Time to update the map, Tesh?

Oh, what the heck, let's get wildly speculative...

1)
I would assert that today's movement was intentionally slight, a deliberate perturbation to confirm data on Spirit's mechanical and frictional state. This includes noting if the rock underneath is in significant contact (is it a pivot?). The response from this perturbational movement will help determine the next movement set. Mechanical engineers love perturbation theory. ...And the principal of virtual work...

2)
Digging myself deeper (not Spirit, hopefully), the JPL guys in the sand pit found a "stick-slip" type of behavior, so that a slight movement below a certain threshold (.5 mm?) would not cause the sandy dust under the wheels to break free and slip. Furthermore, the diurnal thermal cycle causes a "reset" of the grain inter-particle friction, so the slight stick-before-slip can be repeated the next Sol. So, we can expect Spirit to explode out of its hole at 1.5 cm/month.

3)
The right front wheel no longer turns, but it can be rotated left and right. In fact, if this is done while the other wheels are pushing, it reduces somewhat the effective drag that dead wheel provides. Helps Spirit shimmy out of there...

--There, now I feel better.

According to the update:

"A tight limit on vehicle roll and pitch of less than 1 degree change was set for this first drive. As the rover began its first move, it sensed that its roll was outside the allowed limit and safely stopped the drive."

Less than one degree pitch or roll. I don't think they expected the drive to last very long with those restraints -- a few seconds at the most.

The specific pitch or roll movement triggering an abort would give them some valuable information. According to the update it was a roll that shut the drive down. To the left or right? It seems that a roll to the right would be preferable, but then it would still need to be determined if it was because of the left wheels rising or the right wheels sinking.

Patience fellows.....the next drive will be huge...like 2 seconds......

BTW, what was the normal tilt and roll limit, and the ultimate limits which would make it fall over ?

Did the rovers tilt change during the drive attempt? I don't understand how the abort could have been triggered during the first second of the wheels turning.

Sunspot,
My impression from the line:

"As the rover began its first move, it sensed that its roll was outside the allowed limit and safely stopped the drive."

Is that the roll of the vehicle was *already* out off allowable drive limits before even starting the drive. i.e there was a mismatch between the actual roll of the vehicle and what the drivers thought it was when setting the limits. As soon as the first roll limit check was made after the drive started, this was discovered and the drive stopped.

Using 145cm wheelbase and 1 degree angle I get 2.5cm* change in elevation from one side to the other. We don't know what the limit was set at, one half degree? One tenth? It's not hard to imagine a vehicle the size of Spirit settling one centimeter on one side as soon as its wheels start turning. No need for that to take one whole second.

*EDIT: correcting the wheelbase to 106cm, the calculation gives 1.85cm change in elevation.

Not knowing which direction the tilt actually went at this stage (did someone work it out?), but...
...to visualise a full 1 degree tilt (the real tilt was less) in 1 second, here's a animation.
Click to view attachment

EDIT: 11 degree angle at start, rotating by 1 degree.

I got the impression it was 1 degree from where they were at, they already said she was at a small of degree roll on the edge of the crater didn't they ?

The change in the Z-axis during the drive, measured by the rover itself, was only 1mm.

A polite reminder repeated from earlier...

This is a warning. Progress may be negligible, it may be significant, it may be entirely invisible and non existent.

DO NOT attempt to draw too many conclusions, positive or negative, regarding the efficacy of the technique, the likelihood of a final extraction and so on. People using words like 'never' or 'certain' or 'wrong' will have firm forum wrist slappings.

Many of you were not around during the month of extraction at Purgatory Ripple. The old hands know how frustrating this process is going to be. Be patient. Do not jump to conclusions. Do not resort to superlatives.

Warning over.

From the original press kit, way back when:
"The distribution of mass on the vehicle is arranged... to tolerate a tilt of up to 45 degrees in any direction without overturning, although onboard computers are programmed to prevent tilts of more than 30 degrees."


Spirit is now maybe 11 (12?) degree tilt after recent movements.

Interesting. We don't know what fraction of one degree tilt was set as a limit, maybe just a tenth of a degree.

Plugging in wheel base 145cm and tilt 0.1cm, I get an angle of 0.04* degrees. If the Z axis change is from the center of the wheel base, the total from one side to the other would be 2mm for close to a 0.1 degree change.

edit: Maybe Astro0 should adjust his animation so it's not so scary.

*EDIT: correcting the center line wheelbase to 106cm, the angle is 0.054. Still approximately 0.1 degree total change.

My original interpretation was that the change in roll after starting the drive exceeded the limit. But this quote from the update supports your conclusion:


Still, answering Centsworth's question from above, there definitely was a roll during the short drive attempt. Compare hazcams from 2088 and 2079. In both front and rear cases, the horizon tilts between 2079 and 2088, which means there was a roll. The direction of roll that causes that change in horizon tilt is down on Spirit's left side/up on her right side. My guess would be down on the left rather/more than up on right. So that means she's sunk a bit deeper into Scamander on her left.

But from those same hazcams you can also estimate the angle of the roll. In both cameras, the horizon rises by about 1 pixel on Spirit's left side from 2079 to 2088. It's unclear exactly what to take as the long arm of the rotation, plus there are significant distortions in these hazcameras, but that gives roughly a roll angle of 0.05 to 0.1 degree. That agrees surprizingly well with Centsworth's estimate of the roll based on the drop in z above.

One last comment that no one's made yet, before I blow Emily's analysis-to-drive-duration ratio completely through the roof: I definitely see the drive stopping prematurely as good news. The images showed very depressing progress had the drive proceeded for the planned 5 metres. But very little progress in less than one second is just fine.

Another armchair rover driver here. In that one second of "activity" I doubt the wheels turned even the 6.3% of one full rotation. I would guess the very cautious pitch threshold was reached the moment the wheels started to spin, and thus the movement was aborted. I'm thinking its an actual tilt vs. what JPL told the computers the tilt was abort. Either way, this is an easy fix.

I think we can answer that. Looking carefully at the LF and LR wheels in the hazcams, we can identify features on the wheels (or in the powder caked on them) between sols 2079 and 2088:
Click to view attachment Click to view attachment
It looks like the wheels turned only about half a wheel cleat. With 29 cleats all around, that gives a rotation of about 6 degrees, or about 1.7% of a full rotation. If 6.3% corresponds to one second at full speed, then this drive was about 1/4 second at full speed. Longer if slower.

"Progress ?" was there any forward progress, or did the wheels just spin ?

Given that they are expecting to drive many meters to get the couple of feet they need to get out of the trap, I think it's safe to say that the one second drive was all wheel spin, if such a minuscule turn can be called "spin"

Although full speed is quoted as 5 cm per second, according to this paper "Straight line driving speed is set to 3.75 centimeters/second...." (page 2). Who knows what the speed was set to.

Are you sure about that? When I compare front hazcams from 2088 and 2079, I reach the opposite conclusion. Also, looking at the animation, it does not appear the LF wheel has sunk deeper; if anything, it seems (even after discounting the deceptive effect of the shadows) to have risen slightly.

Pretty sure. Flip between the 2079 fhaz image:
http://marsrovers.jpl.nasa.gov/gallery/all...04P1214R0M1.JPG
and the 2088 image:
http://marsrovers.jpl.nasa.gov/gallery/all...05P1214R0M1.JPG
Make sure the images are aligned exactly (full screening both should work). Notice that the horizon along the left side of the image rises between 2079 and 2088. It's only by a tiny amount, around 1 pixel. If you try to picture the camera fixed to the rover, if the left (west) side of the rover sunk down a bit, you'd expect the left side of the horizon to rise a bit in the front hazcam view, and that's what we see.

I agree about the LF wheel looking like it's risen a bit. But remember that with the rocker-bogey system, one wheel rising doesn't mean rover rising.

I used an incorrect wheelbase dimension in my previous calculations. I have added edits.
The total calculated tilt change of about 0.1 degree remains the same.

The paper I quote in my previous post says, "Wheel baseline is roughly one meter side-to-side..." And in a post by Doug: "Wheel width is 160mm - and the distance between the insides of the wheels is 900mm for the front and rear wheels..." That would give 106cm between wheel center lines, not the 145cm I originally used.

So, the less 'scary' version based on everyone's calculations above (ie: 0.1degree after .25secs)
Almost imperceptible!
Click to view attachment

Let's hope that's closer to reality.

Sol2088 Navcams taken at 13:37.52 and 13:41.40 local time.
Click to view attachment

According to marssciencegrad it looks like there'll be another drive attempt on sol 2090. If it's around noon or 1pm local Spirit time, that'd be November 19th, 4 - 5 am California time, or 12:00 - 13:00 UT.

Second drive should be similar to the first:

http://marsrovers.jpl.nasa.gov/spotlight/20091118a.html

The second commanded drive is similar to the first commanded drive.

I think we're probably hoping the actual drives are a bit different

Some data flowing in the tracking web.
Dunno how many steps were executed but the net advance seems to be about 1cm.

That would be fantastic, considering the tilt restraint that must still be pretty tight.

I do have basic questions here that apparently have not been addressed since we all thought we'll drive backward.
Assuming the crust will not broke more, what is the distance for Left front to reach firm terrain? (Same question for Middle an Back weels from both sides)
I mean, once LF will be in firm terrain the traction could improve a bit and so on when more weels will meet it. I'll also say the Middle Right will be the first to get to firm terrain, correct?
How this compare to purgatory?

My guess is about one meter before the LF gets out of churned up soil, if they stay on the tracks. As said by many, there are so many differences that a comparison with purgatory would not mean much.

this is reminding me of something a few years back ( well many )
seeing a Corvette balancing on a OVERLY tall speed bump . it bottomed out on it and had the back wheels up in the air about 1.5 in.

Since the poor RF wheel is stuck with undesirable bulldozing up front on the solid terrain, does anyone have any idea what is its' lowest-drag profile? It's currently set square with the other wheels but one wonders if a bit of toe-in or toe-out might slightly reduce drag/pile-up and maybe induce a bit clockwise twist to the whole rover movement. The equation gets even more interesting (complicated) when the rover does a 'crabbing' move...

We've got movement! Sol 2088 to 2090:
http://qt.exploratorium.edu/mars/spirit/fo...05P1214R0M1.JPG
http://qt.exploratorium.edu/mars/spirit/fo...06P1214R0M1.JPG

Not only movement but I think the LF wheel has moved up a little bit so it's not digging in the soil.

The RF wheel looks to have moved ahead in relation to the small rock next to it by as much as a cm*. I can't wait for the flickers!

*edit: Compared to 11/17/09 hazcam.

To play devil's advocate, one reason why the horizon would drop in the front camera view would be if the rover were sinking in back. I'm having a hard time judging what's going on in the rear hazcam views -- anyone care to interpret?

Attached is fhaz animation, I'm working on rhaz.

Rear Haz
Click to view attachment

Doesn't look too bad.
I think that there's a view from the MI that might show something. Will upload in a second.

I can see quite some differences in LF .
There are little agregates (?) on the rigth of the soil in front of it that end up near the edge of the soil in front of it while they were further (sorry, I miss words to be precise enough in my description); And, I agree Eduardo, I can see it (LF) moving up quite a bit. It seams that some soil in front of LF is now under it, i.e. the weel went up a bit.

MIs sol2088-2090 view to LR wheel.
Click to view attachment

It suggest to me that the right-middle wheel maybe lifting.
There are some pancams of both middle wheels that need to be compared.

Regardless of the horizon, in the last drive the front left cleared a lot of sand and the front right pushed forward noticeably in relation to that small rock next to its front tread.

Centred MI view and Navcam view.
Click to view attachmentClick to view attachment

Do I see the pointed rock fading into the distance?

This looked extremely promising, though I guess we'll have to wait upon the official word on how they thought it went.

My first contribution:



Movement for sure!

Looks encouraging!

Seeing the stuck right wheel slide made me wonder if the tread on the bottom of the wheel has worn off by now? Has this been discussed? According to this quote I found on a NASA site "At 26 centimeters in diameter (a little over ten inches), these aluminum wheels are twice the size of those on Sojourner" it appears the wheels are aluminum which is soft. Spirit has been dragging the wheel for a while and sometimes over hard material such as on home plate. Is it possible that the tread on the bottom is now completely worn smooth? Did NASA use a worn tread right front wheel when doing sandbox tests? (I suppose if the treads are caked with material anyway it might not make much difference, but for future mobility over hard material it could be adventageous if the tread is now gone) (at the very least it is good the wheel froze without the criss-cross-lander-attachment-point-tread pointing down!)

Had the same thought.

There appears to be a 1 pixel drop on the left side of the horizon in the front hazcam views between 2088 and 2090. That corresponds to a pitch up at the front by about 0.1 degree. There's a corresponding lift of the right side of the horizon in the rear hazcam views. Considering that that may have more to do with rocker/bogey than wheels, that sounds like nothing to write home about to me.

The movement of the horizon in the front hazcams appears to be mainly a shift to the left. That means the rover is yawing clockwise as viewed from above, which is what you'd expect with the dead RF wheel. Again though, the yaw angle is very small.

MIs are consistent with us moving forward. It's really good to see LF eating into the soil in front of it, and the RF make progress too.

Tweaked rear hazcam comparison, 2088 to 2090:
Click to view attachment