FREE SPIRIT - Extrication FAQs
FREE SPIRIT - Extrication FAQs
Jul 2 2009, 11:34 AM
Joined: 8-February 04
Member No.: 1
Thanks to Paolo & Astro0 for contributing / editing / writing this.
FAQ - GETTING SPIRIT UNSTUCK
UPDATED AS AT 2300UTC - 21 JANUARY 2010
Due to the number of 'repeat' Frequently Asked Questions in the 'Spirit Getting Unstuck' thread, here is a compiled list of those FAQs. We recommend that you read them BEFORE asking something that has been answered previously.
BUT FIRST - AN IMPORTANT POINT THAT YOU SHOULD CONSIDER BEFORE ASKING ANYTHING:
"They will test what they can see, either from imaging and telemetry. They cannot simulate what they cannot see."
HOW DO THEY COMPENSATE IN THE TEST BED AT JPL FOR THE .38 GRAVITY ON MARS?
There are two vehicles they can use for testing:
SSTB1 (Tee-Bee) a full size replica of the MER vehicles, minor some minor differences (no solar panels, some temperature probes are missing...) which of course has the same mass as MER and higher weight on Earth than MER has on Mars;
and SSTB Lite, a stripped down vehicle with same wheel size, actuators and suspension system, same WEB size but major components like the IDD and others are missing. This vehicle has a weight on Earth that is similar to the weight of MER on Mars. For Purgatory and for this event they are going to use the SSTB1, not the SSTB Lite. During Purgatory testing it was found that the SSTB1 behaved more similarly to the MER vehicles, possibly because both the SSTB1 and the soil were subject to the same gravity vector.
HOW DO SIMULATE THE SOIL THAT SPIRIT IS STUCK IN?
As soil simulant they have settled on is 1 part of Perma-Guard food grade diatomaceous earth and 2 parts of Lincoln 60 Fire Clay (by weight).
The reason they chose DE is because it is a fine powder, it does not ignite and is readily available. The feel from the pool grade DE is different from the food grade. The first is more abrasive, the second one is more chalky. There is absolutely no evidence that the material on Mars is DE of course.
They really can't try to replicate the exact soil characteristics that have been measured with spectrometers and MIs but try to replicate how the vehicle behaves during the embedding event and hope it will be representative enough for testing the extrication maneuvers.
They have used some of the information recevied on particle size, but ultimately the vehicle is the best instrument they have to select the soil simulant.
DOES THE CONSISTENCY OF THE SOIL CHANGE WITH TEMPERATURE?
COULD THEY DRIVE AT NIGHT WHEN IT IS COLD AND THE SOIL FROZEN - the temperatures on Mars are cold anyway - certainly colder at night. More data needs to be collected, but driving at night has been considered.
HOW CAN THEY BE SURE THAT THE SIMULATIONS IN THE TEST BED WILL ACCURATE?
They try to reduce the variables as much as they can. Although the soil under the right side is different from the soil under the left side of the rover, they have a soil compound that can reproduce the two soil mechanical behaviors depending on how the simulant is tamped down, and this can be reproduced consistently. Since the soil is all the same there is no need to keep two different soil simulants separated. The sandbox is in an air-conditioned area and the rover will be placed in the same configuration each time in the 8'x12'x2' box set at about 12 deg roll. They don't know how accurate the testbed simulation will be. One thing that can be done is to issue the same commands that were executed on Mars and compare the testbed results with the data from Mars. Moreover, what the testbed will be used for is a differential study. They just need to compare the different extrication strategies that have been through of and find out the best one. They do not need 100% accuracy, just enough fidelity to find the most efficient maneuvers and the one to avoid.
WHAT POSITION DID THE ROVER START ITS EXTRICATION ATTEMPTS IN?
Brief description of the rover attitude:
1) the rover is aligned more or less north-south (front of the rover north)
2) the rover is on a slope about 10-12 degrees, left side of the rover is lower than the right side
3) pitch is almost zero
The soil under the left side is cohesionless, the soil under the right side seems to provide more traction.
The two middle wheels are only partly embedded, the RF is on top of the surface, LF, and rear wheels are fully embedded.
I HEARD THAT THE ROVER IS SITTING ON A POINTY ROCK. WHAT WILL THAT DO?
There was some concern that the rover was "high-centered" which usually means that the vehicle's center of gravity is being supported on something other than the drive surfaces, so there's little or no traction.
There is a rock underneath the rover which is just touching the underside of the main body. It is thought though that this rock is not fixed on solid ground and will probably move as the rover shifts.
COULD THE ROCK PUNCTURE THE ROVER AND CAUSE DAMAGE?
If the body of the rover was punctured, there is about 1-inch between the shell and any internal electronics/mechanisms.
DOES THE SOIL CAKING ON THE WHEELS CAUSE A PROBLEM?
When the wheels are caked, they don't sink. It is clean cleats that cause soil to be removed from underneath and therefore you get sinkage. Moreover, when the wheels are caked, soil from the front of the wheel gets moved under the wheel, not behind (there is not enough friction to move it up from underneath the wheel). This causes lift, raising the wheel. When the wheel raises above the hubcap, cleats start to clean and get more traction.
CAN THE ROBOT ARM (IDD - Instrument Deployment Device) BE USED TO RESCUE THE ROVER BY...
UPDATE - As for other techniques to consider for extrication, the rover team has examined the two options that would use the robotic arm: pushing with it and re-sculpting the terrain by the left-front wheel. The assessment of pushing with the arm reveals that only about 30 newtons of lateral force could be achieved, while a minimum of several hundreds of newtons would be needed to move the rover. Further, such a technique risks damaging the arm and preventing its use for high-priority science from a stationary rover. The other technique of re-sculpting the terrain and perhaps pushing a rock in front of or behind the left-front wheel is also assessed to be of little to no help and, again, risks the arm. There is also a large risk of accidentally pushing the rock into the open wheel and jamming.
PUSHING - The IDD is about 2-3 Kg and about 1m long, the rover is about 185Kg. You do the math. The wrist/turret are much much weaker - but the Elbow can manage 20 Nm force, it can handle 40 Nm static, the Shoulder's two motors about 45 Nm.
If you could brace the IDD into the ground, you could apply, say, 30 Nm at the shoulder without getting near the static limit of the elbow. You would have only about 60 N of force at the IDD attachment point though - as each part of the arm is about half a meter long. The rover, at 185kg, has a total force from gravity of about 697N. So the arm would only ever be able to take about a tenth of the weight of the rover.
PULLING - Same as above.
TIPPING - Changing the Centre of Gravity of the rover by extending the arm thereby shifting the weight on the wheels. See above. They are going to try the idea in the test bed though and see what happens.
DIGGING - The arm has a limited degree of motion and it has no 'shovel' to use. The IDD could only push against soil (which it has done in a downward motion when analysing soil), however it does not have the strength to move soil or rocks anywhere and a more likely scenario is that you would clog the key 'science' instruments and render the IDD useless. Moving aside soil could also snag the arm on a hiddden rock and that would be even worse.
USING IT TO MOVE ROCKS UNDER THE WHEELS TO GAIN TRACTION?
Not impossible, but again the IDD isn't that strong. Lots of little rocks maybe, that unfortunately are not readily stacked about anywhere.
USING THE IDD FOR TRACTION - The rover cannot drive while the IDD is in motion at the same time. Consider this also, if the IDD was lodged against the ground, driving over it, the wheels could get a good grip - BUT, then the IDD (the main science instrument) would be useless.
DRIVING TECHNIQUES TO RESCUE THE ROVERS?
PUTTING THE ROVER IN REVERSE AND DRIVING 'OUT' OF THE SAND TRAP - This has worked before at other locations when both rovers have been 'bogged', but the soil here is very different and needs further analysis. It will be tried at some point.
TURNING THE WHEELS - The wheels are capable of turning, but the maximum is 60deg. They did turn them 30deg and drove forward and made some slowly diminishing progress. Turning in place was tried (clockwise) but was unsuccessful. They might try to revisit this technique.
DRIVING THE ROVER FORWARD (FAST) - A car or truck is not really representative of the type of motion they have on the rover. It is mostly a sequence of quasi-static moments rather than a dynamic event.
TURNING THE WHEELS SLOWLY - the wheels already turn slowly. During each command they ramp up quickly to the steady speed that is required for the motion and just prior to the end they ramp down. Just like any servo motor. But during an arc, say to the left, the left side wheels move slower than the right side, simulating a mechanical differential if you want.
They typically "lock the differential" so to speak since there is one motor per wheel. They can control each wheel independently but that has to be preprogrammed. They cannot control the wheel torque dynamically though. For that they would need to sequence one command, downlink the telemetry, see the wheel current (proportional to torque) and change the parameters accordingly. That would be slow, but possible.
FREE-WHEELING - No. The wheels cannot be set to free-wheel either individually or in combination.
VIBRATING THE WHEELS - The gear ratio in the steering actuator makes it impossible to vibrate the wheel with any significant amplitude.
TWISTING THE WHEELS LEFT AND RIGHT (scooping soil and gaining traction on a new surface) - This has already been tested. Wiggling the wheels causes the rover to sink in fast.
ROCKING THE ROVER - Unfortunately, the interval between commands is one second. That limits the response time. In addition, sensors are queried every 1/8 sec. Way too slow for any real-time control of the vehicle. Some serious coding would be needed, tested on the ground, uploaded on the vehicle and verified on Mars before they would even attempt to use it.
TRYING THE STUCK RIGHT-FRONT (RF) WHEEL TO SEE IF IT WILL SUDDENLY WORK - Believe it or not, that is actually already on the list of things to try, but not at the top - in fact it's so far down the list that they can't see it. In order to save mass and reduce complexity, some of the actuators (~=motors) share motor controllers (electronics), for example IDD and mobility. This means that if an attempt at applying voltage to the RF wheel results in damage the motor controller they potentially lose the use of another actuator. In addition, given the fact no-one is 100% sure about the type of fault on the RF actuator, sending power down a line that might be shorted some unknown place might ultimately send power to a device that might get damaged. What MIGHT be tried is sending a voltage that is low enough not to cause damage but high enough to verify whether the motor is OK or not.
In addition, and this might be stating the obvious, one advantage of not turning the RF wheel is that it won't sink! And in the current predicament it is quite helpful.
DRIVING TECHNIQUES BEING ASSESSED ON THE TEST BED ROVER AT JPL
As a first step, the Rover team is testing at least 11 driving manouevers on the Test Bed Rover (also known on UMSF as TeeBee).
None of these techniques have been tried on Spirit yet. They are simply assessing each technique at this stage.
Manouevers they want to use include driving:
1) Straight forwards; 2) Straight backwards; 3) 'Crabwalk' - driving upslope with the wheels turned at 60 degrees to the right; 4) 'Crabwalk' - driving upslope with the wheels turned 20 degrees to the right; 5) 'Crabwalk' - driving downslope backwards with the wheels turned 60 degrees to the right; 6) 'Crabwalk' - driving backward with wheels turned 20 degrees to the right; 7) a tight forward right arc; 8) a clockwise turn in place; 9) a counterclockwise turn in place; 10) 'Crabwalk' - driving forward with wheels turned to the left; 11) driving while steering. Some of the maneuvers might be repeated.
COULD THEY FOLD AND UNFOLD THE WHEELS LIKE THEY DID AT THE START OF THE MISSION?
No. The wheels are locked in place. Their only movement comes from the steering actuators and rotation motors in each wheel and the movement of its rocker-bogie suspension system.
WHAT ARE THE TIME LIMITS FOR TAKING PICTURES WITH THE ROVER CAMERAS?
According to the Pancam technical paper, the integration time for the CCD's can be between 0 - 335 seconds at 5.12ms steps.
|Lo-Fi Version||Time is now: 18th May 2013 - 05:29 AM|
RULES AND GUIDELINES
Please read the Forum Rules and Guidelines before posting.
OPINIONS AND MODERATION
Opinions expressed on UnmannedSpaceflight.com are those of the individual posters and do not necessarily reflect the opinions of UnmannedSpaceflight.com or The Planetary Society. The all-volunteer UnmannedSpaceflight.com moderation team is wholly independent of The Planetary Society. The Planetary Society has no influence over decisions made by the UnmannedSpaceflight.com moderators.
SUPPORT THE FORUM
Unmannedspaceflight.com is a project of the Planetary Society and is funded by donations from visitors and members. Help keep this forum up and running by contributing here.