press, 1991).
7. “The Relative Biological Effectiveness of Radiations of Different Quality,” National Council on Radiation Protection and Measurements Report No. 104 (1990).
8. J.W. Wilson et al., NASA Technical Paper 3682 (Hampton, VA: Langley Research Center, December 1997).
9. J.E. Turner, Atoms, Radiation, and Radiation Protection, 2nd edition (New York: Wiley-Interscience Publishers, 1995), pp. 160–161, 350–351.
10. F.A. Cucinotta et al., Adv. Space Res., 17 (2) (1996), pp. 77–86.
11. S. Guetersloh et al., Nucl. Instr. and Meth. in Phys. Res. B., 252 (2006), pp. 319–332.
12. R.K. Kaul, A.F. Barghouty, and H.M. Dache, “Radiation Transport Properties of Polyethylene–Fiber Composites,” Proceedings of the Conference on Microgravity Transport Processes in Fluid, Thermal, Biological, and Materials Sciences (New York: Annals of the New York Academy of Sciences, 2004), pp. 138–149.
13. J. Miller et al., Rad Res., 159 (2003), pp. 381–390.
14. J.W. Wilson et al., NASA Report #1257 (Washington, D.C.: NASA, 1994).
15. P.M. O’Neil, Adv. Space Res., 37 (2006), pp. 1727– 1733.
16. F.C. Campbell, Manufacturing Technology for Aerospace Structural Materials, 1st edition (Amsterdam, The Netherlands: Elsevier Ltd., 2006).
17. D.M. Choi et al., Polymer, 38 (25) (1997), pp. 6243–6249.
18. Reusable Launch Vehicle Technology Development and Test Program, report submitted by the Committee on Reusable Launch Vehicle Technology and Test Program (Washington, D.C.: National Academy Press, 1995).
19. ASM Materials Handbook, Volume 20 (Materials Park, OH: ASM International, 2001).
20. Aluminum Standards and Data (Arlington, VA: The Aluminum Association Publication, 2006), pp. 2-2–2-3.
21. G. Johnson and T. Holmquist, J. Appl. Phys., 85 (12) (1999), pp. 8060–8073.
22. A. Tasdemirci and W. Hall, J. Comp Mat., 39 (11) (2005), pp. 981–1005.
23. J.E. Pavlosky and L.G. St. Leger, NASA Technical Note # TN D-7564 (Washington, D.C.: NASA, January 1974).

S. Sen, principal scientist for the NASA Program, is with BAE Systems, NASA Marshall Space Flight Center/EM 30, Building 4464, Room 111A, Huntsville, AL 35812, USA. E. Schofield, materials engineer, and J.S. O’Dell, project engineering manager, are with Plasma Processes Inc., Huntsville, Alabama. L. Deka, materials engineer, and S. Pillay, assistant professor, are with the Department of Material Science & Engineering, University of Alabama at Birmingham. Dr. Sen can be reached at (256) 544- 8264; fax (256) 544-6660; e-mail