Help - Search - Members - Calendar
Full Version: Skycrane
Unmanned Spaceflight.com > Mars & Missions > MSL
Pages: 1, 2, 3, 4, 5, 6
cIclops
In his recent long piece on spacedaily.com Bruce Moomaw reports:

... JPL's Pete Theisinger and Mike Meyer (chief engineer and scientist of the 2009 Mars Surface Lab) confirmed that the current plan is to have MSL land using a strange, Rube Goldbergian system called "Skycrane".

Skycrane is a very innovative landing system and now that MSL appears to be planning to use this approach there are lots of open questions, such as: How exactly does it work? What advantages does it have over legs, airbags or pallets? What the heck is it anyway?

Some background info:

JPL note

Space.com article
djellison


Essentially the whole things enters in an aeroshell as usual - drops the heatshield, drops the backshell and enters a viking like terminal des. phase - but comes to a hover about 5m off the ground. The paylod ( rover in this case ) is then dropped down on cables, the cables are cut, and the crane flys off.

Some people think it's stupid and will never work

10 years ago - people thought the same thing regarding the three-body chute-backshell-airbag mechanics for MPF and MER smile.gif

Doug
remcook
If you look at planetary missions, and especially landings, the scenarios always look highly unlikely to me. The margins are terribly small (for instance reentry angle, or time to burn your rocket). But somehow most things just work smile.gif
cIclops
QUOTE (remcook @ Mar 12 2005, 11:49 AM)
If you look at planetary missions, and especially landings, the scenarios always look highly unlikely to me. The margins are terribly small (for instance reentry angle, or time to burn your rocket). But somehow most things just work smile.gif

They "just work" because of the enormous amount of effort, experience and knowledge put into their design, construction and operation. The MERs cost more than $800 million and took several years to build based on the pathfinder design, which in turn was based on Viking, and in its turn based on Surveyor and Ranger technology.

Most landers haven't worked but much was learned from those failures. For example MPL, Deep Space 2 as well as several Surveyor and Ranger lander probes. And I won't mention the numerous Russian lander failures ...
remcook
obviously.

that was kind of my point: it's amazing what people can do. So although the skycrane looks very difficult, I wouldn't say it is impossible. There will be an Earth-based test at some point I suppose.
dvandorn
QUOTE (cIclops @ Mar 12 2005, 05:07 PM)
Most landers haven't worked but much was learned from those failures. For example MPL, Deep Space 2 as well as several Surveyor and Ranger lander probes. And I won't mention the numerous Russian lander failures ...

Since you don't want to get into Soviet/Russian lander failures, looking at just American attempts, there have been three loss-of-vehicle failures in thirteen attempts at lunar/planetary soft landings(*). 10-3 is a pretty decent track record.

Now, if you add in American attempts at hard landers, at which we're zip out of five(**), you're still talking seven failures out of seventeen attempts. 10-8 is still a winning record.

Finally, if you include American manned soft-lander attempts into the total -- one failure (in which the landing itself wasn't even attempted) out of seven tries(***) -- you achieve a whopping sixteen successes versus only nine failures. Winning nearly twice as often as you lose will get you thrown out of any casino in Vegas. And if you look just at American soft-landers, both manned and unmanned, the record is a phenomenal 16-4.

So, I'm thinking maybe NASA is a little better at landing probes on other planets than you suggest... *smile*... At the very least, "most landers haven't worked" is an inaccurate assessment of American lunar/planetary landing attempts.

All of that said, I agree with your main point. We learn from our failures just as much as we learn from our successes. In fact, we learn the things we want to learn from our successes, but we more often learn the things we need to learn from our failures...

* - American soft-lander failures: Surveyors II and IV, Mars Polar Lander. Successful soft-landers: Surveyors I, III, V-VII; Vikings 1 and 2; Mars Pathfinder; MERs A & B (Spirit and Opportunity).

** - American hard-lander failures: Rangers 3-5; Deep Space 2 (two hard landers). No successful hard-landers.

*** - American manned soft-lander failure: Apollo 13. Successful manned soft-landers: Apollos 11-12, 14-17.

-the other Doug
tedstryk
Of course, the Pioneer Venus Day Probe could be considered an accidental success at a hard landing.
OWW
You could say the Galileo probe was a successful 'lander' too. smile.gif And what about NEAR on Eros?

Back on topic. What I worry about with that skycrane landing is the wheels of the rover. Are they protected in some way at the moment of touchdown? It seems to me only the slightest horizontal velocity will tear a wheel off or at least bend it into an unusable piece of metal.
djellison
I think the idea is that the crane will hover stationary - although anyone who's ever flown a helecopter simulator, or been lucky enough to fly a real helecopter will know - it's damn hard to hover on the spot in a cross wind

Doug
cIclops
QUOTE (djellison @ Mar 13 2005, 02:22 PM)
I think the idea is that the crane will hover stationary - although anyone who's ever flown a helecopter simulator, or been lucky enough to fly a real helecopter will know - it's damn hard to hover on the spot in a cross wind

Doug

Yes but the Martian atmosphere is very thin and MSL/Skycrane won't have much surface area so the jets should easily be able to compensate for drift during the short hover time. The really hard part IMO will be ensuring that MSL is placed in a safe location.
djellison
The atmosphere was thing - yes - but the wind speeds can be very high - and I can imagine that Skycrane would still offer considerable X-section.

Take MER - Granted - much bigger X-Section whilst being much lighter - but some quite high speeds before using DIMES and the sideways motors before touchdown

Doug
BruceMoomaw
Theisinger says that winds are not a problem -- even Martian winds of 50 kph or so produce almost no horizontal effect on the rover. The real worry is slight instabilities in the attitude control and horizontal hover stability of the lander itself, and the possibility of a slightly uneven unlatching of the rover from the lander's bottom -- any of which could set the rover swinging. The ground tests this year will be absoltuely crucial on whether to switch to a pallet lander design instead.
cIclops
QUOTE (BruceMoomaw @ Mar 13 2005, 08:17 PM)
The real worry is slight instabilities in the attitude control and horizontal hover stability of the lander itself, and the possibility of a slightly uneven unlatching of the rover from the lander's bottom -- any of which could set the rover swinging.

Holding the rover tight to the frame of the Skycrane by the winch cable should damp any oscillation during unlatching.
OWW
By the way, why lower the thing from a tether in the first place? Wouldn't it be less risky to keep the rover attached to the skycrane and at the moment of touchdown fire some pyros and let the crane fly away? Lots of dust on the rover, but that's no problem since it's nuclear powered.
This way the final descent will be more precise and you keep the advantage that the rover is on the surface immediately.
BruceMoomaw
Ciclops may have a point, but Theisinger did tell me he was worried about instabilities during unlatching. As for ObsessedWithWorlds' idea (can I just call you OWW from now on?), I imagine the problem is that the lander -- unless it had control of its descent speed during the last few feet of its descent too fine to be feasible with any confidence -- might actually end up pounding the rover into the surface with its own weight, thereby damaging its wheels or suspension, before unlatching from the rover and flying back up off the surface.
erwan
Such a conception with a long cable, whose lenght may be adjusted (radar sensor...) isn't an additional way to control the descent velocity of the rover, on the last meters before touchdown?
Chmee
My concern with the Sky Crane concept is that it is not as robust in landing in uncertain terrain. With the MER airbags, almost regardless of where the airbag first lands, it will roll to a more steady/level position. It can be used in a variety of terrains.

Contrast with the Skycrane where landing on the side of a crater or large rock would leave the lander/rover in a dangerous position. This nearly happened with Viking 1 with a very large rock nearby. I can see the Skycrane working well in Meridiani Planum, but would likely fail at the Mars Pathfinder site.

I understand that this new landing concept allows a much larger payload, but I think the risk is much great with such a system.
cIclops
Yes. The airbag system worked well and can cope with a more rocky environment. I think MSL will need some scene analysis capability to select a clear landing area or it will have to be restricted to safer and probably less interesting terrain.

Given the importance and cost of MSL, NASA will be looking to reduce risk not increase it which will mean a safer landing system as they want to explore more dangerous areas. It will be fascinating to see how they solve the problem smile.gif
Chmee
Of course, I just thought of this. They could land in a relativley safe area and DRIVE to the rougher area ohmy.gif . Heck, the MSL is nuclear powered so it should be able to go much farther (maybe faster?) the the MERs.

But trying to figure out what a relatively safe area to land in from orbit has proven tricky.....
djellison
QUOTE (Chmee @ Mar 17 2005, 10:03 PM)
But trying to figure out what a relatively safe area to land in from orbit has proven tricky.....

Until MRO gets there
smile.gif

Doug
BruceMoomaw
Well, the airbag system has turned out to have a large array of very serious problems -- which became apparent as soon as Pathfinder flew. (It was reused for the MERs only because those rovers had already been designed and, on balance, it was the cheapest way to get them down to the surface.)

(1) It's alarmingly heavy -- something that became apparent on Pathfinder when they had to use 5 bag layers rather than the two planned one, to prevent the bags from ripping on impact. (The MERs, with their higher weight, produced even bigger troubles in designing non-ripping bags.) It is, in fact, a far heavier system than the more complex full soft-landing systems -- which makes a big difference in launch vehicle costs.

(2) It's extremely vulnerable to dangerous crosswinds. Without that Rube Goldbergian DIMES system that was added literally at the last minute to partioally cancel out horizontal velocity, Spirit might very well not have survived its landing -- and even with that system, this very seriously restricts an airbag system's variety of landing sites.

(3) It does not allow anything like a pinpoint landing, simply because after touchdown the lander is likely to roll for hundreds of meters across the surface (and sometimes over a kilometer). This will make a lot more difference in the near future, when we start trying to make pinpoint precision landings to reach otherwise unattainable but scientifically valuable landing sites immediately surrounded by dangerously rugged terrain. (The fact that a guidance system will be added to Mars landers, starting with MSL, to allow them to actively identify and dodge patches of dangerously rugged terrain during landing also decreases the relative utility of an airbag system in that respect.)

So the widespread trumpeting about the supposed brilliant success and high reliability of the airbag system that we keep hearing from the press is -- as with so many space- and technology-related issues -- based on serious ignorance. (Another spectacular example is the endless blather we keep hearing about the supposed superiority of repairing Hubble as opposed to simply launching a new one -- but that's another story, which I won't get into now.) Regardless of whether Skycrane itself finally proves practical, the sooner we switch to full-fledged throttled-rocket soft landing systems and away from airbags for all but the smallest Mars landers, the better.
centsworth_II
QUOTE (BruceMoomaw @ Mar 17 2005, 10:16 PM)
...the sooner we switch to full-fledged throttled-rocket soft landing systems and away from airbags for all but the smallest Mars landers, the better.

And if manned missions are the ultimate goal, we might as well start getting experience with powered landings. Can you imagine a manned landing using air bags?
djellison
QUOTE (centsworth_II @ Mar 18 2005, 06:50 AM)
Can you imagine a manned landing using air bags?

Well, Hollywood can. smile.gif

Doug
erwan
Bruce: thanks for your interesting last comments; i understand airbags are now an "out to date technology", adapted for landing of the first martian rovers generation, but not for the future missions; reading the JPL notes Cyclops linked on the first post of that thread, i think we may add to the disavantages you commented:
-Airbags add big volume/geometric constraints on the payload (the tetraed geometry, plus all the harware needed on MERs to deploy before egress)
-maybe the most important one: JPL's engineers argue they couldn't conceive airbags for heavier payload than MER+landing platform.
And obviously your comment, Centsworth II...
djellison
Put it this way - the roer was around 185kg. the entry-vehicle (minus cruise stage) was around 800kg

Not ideal really

Anyone got figures for Viking?

Doug
chris
From http://nssdc.gsfc.nasa.gov/database/Master...og?sc=1975-083C

The fully fueled orbiter-lander pair had a mass of 3530 kg. After separation and landing, the lander had a mass of about 600 kg and the orbiter 900 kg
erwan
Viking entry vehicle (lander + heatshield + chute) was 1193 kg for a 598 kg (fueled) lander.
erwan
Chris: our sources are the same; orbiter weight you reported is dry; the orbiter carry around 1400 kg fuel. Thus fueled orbiter account for ~2300 kg in the total mass of 3530kg; the remaining 1193 kg = entry vehicle...
djellison
The killer with the Airbag system is the 300+ Kg of lander structure that serves no purpose after deployment I guess.

Doug
erwan
Sure you're right, Doug; it's pleasant for me, since the beginning of that thread, to probably better understand the need for skycrane / new EDL technology.
Chmee
Of course with the Skycrane, if it really is going to hover around and locate a safe place to drop MSL, would need a significant amount of fuel capacity and this would add much weight. So what does the EDL to Rover weight ratio then become? Maybe as bad as the MERs

I realize that the Airbag system has reached its limit for mass, but it is hard to argue with success: 3 for 3 attempts with very different landing condiitons.
djellison
QUOTE (Chmee @ Mar 18 2005, 06:31 PM)
I realize that the Airbag system has reached its limit for mass, but it is hard to argue with success: 3 for 3 attempts with very different landing condiitons.

I dont think anyone is saying "airbags are crap" - I think people are saying that it's simply impossible to scale it up to the capacity needed for more able missions. Not only that, but the Airbag system is highly limited as to the terrain it can visit. Skycrane or similar isnt going to be as limited.

Doug
lyford
Well, call me old fashioned, (or a broken record - which is old fashioned now that I think about it...) but I still have soft spot in my heart for the Tumbleweeds. I hope they make a fleet of these for a Scout mission, but I doubt it'll happen. Of course you would need the communications orbiter network up and running first, but it's the ultimate efficiency: the airbag IS the rover - no lost weight!

Of course, it's the very antithesis of a controlled landing and directed mission since it's at the mercy of the winds, but for mapping, magnetronomy, sounding and "ground truth" observations of harsh areas it may be just the ticket.
lyford
Oooh - found a better Tumbleweed site.

Apparently I am not the only one thinking these fellas might work. Lots of good papers in the resources section....

And quite a selection:


Now I will go back on topic and kvetch about the Skycrane.... smile.gif
BruceMoomaw
To Chmee: The landing hazard-avoidance system they're planning for every (fairly big) Mars lander starting with MSL doesn't "hover around and look for a safe pladce to drop MSL": it starts examining the roughness and slopes of the terrain even while the lander is still dropping via parachute, and then as soon as the chute is cut loose and the soft-landing engines start up it angles toward the safest-looking spot -- readjusting itself later if any new indications come in. Thus it utilizes far less fuel than if it came down and then started moving sideways looking for a safe landing site -- and the total weight of fuel used is still far less than the weight of an airbag system.

By the way, the US has been toying with the idea of a descent-camera controlled harzard avoidance system for a very long time -- starting with plans to put one on any Viking followups. One was originally planned for Phoenix, before it was decided that the system was too heavy and power-consuming for that mission, since it will land on a very flat and obstacle-free kind of terrain anyway. And the Selene-B lander that Japan is tentatively planning as a followup for their big 2006 Selene-1 lunar orbiter would be flown largely to test such a system. But MSL will utilize a more complex type using high-resolution phased-array radar, which can identify not only rough terrain but dangerously steep slopes of the type that a camera system might overlook.
remcook
about the tumbleweed: I always wondered how you can make sensible pictures from these things. It must be one hell of a job to keep it stable and the outside must be transparent. It's not made for picture, but a thing that makes pictures is simply much easier to sell.

and what usefull things can it so except measure pressure and temperature? Nothing with a high data rate, because you won't be able to put on a high gain antenna. It's also a though job to figure out where it is (if you don't have pictures even more). You could have a sunsensor but the stability is an issue again. To obtain position accurately you'd need a GPS-like network of 3 satellites and you will only get a rough estimate.

So it's good to study dynamics, but there are maybe easier ways.

the site mentions these stop-and-go concept which seem better for pictures and higher data rates, but how would you keep your instruments from falling apart?

I think I'm more in favour of balloons cool.gif
lyford
QUOTE (remcook @ Mar 19 2005, 02:26 AM)
about the tumbleweed: I always wondered how you can make sensible pictures from these things.

There are designs that allow the bags to deflate and when they want to stop and extend a sensor mast periscope style and take pix and or readings. The instruments are suspended in the center in a shock absorbing cage.

I will post more detail over in this new thread. Just trying to keep the signal to noise ratio high. tongue.gif

PS. Balloons is good too. wink.gif

And now let us speak of it no more here... back to your regularly scheduled MSL coverage.
DaveM
Many of those Tumbleweed concept images were created in my undergrad senior design project. We built a 2 m diameter scale model of the "boxkite" design called TED (Tumbleweed Earth Demonstrator), though based on our calculations the real one would have to be 6 m in diameter. We even fitted a camera on TED and got some fairly decent images, though for science it would be quite limited and not very predictable. It was fun seeing that thing roll around when the wind blew, though. It was a fun concept to work on, but I kind of like balloons myself. Though you have to have enormous balloons to be able to carry a significant payload mass. One of the advantages of Tumbleweeds is that the mass constraints are not as bad as on balloons.


mars.gif+wheel.gif
cIclops
QUOTE (BruceMoomaw @ Mar 14 2005, 12:29 PM)
I imagine the problem is that the lander -- unless it had control of its descent speed during the last few feet of its descent too fine to be feasible with any confidence -- might actually end up pounding the rover into the surface with its own weight, thereby damaging its wheels or suspension, before unlatching from the rover and flying back up off the surface.
*


Armadillo Aerospace have recently demonstrated a very precisely controlled landing technology which shows that it is feasible. See this test flight video (7MB)
lyford
QUOTE (cIclops @ Mar 25 2005, 09:57 AM)
Armadillo Aerospace have recently demonstrated a very precisely controlled landing technology which shows that it is feasible. See this  test flight video (7MB)
*


Fake! You can see the strings! (A joke! I kid the fine people because I love...) laugh.gif
This reminded me of all the 50's sci fi movies where the big V2 style rocket backs itself down on to the moon...

Obviously they have success with a slow landing from a hundred feet on a flat pad (I couldn't find any info on the guidance control or nav they are using... GPS?), but I think the kicker is going to be the real time hazard recognition - finding a flat landing zone and going to it autonomously before the fuel runs out...

But I am glad to see the hours I spent playing Doom have led to actual benefits! tongue.gif

And now for something completely different. (2.3 MB Quicktime)
erwan
Hi Lyford: impressive link. Remenbering issues with LEM demonstration vehicle - Armstrong aboard - before Apollo 11. Finally everythings goes perfectly with the spider vehicle for successive Apollo missions!

And now for something completely different. (2.3 MB Quicktime)
*

[/quote]
lyford
QUOTE (erwan @ Mar 25 2005, 11:04 AM)
Hi Lyford: impressive link. Remenbering issues with LEM demonstration vehicle - Armstrong aboard - before Apollo 11. Finally everythings goes perfectly with the spider vehicle for successive Apollo missions!

I never saw any video of Armstrong's trainer accident, but I also remember that they had to "fly around a bit" using up precious fuel on that first landing since the planned spot seemed too hazardous:
QUOTE
While Armstrong was maneuvering to avoid a boulder field, alarms sounded in the lunar module indicating that the computer was overloaded. Mission Control quickly told the crew to proceed. Then, as fuel was running low, a dust cloud obscured the surface and Armstrong had to touch down without a good view of his landing spot.

The MSL lander will have to be able to make this judgement in real time, but hopefully a hazard free site will be chosen with the MRO super resolution pics.

And now, a bit of nostalgia. (2.6MB quicktime) Long version here.

And since this is unmannedspacefilght.com, I apologise for that shameful display of non robotic flight, and for penance offer up this "lunar landing" movie from theRanger program:



SPLAT!
erwan
Lyford: I readed (where?, i will try to find a link)) that before real Apollo missions, Armstrong (as an 'essay pilot', is it correct in english?), see the death during a training fly with a LEM demonstration vehicle - some issue requering the "ejectable seat". Moreover, there was successive serious issues with LEM precursors, stressing the entire Apollo team. Not seen any video though...
lyford
QUOTE (erwan @ Mar 25 2005, 11:48 AM)
Lyford: I readed (where?, i will try to find a link)) that before real Apollo missions, Armstrong (as an 'essay pilot', is it correct in english?), see the death during a training fly with a LEM demonstration vehicle - some issue requering the "ejectable seat". Moreover,  there was successive serious issues with LEM precursors, stressing the entire Apollo team. Not seen any video though...
*

Here's the story. Astronautix to the rescue! And "test pilot" is the term I believe you are looking for in English.

And now back to unmanned flight. wink.gif
erwan
Lyford: next, your comment about Apollo 11 real landing is nice; may i add that plus the boulders area, requering Armstrong piloting, a global alarm sounded on the LEM, because descent control software was overflooded with datas rate and crashed. A young NASA engineer (name?) had the responsability to evaluate the issue in the minute, and responded to LEM team: go ahead, no problem...
erwan
Thanks for the link, Lyford; now back on topic!
erwan
Back to skycrane: a kind of 'LEM like', but with full automated descent, and not any real time human control; just to consider the level of informatic upgrade since 1970s; i wonder if Apollo computing power was not lower than an up to date personal scientific calculator....
cIclops
QUOTE (lyford @ Mar 25 2005, 06:47 PM)
Obviously they have success with a slow landing from a hundred feet on a flat pad (I couldn't find any info on the guidance control or nav they are using... GPS?), but I think the kicker is going to be the real time hazard recognition - finding a flat landing zone and going to it autonomously before the fuel runs out...


... and if you want to know what happens when the fuel does run out see this page from John Carmack's blog and the video (4MB) at the end.

Yes the vehicle uses GPS which is unlikely to be available on Mars, however careful site selection and radar ranging together with an Apollo style contact strip for the final meter should cut it.
erwan
Cyclops: "uses GPS which is unlikely to be available on Mars" I readed (where?.... Lyford Help please!) that: about the future human space US missions ( first Moon, next Mars), a system analog to GPS (or instead: beacons previously released on the lunar soil, instead of satellite constellation) is likely programmed for the lunar step of the program. Thus, although certainly not available for MSL 2009, i guess humans will not land on Mars before a kind of GPS system will be avalaible there!
lyford
QUOTE (cIclops @ Mar 25 2005, 12:15 PM)
... and if you want to know what happens when the fuel does run out see this page from John Carmack's blog and the video (4MB) at the end.

Yes the vehicle uses GPS which is unlikely to be available on Mars, however careful site selection and radar ranging together with an Apollo style contact strip for the final meter should cut it.
*


Ouch! Any paragraph beginning with the title "Complete Loss of Vehicle" is a sad read... chalk one up for unmanned flight: At least no one was hurt - though some debris seems to get too close to the camera, which is also hopefully "unmanned!"

Love the view from the rocket going up, nothing beats it.

As far as the last meter down, I am just worried about a Big Joe issue like on Viking 1. If it had just come down a few meters over from where it landed, it wouldn't have survived. Of course, MRO should identify all hazards down to under a meter, but that could still ruin the lander if it's unlucky enough to hit it with the contact strip first and cut engines before touching down. And even a palette style lander may end up so tilted as to hinder egress.

The Skycrane may actually help in this since the rover suspension should be able to handle uneven terrain... While admittedly a long shot, the worst would be something like this, MSL just teetering back and fort, spinning its wheels:



"Skycrane reports successful separation - engine cut off - all systems nominal - rover reports zero wheels on Mars." mad.gif
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2014 Invision Power Services, Inc.