Kirk Sorensen NASA Marshall Space Flight Center The recent announcement of a return to the Moon and a mission to Mars has made the question of human response to lower levels of gravity more important. Recent advances in tether technofogy spurred by NASA‘s research in MXER tethers has fed to a re-examination of the concept of a variabie-gravity research Facility (xGRF) for human research in tow Earth orbit. Breakthrozlghs in simplified inertial hacking have wade it possible to consider eliminating the ‘Wespun‘’ section of previous designs. ‘This, in turn, improves the prospect of a facility based entirely around a tether, with the human module OR one end and a countermass on the other. With such a configuration, propellantless spinup and spindown is also possible based on the conservation of angular momentum from a gravity-grmBent configuration to a spinning configuration. This not only saves Iarge amounts of propellant but vastly simplifies crew and consumable resupply operations, since these can now be done in a microgravity configuration. The importance of the science to be obtained and the performance improvements tn this new design argue strongly for further investigation. Introduction The desire to send humans to explore and settle the Moon and Mars is constzained by a very practical question, currently unanswered: what happens to the bwnan body in gravity levers Less than 1 g? The experience of long- duration exposure to microgravity, on space stations from Skylab to Mir to the International Space Station, has demonstrated that such exposure has undesirable effects on humans. Bone loss, himme system suppression, muscle atrophy, and a host of other medical conditions appear during such flights and make readaptation to Earth life more difficult. To counteract these problems, astronauts currently undertake a strenuous program of exercise while in space, to prepare their bodies for eventual return to the Earth. This exercise regimen has been shown to improve their recovery times after Earth return, but still leaves them typicalIy in an immobilized condition upon landing, although they do recover over time. The concept of using rotation to simulate artificial gravity has been proposed €or many decades. Not only could a properly-rotating space station provide artificial gravity to its crew, but by varying the angufar rate of the &ciiity, it would be possible to simulate different levefs of artificiai gravity, thus allowing the crew io simulate the gravity levels of the Earth, Moon, or Mars, and test themselves and their equipment €or operation in these environments. ZERO G I x X Figure 1: A comparison of artificial gavity and zero-g facilities from an early space station study (“Preliminary Technical Data for Earth Orbiting Space Station”, NASA MSC-EA-R-66-1, November 7, 2966.) Approved for public release; distribution is unlimited. As space stations were first examined seriously in the 1960s, both “zero-g” and artificial gravity concepts were actively pursued, since planners were not sure whether humans could survive for extended periods without the presence of gravitational acceleration. In such trade studies, the operational challenges of the artificial gravity facilities were seen as greater than the “zero-g” facilities; they complicated basic operations such as solar array pointing, communications, rendezvous and docking, ingress and egress, and attitude control. This created a strong incentive for mission planners to assess whether a “zero-g” facility was adequate for their needs, and whether the human organism could be made to adapt to the environment rather than the other way around. Human adaptation to microgravity has been fraught with difficulty and surprises. One of the early surprises was that humans seemed to fare quite well in the microgravity environment-after an early period of “space sickness” most astronauts became quite adapted to microgravity and even enjoyed it, as was shown on the American Skylab space station and the early Russian Salyut stations. But an insidious change was taking place in the astronauts’ bodies that threatened their return to Earth. Their bones, unloaded from the stresses of operating under gravitational loads, were leaching calcium from the bone structure into the blood, where it was excreted. This process of decalcification appeared to continue throughout the duration of the space mission, and leR the astronauts dangerously weak when they returned to Earth, much like elderly persons who suffer from osteoporosis. It was clear that the load-bearing bones need to experience stress during the space mission in order to slow the loss of calcium. Strenuous exercise was proposed as a remediation measure during long space flights, and Russian experiences on the Salyut 6 space station in the late 1970s showed that 2-3 hours a day of strenuous exercise could significantly reduce the effects of decalcification on bone structure. Indeed, it has now become common practice on later space stations, such as Salyut 7, Mir and the International Space Station, for crews to engage in strenuous exercise each day to combat the effects of microgravity. But this countermeasure has a limited effectiveness and comes at a great cost. The exercise regimes have not been shown to stop or reverse the effects on bone structure, only to slow them, and there are other effects to the body from prolonged microgravity exposure, including immune system suppression and alternations in blood flow patterns that are not counteracted by the exercise regime. The great cost of the exercise is in useful crew time, a significant fraction of which is spent m preparation, execution, and cleanup from the exercise regime. As we ponder missions to the Moon and Mars which may last months and years, the cumulative loss of valuable crew time is a significant concern. Consider a crew on the surface of Mars. Were they to engage in the same type of exercise countermeasures used on the International Space Station, they would consume about 2-3 hours per day of waking time in activities related to exercise. If we assume that they are awake 16 hours per day and that roughly 8 of these hours are engaged in meal preparation, hygiene, and crew personal time, we see that only leaves about 5-6 hours per day for scientific observations and extravehicular activities. The “cost” of all this exercise time is not insignificant. Figure 3: Knowing the human response to low gravity could mean the difference. between exploring or exercising! But on Mars, with its gravity 318 that of Earth’s, is it necessary to engage in such a level of exercise? Is the gravitational environment sufficient to produce the needed stresses and eliminate the need for exercise as a countermeasure? And if so, how could we know before we go? Where could we simulate a gravity level similar to Mars and assess human performance? Unfortunately, it is not possible to simulate such a gravitational environment for extended periods of time on the Earth. Brief intervals could be simulated by high-altitude aircraft flying near- parabolic trajectories, and inclined bedrest could simulate (poorly) the loading patterns on the body, but there is no way to simulate Mars gravity for months at a time in a way that specimens could move about and engage in normal activities. In Earth orbit, in a rotating facility, such conditions could be accurately simulated for \ extended periods of time. The cost of doing so would be significantly less than sending crews to Mars and “finding out” what happens to them in that environment. Indeed, we would never threaten the health of a crew on Mars by asking them to refrain from exercise countermeasures to use them as a “control group” to assess response versus a group that exercises. But without such scientific investigation, we will be forced to resort to the application of microgravity countermeasures, neither knowing if they are too much or too little for the “mini-gravity” environment of the I Moon and Mars. Furthermore, essential I I li engineering data on the performance of ’ equipment and life-support systems in lower Gravity level 1/39 1/69 og levels of gravity could be obtained. Figure 4: At the risk of oversimplification, we know that Earth gravity is fine and zero gravity is not, but what about in- between? Previous Design Concepts and Issues for Varbble-Gravity Research Facilities In a sense, we return to the “fork-in-the-road” taken in the mid-l960s, when microgravity and artificial gravity facilities were evaluated one against another. The microgravity space station path has been taken to its logical conclusion for the past 30 years and found wanting-our countermeasures to long-term human microgravity exposure are insufficient to insure the long-term well-being of the crew. Thus let us examine the other, “complicated” option, artificial gravity, and see if we can get it to work acceptably. Inherent in any design for an artificial-gravity facility is the need for a facility to rotate sufficiently to produce the desired acceleration. The acceleration (a) is a function of both the moment arm (r, the distance from the rotational center) and the angular rate (a). With these two variables, one can be traded off against another to achieve the desired acceleration. Obviously, it would be attractive to increase angular rate to decrease the rotational moment arm, but there are real constraints on this-angular rate cannot be made so high that the crew feels significant discomfort or sicleness. Several studies have attempted to quantify the maximum angular rate that can be tolerated by humans over an extended period of time. The results have been inconclusive and have ranged from 2-6 RPM, mainly due to the difficulty of accurately measuring human response to the angular rate. But given the maximum angular rate that can be tolerated and the design level of acceleration, the basic length of the moment arm can be quantified, and that fundamental length constitutes a sort of “yardstick” in the design of artificial gravity facilities. We see in Figure 5 a number of artificial gravity station concepts, each of which is fundamentally sized by the length of the rotational moment arm. Each of these is also an example of a configuration with a fixed moment-of-inertia, which is a basic consideration in determining the spin-up and spin-down of such a facility. The design of the facility is further complicated by all the “irreducible complexity” of a manned space station. Obviously, there must be stability, power, energy storage, and crew exchange capability. There must be the capability to communicate with the ground and/or data relay satellites. All of these tasks are made much more difficult, at least on first examination, by a space station that is rotating. Let us consider power, communications, and attitude control. All of these have the common element of requiring some type of pointing. The power system (ostensibly a solar array) must point toward the Sun. The communications antenna must point toward the &round or to a data relay satellite. The attitude control thrusters must Table 1: The length of the moment point in some inertial orientation to create the necessary torques and translations to arm for a given gravity level stabilize the facility. On a rotating space station, all of these pointing directions are “anyda ra&ng~ujlacr~i rn a tthee r espizrees oenfats rao tating in constant movement relative to the facility itself and require agile tracking facility. systems. Then there is the issue of getting crew on and off the station. If, as it will be shown, there is a substantial penalty €or spinning up and spinning down the facility using thrusters, then the alternative is to c direct the crew and supply transfer to a location on the facility that is either not rotating or rotating very slowly. Thus we see the common feature of a “despun,’ or slowly-spinning hub on many of these rotating station concepts. But the hub must then be connected to the pressurized modules that are under artificial gravity-either by a long pressurized tunnel or by some sort of elevator car. Either option is more complex than simply docking the crew and supply craft directly to the pressurized module-but that requires spin-up with trackable N0 and Spin-down. 0 3 8 There is also the significant question of crew escape. If the crew for 2 some reason needs to leave the facility and return to Earth, on very short notice, how do they do so? Is there an emergency-return vehicle? Where is it located-at the hub? Do they have to “climb the ladder” or ride an elevator to get there? What if one of them is severely injured and requires immediate medical attention? Do they despin the station, and how long will that take? So this is a classic case of “picking your pain” in engineering- which design option entails less cost and risk? Until very recently, it would have been the “classic” design that has been seen in the Spent Energla stages previous pictures. But very recent developments in tether as countermasses technology, energy storage, tracking systems, and manned vehicles make a new and much simpler option finally feasible. Figure 6 An artificial-gravityf acility concept developed at the 1989 International Space University summer session. Analysis of a Simplified Facility In the interest of analyzing some different options, let us examine a facility that is essentially a dumbbell, with a manned module on one end and a countermass on the other. Let us further assume that the two masses are connected by a massless structure such as a truss or tether. Given the design moment am( ro) and the total mass (MJ of the facility (manned module mass plus countemass), it is usefbl to define a countermass fraction (y)-a fraction of the total mass that comprises the countemass. With this definition, the total separation length (8) between the manned module and the countermass can be found as a function of the countermass fraction and the design moment arm alone. m, = (1 - y)M M=m,+m, = yM TIZ, Figure 7: A “dumbbell” model of the facility7c onnected by a massless structure, showing the relationships between the total mass, pressurized module mass, moment arm lengths, and total length. The basic question before us is how should the facility spin up to the desired spin rate? Based on the basic equation ! for angular momentum: c It is clear that if a given value of co (angular rate) is the goal, then there are two ways to achieve it. One is to alter the total angular momentum of the system, leaving the moment-of-inertia fixed, until the desired angular rate is achieved. The other assumes that the initial angular momentum is greater than zero, and proceeds to alter the moment-of-inertia until the desired angular rate is achieved. In simpler terms, the first technique uses rockets to spin up the facility and change the angular momentum, whereas the second retracts tether to change the moment-of- inertia and increase the spin rate, much like a figure skater pulling her arms in during a spin. Figure 8: A recent concept for a Mars exploration vehicle using nuclear electric propulsion and artificial gravity. It has a pressurized module on one end and the nuclear reactor, power conversion system, and radiators on the other, It also has a fixed moment-of-inertia and relies on propulsive spinup to achieve the angular rate necessary to create the desired artificial gravity at the pressurized module. Constant Moment-of-Inertia, Variable Angular Momentum (propulsive maneuvers to change angular rate) Let us first examine the case of the fixed moment-of-inertia and the propulsive spin-up. This technique has the advantage that it does not require any initial rotation rate to begin. But what is the proper length of the tether and should the countermass be large or small? Before examining the equations, simple reasoning says that if the countermass is small, the separation length will be large, and so will the moment-of-inertia. But if the spin-up thrusters are located on the countermass, then they will have a long moment arm with which to change the angular momentum. So which effect will dominate? Let us solve for the moment-of-inertia of the system in terms of the pressurized module mass (m), the design moment arm (r,,), and the countermass fraction (y). I=moro2 +mlq2 =mO52 +mo (3) Y As expected, the moment-of-inertia is inversely proportional to countermass fraction, so as the countermass gets smaller (which is desirable) the separation length increases and the moment-of-inertia increases. Now let us define the angular impulse (J) as the integral of propulsive force over time multiplied by the length of the moment arm. The angular impulse is also the change in angular momentum and, for a facility with a constant moment-of-inertia, proportional to the change in angular rate. J = r IFdt = IAo = 1v-I (4) If we assume that the spin thrusters are mounted on the countermass so as to maximize their effectiveness, we can substitute the definition of the countermass moment arm (rl) into the equation and simplify. y ) q = ro( If we further assume that we begin with an angular rate of zero, then the design angular rate (a) is essentially the same as the change in angular rate (Am); this allows fwrther simplifications in the equation by substituting the definition of the design moment arm (ro = ah2). Therefore, if we assume that the mass of the pressurized module, the desired acceleration, and the design angular rate are given, the only variable we can use to minimize the propulsive impulse required to spin-up the facility is the countermass fraction, which this equation indicates should be minimized. This is a somewhat surprising result, as it might have been expected that the moment-of-inertia, which was quadmtic with length, would dominate over the spin-up torque, which was only linear with length. Nevertheless, any facility that uses spin-up thrusters must have a "smart" countermass and deal with the issues of resupply of propellants to the countermass and active control. Constant Angular Momentum, Variable Moment-of-Inertia (tether reeling to change angular rate) Let us examine the other case, altering moment-of-inertia and conserving angular momentum to achieve the desired spin rate. Let us again r e mt o the equation for moment-of-inertia, this time solving it in terms of total mass and total separation distance. This form of the moment-of-inertia equation can be shown to be identical to the previous derivation, in the limiting case of a separation distance required for the design moment ann and countermass fraction. But the real utility of this new form of the equation is in cases where the separation distance is varied to achieve a variable moment-of- inertia. This equation also shows the quadratic dependence of moment-of-inertia on total tether length. If we assume that the facility is initially oriented along the radial direction in its orbit around the Earth (gravity-gradient configuration) then it will have some rotation rate commensurate with that orbital period, and hence, some initial value of angular momentum. Let us assume that the subscript 0 refers to the fully-extended length and the subscript 1 refers to the length in the retracted, design condition. Assuming that angular momentum is conserved during tether reeling (which can be achieved through proper timing of tether extension or retraction) then the ratio between initial angular rate and final angular rate can be shown to be a ratio of initial and frnal length. The subscripts 0 and 1 again correspond to the extended and retracted configurations. Hence, for a desired angular rate, this equation can be used to assess the amount of tether that must be retracted to achieve that rate for a given initial anguIar rate. The angular rate ratio will then be the square of the retraction ratio (initial lengthhetracted length). Thus, within these simplifying assumptions, the spinup of a rotating facility is independent of its totaf mass and its mass distribution, but only depends on the retraction ratio. I. When the tether is fully extended and the facility is in a stable, radial orientation, begin 2, As the tether retracts, retraction of the tether. angular momentum is consented and the facility begins to rotate. 3. Retraction continues until the facility is spinning at the proper rate to yield the desired artificial gravity levels. Figure 9: The facility can be spun-up through a tether retraction technique. 3. When the tether is fully extended the facility is stabilized in a gravity- 2. The tether is extended gradient configuration. and angular momentum is conserved-rotation rate slows quickly. 1. To stabilize the facility far docking and other "zero-g" activities, all the operations proceed in reverse. Figure 10: The facility can be de-spun in a similar yet reverse manner. The significance of this equation is in its independence on total mass and mass distribution. This means that no matter what the total mass of the facility, no matter what the countermass fraction, no matter what the final length, this spinup technique could be used. This gives the designer an enormous amount of additional flexibility in the design of the facility, since now all of the “smart” mass (pressurized module, power system, tether reel and deployer) can be located at one end of the facility and the countermass can be “dumb” mass, such as the spent upper stage of the rocket that launched the facility initially. Design Concept and Concept of Operations of the simplified xGRF It is clear that the resources to pursue a variable-gravity research facility in space will not be available unless the costs of such a facility are reduced dramatically. And in order to reduce costs, the mass and complexity of the facility must be reduced dramatically. A tentative goal for this study was to reduce the mass of the xGRF to the level where the facility might be launched on a single flight of a large rocket such as the Delta IV Heavy, which can carry a 24 metric tonne payload to low Earth orbit. Of course, it is not sufficient to merely hope for cost and mass reductions-there must be real design innovations that “change the equation”, so to speak, and permit a design solution that is actually simpler and cheaper. The main advantages to pursue in this design for mass and cost reduction are 0 Simplified spin control through tether reeling (variable moment-of-inertia) The elimination of any despun elements through the spin control technique and the use of unique pointing mechanisms (Canfield joints). 0 Simplification of crew exchange, consumable resupply, and emergency escape, through the use of standard “zero-g” docking made possible through the ease of spinup and spindown. The design concept for consideration is a facility that is much like a simple dumbbell connected by a variable-length tether. On one end of the facility would be the pressurized module where the crew would live and conduct their experiments. For mass savings, this module would probably be an inflatable pressurized module based on the “Transhab” technologies developed at Johnson Space Center in the late 1990s and now being pursued by Bigelow Aerospace. At the one end of the pressurized module would be a docking port for an unmanned logistics module or a manned capsule such as the proposed Crew Exploration Vehicle. The docking port would be designed for low-gravity operations but would include “hard-docking” mechanisms that could firmly attach the crew transfer vehicle to the facility while it is spinning (during artificial-gravity operations). This continuous connection to the facility allows the crew transfer vehicle to serve as an “escape module” for the crew in the event that they needed to leave the xGRF and return to Earth rapidly. In order to pursue this simple mode of rendezvous and docking operations,,it is necessary to have a mechanism for spinning up and spinning down the facility rather quickly. A spinupispindown technique based on tether reeling has been proposed which accomplishes this need without the consumption of propellant. Furthermore, this technique can reduce spin rate very rapidIy in the event an evacuation is required as well as generate emergency power, if needed, during such an evacuation. While angular momentum is conserved, rotational energy is decreased, and electrical power would be generated by the reeling motor as the tether is reeled out. To reel the tether in, rotational energy must be added, and this energy (to drive the reel) would be provided by flywheel energy storage systems which would be charged by solar panels. Another problem with such a simple facility that has been raised in past studies is the difficulty of tracking inertial and pseudo-inertial targets from a rotating facility. To generate power, solar arrays are required to track the Sun. Communications probably must track a pseudo-inertia1 target such a TDRS satellite in geosynchronous orbit. And thrusters would need to be able to point at inertial thrust vectors during rotating operation. Figure 11: xGRF concept Solar arrays, communications systems, and thrusters would all need to transfer power, data, and fluids across rotating joints or slip rings if they were to be mounted near the pressurized module and yet fi.~lfiltIh eir duties. These fluid, data, and power transfers could become very unpleasant development problems. Fortunately, several recent innovations make such operation possible without any sort of rotating connections. The key is a unique mechanical joint developed by Dr. Stephen Canfreld of Tennessee Technological University. The “Trio Tri-Star Carpal Wrist Robotic Joint”, mercifidly referred to simply as the Canfield joint, is a “parallel” mechanism that allows a structure to point at an inertial target while being mounted to a rotating platform. Fluid, power, or data connections can be made from one side of the joint to the other because the two sides of the joint do not differentially rotate, even though the joint itself may be bent 90 degrees or more. The kinematic behavior of the Canfield joint is extremely difficult to describe in words, and not much easier to discern in still pictures, but suffice it to say, seeing is believing-the joint is truly capable of smooth pointing and orientation within a hemisphere. Figure 12: A series of still pictures showing a typical reaction-control thruster mounted on a Canfeld joint, capable of pointing any direction within a hemisphere. The parallel structure of the joint not only gives it tremendous strength, but three degrees of freedom-altitude, azimuth, and “plunge7’. The first two degrees are sufficient to allow it to point to any direction within a hemispheric workplane. And unlike other structures, the joint can access this entire workspace without encountering mathematical “singularities” that make solution of the base angles impossible. The joint is driven by three actuators on each of the three base legs. Any commanded direction corresponds to a unique solution of the three angles between the base and the base legs. Figure 13: A unique arrangement of solar arrays mounted on Canfield joints enable constant solar tracking at any beta angle on a rotating facility with a minimum of bending loads across the array. As shown in figures 10 and 11, liberal use of Canfield joints on the xGW will enable solar tracking, high-data-rate communications, and even active propulsive stabilization to take place during the rotation of the facility. This, combined with the ease of spin rate change made possible through tether reeling, enables the deletion of the despun