POLYURETHANE CHEMISTRY IN SPACE

Jan-Michael Gosau, Jeremy Barlow, Tyler Wesley and Ronald E. Allred Adherent Technologies, Inc. Development Laboratories 11208 Cochiti SE Albuquerque, NM 87123

ABSTRACT
Future space exploration past the earth orbit has a significant need for manufacturing-in-space beyond simple assembly of prefabricated parts. The two main areas of research right now are large structures and in-situ resource utilization. The next generation of very large aperture antennas will exceed the size achievable with conventional folding mesh technologies, and new concepts are needed to support football field size structures. Envisioned lunar bases will need to have to deal with the omnipresent regolith, both as a resource and as a hazard. Adherent Technologies, Inc. has, under NASA funding, developed technologies to address both problems using the formation of polyurethanes in a vacuum environment. Large inflatable structures can be stabilized by the formation of polyurethane foams of controlled density. Regolith on the moon can be used as a building material by binding it into bricks with nonfoaming polyurethane mixtures, and the regolith dust can be controlled by the application of thin polyurethane layers. The paper describes the challenges of performing chemistry in a vacuum environment, especially of obtaining controlled foaming of polyurethanes, and the solutions developed.

1. INTRODUCTION
Building large structures in space or converting large quantities of extraterrestrial material into useful building blocks by chemical means ultimately requires compatibility of the process with the vacuum environment. NASA mission scenarios include football field size relay antennas to be deployed in interplanetary space and the use of large quantities of lunar regolith for habitat construction on the moon. Both applications are of a size that make it impractical to conduct them in a pressurized module. Unfortunately, most chemical operations on earth are either based on volatile organic solvents or water, both of which are incompatible with the application scenario. Urethanes are formed readily by the reaction of hydroxyl and isocyanate moieties. Using at least difunctional monomers, polymeric materials with a large range of potential uses are formed. Depending on the conditions, hard rubbers, used as wheels to transport heavy loads, or lightweight foams for insulation can be obtained from the same base ingredients. For the use in a vacuum environment, the availability of oligomeric precursors is important. Low molecular weight components would immediately evaporate, changing the stoichiometry of the reaction

and potentially contaminating a space environment. High molecular precursors have a much more limited range of properties. This paper describes first the use of polyurethanes as a solid elastomer to consolidate lunar regolith into a stable building material with minimal use of earth supplied materials. During lunar exploration, regolith is both the major available resource and a substantial obstacle in establishing a long term presence. The fine surface dust is highly abrasive and can both be a health hazard for the crew and the main contributor to equipment failure. On the other hand, regolith is the only material readily available to build protective shelters for the crew strong enough to prevent exposure to lethal doses of high energy particles during solar eruption events. Estimates show that 10 ft of solid cover is desired for long term exposure protection [1]. To achieve this heavy cover, methods have to be developed to convert regolith into a stable building material with minimal need for terrestrial supplies. The technology should also be applicable to stabilize regolith in place, for example on excavated slopes or around air locks and landing fields, where dust is disturbed by gas flows. Typical approaches used in terrestrial applications would rely on water to trigger cement forming reactions; this is not acceptable for lunar applications. In the second part, the paper deals with the use of polyurethane foam to stabilize large inflatable structures in space. Large aperture antennas are of interest to NASA for applications in establishing high-speed communication relays for interplanetary missions. Initial design goals include 20-meter apertures. To minimize mission costs at these large diameters, weight must be severely restricted. Current large aperture projects like the James Webb Space Telescope (JWST) have a target areal density of 1015 kg/m2 -- an areal density of 1 kg/m2 is desired for new large aperture concepts. Combined with the desired low mission weight is a high surface and form accuracy. ATI has studied two technologies for ultra-light space structures and antennas. The first, the Rigidization on Command (ROC) concept, uses UV-curing resins to stabilize inflated structures [2 810- 11]. This removes the need for make-up gases inherent in typical inflatables. This technology is suited for extremely low areal density construction, but the thermal and mechanical dampening in these structures is low. ATI has also developed a self-deploying foam antenna concept at small aperture sizes under funding from the NASA SBIR/STTR Program [12]. A 0.6-meter antenna was fabricated and tested at NASAs Jet Propulsion Laboratory. An RMS surface error of 28 mil was demonstrated as was some beam focusing despite relatively high RF losses. These foam antennae have an areal density of approximately 5 kg/m2, which was achieved with a foam density of 48 kg/m3 (3 lb/ft3). While the areal density of these foams was as low as any state-of the art-solution, it was still a factor of 5 above the threshold for very large aperture designs. We, therefore, explored the usage of the foam technologies in combination with ROC systems to achieve the required low areal density antenna structures. The new foam materials would be used as a basis for an in-space foamed structure, using the ROC technology to provide the necessary shape control. The advantages of the new system include low areal density, high volumetric packing efficiency (with a target density of 10 kg/m3 the theoretical packing density is 100:1 from the liquid precursor to the foamed structure), and self-deploying characteristics.

2. EXPERIMENTAL
All chemicals used in the production of the urethane forming reactions were used as is from their suppliers without purification or analysis. Once apportioned, the chemicals were stored at room temperature under a blanket of argon to reduce water absorption. Regolith simulant JSC-1A was provided by NASA JPL. Due to its large surface area, the simulant required extensive drying before use. DSC analysis showed irreversible changes to the material above 250C, drying temperature was therefore limited to 200C in vacuo. Components (component A being the polyol mixture, component B being the neat isocyanate) were prepared by carefully measuring the required equimolar amount of each compound in the component one at a time. Once all of the compounds that comprise the component had been combined, the component mixture was placed under argon and sealed hermetically before being stirred magnetically to homogenization. The compounds present and their relative amounts were varied as needed to obtain the required properties of the polyurethane end product. Thermogravimetric analysis (TGA) was performed using a 2050 thermogravimetric analyzer manufactured by TA instruments. Differential scanning calorimetry (DSC) was performed using a 2920 modulated DSC manufactured by TA instruments. In order to study the formation of urethanes in vacuo and on a high surface area substrate it was necessary to construct a test chamber which allowed blending of substrate and urethane mixture under those conditions. A suitable test chamber, allowing the introduction of urethane either via a magnetic release or by mixture in situ was designed and is shown in Figure 1.

Figure 1. Magnetic pouring device positioned on bell jar base (glass bell removed for clarity)

3. RESULTS AND DISCUSSION Lunar Regolith Stabilization

The starting foam formulations used in this program were originally developed for the stabilization of asbestos prior to removal [13]. Figure 2 shows penetration test samples of ATI polyurethanes with different diluting agents on two inch cubes of dense fiberboard. The material is completely soaked through by the resin. The solidified fiberboard, usually crumbling under hand pressure, needed to be cut by a band saw to show the cross sections. The material was successfully tested for asbestos stabilization during removal in ship demolition operations [13].

Figure 2 a and b. HCFC (2 a) and pentane (2b) formulations on 2 cube of high density board While the original formulations had shown great potential they depended on low molecular weight dilution agents to achieve any penetration. For the lunar application, low viscosity and slow cure was needed to enable similar results. Initial experiments using dried sand as a substrate showed penetration depths of up to 2 cm before the progressing resin cure stopped any further advance. The penetration of the same resins into regolith simulant was much less pronounced; due to the very fine particle size the material compacts very well and the high surface area further reduces the ability of the resin to flow. Also, drying the regolith simulant prior to resin application was essential to suppress foaming, which also reduced penetration depth. After it was established in small scale experiments that the urethane components could be successfully combined with the surface material under vacuum with little to no cell formation, the experiments were expanded such that the surface material was spread over an aluminum pan. These experiments produced samples such as the one shown in Figure 3.

Figure 3. Surface stabilized sample produced by pouring the urethane components over dried regolith under vacuum. The sample needed to be broken to be pulled from the pan. By the end of this part of the program, approximately 2 grams of surface material could be stabilized for every gram of urethane applied to it. This produced a 2 mm thick, very hard surface layer, that could not be ripped by hand, but needed to be cut using shears. As is, it is suitable as a coating to form lunar roads in heavily trafficked areas. Building Brick Formulations Although there are two ultimate end-uses for the urethane material, the overall function of it largely remains similar: to stabilize loose lunar surface material via incorporation into a urethane matrix. That being the case, extrapolating the function of the urethane from a stabilization role to one of binding surface material together was not altogether unfeasible. When the urethane formula designed for surface stabilization was used as the initial formulation for experiments involving the fabrication of a block, it performed quite well. Reformulations were mainly focused on the amount of catalyst to adjust the cure time. The modifications were made so that more time was available for combining the regolith and urethane components before onset of cure. The first experiments were made using sand to preserve the limited amount of regolith simulant on hand. The experiments were varied in their relative surface material to urethane component ratios. The experiments started using a relatively concentrated mixture of five parts sand to one part urethane components (mass to volume). A polytetrafluoroethyelene (PTFE) mold was used to form the material. The resulting bricks were so robust that the relative ratios were diluted to

10, then 15, and finally 20 to 1 ratios respectively. In every case, a robust brick was made (Figure 4).

Figure 4. Sand-based brick at 20 to 1 ratio The sand-urethane mixtures were easily blended by hand, and the resulting bricks were fully uniform, and required a band saw to section. Using these results, we started using JSC-1A regolith simulant under identical conditions. Development of a suitable mixing method for this material was a non-trivial matter. Blending low volumes of urethane mixture with the regolith lead to clumpy mixtures that cured before the regolith and urethane were completely mixed. Even the slowest possible formulations were not homogenously blended. Other changes to the formulation were also unsatisfactory in their result the bricks could not be removed intact from the mold. In order to give the reactants sufficient time to fully coat the regolith, the individual components were blended with half of the regolith respectively. By using a ball mill type set-up to blend the materials under argon, homogeneous mixtures were obtained. The treated regolith was then pressed into the PFTA brick mold and allowed to cure (Figure 5).

Figure 5. PTFE mold with regolith brick during cure The time allotted for curing was based - in part on the approximated cure time given the amount of catalyst in the urethane component system. Most blocks were allotted 8 to 10 hours to cure. The resulting blocks were more structurally sound than those obtained by blending the urethane mixture with regolith, but still not at the desired strength. A 3 m stack of bricks exerts approximately 1.5 psi pressure on the bottom layer under lunar conditions. Unfortunately, neither changes to the drying process nor further experiments with varying degrees of isocyanate excess, presuming that the chemically labile isocyanate compound was interacting with the surface chemistry of the regolith, were successful; the consolidated brick showed clear signs of poorly reacted material. The breakthrough was finally achieved when the regolith was first coated with the polyol components, allowed to blend well, followed by the addition of the isocyanate just prior to consolidation in the mold. The resulting bricks were easily removed from the mold and showed good mechanical properties (Figure 6).

Figure 6. Regolith block after curing in the mold These bricks were made with a ten to one ratio of regolith and polyurethane. Attempts at higher dilution ratios using the mold technique were unsatisfactory. Mechanical testing was performed using the ASTM D1633 compression test as a guideline. A typical stress-strain curve is shown in Figure 7.

Figure 7. Compression Stress-Strain Curve for a regolith-urethane composite block

The measured maximum stress of 800 psi is sufficient to support the weight of a 500 m stack of bricks under lunar conditions, plenty of strength to hold up the sites of the most ambitious excavations currently proposed. In addition, the large area under the stress-strain curve in Figure 7 indicates a high level of toughness. To improve the mechanical strength of the building bricks at lower urethane loading, we experimented with compression molded samples. Using a inch diameter round mold, loaded to 2000 lbs (10,000 psi), very strong composite samples were obtained. Figure 8 shows one of the samples, and the accompanying strain-stress curve of the compression test. The sample is stable up to 1300 psi.

. Figure 8. Compression molded 20:1 sample (left) and compression test curve (right) Inflatable Antenna with Foam Support Literature data indicates that it is hard to predict the foaming behavior of standard foams under reduced pressure [14]. Our design fills a correctly shaped double layer membrane with highexpansion polyurethane foam. This foam is made from open-cell material, allowing all gases used during expansion to rapidly diffuse out of the foam after deployment, greatly reducing the tendency to absorb heat under solar irradiation and thereby reducing the chance for thermal deformation due to changes in orbital solar exposure. Figure 9 depicts the deployment process. A three-layer membrane is deployed in space, having two separate chambers, and having an integrated metallized, rigidizable surface to form the future RF-reflective coating.

Figure 9. Foam stabilized inflatable large aperture antenna The first chamber is inflated to give the antenna its desired shape. The shape is then frozen by rigidization to prevent possible distortion of the antenna during the subsequent foam stabilization process. Then, the second, smaller chamber is filled with foam to permanently stabilize the membrane shape and protect it from damage. After the foaming process is complete, the inflation gases are allowed to diffuse out, removing the source of most thermal distortion potential typical for a classic inflated structure. From the experiments discussed in the first part of this paper, it was clear that polyurethanes could be successfully cured in a vacuum environment. The critical part for this application was the controlled volume expansion to give a low weight structural foam. Initial experiments showed that a completely unrestrained system foamed highly inconsistently, with volume variability of 500% depending on the speed of skin formation. If the foaming took place in a perforated mold however, the ratio of foam expansion became very controllable. Adding a third of a percent of water to the urethane mixture provided the desired foam density of 10 kg/m3 when expanded in vacuo. Small scale laboratory experiments using a static mixer to inject a foam mixture into perforated latex forms produced shapes with a 10% volume control. The problem in manufacturing larger parts was not, as initially assumed, the vacuum environment but the mixing of components of different viscosities, especially when long feed lines to the static mixer were needed. The antenna program was accompanied by extensive computer modeling, conducted by Paul McElroy, to show the viability of the concept for very large structures. Prototype design and construction was performed in cooperation with ILC Dover, the leading manufacturer of space inflatables. The program was managed by Steve Scarborough and the lead engineer was Dan Gleeson. The prototype antenna built for this program was strictly a technology demonstrator and not tested for RF performance. We, therefore, decided on a simple spherical design for the mold. Figure 10 shows the mold design drawings. The top bladder is made from Mylar, and the foam containment bladder is constructed from clear polyfluoroethylene (PFE).

Figure 10. Demonstration antenna mold design, unfilled (left) and after foam deployment (right) ATI has developed nine different UV-curable resins for ROC applications [8]. While the epoxy acrylates are stronger and stiffer, the urethane acrylates are still reasonable and show higher strain to failure. Also, the urethane acrylates have superior cold cure, requiring less heating in a space environment. ATI-ROC-U945X20, having a high Tg and the best cure properties, was chosen as resin for the prototype antenna ROC layer. The resin was applied to a layer of e-glass fabric and sandwiched between an aluminized Mylar foil and clear Mylar. The finished mold can be seen in Figure 11.

Figure 11. Appearance of the prototype mold

The ROC layer was then UV cured to produce the reflector part of the antenna. Figure 12 shows a schematic drawing of the injection set-up in the vacuum chamber. It consists of two injection lines leading to the mixing head, vacuum controls for the chamber vacuum, and a separate control loop for the pressure inside the inflated top half of the mold.

Figure 12. Prototype manufacture set-up The mold was then mounted inside ATIs vacuum chamber. The pressure in the inflated top chamber of the mold was regulated to 0.4 torr above the pressure in the vacuum chamber (approximately 1 torr). Foam injection was performed using a mixer head containing a 15 cm long static mixer of 12 mm diameter. The volume of the foam chamber in the mold was calculated to be 20 l. We injected 105 ml of each component into the mold for a total of 237 g of foam. The resulting foam-filled mold is shown in Figure 13.

Figure 13. Inflated mold after foaming in vacuum The demonstration antenna was then removed from the vacuum chamber, and the top balloon was deflated, resulting in the finished antenna shown in Figure 14.

Figure 14. Foamed antenna in final configuration The finished weight of the antenna was 654 g. The areal density of the antenna was 2.25 kg/m3, due to the 25 cm strong foam backing in the center. Future models will incorporate baffles to prevent the deformation of the mold into an excessively baggy shape under gravity.

4. CONCLUSION
Vacuum processing of liquid polyurethane precursors has been shown to be a viable method to tackle some of the biggest problems in lunar regolith handling. The formulations and methods presented in this paper allow for the use of urethanes to stabilize loose regolith during excavation or in high-traffic areas. They can also be used as dust mitigation coatings. By blending the urethane with regolith, it is also possible to produce building blocks from excavated regolith. These blocks can be stacked up to a predicted height of 500 m before their own weight would crush them under lunar conditions. This allows them to be used as retaining walls or in the building of covered radiation shelters. The methodology developed for the production of foam stabilized inflatables in a vacuum environment was a major technological breakthrough. It allows the production of very large but lightweight structures in space with permanent stabilization against pressure loss and vibrations. By meeting the design goal of 1 kg/m3 for a very large aperture antenna, the technology could become an enabling step for the interplanetary antenna program. Computer modeling has given vital input for the future design decisions and the manufacture of the technology demonstrator gave us insight in future manufacturing-in-space applications.

5. ACKNOWLEDGEMENTS
The work on lunar regolith treatment was funded by NASA under an SBIR Phase I Grant Lunar Regolith Excavation and Material Handling Contract No. NNX08CD28P. We wish thank Kenneth R. Johnson of NASA Jet Propulsion Laboratory for his support during the program and the provided JSC-1 simulant. The antenna work was funded by NASA under an SBIR Grant Foamed Antenna Support for very large Apertures, Contract No. NNX08CC67P, and we thank the TPOC, James Gaspar of the NASA Langley Research Center, for invaluable discussion throughout the program.

6. REFERENCES

1 THE LUNAR ENVIRONMENT, D. Vaniman, R. Reedy, G. Heiken, G. Olhoeft, W. Mendell in The Lunar Sourcebook Cambridge University Press 1991 2 R.E. Allred, L.A. Harrah, A.E. Hoyt, P.M. McElroy, R.A. Wise, and M.C. Lou, Inflatable Spacecraft Using Rigidization on Command (ROC) Concept, Proc. 41st AIAA/ASME/ASCE/AHS/ASC SDM Conf., Atlanta, GA, April 3-6, 2000, paper 2000-1637 3 P. M. McElroy, R. A. Wise, R. E. Allred, L. A. Harrah, and A. E. Hoyt, Orbital Thermal Analyses of Rigidization on Command (ROC) Materials for Inflatable Spacecraft, 31st Intl Conf. on Environmental Systems, Orlando, FL, July 2001, Technical Paper Series 2001-01-2220

4 R. E. Allred, A. E. Hoyt, P. M. McElroy, S. Scarborough, and D. P. Cadogan, UV Rigidized Carbon-Reinforced Isogrid Boom for Gossamer Applications, Proc. Space Technology and Applications Intl Forum-STAIF 2002, Albuquerque, NM, Feb. 3-6, 2002, pp. 406-413 5 R. E. Allred, A. E. Hoyt, P. M. McElroy, S. Scarborough, D. P. Cadogan, UV Rigidizable Carbon-Reinforced Isogrid Inflatable Booms, Proc. 43rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conf., April 22-25, 2002, Denver, CO, Paper 2002-1202 6 A. E. Hoyt, L. A. Harrah, R. E. Allred, and P. M. McElroy, Rigidization-on-Command (ROC) Resin Development for Lightweight Isogrid Booms with MLI, 33rdt Intl Conf. on Environmental Systems, Vancouver, BC, July 2003, Technical Paper Series 2003-01-2342 7 D. Cadogan, T. Smith, R. Lee, S. Scarborough, and D. Graziosi, Inflatable and Rigidizable Wing Components for Unmanned Aerial Vehicles, Proc. 44th AIAA/ ASME/ASCE/AHS Structures, Structural Dynamics, and Materials Conf., Norfolk, VA, April 7-10, 2003, paper 2003-1801 8 A. E. Hoyt Haight, L. A. Harrah, and R. E. Allred, Light Curing Resin Systems for Use in Inflatable Wing Applications, Proc. 50th Intl. SAMPE Symp. and Exhib., Long Beach, CA, May 1-5, 2005 9 R. E. Allred, A. E. Hoyt, L. A. Harrah, P. M. McElroy, S. Scarborough, D. Cadogan, and J. W. Pahle, Light Curing Rigidizable Inflatable Wing, Proc. 45th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conf., April 1922, 2004, Palm Springs, CA, paper 2004-1809 10 A. E. Hoyt Haight, R. E. Allred, L. A. Harrah, P. M. McElroy, S. E. Scarborough, and T. Smith, Design and Fabrication of Light Rigidizable Inflatable Wings, Proc. 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conf., May 1-5, 2006, Newport, RI, paper 2006-1695 11 A. E. Hoyt Haight, R. E. Allred, and L. A. Harrah, Design and Fabrication of Light Rigidizable Inflatable Wings, Proc. 47th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conf., May 1-3, 2006, Newport, RI, paper 51026 12 Self-Deploying Foam Antenna Structures, Contract No. NAS5-02042, NASA Goddard Space Flight Center 13 J. M. Gosau, T. Shkindel, and R. E. Allred, Encapsulating Foam System for Asbestos Remediation, Proc. 50th Intl. SAMPE Symp. and Exhib., Long Beach, CA, May 1-5, 2005 14 M. Schoenfeld, M. Tinker Proceedings 44th AIAA/ASME/ASCE/AHS Structures, Structural Dynamics, and Materials Conference, Norfolk, VA, April 2003.