Search

This report documents findings and recommendations by the Ultrafast Computing Team (UCT). In the period 10-12/98, UCT reviewed design case scenarios for a supersonic transport and a reusable launch vehicle to derive computing requirements necessary for support of a design process with efficiency so radically improved that human thought rather than the computer paces the process. Assessment of the present computing capability against the above requirements indicated a need for further improvement in computing speed by several orders of magnitude to reduce time to solution from tens of hours to seconds in major applications. Evaluation of the trends in computer technology revealed a potential to attain the postulated improvement by further increases of single processor performance combined with massively parallel processing in a heterogeneous environment. However, utilization of massively parallel processing to its full capability will require redevelopment of the engineering analysis and optimization methods, including invention of new paradigms. To that end UCT recommends initiation of a new activity at LaRC called Computational Engineering for development of new methods and tools geared to the new computer architectures in disciplines, their coordination, and validation and benefit demonstration through applications.

Calculating the line of intersection between two three-dimensional objects and using the information to generate a third object is a key element in a geometry development system. Techniques are presented for the generation of three-dimensional objects, the calculation of a line of intersection between two objects, and the construction of a resultant third object. The objects are closed surfaces consisting of adjacent bicubic parametric patches using Bezier basis functions. The intersection determination involves subdividing the patches that make up the objects until they are approximately planar and then calculating the intersection between planes. The resulting straight-line segments are connected to form the curve of intersection. The polygons in the neighborhood of the intersection are reconstructed and put back into the Bezier representation. A third object can be generated using various combinations of the original two. Several examples are presented. Special cases and problems were encountered, and the method for handling them is discussed. The special cases and problems included intersection of patch edges, gaps between adjacent patches because of unequal subdivision, holes, or islands within patches, and computer round-off error.

Aero-assisted orbital transfer vehicles are analyzed. The aerodynamic characteristics over the flight profile and three- and six-degree-of-freedom performance analyses were determined. The important results, to date, are: (1) the aerodynamic preliminary analysis system, an interactive computer program, used to predict the aerodynamics (performance, stability, and control) for these vehicles; (2) the performance capability, e.g., maximum inclination change, maximum heating rate, and maximum sensed acceleration, can be determined using continuum aerodynamics only; (3) guidance schemes can be developed that allow for errors in atmospheric density prediction, mispredicted trim angle of attack, and off-nominal atmospheric interface conditions, even for vehicles with a low lift-to-drag ratio; and (4) multiple pass trajectories can be used to reduce the maximum heating rate.

This paper investigates the use of multiple-pass trajectories for aeroassisted orbital transfer vehicles (AOTV) as a means of reducing the severe aeroheating environment and the extreme sensitivity to off-nominal trajectory and atmospheric conditions. One-, two-, and three-pass trajectories were calculated for AOTV's with ballistic coefficients ranging from 5 to 150 psf. Compared to the single-pass case, the maximum heat rate was reduced by 30 percent for two passes and 45 percent for three passes, while the maximum acceleration was reduced by 40 percent and 55 percent, respectively. The sensitivity of the trajectories to variations in atmospheric and orbital parameters was not significantly reduced.

It is pointed out that early recognition of operational requirements and assessment of their effects provides the best chance of designing an economically viable future space transportation system (FSTS). Operational factors which may influence configuration design include fleet size, operation mode, refurbishment, and the resource requirements. FTST has a simplified operational role compared to the Space Shuttle, which, in addition to transportation, must perform experiments, support payloads, and stay long on orbit. In the future a space station will perform many of these tasks. The FSTS comprises a two-stage fully reusable launch vehicle designed to carry 150,000 lb to a space station, off-load, and return. It would always be launched fully loaded, and its cargo would be processed and redistributed at the space station. Attention is given to ground servicing, flight operations, rendezvous-compatible orbits, launch windows, standard trajectories, entry windows, operational costs, the mission model, and resource requirements.

This paper presents the results of a comparison of space vs. ground basing of orbital transfer vehicles (OTV). The comparison was done assuming an 11-year mission model beginning in 1995 and averaging over 100 MT of payloads per year to GEO. When analyzed from a total transportation standpoint, the launch system employed had the greatest impact on the basing mode comparison. A launch fleet consisting of both a basic STS and a Shuttle derivative cargo vehicle provided the least cost. Only a small advantage (3%) was found in flight performance for the space based (SB) OTV once it incorporated the necessary provisions for space debris protection and on-orbit maintenance. Propellant storage and transfer losses associated with the SB OTV amounted to 12% of the actual flight propellant requirement. Turnaround of the SB OTV required a crew of 3 and 40% duty cycle. The life cycle cost comparison including all involved space elements indicated less than a 10% difference between basing modes. The SB OTV however is judged to provide more flexibility in launch manifesting, simplified recovery operations and greater potential for improvement.

Many studies are being made to determine the most efficient method for the transfer of large space structures from low earth orbits (LEO) to higher altitudes, principally geosynchronous orbits. In this paper, towing is proposed as a possible means of orbital transfer of large space structures. A potential towing arrangement is identified and some of the systems sized for the transfer of a 2460-ft (750 meter) diameter antenna weighing approximately 154,000 lbs (70 metric tons). A savings of approximately 23 metric tons in propellants and structure resulted from the use of a towing arrangement using six towlines in parallel when compared to the use of a single tug hard mounted to the structure being towed. The towing concept appears practical and affords a means of applying loads using multiple towlines to a wide variety of structures without special adaptors.

Computerized structural sizing techniques were used to determine structural proportions of minimum mass tetrahedral truss platforms designed for low Earth and geosynchronous orbit. Optimum (minimum mass) deployable and erectable, hexagonal shaped spacecraft are sized to satisfy multiple design requirements and constraints. Strut dimensions characterizing minimum mass designs are found to be significantly more slender than those conventionally used for structural applications. Comparison studies show that mass characteristics of deployable and erectable platforms are approximately equal and that the shuttle flights required by deployable trusses become excessive above certain critical stiffness values. Recent investigations of eractable strut assembly are reviewed. Initial erectable structure assembly experiments show that a pair of astronauts can achieve EVA assembly times of 2-5 min/strut and studies indicate that an automated assembler can achieve times of less than 1 min/strut for around the clock operation.

The sensitivity of the trajectory of an Aeromaneuvering Orbit-to-Orbit Shuttle (AMOOS) to unpredictable variations in atmospheric density is shown and a compensating guidance law is developed. The sensitivity was determined using a varying atmosphere model whose statistical properties matched those of actual atmosphere data. This study showed that a compensating guidance law is crucial to the feasibility of AMOOS. Using an optimal regulator approach, a linear feedback law for bank angle was developed and checked out with the atmosphere model. This law uses the altitude profile of the trajectory to control the amount of energy dissipated within the atmosphere.

One candidate for a reusable upper stage to be carried by the space shuttle is an aeromaneuvering orbit-to-orbit shuttle (AMOOS). This concept uses the drag of the vehicle during a pass through the atmosphere rather than the propulsion system to slow the vehicle on a return from a high energy orbit. The nature and magnitude of the sensitivity of AMOOS to uncertainties in the properties of the atmosphere are shown. Various guidance schemes for correcting for the effects that the unpredictable variations in the atmosphere have on the trajectory are discussed. For the mission studied here, a payload retrieval from geosynchronous orbit with aerodynamic plane change, a linear feedback guidance scheme was developed. A relatively simple heuristic law was used to demonstrate the concept. Using optimal control theory a feedback law was developed analytically. Testing with a large number of different atmospheres showed this law to be a feasible means of controlling the AMOOS trajectory. Refinements to the technique offer promise of significant improvement, and these are discussed.

A study is presented that identifies the potential performance capability of the production shuttle orbiter for powered flight tests using several propulsion systems following vertical takeoff and using J-2 rocket engines following air launch. Of the approaches considered, the air-launched orbiter equipped with J-2 rocket engines appeared to have the highest potential for early shuttle development flights. With this approach, mach 4 appeared attainable using 45.5K kg (100K lb) of internal propellant. Several issues were identified that require resolution to prove feasibility.

A preliminary automated structural sizing procedure suitable for conceptual design and early tradeoff studies of large truss platforms configured for shuttle transportation to LEO is discussed as well as some orbital transfer design considerations. Platforms that are sized to withstand orbital transfer loads for the LEO to GEO maneuver are compared to platforms sized only for LEO application. It is concluded that for platforms supporting low mass distributed payloads, platform and strut frequency requirements are strong design drivers for LEO applications. The struts are found to be extremely slender, thinswalled, and of small diameter. If full advantage is to be taken of these minimum mass designs, a manufacturing capability must be developed for long straight struts. For platforms that are to be transferred from LEO to GEO in a deployed state, the orbital transfer loads become design drivers. However, even for an initial thrust-to-weight ratio equal to 0.1, a platform on the order of 100 m diameter appears packageable with its OTV in one shuttle flight, and larger platforms appear possible at lower thrust-to-weight ratios.

The structural proportions of minimum-mass tetrahedral truss platforms designed for low earth and geosynchronous orbit are determined by means of computerized sizing techniques, taking into account multiple design requirements and constraints. Strut dimensions characterizing minimum mass designs are found to be significantly more slender than those used for conventional structural applications. It is also shown that the number of shuttle flights required by deployable trusses becomes excessive above certain critical stiffness values, and that an automated assembler can achieve rates of 1 min/strut, by comparison with 2-5 min/strut for two astronauts using manual labor.

Structural optimization studies are made using mathematical programming techniques to examine minimum mass structural proportions of deployable and erectable tetrahedral truss platforms subject to the integrated effects of practical design requirements. Considerations integrated into the optimization process are: 1) lowest natural frequencies of the platform and individual platform components (struts); 2) packaging constraints imposed by the Shuttle cargo bay capacity; 3) initial curvature of the struts; 4) column buckling of the struts due to gravity gradient, orbital transfer, strut length tolerance, or design loads; and 5) practical lower limits for strut diameter and wall thickness. Ultra-low mass designs are shown to be possible with strut proportions much more slender than those conventionally used for earthbound application.

The basic AVID (Aerospace Vehicle Interactive Design) is a general system for conceptual and preliminary design currently being applied to a broad range of future space transportation and spacecraft vehicle concepts. AVID hardware includes a minicomputer allowing rapid designer interaction. AVID software includes (1) an executive program and communication data base which provide the automated capability to couple individual programs, either individually in an interactive mode or chained together in an automatic sequence mode; and (2) the individual technology and utility programs which provide analysis capability in areas such as graphics, aerodynamics, propulsion, flight performance, weights, sizing, and costs.

A space-based orbital transfer vehicle has been sized for a 50-metric-ton payload delivery from low-earth-orbit to a geosynchronous orbit. Space basing effected substantial reductions in cryogenic insulation, tank, and body structure. The tank and body structural masses are shown to be lower for space basing because of the larger difference in acceleration loads between the on-orbit case (0.2 g's) and delivery (3.0 g's), the latter applying to ground-based vehicles which are delivered to orbit fully loaded with propellants. Insulation masses are lower because of the absence of an atmosphere and the attendant heat transfer losses. Insulation systems masses are also reduced because of the elimination of the problem of liquefaction and freezing of moisture on the tanks.

A comparison is made between a space-based and earth-based orbital transfer vehicle (OTV) for use in the Shuttle era and beyond. The space-based vehicle alleviates the limited capability inherent in an earth-based OTV whose design is constrained by a particular launch vehicle. Several sizes of space-based OTVs were generated and compared with an earth-based system for a number of mission scenarios with varying types and levels of traffic. The space-based OTVs showed substantial cost savings for each scenario, with the smallest space-based vehicle showing the largest saving. The space-based OTV retains the cost advantage even if the number of missions or the Shuttle cost-per-flight is drastically reduced.

A concept for a lightweight space-based orbital transfer vehicle (OTV) featuring thin, spherical, pressure-designed aluminum liquid-oxygen and liquid-hydrogen tanks and a truss structure of composite materials is used as a baseline design for a large-cargo OTV. Vehicle sizing, fleet analysis, and parametric cost analysis are used to evaluate the effects of engine technology, vehicle staging, and high-thrust versus low-thrust transfer. Results indicate that Earth-to-orbit launch costs and OTV engine performance are strong drivers in orbital transportation cost and that there is no significant benefit in staging vehicles. At the low values of Earth-to-orbit cost representative of advanced launch vehicles, low-thrust and high-thrust OTV's are competitive.

Techniques for sizing electrically or chemically propelled orbit transfer vehicles and analyzing fleet requirements are used in a comparative analysis of the two concepts for various levels of traffic to geosynchronous orbit. The vehicle masses, fuel requirements, and fleet sizes are determined and translated into launch vehicle payload requirements. Technology projections beyond normal growth are made and their effect on the comparative advantages of the concepts is determined. A preliminary cost analysis indicates that although electric propulsion greatly reduces launch vehicle requirements substantial improvements in the cost and reusability of power systems must occur to make an electrically propelled vehicle competitive.

The results of a parametric study of ascent performance are presented for a vertical take off, horizontal landing, single stage earth-to-orbit transport vehicle. Two dual fueled concepts, series burn and parallel burn, were investigated, both of which utilized dual position rocket nozzles. The analysis was made by systematically varying a set of propulsion similarity parameters, initial thrust-weight ratio, the proportion of the thrust due to dual position nozzle engines, expansion ratios of the rocket nozzle, and the relative split between the two fuels, hydrogen and hydrocarbon. The data are presented as a series of curves of mass ratio plotted against each of the similarity parameters for various combinations of the other similarity parameters.

Ascent performance characteristics were studied for a vertical-take-off, horizontal-landing, single-stage-to-orbit transport vehicle powered by hydrogen fuel rocket with a mixture of fixed- and dual-position nozzles. The analysis was made by systematically varying two sets of trajectory similarity parameters based on the propulsive and aerodynamic characteristics of the vehicle and by calculating a trajectory for each combination of the parameters. The propulsion parameters were the initial thrust-weight ratio, engine combination, and the two expansion ratios of the dual-position rocket nozzles. The aerodynamic parameters were the ratio of reference area to initial weight and the ratio of maximum allowable normal force to initial weight. A first-order analysis was carried out to determine the effect on the performance of including the engine mass penalty. This analysis indicates that the configuration with the lowest initial mass for a given payload requires all dual-position nozzles with initial expansion ratio of 50 and a final expansion ratio of 150.

A computerized preliminary design procedure has been used to determine rocket-engine bell-nozzle configurations which achieve either minimum gross mass or minimum dry mass on an advanced technology winged single-stage-to-orbit launch vehicle. This procedure evaluates the trade offs between the optimized ascent flight trajectory performance and the flight vehicle sizing driven by the engine mass and propellant requirements. Numerous mission, flight, and vehicle related requirements and constraints are satisfied in this process. Propulsion system configurations which incorporate dual-position nozzles are evaluated and compared to configurations using all fixed nozzles.

Space shuttle trajectory-shaping optimization considering the vehicle's thermal environment and the resulting requirements of the thermal protection system (TPS) are discussed. Optimization studies have been conducted yielding results which are generally applicable to shuttle orbiters and are independent of evolution and redirection of the space shuttle program. Two investigations of optimal trajectory shaping, with different methods of considering the thermal environment are presented.

A pseudo-one-dimensional model of the supersonic combustion ramjet engine cycle is revised on the basis of recent (additional) data obtained from combustor tests. The data were generated in a simple nonreacting system which produces shock structures and shock/boundary layer interactions analogous to those observed at the entrance of supersonic combustors. It is shown that the revised model provides better descriptions of the wall pressure distributions and the overall shock pressure rises for the available test data.