I guess the soil crashed and filled the gap opening under the probe by the hammering. The properties of the regolith remind me of what Arthur C. Clarke wrote in "A Fall of Moondust": "This dust has the worst properties of solids and liquids, with none of their advantages. It won't flow when you want it to, it won't stay put when you want it to." Maybe one has to think of it as a very viscous fluid rather than a proper solid. I don't know whether the change in the season and temperature has affected the properties of the soil in any way.

It's worth remembering that no one would have proposed doing the heat flow experiment this way if there had been a more straightforward drilling technique that would fit within the mission constraints.

As was pointed out earlier, even with many fewer constraints, the Apollo heat flow experiments didn't work well the first two times and Apollo 17's only got down about 2.5 meters. https://www.hq.nasa.gov/alsj/a15/a15carrier.html

Exactly. This is alien soil on an alien planet. We have no idea what to expect. And of course, the mole has no way of 'knowing' if it's moving in the wrong direction, so it just kept on unearthing (unmarsing?) itself until the commanded strokes ran out, right?

Not sure I can accept the explanation that high frequency drilling is the problem. Soil falls in front of the rebounding mole? That takes time, even if only a tiny increment of time, so I expect it is less likely with faster operation. But what do I know? (spoiler: nothing). Only testing is reliable, and even that is difficult when considering another world. Do we know anything about other kinds of percussion drilling that might be significant here?

Phil

While not really percussion drilling, why not a self-contained pile driver? There would be no recoil. The mole would need to be longer and of a larger diameter of course to accommodate a weight.. And its effectiveness would diminish if it turns away from the vertical. But there would be no back-filling of material at the front tip. No vibrations to "break friction" on the sides. 100% reliant on gravity to generate the downward motion of the internal impactor. It would still be all self-contained, using mass instead of springs. And it work at 1 hit per second more or less.

Are you sure you know how the mole works? See http://esmats.eu/amspapers/pastpapers/pdfs...grygorczuk2.pdf and https://www.geomechanics.caltech.edu/public...shall-2017a.pdf for details.

One assumes that if they could have gotten a more robust solution into the mass/volume/power box they were in, they would have done so.

the difference between the successful sessions and this one was the scoop directly touching the mole vs pressing on the ground. maybe it isn't so much backfill, but the mole bouncing off the far wall, rebounding, and catching because its tilted over and levered against the far wall? maybe touching the mole stopped much of the rebound, but now it was one step forward and two steps back? the scoop was definitely pushing into a crumbling hole more than ever

the good news is the retreat slowed down in the second 150 hammers, vs the first 150 where it popped almost all the way out

Last week's Mars Society talk by Tom Hoffman is up on youtube:
https://www.youtube.com/watch?v=Q1yJPFRuv74

Mole / Pit of Doom discussion is around 18 minutes in.

Interestingly at about 21:45, he says "We need friction ... because we have the 100N force down and about 7N force back up. We did some experiments on earth where if we had the right conditions the mole could either bounce or come back up."

Holey Moley!
Looking at the images on the Insight site, the mole just popped out!
(a cascading backfill from the hole side material to the conical tip, all the way to the surface?)

I see a shovel on that arm, maybe dig a good-sized hole/trench and then put mole in there?

Wow, that was unexpected!

What proportion of the mole is still under ground?

Half is still below, according to the bold text here: https://www.jpl.nasa.gov/news/news.php?feature=7519

NASA page essentially the same details.
It seems Mars will just certainly not have any of this kind of thing going on!
Gif animation (5fps) of unsettling shennanigans seen on Sol325
Click to view attachment

Drilling alters the frictional properties of the soil.

Could be that impacts are heating up the soil and lubricating the mole.
e.g. friction generates heat, heat melts ices, melted ices and clay becomes a slippery mud.

Could be that repeatedly hammering into a conglomerate-duracrust horizon grinds the soil into a finer particles that act like ooblick and push the mole backward.

Could be like driving across a dry lake after a rain, clay covered by water is structurally stable, but if you stop and add too much force, you mix the clay and water into mud and you loose friction.

For what it's worth (meaning: probably nothing), here's a quick impression of the probe's situation.

In the new DLR blog entry from Tilman Spohn he's still positiv. Nothing is mentioned about a technical error. The mole itself seems fine. They want to inspect the pit and then hope to drive the mole back in. But how?

sounds like they want to be sure it wont tip over, then remove the scoop to inspect the hole, then simply set up for another pinning operation.

I wonder if they also need to correct it to the vertical, as it has already tipped over quite a bit and i doubt there is any intrinsic physics involved in the design or process that should make it auto-correct to the vertical, else any diversion from vertical might create an erosional feedback loop to that side. It seems the scoop will need to be positioned on the opposite side in order to pin it vertically. The worst case scenario is like hammering a nail where the initial angle is critical, even without hitting other obstacles on the way down, left to its own devices in this type of regolith the path of least resistance might ultimately result in a wide circle back to surface given no other inputs, but presumably they can locate the mole position using data from SEIS. From the looks of the current angle, the mole would seem to be ultimately headed somewhere beneath SEIS, if that matters, but in the best case scenario the ultimate depth being reduced by the length of the tether at that angle. how much better science can be accomplished at full depth vs the 'acceptable' 3 meters?

I think the quality of the heat probe data as a function of depth could only be answered, precisely, if we already knew some of the things that the probe is there to measure.

The experimental design is to probe the local thermal inertia at 50 cm intervals, so the number of data points with a 5-meter penetration would be ten. Noise must certainly be much greater near the surface due to diurnal variations, which are great on Mars, so the least noisy measurements would be the last (deepest) two.

The design min is 3m and the only difference I can recall from the briefings was at 3m it would take quite a bit more time to achieve the results.

The arm has just been raised a short distance above the surface. Another rise and we should see the state of the hole again.

Phil

Scoop pulled away from the Mole on Sol 332
Waiting to see the next move
If me, I'd go for filling the pit, but what do I know?
Click to view attachment

Looks like it may have dropped down a bit after removing the scoop.

So, Mars has a duricrust, caliche, or hardpan layer that impedes probes.

Hmm, what could we do about that?





Perhaps the next Mars mission should make those "entry ballast masses" do double duty as mini "rods from god"(RFG).
Make the ballast into long narrow flechettes.

Heck, how about adding some electronics to the tail end of the weights, has NASA asked the Air Force to share some of the high-g electronics from the "MOP" (Massive Ordinance Penetrator) development?



Question - could we have skipped the entire "mole" part of the mission by using a "rod from god" impacting Mars?
If bombs dropped in dense Earth atmosphere penetrate to ~9 meters, a "rod from god" in Martian atmosphere should go deeper. Add a temperature sensor on the tail end,. Add few hundred meters of trailing TOW-wires to the RFG, and use that for radio communication back to the main probe.

^ Deep Space 2.

An impactor would have to leave a hole behind it that collapses, which is automatic if the depth and width would exceed the angle of repose. I'm not sure, however, if that would alter the requirements of the experiment by reconstituting the vicinity of the probe with broken-up ground instead of the native regolith in its natural state. Also, I wouldn't take for granted that antenna lines left outside of the impactor would survive the impact.

It seems to me that an appropriately-designed impact penetrator could meet the requirements but it wouldn't be automatic.


With both Mars and the Moon (see: Apollo) the problem is that the subsurface mechanical properties are not well known. And, FWIW, the Moon and Mars are themselves potentially quite different from one another, so the specific lessons from Apollo don't aid designing for Mars.

If Insight's mole never achieves its goal, we have learned a lot about this site for any future attempts, although the next try might be a very long time in coming. Another Discovery mission, Contour, utterly failed, and there hasn't been any re-fly since.

This was from 2015, but I don't recall anything in the Mars 2020 rover. It seems like it would be near perfect for a University program to specialize in penetrators like this - the small form factor means the cost can be relatively contained, testing would be cheap, and once there is a standard "mass balance penetrator" design, it can start being incorporated into missions.

https://www.nasa.gov/content/nasa-announces...mass-challenge/

Penetrators were all the rage in the mid-70s to mid-90s (and flew on the Mars-96 mission), but the devil is in the details and it's not as easy to get anything useful out of them as might be supposed at first glance. https://ntrs.nasa.gov/archive/nasa/casi.ntr...19750004806.pdf

Maybe there's another thread where this could be discussed. InSight doesn't have a penetrator so it's off-topic here.

How does all those "unscheduled" actions affect the seismometer data collection ?

Some years ago I wrote a historical survey of penetrator missions and proposals. I think they really only make sense at Mars, but even then, you need a programmatic commitment to launch 'many': as there is an irreduceable terrain risk, so a 'mission success = 8 out of 10 vehicles return data' paradigm needs to be adopted.

https://www.lpl.arizona.edu/~rlorenz/penetrators_asr.pdf

"How does all those "unscheduled" actions affect the seismometer data collection ? "

Not much at all. We know when the arm is moving or mole hammering is taking place, and those things only happen during very limited times. It would be easy to take them into account.

Right now it looks like the scoop is being repositioned for another pinning and hammering session.

Phil

The mission have updated their FAQ about InSight's Mole Link

They also added a new animation "InSight's Arm Camera Stares Into the Pit" link

thanks for that, lots of good info in there! oddly all the images are partially obscured by a dark elliptical segment at upper left.
haven't made it through all of it yet but recollecting how sadly after two decades there still haven't been any follow-up efforts after DS2 failed to phone home, which combined with the whole MPL debacle seemed to push the whole faster/cheaper/riskier approach off the table, unfortunate since it seems such probes can be done cheaply enough even as low-key evolutionary engineering studies we could have had quite a scientific bonus in the fraction where they may have survived.

Perhaps a mole-like device would work fine if the "hammer" and the "spike" were separated, at least until the "spike" part has been securely buried. After that, the "spike" could morph into a separate drill itself. (It has been speculated that the current problem is due to an infill of soil when the drill is raised in between strokes.)

The ICC takes images up to several times a day which can be searched for evidence of clouds and dust devils. This composite shows changes in the sky in recent sols, using images with similar lighting.

For each pair of sols I merged the later image with the negative of the earlier sol and greatly increased the contrast of the result. The surface looks really bad because the shadows are not exactly the same, so I replaced the surface with an unprocessed image. The sky sections show some markings through the noise, which I have blurred to show the markings better. I did it for two image pairs to make sure the changes were not the same each time, so, at least part of what we see should be real.

Phil

Click to view attachment

Nice work Phil

Meanwhile on sol 339, the scoop moves closer to the mole. Likely setting up for another pinning session.
GIF using 2 processed frames acquired shortly before local noon. GIF reduced to 800x800 to speed loading time.
Click to view attachment

While we wait for action on the pin, here's a picture taken on sol 343 (yesterday as I write this). Note the disturbed surface indicated by arrows. It was made on sol 318 when a section of the tether was briefly in contact with the surface.

Phil

Click to view attachment

New images from Sol 346 just arrived. It seems the movement caused the mole to come out a bit more.

Click to view attachment

I don't see it getting higher out of the ground; seems more like a sideways/translation movement, with a bit of twisting.

Were seismogramms recorded during penetration attempts?

sol 349 : Another pinned hammer session. Looking good, mole heading into the regolith
Processed GIF (800x800). More frames likely to be in the pipeline...
Click to view attachment

This tweet seems to show quite a bit more progress. https://twitter.com/NASAInSight/status/1197594417667772416

Yes (for some of the hammering sessions): Read the DLR Mole log book / blog on this link then search for 'SEIS' You will find the teams interpretation of the data from SEIS and how they used to understand what was happening under the ground. AFAIK this will have continued through all hammering sessions, but you'ed have to check the data. The early data from SEIS is already in the mission PDS, but looks like the data for the HP3 observations by SEIS will only be added at a later date, See the PDS mission link. This states the next PDS (release 3) is currently scheduled for January next year.

The sol 370 night ICC views show some faint features, such as the SEIS shield and tether field joint:
Click to view attachment
The horizon may be faintly visible too. This frame was taken after 9pm local time, so it can't be residual sunlight. One guess is Phobos moonlight. I'm not sure if it was up at the time or if it's bright enough to create a visible glow to ICC...

That's a nice catch. I applied some tricks to increase the signal to noise and came up with this. It is clear that the light is coming from a point source, so Phobos almost certainly is correct. There is a full Phobos at least once and often twice each night on Mars. Phobos must be overwhelmingly the brightest thing in the martian night sky, so if this turned out not to be Phobos lighting this image, then Phobos-light images would have to be much brighter still.

Just for those interested, a SEIS News section has been added to the english version of the SEIS website. Better late than never I guess. For the moment eleven articles, selected amongst all the original content, have been translated in english. Should be better than Google translation, at least I hope. Please don't hesitate to tell me if you spot anything suspicious or incorrect.

The SEIS News section will work mainly as a partial archive, and will be not fully synchronized with the french section. It means that some french articles (old or new) will be not automatically translated and added to the english section.

However, the next thing to come will be a ~10 pages general article summarizing all the SEIS scientific results obtained so far. It will be normally published in January, with the official release of the first batch of "big" scientific papers. We will do our best to publish this article in english and french at the same time.

My version of the night imaging. The three best images cleaned and merged.

Phil

Click to view attachment

Nice job pulling out those details, Phil. I think we can even see the shadow of the SEIS shield to its right now!

The illumination on the shield, the fact that the sky appears to be brighter towards the left of the frame, and the putative shadow all point to illumination from the left, ie from the east.

And the mole is on the move again on Sol 373 Digging deeper into Mars...
Click to view attachment

Now how will they prevented what happened last time, once it got too low for the arm to keep pinning?

A tweet posted to the Insight Twitter account today states: "The @NASAInSight seismometer has discovered a strange, continuous signal at 2.4 Hz, apparently not related to the lander or weather activity, but excited by a lot of #MarsQuake. This puzzling resonance acts as a natural seismic amplifier!"

Likely associated with this abstract from the recent AGU