An illustration of an artifical gravity test involving a Crew Dragon spacecraft and Falcon 9 second stage.

How to test artificial gravity

by Joe Carroll
Monday, December 9, 2024

But first: Why does artificial gravity (“AG”) matter?

Most people interested in human expansion beyond Earth suggest settling the Moon or Mars first. But we don’t know how much gravity we will need, to avoid the novel health issues we keep finding after long stays in microgravity.[1-7] Many of those problems were found only after 6- to 12-month stays in microgravity, so it seems prudent to test human health at both gravity levels, possibly eventually in multi-year tests.

The five other large moons in our solar system all have about our Moon’s gravity (0.13 to 0.18g, versus 0.16g). And Mars and Mercury have the same gravity, 0.38g. The only other bodies between 0.09g and 2.3g all are within 12% of 1g: Venus, Saturn, Uranus, and Neptune. All four are all hard to live on or return from. So, our solar system offers only lunar and Martian gravity (and with difficulty, terrestrial gravity) for settling.

We know that humans thrive in 1g but not in microgravity, but we know nothing about health in between. In 2021, I suggested how to clarify human futures beyond Earth.[22] We can determine our “gravity prescription” for good health by having people live in a long dumbbell-shaped station in LEO.[20] It can offer lunar and Martian gravity at opposite ends, while using a slow enough spin to avoid sustained negative sensory effects of spin. This can tell us whether and how we might handle long-term living at Moon and Mars gravity, at far lower cost and risk than by actually living on those bodies. And if we do end up needing more than Mars gravity, we must live in AG, in free space.[19] AG tests then become even more relevant. Besides better health, AG offers other benefits to space habitats. Even Moon-level AG may improve design, operations, life support, training, and many other aspects of space facilities.[6]

For any given gravity level in AG, spin rate drives facility radius and cost. Most advocates of AG suggest the highest spin rates they feel reasonable, to minimize the facility radius.[13-16] They usually base this on the responses of people in rotating-room tests. But all such tests are of uncertain relevance, as discussed below.

Sensory effects in rotating rooms vs AG

Early in the space age, Graybiel et al. tested six young soldiers in a rotating room over two days.[8] After that, the rotation stopped and they returned to the room for tests on a third day. The room was round, windowless, and 15 feet [4.6 meters] in diameter, and smoothly rotated about a vertical axis, at rates from 1.71 to 10 rpm. Some subjects reported an initial mild general malaise, nausea, or headaches even at 1.71 rpm, and one vomited at 2.21 rpm. The frequency and severity of reported negative reactions increased at 3.82 and 5.44 rpm. But eventually most of the subjects did fully adapt, even to spins up to 5.44 rpm. One subject had minimal negative reactions even at 10 rpm. And a control subject who lacked normal vestibular function had little response to rotation. Some later studies tested longer adaptation periods, including up to two weeks of sustained rotation.[9]

Some later AG proposals added other constraints.[10-11] One is setting a maximum allowed relative change in gravity from head to foot when standing. This directly sets a minimum spin radius. Another is constraining the change in relative weight when walking with or against the spin. This in turn sets the minimum ratio of tangential spin velocity to assumed walking speed at that radius.

John Charles, then chief scientist of NASA’s Human Research Program, told me that ground-based vertical-axis rotating room tests may not be relevant for estimating the response to the spin of orbiting AG facilities. He said that in rotating rooms, the spin axis aligns with gravity, but in AG, they are at a right angle. A footnote in the Graybiel paper also notes this.[8] I thought this was a minor subtlety at first, but eventually realized that this difference causes completely different sensory effects from vertical-axis rotating rooms. This is worth discussing in detail.

The sensed effects of spin in different spin axes

In a vertical-axis rotating room, standing everywhere but the spin axis requires you to tilt. But each point has a fixed tilt and direction. Horizontal motion adds a horizontal Coriolis force at a right angle to your motion. Facing in any direction, you feel the same amount and direction of motion-induced force in your body coordinates. Walking or swinging your arms causes aiming errors initially. But this effect is consistent, so people adapt to it after several iterations. But if the rotation stops or reverses, people retain their recent spin adaptation. The result is an opposite aiming error at first, as seen the video for Ref. 12.

Compare this to AG. Coriolis effects still make all motion in the spin plane cause a force in the spin plane, at a right angle to the motion. But in AG, that plane is vertical. Horizontal motions cause vertical forces, and vice versa. And unlike rotating rooms where effects are fixed unless you change the spin, the felt force in AG reverses each time you turn around. Adapting to such changes may take far longer, and we may never fully adapt to them. Trying to simulate this in a rotating room requires either smoothly reversing the spin, or lying down, moving your limbs, rolling over, and then moving them again.

You are also directly sensitive to spin even if you don’t move. Room rotation rates down to about 0.5 rpm may be felt if people stay still, but they tolerate it. They also adapt to changes in rate and even direction but that takes time. But in AG, the direction of felt rotation depends on which way you face. Turning around reverses the direction felt. Another turn flips it back. As with ground tests of Coriolis forces, the analogs are reversing the spin, or lying down and rolling over.

But there is one sensory effect of AG that we can test fairly well on the ground. It is a significant weight change that occurs when you walk with or against the spin direction in AG. I discuss this next.

How to test AG weight changes caused by walking

People feel an approximately 5% transient change in weight when an elevator starts or stops. We often stumble a bit if we are moving around then. In AG, walking either with or against the spin changes your spin rate and hence your centrifugal force and weight. If you walk with or against the AG spin at 1 meter per second, you feel a weight change of just 2.1% of Earth gravity per rpm. But at 10 rpm, that is a weight change of 21% of Earth gravity, or 55% of felt gravity in Mars AG.

A Vertical Motion Simulator (VMS) at NASA Ames can provide these changes in felt gravity. The VMS allows a vertical stroke up to 18 meters. This is more than enough to provide the same weight changes as AG, when people walk around. The VMS has fove interchangeable cabins. The largest one has a 1.8 x 3.7 meter floor.

Consider two people each walking around a 1.2 x 3.0 meter elliptical path, as illustrated above. Moving spots of light on the floor or wall can pace them to allow smooth sinusoidal motion in the long axis. If each cycle takes 8.5 seconds, peak walking speed is 1.0 meters per second. If the cabin oscillates up and down with strokes of 0.76 to 7.6 meters, it causes the same Earth weight changes as walking the same path in AG at spin rates of 1 to 10 rpm.

This test is open-loop, with walking to fit cab oscillation. At higher cost, one can move the cab in response to sensed one-axis motion of one test subject. But open-loop tests are easier, can test two people at a time, and may be enough to limit the acceptable AG spin rates based on this constraint. The VMS can test many people to quantify their sensitivities, adaptation times, and any residual sensitivity. I don’t know a good way to test other felt AG effects other than in orbit, by actual AG spin tests somewhat like the test done on Gemini 11.

Gemini 11: a precursor tethered spin test in LEO

The Gemini spacecraft program actually began after Apollo, to answer time-critical questions that Mercury could not test. Early missions answered most of them, so NASA added new tests to the last few missions. This included passive tethered stationkeeping, using slow spin on Gemini 11, and gravity gradients on Gemini 12.[17]

The Agena’s docking collar on Gemini 11 held a 30-meter flat coil of seatbelt-like tape. During an EVA, the crew attached its free end to Gemini’s docking bar. Later they released Agena and used Gemini’s thrusters to move away and pull out the tape. Then they used other thrusters to spin up to 0.11 rpm. Later they went to 0.15 rpm. The crew did not feel the less than 0.0004g accelerations. The tape tension also caused Gemini to slowly rock back and forth. The crew found the test uneventful enough that they ate lunch during the test. After three hours they released the docking bar, and the Agena and tape drifted away.

The Gemini 11 mission movie is on YouTube.[18] The tether test starts at 10:50. The narrator said it was a test of passive stationkeeping. The Gemini 7 crew had already spent two weeks in microgravity, longer than any planned Apollo mission, so AG itself was of little interest.

Possible AG spin tests on Crew Dragon

Using spinning more than ten times faster, Crew Dragon allows similar tests, but at lunar and Martian gravity. The tests can be simpler than on Gemini, if they use the Falcon second stage as counterweight. This avoids any need for rendezvous, docking, or spacewalks. Any mission with enough free time and RCS margin can do these tests. One might do them on flights for Axiom, Polaris, Vast, Fram2, NASA, and future customers. Releasing data on crew responses to spin could greatly speed AG development.

One can use a “nose-up” Dragon gravity alignment, as before launch. But that requires attaching a strong tape or bridle to Dragon’s nose, before or after launch. Either option seems far more challenging than attaching a bridle inside the trunk, before launch. I suggest starting with nose-down AG spin tests. If there is enough interest in nose-up tests, those details can be worked out then.

Crew Dragon spin tests can focus primarily on finding the spin rate at which unpleasant effects start to occur, and how fast and complete any adaptation is to that spin. During a spin test, the crew can stand, sit, lie down, turn around, look out at Earth, and try other activities while facing in different directions. There is little room to walk, but the VMS can test the effects of walking in AG.

The most important goal from Crew Dragon AG spin tests is to find suitable spin rates and radii for early crewed AG facilities. Another useful goal is to learn whether time and/or activity in low AG either eases or hinders any later adaptation to microgravity. If it eases adaptation, AG spin ops might become regular parts of future crew flights. Another goal might be quantifying correlations between crew responses to ground tests and spin in AG. Any such correlations could let tourists know how well they are likely to handle spin in AG, before they pay for a flight. This should be invaluable for large-scale AG tourism.

AG spin tests of Dragon with attached Falcon stage

The simplest spin test just spins Dragon. But retaining the Falcon stage shifts the center of mass about three meters aft. This allows a longer crew spin radius and is worth doing, before doing later tests with Falcon on a long tape or tether.

Spin will pull the crew away from their seats. Then can unbuckle either before or after the spin starts, and can sit or stand on the Dragon “ceiling.” It will serve as their floor during the spin test. The crew should have roughly a four-meter average spin radius both in their seats, or standing on the ceiling.

Each test can start at the usual 200 x 200 kilometer Crew Dragon insertion orbit or any other low orbit. Normally, Dragon and Falcon separate then. But here Dragon starts to spin, with the Falcon stage still attached. The spin axis should point near the sun. Then Dragon’s solar arrays can provide full power, and the radiators can avoid the sun. This allows normal power and thermal performance.

Long pauses during spin-up let the crew try various activities and describe their reactions. When unpleasant effects start to occur, they can maintain that spin rate to see how fast and how thoroughly they adapt to it. If they adapt well enough, they may increase the spin, but smaller steps and with longer adaptation times at each step.

But at any time, the crew can end the spin test by releasing Falcon and despinning Dragon, in either order. We don’t know how fast a spin the crew may be willing to test. But being active in a spin of about 3 rpm that gives only 0.04g may be useful, if it finds sustained low-level AG helps crews adapt to microgravity. It can also start to correlate both similarities and differences in crew responses to low-level AG vs 1g rotating-room tests. Spin tests might be repeated on later missions, using chairs, hammocks, and other props to test a wider range of situations. But even without such props, testing another crew or a new spin-up sequence will be useful, and has very low added cost.

AG spin tests with an inflatable room

Adding a “popout” inflatable room to a Crew Dragon can give the crew much more usable pressurized volume, for both microgravity and AG spin tests. There is limited room inside the nose cap to stow it. But it might expand to as much as 2.2 x 3.6 meters (15 cubic meters), as shown at right. The roughly three meters it can add to the crew spin radius may be most useful in tests with an attached Falcon. The added room may also enable more activities in tethered tests.

Parts of the normal docking interface can securely hold and then reliably release this hardware, to allow later docking and nose cap closure for reentry.

AG Dragon spin rate tests using a tethered Falcon

Spin tests with an attached Falcon stage limit the crew spin radius to about four meters, or seven meters with an inflatable. This requires a fast spin to get much gravity. The table below shows spin rates and velocities using both attached and tethered Falcons. Longer tethers allow slower spin at the same gravity. A tether system can be lighter, simpler, and more reliable if the full length deploys at low tension. Then different spin rates during one flight test will all occur on one row of this table:

To generate this table, I assumed an average crew center of mass during spin tests one meter forward of Dragon’s center of mass, and a Dragon mass 3.5 times that of the Falcon stage, after Falcon vents its remaining propellant.

Tethering Falcon to Dragon causes new complications but allows much lower spin rates at each gravity level. This lets us quantify how much the AG level affects the crew’s tolerance of spin. RCS margins or safety may constrain the spin-up velocity. I quantify them later.

As in AG spin tests with an attached Falcon, “nose-down” gravity is easier to do. Releasing Falcon and thrusting away can pull a tape from the trunk. I suggest a 36-meter crew spin radius in the first test. Then 0.04g, Moon, Mars, and Earth-level AG need spins of 1, 2, 3, and 5 rpm. Later tests can use other combinations of length and spin.

A scenario for tethered AG crew spin tests

As shown below, Dragon might pitch up about 60 degrees, release Falcon, and fire four aft Draco thrusters for about two seconds to move away at about 0.2 meters per second. Payout of a roughly 150-meter tape takes about 12 minutes. Near the end, the tape pulls a bridle from the Dragon trunk, and Draco thrusters start the spin-up. The best thrusters for spin-up also cause inward thrust. That brakes the deployment, but then complicates the spin-up a bit.

Initially, Dragon can tilt to null out the inward part of the spin-up thrust. But as tension rises, the bridle will orient Dragon radially. Pulsing the spin-up thrust can keep the average inward thrust below the AG force. Once the spin exceeds about 0.3 rpm, the AG force exceeds even a sustained spin-up thrust. But using well-timed pulses can also damp attitude oscillations, like those seen in the Gemini 11 movie.[18] And good timing of the pulses can also provide useful orbit boosting above the low SECO orbit.

With a 36-meter crew spin radius, each 1 rpm spin change requires 16.5 meters per seconds of relative tangential ΔV, from either Dragon or Falcon. Dragon has 3.5 times Falcon’s mass but its Dracos have five times higher specific impulse than Falcon’s cold gas thrusters. Here I assume use of Dragon thrusters for both separation and spin-up.

To use the four axial Draco RCS thrusters mounted there, Dragon must open its nose cap. But spin tests just use the 12 aft Dracos, so Dragon can keep its nose cap closed until after the tape is released.

I assume that with cosine losses, the two best-oriented 400-newton Dracos give 320 newtons of useful transverse spin-up thrust each. Then each 1 rpm of the spin-up takes six minutes of firing. Thrusting in only part of each rotation, and just in the best parts of each orbit, allows effective spin up as well as orbit boost. Long pauses after some pulses allow more time for crew to adapt to the higher spin rate and do more tests. If second stage engine cutoff is at 200 x 200 kilometers, a 3 rpm spin-up can boost the orbit to 260 x 260 kilometers. If all pulses are done near one location in the orbit rather than two, spin-up can boost the orbit to 210 x 310 kilometers. That may be useful.

Dragon ends the test by releasing the bridle. If release is at apogee of a 210 x 310 kilometer orbit, with Falcon moving aft then, release can boost Dragon to 248 x 310 kilometer, while slinging Falcon into a roughly 85 x 310 kilometer targeted reentry trajectory. A lower SECO could let Falcon drop even lower. Falcon’s 3 rpm spin should help reduce reentry dispersions.

Spin tests docked with another module

Another spin test option is to first dock with another previously-launched module, like Haven-1. This allows even more usable space than an inflatable like that shown earlier. But it also doubles the spin test mass. For the same AG level and spin rate with Dragon’s Falcon stage as the counterweight, we need about 1.8 times the tape length and 3.6 times as much Dragon RCS propellant. But each flight provides a new Falcon stage and tether. This lets the tape length and strength be customized for each test.

Safety issues

A critical safety issue is to ensure that a Dragon launch abort can occur without a strong tape tying Dragon to Falcon. I suggest stowing the tape and bridle in the Dragon trunk, with the free end only weakly held by Falcon. After SECO, an actuator on Falcon can securely clamp that end.

A second issue is accidental tape severance under load, which causes fast recoil and may cause the tape to foul on part of Dragon. In the Gemini 11 mission movie, the crew said that after they released the tape, it slowly wrapped and unwrapped around Agena. Here a tape cut at far higher tension will wrap the tape around Dragon far faster. If the tape snags then, it could keep Dragon from ejecting its trunk, until RCS thrust or initial reentry heating can melt the tape. To prevent snags, tape should probably be smooth, slick, and stiff. Also, Dragon’s nose cap should stay closed during the test, and any other exposed Dragon features that could cause a snag should be taped over or otherwise modified.

Note that adding an inflatable volume or docking to another module require that the nose cap be open during the test. Both of these cases also provide additional places for a tape to foul on.

A tape wrapping around Dragon wraps in a direction is opposite to that used in yo-yo de-spin maneuvers. But the continued Dragon spin eventually gives a useful centrifugal peel force, as on Gemini’s Agena. And Dragon’s RCS can spin Dragon far faster when Falcon is not attached. The RCS might also create a wobble, or even melt the tape. Using too much RCS may preclude later parts of a mission, but that may be found acceptable if it lets a crew return safely, after an extremely unlikely but not provably impossible sequence of events.

High drag in AG test orbits of 200–260 kilometers greatly shorten debris transit times and populations of untracked orbital debris. This greatly reduces the chances of a tape cut, or a leak in an inflatable that is too large for the crew to immediately patch.

Multi-strand tethers can preclude severance, but only if enough strands are kept apart far enough to preclude a full cut by a several-centimeter untracked object. And if a multi-strand tether is cut, it is likely to offer far more opportunities for snagging.

Severance or inadvertent release of a tape when Falcon is rotating forward could also sling the Falcon and its tape high enough to reach ISS. A Mars-gravity 3-rpm spin could boost Falcon apogee by up to 135 kilometers.[23] A worst-case cut at 260 kilometers could sling Falcon to 395 kilometers. And release at perigee of an eccentric orbit may let it reach much higher. But a suitable spin axis tilt may be able to preclude contact.

Crew incapacitation by spin can be limited by increasing spin in small steps, and waiting until there is enough adaptation before further spin increases. Other safety issues and complications will come up during development, and may require changes or constraints in the scenario above.

Reasons for selecting other key features

Sustained tests at both Moon and Mars gravity are useful because Moon and Mars gravities are the two most relevant non-Earth gravity levels in this solar system. Later flights with different tape lengths can quantify sensitivities to spin rate vs AG level, scheduling of spin up, and other relevant factors.

These spin tests use the spent Falcon stage as the counterweight, so they must be done early in a mission. Spin tests after Falcon departs put the crew only about one meter forward of the Dragon center of gravity. That seems of little value for crew tests. But that may allow useful low-gravity AG lab or fab tests on Cargo Dragon flights.

It would be more useful to attach a bridle to Dragon’s nose rather than its trunk. Then the gravity direction fits the crew seats and cabin layout better. But it seems far easier to stow, deploy, and release a tape from the trunk than to attach a bridle to Dragon’s nose in orbit and later reliably release it. How to do “nose-up” tests can be worked out later if there is enough interest in that orientation.

If desired, Falcon ullage gas and liquid oxygen boiloff can allow some of the spin-up while the Falcon’s avionics batteries stay alive.

For the best Dragon solar array and radiator performance, the spin axis should aim near the Sun. Then Dragon’s solar arrays can face the Sun and the radiators can avoid it. But this does assume that when the RCS tanks are nearly full, Dragon’s stable flat-spin axis puts the solar arrays and radiators on opposite sides of the spin plane, not straddling it. If not, then we may have to rearrange cargo, add a reaction wheel, actively maintain an unstable balance, or limit spin tests to a few hours.

Spin-up thrust can provide efficient orbit boosting during part of each spin in any spin axis, at least near one or two points in each orbit. And as noted earlier, slinging Falcon aft from apogee of an eccentric orbit can cause a targeted reentry, somewhat like the SEDS-1 targeted reentry tether test.[24] One might even disperse small reentry test payloads from the spinning Falcon, before it reenters.

To do spin tests while docked with a module like Haven-1, I suggest discarding that module’s Falcon stage as usual, and using each visiting Dragon’s Falcon stage as counterweight. This allows a different tether length, strength, and RCS test budget for each flight. But using the Dragon’s Falcon stage as counterweight does require that Dragon approach and dock with the module with that Falcon stage still attached. It may be best to vent any residual fuel before that, to prevent complications from sloshing.

Crew Dragon customers may want to add AG spin tests to many flights. They may want to test different spin rates and schedules, activities, and crews. I don’t know how much each test will cost, but I think it is far better to invest in AG spin tests before we try to do any detailed designs of crewed AG stations. Above I baselined a 36-meter crew spin radius. Tests in the VMS and Dragon can guide us to better tests for later AG spin tests, and eventually to a good radius and design for the first long-term crewed AG station.

Possible payload mass penalties from AG spin tests

The main payload mass penalty in tethered spin tests will usually be from RCS use. Consider a 3-rpm spin at Mars AG, with 3.5:1 mass ratio. It needs about 0.2 meters per second for Dragon to separate, plus 49.6 meters per second tangential ΔV to spin up, each at assumed 0.8 cosines for the best-oriented thrusters. At 300 seconds of specific impulse, this takes 296 kilograms of Draco RCS propellant. If the pulses occur when the thrusters aim furthest aft, the boost impulse may be about 90% of the spin impulse. This can boost the average altitude of the Dragon and Falcon assembly by 60 kilometers. The boost can be circular or eccentric, as desired. If Dragon releases the bridle at apogee, when Falcon spins aft, it can boost Dragon from 210 x 310 kilometers up to 248 x 310 kilometers.

To compare, using the axial thrusters with no cosine loss to boost just Dragon from 200 x 200 kilometers to 248 x 310 kilometers takes 220 kilograms of RCS. So added RCS use for this case is 296-220 kilograms or only 76 kilograms, but only if such an orbit boost is needed. For spin-up with Falcon attached (and spin-down after release), net RCS per rpm is smaller.

There is also a small RCS cost for reboost due to a large drag area and low average altitude. The full spinning assembly may have an average coefficient of aerodynamic drag of about 160 square meters, versus 40 square meters for Dragon itself flying end-on. The added drag at a 230-kilometer average spin test altitude should cost less than one kilogram per hour of added RCS.

There are other payload mass penalties. Spectra and Dyneema have the highest strength/weight of any commercial fiber, and were used in most successful tests of long space tethers.[24-27] A very low ratio of absorptivity to emissivity keeps them very cold. This further increases strength and toughness and greatly reduces creep. But they must avoid RCS plumes, unless we actually want to blow away or melt a tether. With a 14-tonne Dragon and 3.70 meters per second squared Martian gravity, tension is 52 kilonewtons. With SF=6, the mass of a 150-meter Spectra or Dyneema tape is less than 30 kilograms. If sized just for Moon gravity, it is less than 13 kilograms. Support hardware may add 10 to 20 kilograms. Tests using an inflatable volume and its attach interface will require extra hardware.

There is one more mass penalty: added items needed to make low-gravity spin tests useful. Food, water, and air needs will not change if a test ends during planned phasing to reach ISS. But the crew will need some hardware for useful spin tests, and added supplies on non-ISS missions. AG spin tests may also lead to later tests that need more supplies, test props, and perhaps some AG lab experiments.

RCS and tether sensitivities for other spins and lengths

With a fixed tether length, RCS mass penalties and altitude boosting both scale with rpm, but the tether loads and mass scale with centrifugal force and hence rpm squared. But if you compare different spin rates and lengths with the same gravity, tether mass scales with roughly 1/rpm², and spin-up ΔV scales with 1/rpm. If RCS use or worst-case Falcon altitude are too high for a desired case, one can reduce RCS needs and test altitude by using a shorter tether and faster spin for the same maximum gravity. One can reduce RCS use by retracting tether after spin-up, but at higher hardware cost and failure modes.

AG spin tests with Dragon docked to another module will be dedicated missions. If that module has the same mass as Dragon, tape length grows by 1.8 times, and tape mass and RCS use both grow by 3.6 times. And RCS use won’t displace any later boost, except for any module re-boost. Such AG tests will have higher costs, unless Draco thrusters are added to Falcon to allow spin-up from the lighter end.

What gravity levels should early AG space stations use?

The most important sustained early AG tests should focus on the health of humans in Moon and Mars-level AG. Such a facility can also do other biological tests at those gravity levels, including animal gestation.[21] It also seems prudent to learn the effects of lunar and Martian gravity on crop-based ecosystems, in AG in LEO, before we try to settle the Moon or Mars. And we can get far better test results, with fewer complications, if the tests don’t also include high GCR doses, abrasive regolith, or toxic dust.

I have not found clear early payoffs for AG levels between Moon and Mars gravity, or between Mars and Earth gravity. But if AG tests at Mars gravity show that we really need higher average gravity, then living at least most of the time in AG at 0.5 to 1g will have to be the focus of any realistic plans for human expansion beyond Earth.

Much lower gravity levels may have some value, for lab & fab more than crew. The figure above[23] suggests that many processes have different minimum and maximum gravity tolerances, with some overlap. Slow-spinning crew-tended AG can offer low enough gravity near the spin axis for some “microgravity” processes, while easily settling stored fluids in nearby tanks. Useful testing of low-g AG processes might be done on cargo Dragons, using payloads located both along and near the spin axis.

Implications of AG spin tests

As discussed earlier, there are large differences between the felt effects of vertical-axis rotating room tests and AG. Those differences could plausibly allow higher rpm in AG than in rotating rooms. But AG will probably require slower spins, because each time you turn around, spin sensory effects stay the same in rotating rooms, but reverse in AG. That should affect adaptation time and completeness.

But we don’t know! This is a critical uncertainty preventing prudent investment in commercial AG stations. In addition, the thresholds and severities of reactions to AG spin may vary with the individual, spin rate, gravity level, facility architecture, type of activity and motion, adaptation time, alertness, and motivation. We may need only typical responses to make a good business decision, but we will need to know the sensitivity to many of those details to design a competitive facility.

Weight-change tests on the VMS plus AG spin-rate tests on Crew Dragon can enable realistic AG designs, and perhaps even predict responses of individuals, if we can correlate AG spin test data with ground test data. AG with “slow enough” spin rates is also useful for all long crew trips, as well as LEO resorts and spinning free-space settlements and colonies. Long crew stays in Moon and Mars-level AG will do much to clarify health constraints on human future expansion beyond Earth.[22]

To experience sustained free fall, you must be in space. Sustained lunar and Martian gravity are also available only beyond Earth, either on those bodies or in AG. To date only a dozen astronauts have experienced lunar gravity, and then for only one to three days each. And nobody has yet experienced sustained Martian gravity, real or in AG. Parabolic flights can give 30 seconds of microgravity for people floating inside an aircraft, even if there is some turbulence. But parabolic arcs providing lunar or Martian gravity give bumpier rides.

AG can offer both free fall and sustained lunar and Martian AG. Nearly all tourists in an “AG resort” will try free fall, but most may spend far more time in lunar and Martian gravity, which are even more novel, and easier to adapt to. In contrast, both real gravity and free fall offer only one gravity level, not three. But AG tourists will want any spin adaptation to be brief. They will also want a slow spin (i.e., the longest affordable radius.) This can reduce all spin sensory artifacts. There will only be AG tourists if they really enjoy their stay in AG, and don’t just tolerate it. If we can correlate ground test responses with those in AG spin, this could be invaluable for both prospective tourists and resort operators.

Questions for AG spin tests to resolve

1. What is the lowest AG level that allows early activity without the adaptation problems seen in microgravity?
2. Does sustained slow-spinning low-gravity AG either help or hinder a later adaptation to microgravity?
3. What AG spin rates trigger negative reactions, and how fast and how much do we adapt to them?
4. How much does the AG gravity level affect the spin rate thresholds that trigger negative reactions?
5. What are the lowest AG levels for easy walking, sitting in a chair, drinking or eating at a table, etc.?
6. Do people really enjoy Moon and Mars-level AG, if the spin is slow enough to avoid problems?
7. What AG spin rate and maximum gravity might make sense on Havens and larger AG stations?
8. What AG levels and spin rates may be best for lab and fab processes, space resorts, or settlements?
9. Will most tourists adapt to an AG resort’s spin at arrival, or need to stay at a lower spin first?

Conclusion and recommendations

Long crew stays on ISS keep uncovering new health problems in microgravity. Decades of microgravity countermeasure refinements have helped, but don’t fully stop human health degradation.[1-7] In this solar system, the only gravity levels available between 0.1 and 2 g are Moon and Mars levels (six and two bodies, respectively) and Earth (five bodies, with all but Earth being very challenging to settle.)

But we don’t know whether lunar or Martian gravity allow sustained good human health, or what countermeasures may still be needed, or how well they may work, or with what side effects. Testing of sustained health at both lunar and Martian gravity can clarify realistic human futures beyond Earth: eight bodies, or two, or just spinning free-space settlements. We could test health by long crew stays on the Moon and Mars—or far sooner, better, safer, and cheaper in LEO, in a slowly spinning AG dumbbell.

But we don’t yet even know what AG spin rates crews can handle, without queasiness or other discomforts that affect performance or health. We need relevant AG spin tests to find suitable spin rates and facility radii. Crew comfort and performance are likely to drive us toward the longest affordable spin radius, more than the shortest plausible one. Tests in the VMS and Crew Dragon are the first steps. References 20 to 22 discuss how and why we need to find the “gravity prescription,” to ensure good human health beyond Earth.

I have five recommendations:

1. Use NASA’s VMS to test the reactions to weight changes caused by walking in AG.
2. Investigate potential business opportunities for commercial AG facilities in LEO.
3. Get answers to the questions below, to see if tests like those proposed here are workable.
4. Start envisioning what crews can do during spin tests, to make the results most useful.
5. Pursue ways to work on AG spin tests with Crew Dragon customers and SpaceX.

Technical questions for experts

Dragon layout

• Is the cabin ceiling flat and strong enough for the crew to stand on in up to 0.37g? If not, can that be fixed?
• What is the usable volume for crews in “nose down” AG, and can the seats easily move out of the way?
• Can the Dragon/trunk interface handle about 50 kilonewtons tension, both static and during release in AG spin?
• How and where can the trunk stow, deploy, and release a tape and bridle, at loads up to about 60 kilonewtons?
• Can we cover all exposed “snaggable” Dragon details during a tethered spin test, by tape or anything else?
• What size is the “payload envelope” under the nose cap, that could hold a stowed inflatable module?
• Is there a practical way to secure a bridle for Dragon nose-up AG tests, and route it to the Falcon?

Dragon RCS

• Could RCS propellant acquisition be a problem in any AG direction, up to Martian gravity?
• What are the cosine losses and torques for the Dracos that give the most efficient thrust for flat spin-up?
• How much lower specific impulse do the 12 scarfed-nozzle aft thrusters have, versus the four axial nose thrusters?
• How much net extra RCS propellant is available for AG spin tests, on ISS or other missions? (100 kilograms?)

Operations

• How low can the Crew Dragon deployment orbit be, and what altitudes are acceptable during phasing to ISS?
• Does Crew Dragon have any time or other limits on parking at low altitude, other than drag reboost?
• How many kilograms per day does Crew Dragon need for a crew of four for water, food, air, and all other supplies?
• Can Dragon’s nose cap stay closed for hours, until an AG spin test ends and the tape is released?
• What may limit Crew Dragon spin test durations, if long-duration AG tests become of interest?
• Can Dragon safely release an attached Falcon while in a 3–9 rpm flat spin?

Dynamics & control

• Can the inertial platform and controls handle AG & 3–9 rpm flat spins, with or without Falcon’s upper stage?
• In a stable flat spin, do Dragon’s solar arrays stay on one side of the spin plane, or straddle it?
• If the arrays straddle the spin plane, is there a viable fix (reaction wheel, active balancing, etc.)?
• Does a flat spin with solar arrays on one side of the spin plane also allow adequate communications links?
• After SECO, what is the mass ratio of Crew Dragon to a vented Falcon second stage (near 3.5:1)?
• Where is the Crew Dragon CM after SECO, with seated crew, both with and without an empty Falcon?
• Can the controls provide near-optimum automated pulsed spin-up and boost (with crew inhibits)?

Falcon stage 2

• How and where should Falcon weakly hold the free tape end, and later it clamp it (up to 60 kilonewtons load)?
• How long will the standard Falcon second stage batteries keep the avionics alive after engine cutoff?
• How much impulse can a Falcon stage 2 get from its ullage gas, and also with boiloff of minimum LOX?
• Can a roughly 85 x 310 kilometer orbit target reentry of a spinning Falcon well enough? If not, then what perigee?

Random questions for others

• What useful activities and AG tests can crews do in Dragon, with its nose either up or down?
• How much sustained gravity and/or jitter may various useful “microgravity processes” tolerate?
• What gravity levels, spin rates, test durations, and issues are of interest to Crew Dragon customers?
• What useful spin tests needing a centrifuge too large for ISS might be done during Dragon AG tests?
• What other obvious or unobvious “gotchas” do we need to identify, understand, and resolve?

References

Human health in zero or low gravity

1. M. Roach, Packing for Mars: The Curious Science of Life in the Void, Norton, 336 pp paperback, 2011.
2. S. Kelly, Endurance: A Year in Space, a Lifetime of Discovery. Knopf, 400 pp, 2017.
3. K. & Z. Weinersmith, A City on Mars: Can We Settle Space, Should We Settle Space, and Have We Really Thought This Through? Penguin Press, 448 pp, 2023. Some homework for future space settlers.
4. G. Nordley, Surface Gravity and Interstellar Settlement.
5. G. Clement et al, Human Research Program Human Health Countermeasures Element Evidence Report, Artificial Gravity, Ver. 5.0
6. J. Van Loon et al, 2024 preprint
7. Lectures at Baylor Center for Space Medicine.

Tests and analyses of human response to sustained spin

8. A. Graybiel, B. Clark, and J. Zarriello, Observations on Human Subjects Living in a “Slow Rotation Room” for Periods of Two Days, 1960.
9. F. Guedry, R. Kennedy, C. Harris, and A. Graybiel, Human Performance During Two Weeks in a Room Rotating at Three RPM, in NASA and US Naval School of Aviation Medicine Joint Report, 1962.
10. B. Cramer, Physiological Considerations of Artificial Gravity, in Applications of Tethers in Space, vol. 1, p. 3:95-107. NASA CP 2364, 1983.
11. J. Lackner and P. DiZio, Adaptation in a rotating artificial gravity environment, Brain Research Reviews, Vol. 28, Nov 1998, 194-202. Argues that 10 rpm AG may be ok.
12. YouTube video. see 2:10 to 2:50 for adapting to spin, and the aiming error after a stop at 5:40-5:50.

Crewed AG architectures and tests

13. T. Hall. Artificial Gravity. Many useful AG papers and other AG items, all downloadable.
14. R. Olabisi and M. Jemison, A Review of Challenges & Opportunities: Variable and Partial Gravity for Human Habitats in LEO, 2022. An extensive AG literature survey.
15. C. Pengelley, Preliminary Survey of Dynamic Stability of a Cable-Connected Spinning Space Station, Journal of Spacecraft and Rockets 3:10, 1456-1462, 1966.
16. B. Joosten, Preliminary Assessment of Artificial Gravity Impacts to Deep-Space Vehicle Design, JSC-63743, NASA, 2007; 1g at 4 rpm, uses nuclear reactor as counterweight.
17. D. Lang & R. Nolting, Operations with Tethered Space Vehicles, in Gemini summary conference proceedings, NASA SP-138, 1967, pp. 55-65.
18. Gemini 11 mission movie. Tether test starts at 10:40 elapsed time.
19. G. O’Neill, The High Frontier, Human Colonies in Space, Space Studies Institute, 360 pp paperback, 2019 ed.
20. J. Carroll, Design Concepts for a Manned Artificial Gravity Research Facility, 2010.
21. J. Carroll, What Might Partial Gravity Biology Research Tell Us? AIAA, 2015.
22. J. Carroll, “How to clarify human futures beyond Earth”, the Space Review, Dec. 6, 2021.

Space tether analyses and overviews

23. J. Carroll, Guidebook for analysis of tether applications, 1985.
24. J. Carroll, Lessons Learned from Five Orbital Tether Tests, 2024.
25. Wikipedia article on space tether flight tests.
26. M. Cosmo and E. Lorenzini, editors, Tethers in Space Handbook, 3rd edition.
27. E. Levin, Dynamic Analysis of Space Tether Missions, Adv. in Astro. Sci., Vol. 126, Univelt, 453 pp, 2007.

Joe Carroll is a mostly retired aerospace engineer. He developed the mission scenarios and tether systems used on SEDS, SEDS-2, PMG, TiPS, and TEPCE. His email is tetherjoe1@gmail.com.

Note: we are now moderating comments. There will be a delay in posting comments and no guarantee that all submitted comments will be posted.

Home