They've done both of those

I've been eyeing a subtle feature on the southern horizon for a few sols. This is far away, but not Endeavour-far away (plus it's in the wrong direction to be part of Endeavour). Notice the step up from the right to the left (5x vertical stretch, view from sol 1872):
Click to view attachment

It seems that unfrtunately driving backwards does not help, but rest is. I was told that rest allows the lubricant to flow back into the gears. The RF currents are back up again and in a few sols we will be resting again I think. Sigh....

Paolo

She's getting old, when did walking backwards ever help an old lady? - Just let her rest a little...

Non so se tifi per qualche squadra, ma io stavo a San Siro stasera, e mi sembra che Inter sta per vincere il campeonato...

Drive until the axles melt!

Phil

How will this new problem affect the overall mission? Will it drop our chances of getting to Endeavour?

Don't know yet. There is some challenging terrain ahead still. If we lose the RF wheel when we are on the parking lot type of terrain maybe we can still drive her to Endeavor. We got some experience in driving with a stuck wheel. :-(

Paolo

Isn't their another "Endurance" type crater on the route to Endeavour? Maybe that would be a more realistic target.

Forza el AC Milano

Just observing casually, it seems like every time the RF current acts up, it's after a period of driving laterally across drifts. Could be my imagination.

That would make sense, Mike; the drive wheels are probably spinning a bit during dune traverses.

Administrators: is there any mean to hide this post to Ustrax's view?

Hmmm, I remember seeing higher RF wheel currents in VC anulus, and never thought to correlate the wheel currents with ripple crossings. It seems that whenever the RF current is high it is high for entire drives, not particularly during ripple crossings. But one thing that is related to ripple crossing is heading. Since we drive more or less at the same time of day, the Sun will heat one side more than the other. I will try to investigate this. Thanks!

Paolo

which would also be consistent with the observed correlation wrt. to forward/backwards-driving...

IIRC, Spirit RF went off right after a full steering. Any correlation with end of drive manoeuver?

Opportunitys RF wheel no longer has any steering.

Maybe parking at the end of the drive so that the morning sun heats the RF wheel? Of course this could conflict whith communication passes etc...

Can a single wheel be disabled during driving? A sort of voluntary Spiritism (or should that be Spiritosis?) - the idea being to drive on five wheels maybe three drives out of four, allowing Opportunity to make some progress while still resting the bad wheel. Then driving on all six on one drive out of four to make up some extra distance. Those drives could include a direction reversal.

Phil

Yes, in fact Spirit did that a lot when high motor currents first showed up on her back in the first winter.

Yup - the Thanksgiving Pan (sol 300ish?) shows the interrupted drag tracks from doing just that.

Like a lot of good questions, this one's been asked before. Check out Paolo's response here.

To me it seems this would be a really tough call. How do you quantify the advantages of driving faster now, versus extending the life of the wheel to allow greater maneuverability later? Actually, it may be moot at the moment, since negotiating even the current moderate ripples may be a big problem with five wheels. But then I suppose that would be simple enough to test with a short experimental drive using 5 wheels.

Perhaps voluntary 5-wheel driving will become a good option once we get to the flat parking lot terrain.

Hey, when Two Guys from Canada both suggest the same thing, they've got to start thinking it might be a good idea.

That's right, eh, you hosers!

Phil

I agree. In this type of terrain we need as much heading control as we can. Once on the parking lot maybe we can relax a bit. If the issue is the number of thermal cycles, not the number of wheel revolutions, then the answer is pretty clear (use it while you can!).

Paolo

Take off!

(With apologies to non-canucks...)

Any thoughts on why the right front wheel seems to be more vulnerable than the other wheels?

1:6 says that the first wheel to get trouble will be the same on both rovers.

I think the odds are actually lower than that - the front two wheels are likely to be a bit more stressed in general so I'd have thought it was closer to a 50:50 chance. But you are right - it's 1 in 6 assuming all wheels are equal which means we can't read anything into it with just two data points.

Maybe even shorter odds. The corners could be be more vulnerable than the middle wheels.

EDIT Nice one helvick, you beat me to it there.

Assuming there are no dependent variables between the wheels. But I think the front wheels had to do more "unfolding" out of the stowed position versus the middle or rear wheels. Maybe that is a factor? Plus, the weight distribution might not be symmetrical, with one side heavier than the other due to the robotic arm (it comes out from the right, correct?) and the pancam mast (front vs back). If MSL's first wheel failure is the right front, then we know there's a conspiracy.

... so how many wheels need to fail before the rover can't rove ?
Would a Left Rear freeze do that?

I actually went to the ISIL to do some testing to answer this question a couple of years ago. For sure this depends a lot on soil type, but in the testing I have done, if you lose two driving actuators you immediately become a lander. Of all the wheels to lose, the front wheels are the one that cause less trouble (the rocker is longer than the bogie) and they can be seen in the FHAZ (the RR can't be seen very well in the RHAZ). Between the rear and middle, losing one of the middle wheels has the advantage of having the same difficulties in driving forwards vs backward and causes less heading change when trying to drive straight. The loss of one of the rear wheels would probably be worst, less visibility in the RHAZ and driving backwards would probably be a mess since you would immediately pop a wheelie effectively driving only with four active wheels.

As a side note, I also tried to drive the testbed on pavement Spirit-like and do typical Opportunity-like turn-in-place and neither of these activities are look pretty to the eye or to the ear: the grinding noise and the amount of deflection on the rocker is amazing!

Paolo

Thanks for taking the time to answer..
.. so wonderfully informative!
Best of luck MER team..,
Eoin

*Sigh* Youtube was just a little bit too late...

I still have nightmares after seeing the engineering model (or maybe it was a simulation, I don't recall) driving with 5 wheels in the "Five Years on Mars" special. Okay, not literally, but ouch... that was painful to see.

Assuming randomness, though, we would expect to see increased signs of wear, such as occasional elevated currents, on the other wheels before too long after the first, right? Is it reasonable to invoke randomness when one wheel has serious trouble long before the others show the slightest sign of trouble?

My statistics has always been quite sketchy but I believe the probability of two events happening simultaneously is the multiplication of the probability of the two single events. If the probability of losing one wheel is 1/6, the probability of losing the same wheel is 1/36, right?

Paolo

Yes, but the probability of the first event occurring is now 1. So it is 1*1/6.

Remember that the RF steering actuators is stuck at ~-7deg. That adds a bit of stress on the driving actuator so it is possible that we are seeing signs of wear prematurely on that wheel because of the RF steering actuator. The reality of things is that these are all conjectures, we do not have a definite indication of one source of the problem.

Paolo

Paolo, can you talk a bit about the forward plan now? Still trying driving backwards and resting? Maybe trying heating the RF wheel more by avoiding too much shadow as was talked about earlier. And is the current draw up as high as the last time it went up?


Thank you

The increase in drive currents is not as large as last time, but it seems that driving backwards has no longer the beneficial effects that had in the past. It seems rest helps and therefore we will be doing some IDD work shortly (if ll goes according to plan we should be starting either in tomorrow or Friday planning cycle). Since there is no agreed upon theory of why the RF is drawing more current we will probably alternate periods of driving with some rest while this is helping. We will monitor the RF currents and stop for a couple of wheeks when the values are too high.

Just to give you some perspective, the typical current draw from a drive actuator is about 0.3-0.35A, the RF is now about 0.4-0.5A. During the previous instance of higher currents the current went up to 0.6-0.7A. At the current ambient temperature the drive actuator can draw up to about 1.0A without overheating. Therefore, this is a concern regarding longevity of the actuator, not overheating.

Paolo

Is there an automatic check on the drawn current during driving, some routine which stops the rover if the current drawn on any of the actuators reaches a certain value, or is this only monitored on earth after the drive?

Ah, but there are now only 5 other wheels to consider...so 1/5?

Does losing 2 wheels mean losing 2 actuators, or do some wheels not have actuators?

AFAIK, all wheels have driving actuators but only the front and rear wheels have additional steering actuators.

RoverDriver said: It seems rest helps and therefore we will be doing some IDD work shortly (if ll goes according to plan we should be starting either in tomorrow or Friday planning cycle).

So a break for Oppy soon and some IDD work.
Where's a nice spot? Any cobbles about?
Click to view attachment

Here.
Click to view attachment
The mosaic was shot during sol 1873 and Opportunity did a 2m bump on that direction yestersol (1877).
Click to view attachment

Is there a possibility to also take advantage of shorter resting periods, say one sol or two ? (such as the one-sol breaks that are necessary anyway when Oppy is in restricted Sols)

In the past, what has been the observed effect of shorter breaks on reduction the RF current ?

i.e. do you have an (crude) estimate of some kind of "cool-down half live" value like: one sol resting reduces the current by an factor of xx per cent.. which would help to find the optimal (i.e. minimum resting period necessary)

(EDIT: also, if there is a correlation of the wheel wear problem and the total number of thermal cycles (night/day) then those long stops could be counter-productive in the end, just a thought ....)

Of course! We actually have two limit checks. One monitors each wheel istantaneous current draw and if it is above 1A for 1/2sec it stops the rover immediately (you don't want t trip the current limit in case of an occasional spike). The second limit measures the average current draw for each wheel during the last step. If the average current draw is above 0.35A for three ore more wheels it stops the drive. In Purgatory and other embedding events we have observed that the average current draw raises a bit. This check is supposed to trigger whenever we are embedded. So far we have automatically detected embedding events using slip checks, not by menas of the average current limit.

Paolo

Are we there yet? Seems so. The IDD target will be these tiny fragments. It will be a challenging IDD target for sure!

Paolo

We tried to do short breaks (three or four sols) but it did not help. Sigh...

Paolo