The fact is, if we go to all the trouble of sending a probe to Phobos, we absolutely want to land. An orbiter is still invaluable, but landing, while not mission-critical, would vastly increase the mission's science return. Ground-truthing can ensure that data collected from orbit is accurate. Many types of data, such as microscopy (which is important to understanding Phobos's nature and environment), non-optical spectrometry (which can directly measure composition), and regolith properties (critical to know before a crewed mission), can only be taken at the surface. And regardless of whether the landing succeeds or fails, the attempt itself will be a source of data for later, more-capable (and expensive!) missions.
Even our non-lander concept includes a landing at the end of the mission, but this is a why-not, planetary-protection, see-what-we-can-get landing. Designing the probe to reliably survive a landing without unduly compromising the rest of the mission is a different matter. Wheels actually make a landing easier, since landings on Phobos can be at a very shallow angle, to be detailed in a future post (DIFP). Landings are likely to be on the order of a few meters per second; wheels allow us to choose a trajectory that minimizes vertical speed, and the wheels, spun up to match the probe's horizontal speed, can decelerate the probe over time, minimizing stresses. In a worst-case scenario, wheels act as a crumple zone, protecting the rest of the probe and allowing some data to be returned from the surface.
In addition to helping the probe survive landing, wheels would of course allow the probe to rove over Phobos's surface. While many approaches more suited to milligravity have been proposed, such as hoppers, cilia, and legs, these approaches all require the rover to be designed around them. Wheels minimally impact the design of the probe, and can draw from the extensive development of nanorovers for both space and terrestrial applications.
Roving significantly increases science return (details). Not only does it allow fine positioning and investigation of multiple sites of interest, it permits us to investigate the interaction between different regions, which can reveal a great deal about the nature of Phobos. Specifically, looking at the interaction between Phobos's color units (DIFP) can help determine its interior composition.
The mission most similar to our plan is the MUSES-CN nanorover, developed by JPL for the Hayabusa mission. While it was cancelled due to cost and weight overruns, there was plenty of development for us to learn from. Mission details can be found here.
 |
| Credit: JPL |
 |
| Phobos surface. MOC image SP2-55103, credit NASA. |
 |
| Itokawa - credit JAXA, ISAS |
So how fast can we go on Phobos? Acceleration is limited by surface gravity and surface cohesion. We haven't crunched any numbers yet, but the surface gravity near our landing site is about 4 mm/s^2, so acceleration would be some fraction of that. Top speed, however, is much less limited. Obviously, top speed should be kept significantly under escape velocity, to avoid accidentally escaping the body. For MUSES-CN, Itokawa's low escape velocity of 15 cm/s meant that speeds on the order of centimeters per second were risky , while on Phobos traverses of tens of centimeters per second could be possible.
The radius of curvature of the surface must also be taken into account. A rover can go pretty fast inside a crater, but has to slow down on the rim to avoid flying off. However, on the planetary scale, this only limits the rover to about a meter per second, well over expected speeds. Speeds will be more limited at the rim of small craters and grooves, which will have to be taken into consideration. Since impact speeds would approximately equal takeoff speeds, though, ridges could be used to jump large distances quickly, although the risk inherent here, even with reaction wheel control, means that such a gambit would only be attempted late in the mission, after all primary objectives have been reached.
 |
| Credit: D.J. Scheeres |
The primary reason to keep speed down is that the faster the rover goes, the longer it will lose contact with the surface after hitting a bump. This is even less of an issue for us, since our nanorover would have reaction wheels and so would have some control without surface contact. The speed of our nanorover will be constrained primarily by the available motors. We believe speeds on the order of ten centimeters per second to be reasonable, but even speeds of a centimeter per second allow long traverses.
We initially intended to use MUSES-CN's wheels, as they were developed almost to the point of flight-readiness. However, the wheels need to be stowed within the frame to be deployed. Having the wheels fold out was complex and took up a lot of internal volume. We decided against articulated wheels after learning that MUSES-CN's struts were not meant to move to maintain contact with the surface, but only to allow the body to be rotated. The effect of fixed wheels will be discussed in an upcoming post.
We decided to take a new approach to wheels, with the goal of minimizing internal volume. We currently plan to use wheels similar to those used by the LAMAlice nanorover (details), with the primary difference being that our wheels would be slightly larger, at approximately 6 cm in diameter. Other adjustments, such as changing the number or thickness of blades, may be made as we get a better understanding of their effect on performance. The wheels would be stowed inside the nanorover as shown at the right. Our current approach to deployment is as follows:
We initially intended to use MUSES-CN's wheels, as they were developed almost to the point of flight-readiness. However, the wheels need to be stowed within the frame to be deployed. Having the wheels fold out was complex and took up a lot of internal volume. We decided against articulated wheels after learning that MUSES-CN's struts were not meant to move to maintain contact with the surface, but only to allow the body to be rotated. The effect of fixed wheels will be discussed in an upcoming post.
We decided to take a new approach to wheels, with the goal of minimizing internal volume. We currently plan to use wheels similar to those used by the LAMAlice nanorover (details), with the primary difference being that our wheels would be slightly larger, at approximately 6 cm in diameter. Other adjustments, such as changing the number or thickness of blades, may be made as we get a better understanding of their effect on performance. The wheels would be stowed inside the nanorover as shown at the right. Our current approach to deployment is as follows:
 |
| Credit: EPFL |
The drive shaft is a rod with helical flutes, like a drill bit. While within the nanorover's frame, the wheel is prevented from turning by a rod between the spokes, so as the drive shaft is turned, the wheel is pushed outside the nanorover's frame, and the blades spring open. A cap at the end of the shaft prevents the wheel from moving further out, and the open blades prevent it from moving in. Now, when the shaft is turned, the wheel turns with it.
The main difficulty with this design is that the shaft cannot extend outside the frame. Currently we plan to have the wheel attached to a helical spline which remains attached to the drive shaft. Unfortunately, this does somewhat increase the required internal volume, as the shaft must be at least twice as long as the wheel is wide.
Each wheel will be powered by its own motor. Due to the low gravity of Phobos, it's very important for the motor to rotate in very small increments. MUSES-CN's motors could be rotated by a mere 0.044 degrees, moving forward only 23 microns per step! Consistency and control is much more important than power or speed. While we haven't chosen a motor yet, it will likely be very similar to those of MUSES-CN or Sojourner.
It's very difficult to predict how long the probe will survive on the surface. It could fail almost immediately, or remain functional for months. The main life-limiting factor is expected to be thermal cycling, as the nanorover would go through large temperature swings every seven hours. MUSES-CN, with minimal thermal control, was expected to survive about 30 days; our nanorover will likely have better thermal control, but will not be designed as a nanorover, so we really don't know what to expect. We will probably have a better lifetime estimate closer to the mission. This means it's very important to prioritize targets, to make sure the most important targets are investigated first. The landing site will be selected based on ease of landing, but we plan to land close to the Phobos Monolith, which can be seen in the image of Phobos's surface. After that, we plan to traverse the grooves, move around Stickney Crater visiting boulders and craters, and towards the end of the mission investigate landslide features and other riskier targets. All of this will, of course, be DIFP. The mission is unlikely to extend past October 2021, as at this point the Sun will come between Earth and Mars, and communication with Mars will be lost. If most objectives can be met before then, it's unlikely that we will attempt to reestablish communication.
Whew! This was quite a lengthy post. Most posts will be much more readable than this was, but I've been asked multiple times about the wheels, and I wanted to get all the details out there into the public. Thank you for reading!
-Daniel Zsenits
The main difficulty with this design is that the shaft cannot extend outside the frame. Currently we plan to have the wheel attached to a helical spline which remains attached to the drive shaft. Unfortunately, this does somewhat increase the required internal volume, as the shaft must be at least twice as long as the wheel is wide.
Each wheel will be powered by its own motor. Due to the low gravity of Phobos, it's very important for the motor to rotate in very small increments. MUSES-CN's motors could be rotated by a mere 0.044 degrees, moving forward only 23 microns per step! Consistency and control is much more important than power or speed. While we haven't chosen a motor yet, it will likely be very similar to those of MUSES-CN or Sojourner.
It's very difficult to predict how long the probe will survive on the surface. It could fail almost immediately, or remain functional for months. The main life-limiting factor is expected to be thermal cycling, as the nanorover would go through large temperature swings every seven hours. MUSES-CN, with minimal thermal control, was expected to survive about 30 days; our nanorover will likely have better thermal control, but will not be designed as a nanorover, so we really don't know what to expect. We will probably have a better lifetime estimate closer to the mission. This means it's very important to prioritize targets, to make sure the most important targets are investigated first. The landing site will be selected based on ease of landing, but we plan to land close to the Phobos Monolith, which can be seen in the image of Phobos's surface. After that, we plan to traverse the grooves, move around Stickney Crater visiting boulders and craters, and towards the end of the mission investigate landslide features and other riskier targets. All of this will, of course, be DIFP. The mission is unlikely to extend past October 2021, as at this point the Sun will come between Earth and Mars, and communication with Mars will be lost. If most objectives can be met before then, it's unlikely that we will attempt to reestablish communication.
Whew! This was quite a lengthy post. Most posts will be much more readable than this was, but I've been asked multiple times about the wheels, and I wanted to get all the details out there into the public. Thank you for reading!
-Daniel Zsenits
No comments:
Post a Comment