MSL EDL Options

[elakdawalla]

As far as I understand it, the spin of the S/C under parachute (so-called "wrist mode") is actively damped by thrusters.

ChrisC: There were a few hours' downtime today that shouldn't be repeated due to an IP address swap or something or other.

[nprev]

Almost certainly maintenance. It's in the extremely talented hands of one of our Admins.

[Guest_Oersted_*]

As far as I understand it, the spin of the S/C under parachute (so-called "wrist mode") is actively damped by thrusters.

Yes, that is correct. It's described in a section of this (rather hard to find) paper by Steltzner et al:

Mars Science Laboratory Entry, Descent, and Landing System Overview
http://www.google.com/url?sa=t&rct=j&a...cpdnJsZGq91vK3w

P.8: Wrist Mode Damping
Estimating the oscillatory behavior of an entry capsule suspended underneath a parachute is an extremely dynamic and complex problem. Initial conditions at parachute deployment can pump large energies into the capsule wrist mode (rotation underneath the parachute about the capsule center of gravity) which will decay with time. Historical attempts to bound the wrist mode behavior and its time evolution following parachute deployment have failed to bound the behavior observed during flight (e.g. MER-B ).
While the MSL team believes we have attained a deeper understanding of wrist mode dynamics and the energies involved, the sensitivity of subsequent EDL events to high wrist mode energies led to the inclusion of an active wrist mode damping mechanism using the RCS thrusters. In cases where the capsule wrist mode frequency violates the “safe” flight envelope, RCS thrusters will fire to reduce the wrist mode frequency to acceptable levels.
Wrist mode damping is active throughout parachute descent and ensures a safe heatshield separation, good TDS surface acquisition, and a safe backshell separation.

[Phil Stooke]

It's happening again in the run up to MSL... journalists pontificating about the percentage of Mars probe failures.

http://www.theverge.com/2012/8/2/3215863/m...missions-failed

http://www.wired.com/wiredscience/2012/08/...-to-reach-mars/



I thought a look at the data might help. I listed missions by country of origin and success or failure, allowing a partial success category as well. But even doing this gets complicated. I have listed Mars Express and Beagle 2 as separate because it seems unfair to categorize Mars Express as a partial success. I put Mars Polar Lander and Deep Space 2 as separate, for no very good reason. So should all the components of Mars 96 be listed separately? Is Viking two successes or four? And to be more realistic Mars 2 should be counted as a total failure.

Even so, you can see that most failures are Soviet, most successes American. Most failures are early, most successes recent. Most journalist claims are literally true but misleading to make a good story. The worst point is the one about 30 percent of landers succeeding. Six out of seven US landers worked, so to get the 30 percent success rate you have to count every separate lander on every Soviet mission, including launch failures where the lander was never put to the test.


Phil

[RoverDriver]

One thing they don't understand is that we have something nobody else has: ROB MANNING!!!

Paolo

[Hungry4info]

From MSL, Mars is now 0.561° in angular diameter -- larger than the Moon from Earth (0.507° average).

[B Bernatchez]

Are we there yet?

[RoverDriver]

Are we there yet?

Yes we are. Start packing your toys and stuff, we are almost home.

Paolo

[ElkGroveDan]

Present your E-tickets tickets to the lady at the turnstile. Fasten your seat-belts and stand clear of the closing doors. It is important to secure all loose objects, and please keep hands and arms inside at all times.

[John Copella]

So, no L-2 TCM, I'm assuming? (saw no news of it, but maybe I missed it)

[RoverDriver]

...
and please keep hands and arms inside at all times.

You forgot the turret as well. Sorry, couldn't resist.

So, no L-2 TCM, I'm assuming? (saw no news of it, but maybe I missed it)

Apparently they are so dead on that they don't need to do any corrections.

Paolo

[Deimos]

It's happening again in the run up to MSL... journalists pontificating about the percentage of Mars probe failures.

Blame is at least in part misdirected -- that narrative is coming out of NASA HQ. Reporters doing this, I can at least understand. And NASA making sure people understand there are risks, I can understand. Jim Green pushing the 40% story--that seems like a disservice. Christian Science Monitor was working on a story -- the reporter seemed to be calling BS on that narrative.

[jmknapp]

In the press conference going on right now, a reporter tried to pin the panel down on the estimate of the odds, but no one bit. Doug McCuistion, director of the Mars program, essentially said there were too many variables and no way to put a meaningful number on it, and Richard Cook, flight operations manager, said that they've done lots of Monte Carlo simulations of what might happen during EDL, but always assuming that the hardware works.

Since NASA was able to come up with risk assessments for the Shuttle (what was it, something like 1 in 100?), not sure why a ballpark number couldn't be put on this one.

[mcaplinger]

Since NASA was able to come up with risk assessments for the Shuttle (what was it, something like 1 in 100?), not sure why a ballpark number couldn't be put on this one.

IMHO, such figures for complex systems that aren't tested a statistically-meaningful number of times are essentially meaningless. IIRC estimates of shuttle reliability before Challenger varied by several orders of magnitude; the 1:100 was just the one that turned out to be right.

Also, consider that many failures could be caused by bad software.

Reliability engineers argue that the
correctness of a software product is not a probabilistic phenomenon.
The software is either correct (reliability 1.0) or incorrect
(reliability 0). If they assume a reliability of 0, they cannot get a
useful reliability estimate for the system containing the software.
Consequently, they assume correctness. Many consider it nonsense to
talk about "reliability of software."

CACM, 33, 6 (June 1990) page 643.

[diane]

Reliability engineers argue that the
correctness of a software product is not a probabilistic phenomenon.
The software is either correct (reliability 1.0) or incorrect
(reliability 0). If they assume a reliability of 0, they cannot get a
useful reliability estimate for the system containing the software.
Consequently, they assume correctness. Many consider it nonsense to
talk about "reliability of software."

Since software reliability is something I do know about...

Limiting reliability assessments to either 1 or 0 is absurd; in such a case, the reliability must be assumed 0. Complex software systems will not be defect-free. The appropriate questions are whether sufficient testing has been done, whether defects are encountered during a particular operation (EDL, for example), whether the defect causes significant problems, whether there is sufficient error detection and recovery in place, whether redundancies are in place, whether failover mechanisms are in place (and function correctly), whether a failure in one module can affect correct operation of other modules, etc.

I used to work on programming for public telephone switching systems. The watch-phrase was "belt and suspenders" - always have a backup system in place and ready to take over immediately.

I trust that the MSL software engineers have done their work well, and I look forward to great celebrations on Monday.