Shortly before publishing my article on David Kipping’s TARS concept (Torqued Accelerator using Radiation from the Sun), I received an email from Centauri Dreams associate editor Alex Tolley. Alex had come across TARS and offered his thoughts on how to improve the concept for greater efficiency. The publication of my original piece has launched a number of comments that have also probed some of these areas, so I want to go ahead and present Alex’s original post, which was written before my essay got into print. All told, I’m pleased to see the continuing contribution of the community at taking an idea apart and pondering alternative solutions. It’s the kind of thing that gives me confidence that the interstellar effort is robust and continuing.
by Alex Tolley
Dr. Kipping’s TARS proposed system for accelerating probes to high velocity is both simple and elegant. With no moving parts other than any tether deployment and probe release, if it works, there is little that can fail during the spin-up period. There are improvements to the basic idea that increase performance, although this essay will suggest a more complex, but possibly more flexible and performant approach using the basic rotating tether concept.
First, a small design change of TARS to increase the rate of spin-up. The TARS design is like a Crookes radiometer, but working in reverse, with the mirror face of the sail experiencing a greater force than the obverse dark, emissive face. As the tethers rotate, the reflective face increases the spin rate, whilst the emissive face swinging back towards the sun acts as a retarding force. An easy improvement, at the cost of a moving part, is to have the sail reorient itself to be edge-on to the sun as it returns. This is illustrated in Figure 1 below. The rotation can be any mechanism that sequentially rotates the sail by 90 degrees after the tether is aligned with the sun, or other electromagnetic radiation source.
Figure 1. The simplified TARS system with the sail rotating around the tether to reduce the retarding force in the rotation phase.
There are other possibilities to tweak the performance, but at a cost of complexity and added mass.
However, I want to offer an alternative approach that solves some of the limitations of the proposed TARS system.
These limitations include:
- The propulsive force is very phase-dependent as the tether rotates.
- The rotation rate is dependent on the sail aerial density and size<
- The sails add mass to the tether and therefore increase the tether tension, requiring an increased taper
- The TARS rotation must be aligned with the radiation source, limiting the direction it can throw the payloads. This means that a target on an inclined plane to the planets, such as a comet or exoplanet, requires the TARS to take on an inclined orbit, limiting its flexibility.
- The asymmetric forces on TARS change its orbit.
These limitations can be alleviated by eliminating the sails and replacing the rotation with an electric motor, powered by a solar panel. The basic design is shown in Figure 2.
Figure 2. Basic design of a rotating probe launcher using motor-driven tethers.
The tether is powered by an electric motor that requires a counter-rotating wheel or tether (see later) to prevent the system from rotating. This is similar to the power equipment astronauts use in space. The tether is attached to the solar panel by a 3-axis joint to allow full control of the rotational plane of the tether. As the only loads on the tether are its own mass and the releasable probes, the amount of taper should be less than TARS, allowing longer tethers of the same material. The tethers can be flexible or stiff, depending on deployment preferences. Figure 2 shows a preferred arrangement where the tethers form a square, with cable stays to increase rigidity and offset bending during spin-up.
The tether would have 2 releasable probes and 2 small ballasts to maintain tension, or 4 probes. The probes can be released simultaneously in opposite directions, or in the same direction from 1-10 milliseconds apart, depending on the rotation rate. If released in the same direction, the system will tend to be pushed in the opposite direction as the probes released in the same direction would act as propellant, generating thrust in the opposite direction.
A variant would allow for 2 contra-rotating tethers. Because they are mechanically coupled to the same motor, this guarantees that they rotate in synchrony and eliminate the gyroscopic action of a single tether. This removes the need for a counter-rotating disc for the motor, but more importantly, for multiple payloads allows the rotation plane to be changed between payload releases, allowing for different target destinations for the probes to travel in. This would be ideal for a standby to target comets and objects coming from different orbital inclinations, as well as more detailed mapping of the solar system’s heliosphere.
Because the rotation is controlled by a motor, this provides more precise timing of the payload releases. Once the maximum rotation rate is reached, the motor can idle, and the system continue its orbit until the optimum probe[s] release position is achieved, for example when Mars is in opposition. This avoids the continual rotation rate increase of TARS that must release its probe[s] before the tethers snap.
So what sort of rotational speed can a motor provide? The maximum speed for a small motor is 100,000 rpm, or 1667 rps. A much lower speed is achieved by hard disk drives at 7200 rpm or 120 rps.
This translates to:
Because the rotation rate is so fast, any probe release must be timed with very high precision to ensure it travels on the correct flight path towards its destination. While not critical for some missions, encounters with small bodies such as interstellar objects (ISO) like 2I/Borisov will require very high precision releases.
Unlike TARS, the tethers can also be spun down, making the system reusable to reload the payloads. If multiple payloads can be released sequentially like a Pez dispenser, then these can be reloaded periodically when the payloads have been exhausted. With extra complexity, these cartridges of probes could be carried on the system, and attached to the tethers after the rotation has been reduced to zero, making the device relatively autonomous for long periods.
Lastly, because the rate of rotation acceleration is dependent on the motor and power available, the power can be increased with a larger solar array, and the motor torque increased with a larger motor. These are independent of the tether design, making any desired upgrades simpler, or like CubeSats, configurable on manufacture before launch.
This is a great mass driver concept, but it can be made even greater by dispensing with the wasteful spin-up/spin-down process. There is no reason the payloads could not be continuously fed from the hub to the tip along the tether while the tether is under constant rotation. Such a device would have near 100% energy efficiency.
Also, longer tethers are better. They reduce the rotational period and maximum acceleration, and decrease the timing accuracy requirement. So, practical devices will likely involve tether lengths in the 1-100 km range.
The longer the tether, the greater the tension on it. I haven’t thought too much about the angular rotation rate vs the increased tension due to greater mass. I assumed that Dr. Kipping’s design with a 12.1m tether was optimized for this and wanted to keep the alternative version closer to that bound with a 20m tether (10m radius at least for 3 examples).
Yes, one could feed payloads from the center to the periphery. That imposes more complexity and mass, and bending forces on the tether. I think there is also momentum transfer, the moving payload slows down the rotation rate as it travels to the periphery. [See the classic ice skater spin rate slow as the arms are extended.] However, it does offer the capability of having many more low-mass payloads without incurring more tension than if they were all on the end of the tether.
For delicate payloads, including crew, several hundred km are needed, I think. Very feasible, nevertheless.
I think that a crew module is best accelerated by a mass driver. These have appeared frequently in fiction from A C Clarke’s Maelstrom II, to Ian McDonald’s Luna trilogy, as well a popsci such as O’Neill’s lunar mass drivers for colony materials.
The issue of very long tethers, for transport of people and cargo from the Moon to geosynchronous orbit, has been studied, including possible materials. Here, the tension is due to Earth’s gravity tugging on the cable, rather than rotation. A Martian surface-to-areosynchronous orbit cable was indicated in A C Clarke’s The Fountain’s of Paradise that was kindly mentioned to me by frequent essayist and commenter LJK.
I have Space Tethers and Space Elevators (2009) by Michel van Pelt that goes into more detail about how tethers can be used in various space applications. On p20, he talks about tethers as linear launch catapults, but not as a rotation device for a space-based launcher. In this regard, TARS may be a truly novel approach.
If you use carbon nano tubes and the probe can withstand the very high accelerations it may be better to have the probe at the centre of the spacecraft and a dumb mass at the tip. On breaking a nanotube has a wave propagation speed that is around a 1000km/s. In effect you use the rotational velocity and the stored energy in the tensioned nanotube.
@Michael.
I don’t see the connection between wave propagation velocity and the velocity of the cable as it snaps under tension. Are you suggesting that the tip would move at 1000 km/s when a wave in the tether reaches the tip?
AFAIK, the only movement is the lateral movement of the tether as the wave passes. The same as in water, the wave velocity is far greater than the circular velocity of the water molecules. The wave doesn’t mean that anything physical moves in the same direction as the wave velocity. If it did, the tether would immediately snap as the wave started.
It is teh same with the sound wave propagation. The sound wave moves very fast, but the material in the teher does not move at all.
A problem I haven’t seen addressed is the timing of detaching the probe. The timing (=aiming of the probe) will have to be incredibly precise to be sure the probe meets its target instead of passing it millions of kilometers away
Yes. With incredible timing and accuracy for “point” destinations. The probes had best have their own sails that can navigate to overcome this. For destinations such as exploring the heliosphere and ISM, this is not such an issue.
Some tether related thoughts…
For heavier probes, might a Centaur tethered to a science-craft be able to sling a small probe to a target otherwise unreachable by inclination.
An upper stage not able to reach a target on its own can at least yet a probe off to the side.
This is the only way you can “turn” in space without needing the gravity of a large planet, or brute force NSWRs.
Might TARS be a way to slow an inbound object?
A way to reverse a Sundiver?
What would this look like to observers here on Earth?
For the 10 meter line, omega is 2pi * 1000/s radians. Omega squared r is 395 Mm/s^2 = 40 million Earth gravities. Now in biology, centrifuges of 60,000 xg are impressive, but supposedly, Iranian nuclear centrifuges can reach 500,000 xg now.
If 10 meters of the line can withstand an average of 20 million Earth gravities pulling down on its weight, that means that 200,000 km of the line can withstand one Earth gravity of pull. This meets the specs for space elevator line.
Now if it could hold 1 kg at this speed, 1/2 x 1 kg x (62832 m/s)^2 = 1.974 GJ. So the system (if dominated by payload) would store an energy density of 1974 MJ/g, far in excess of any known rocket fuel.
This is the key number allowing the system in this or the previous article to work – yet so far, as far as I know, flywheels have not surpassed chemical fuels for energy storage. Why are people now convinced they can store much more energy than that?
Hi Alex
Whew! Not sure a payload is going to survive the gee-forces of a tip-speed of 62 km/s if the tether is 10 metres in radius. 1000 rps is 6.28 kiloradians per second angular velocity, meaning the gee forces are 40 MILLION gees.