InSight Misson Logbook: Mars Mole Work Suspended for Now

The movement of sand grains in the scoop on the end of NASA InSight’s robotic arm suggests that the spacecraft’s self-hammering “mole,” which is in the soil beneath the scoop, had begun tapping the bottom of the scoop while hammering on June 20, 2020. (Credits: NASA/JPL-Caltech)

DLR Mission Update

In his logbook, Instrument Lead Tilman Spohn who is back in Berlin since April and communicating with JPL via the web, gives us the latest updates regarding the InSight mission and our HP3 instrument – the ‘Mole’ – which will hammer into the Martian surface.

Logbook entry 7 July 2020

On Saturday 20 June 2020 (Sol 557 on Mars), the team completed the ‘Free Mole Test’ announced in my previous blog post. The result was not quite what we had optimistically hoped for, but was also not entirely a surprise. The ‘Mole’ started bouncing in place after making some progress without direct support from the scoop on 13 June (Sol 550).

We have no direct observations of the Mole’s movement or measurements of its progress. Rather, we judge this from the motion of the tether, or more precisely, from the apparent motion of features on the tether with respect to the background. The images clearly show that the tether moved back and forth and eventually stopped altogether, suggesting that the Mole did not dig further down on its own.

The ‘Free Mole Test’ had already started on the previous Saturday, 13 June (Sol 550), but the evidence at that time for the Mole progressing downward during 125 hammer strokes was too ambiguous to be readily accepted. As you can see from the video clip below, the scoop initially went down pushing on the back-cap of the Mole and was further pressed into the sand by the arm. In the middle of the video clip, the scoop stopped moving and the dust particles in the scoop came to rest while the tether kept moving to the right by a few millimetres. This clearly suggested that the Mole had moved away from the scoop and had progressed downwards on its own. The tiltmeter showed the Mole becoming somewhat more upright, which is supporting evidence, and the short-period seismometer showed a change in the frequency characteristics of the recorded hammer signal. The tether apparently reversed motion a few frames later, before going on to move to the right. It then made a very small move to the left again before it finished while moving forward. A careful analysis of the images showed that the net forward motion of the tether (and the Mole?) was two to three millimetres. The integrated motion may have been three to five times as much!

When the team analysed the images on the following Monday, they were pleased to see the Mole moving forward but agreed that at least another hammering cycle was required before we would be willing to conclude that the Mole was in the ground deep enough to dig on its own. ‘Dig on its own’ is admittedly not quite technically accurate, as we were still aiming to provide indirect support by pressing the scoop onto the surface, thereby intending to increase the pressure on the Mole and the friction on its hull. Of course, we could not say how much we helped the Mole that way as we do not know the mechanical properties of the duricrust well enough.

On Sunday 21 June, when we looked at the images that had been sent to Earth after the hammering session on Saturday (Sol 557, 150 hammer strokes), we had to conclude that having the Mole two to three centimetres deeper in and below the surface was not providing the necessary friction, even when helped with pushing on the regolith. The tether moved back and forth and then to the left, reversing much of its forward progress from Sol 550. In the middle of the movie, one can see that the dust particles resumed moving. Two particles seem to even be jumping up some centimetres. But on closer inspection, one can see them rather moving forward from the interior of the scoop in several slides. The moving dust particles imply that the Mole had backed up again and was tapping on the flat side of the scoop from below.

This result of the ‘Free Mole Test’ was, of course, not quite what we had hoped for, but we cannot say that it came as a complete surprise. After all, we are continuing to fight against the missing friction on the Mole hull. The test supports our earlier conclusion that the cohesive duricrust is unusually thick – at least based on what we previously knew about Mars – and that it must be quite rigid. From the Mole’s first backing out (on Sol 322) and the observation that it stopped when it was 20 centimetres out of the ground, some of us (including myself) estimated that the duricrust was 20 centimetres thick (the Mole’s 40-centimetre length minus the 20 centimetres that it had backed out). The present observations at least don’t invalidate that conclusion.

What are our next plans?

First of all, let me say that the team continues to be determined, although we appreciate that the task is not likely to become easier. We have, of course, made a plan for this outcome of the ‘Free Mole Test’. The plan calls for retracting the arm and stereo imaging the pit with the Mole in its interior. We will be interested to see how deep in the Mole really is (it should be a centimetre or so below the surface), whether the morphology of the pit has changed and whether the sand that we had seen in the pit is still there or whether the pit has been drained by the hammering action.

Depending on what the imaging reveals, we plan to see whether or not we can move enough sand into the pit to provide the necessary friction, likely aided by pushing on the sand pile using the scoop to provide additional pressure (see the sketch below). This will be different from pushing on the duricrust surface because the sand has no rigidity and can transfer the force more readily. The scoop will, in addition, guard against the Mole backing out.

Sketch of likely position of InSight mole. (Credit: DLR)

Above is a sketch of the assumed situation in the pit before and after filling the pit as is presently being discussed by the team. During Sols 550 and 557, a large fraction of the pit was empty (left). The stress provided by the push on the regolith with the scoop had to be transferred through the duricrust around the pit to the Mole. Because the pressure generated by the scoop widens and decreases in amplitude with depth, not more than about 35 percent of the load was likely acting on the Mole at a depth equivalent to one scoop width, that is 7 centimetres. If the duricrust is very rigid, this value could be significantly smaller. In addition, approximately 20 percent of the length of the Mole was without any friction. Filling the pit with sand (right) can provide friction to the entire hull and the load can be transferred through the sand filling much more effectively.

Filling the pit will not be an easy task and may take quite some time. (This is the reason for conducting the ‘Free Mole Test’ with­out prior filling of the pit.) A pre­lim­i­nary es­ti­mate be­fore the re­cent push­ing and ham­mer­ing sug­gested a re­quired vol­ume of 300 cubic cen­time­tres of sand, equiv­a­lent to a scrape (or a num­ber of scrapes) with the seven-cen­time­tre-wide scoop for the en­tire 40 cen­time­tres, as­sum­ing that the sand layer is one cen­time­tre thick, as sug­gested by the im­ages.

The team will take a break now to dis­cuss all the is­sues at hand and make the arm avail­able for other sci­en­tific ac­tiv­i­ties. If all goes well, we ex­pect to be back in Au­gust.