Simulation of the Mars Science Laboratory Sky Crane

An anonymous reader points us to Gizmodo for a fascinating video of NASA's Sky Crane. "When I read that the UFO-looking Mars Science Laboratory's aeroshell would use a floating crane — called Sky Crane by NASA — to softly land the rover on Mars, I couldn't believe it. Now, watching this hyper-realistic NASA simulation, I still can't believe how the whole thing works. I don't know about you, but the whole operation mesmerizes me to no end."

Comments

Cool Movie - but bad idea!

[jackb_guppy]

That is so complicated. The "beach ball" idea from the two current rovers was much better.

Maybe if the Sky Crane was a ballon system so it can float around Mars would make this better. But still way too complicated.

Re:Cool Movie - but bad idea!

[Nyeerrmm]

I think the idea is that as you get more massive that doesn't work as well anymore. The weight of the airbags becomes untenable... I haven't looked at the math in few years, so unfortunately I can't be more specific.

Mars is one of the hardest places to land because its atmosphere is so wispy; on Earth some simple parachutes and a well-shaped capsule do the trick. On Venus the atmosphere is so thick anything you drop in will happily land softly as long as it doesn't melt. The moon and other such places you really only have the landing rocket option, which can be heavy but not particularly complicated.

On Mars though, the atmosphere is too thin to allow the capsule to slow it down to subsonic speeds on its own, meaning supersonic chutes are necessary if you want to use the atmosphere to slow you down. If you want to land with a rocket, you run into issues of trying to light an engine with supersonic flow going into the nozzle; trying to light it and flip around I imagine introduces some pretty wretched dynamic and structural problems. That tends to mean a series of parachutes including custom Mach 2 or Mach 3 chutes that would never be needed on Earth, or in this case using an aeroshell as well. Even then, you're still going too fast, so you need to slow down more. As suggested before, the airbags have worked in the past but don't scale well with higher mass vehicles. Thus you really need some kind of rocket (that ignite at subsonic speed); I'm not sure why a sky crane works better than some other system with rockets, I'd imagine its the easiest method of separation and also allows you to use less fuel since the crane itself doesn't have to slow down to a safe speed (i.e. release it down and reel it back up to reduce landing speed.)

Also, they had this option out there three years ago when I worked on a Mars mission for a class, so it's been out there a while and is probably as well developed as a non-tested system can be.

Re:Cool Movie - but bad idea!

[Anonymous Coward]

Airbags don't work on something as big as the MSL rover. It's that simple.

Re:Cool Movie - but bad idea!

[Anonymous Coward]

I'm on the project, and I understand all the reasons why we're doing it this way, but for what it's worth, I think it's as bat-shit insane as the rest of you. One thing no one can argue, though: it's incredibly cool.

Re:Cool Movie - but bad idea!

[LurkerXXX]

Or you need to send down a bunch of pieces which assemble themselves into a larger vehicle.

"Mars Mission. The Beginning of Megatron"

(queue theme music and intro credits...)

Re:Cool Movie - but bad idea!

[Anarchitect_in_oz]

My first thought on why the space crane works is the rocket don't need to change speed. They can keep blasting out the same pressure slowing the craft with the right acceleration the crane has near zero vertical speed at the right height. From then it starts to accelerate upwards. I figure the crane cables are to give enough of a time window between slowly falling and slowly rising to cut the cables. Should massively reduce the control systems needed in the crane.

Re:Cool Movie - but bad idea!

[khallow]

My concern is that the crane is another failure mode. And it's not clear to me what value the crane adds to justify that risk. If the crane fails to deploy or it doesn't retract fast enough, then the vehicle is at the least immobilized, either with a heavy weight of the crane on top of it and/or damaged suspension system from hitting too hard.

Re:Cool Movie - but bad idea!

[The Fun Guy]

The cables allow for some leeway in how hard you can hit the ground and still have a functioning rover. If you come in too fast with a single piece of equipment, the whole thing goes crunch. The rockets and ground acquisition sensors are good, but not perfect.

With the rover on the end of the bridle, this decouples the weight of the scientific payload from the weight of the support equipment needed to ensure a soft landing of the payload. Once the thing is on the ground, you don't need any of the support equipment anymore. You can make the rover tough enough to handle a harder-than-expected landing, even tough enough to bounce a bit. Accelerometers on the rover will send the signal back up the umbilical to give the crane the green light to cut loose and go away. I suppose that you could even have tilt sensors on the rover to tell the crane, "This is a bad place to leave me, pick me up and put me someplace else."

A lot of the weight of the (rover + sky-crane) is up in the crane. This set up will let you have a thinner engineering margin in the crane part, saving weight that can be used for more fuel. As long as the crane can fly, it can take pictures and serve as a supplementary probe.

Re:Cool Movie - but bad idea!

[florescent_beige]

You can see the pedigree of the concept. The airbag system used a brutal retro-rocket on the teather milliseconds before impact to slow the airbag-lander from smush speed to bounce speed.

This is similar. The retro rocket is far more gentle and precise but basically we have a last-second retro rocket on a tether dropping the lander onto the surface.

I presume tradeoff studies were done to find the optimum balance between the amount of crane hover precision and winch control precision for a giving touchdown speed tolerance at min weight.

As noted, there are a lot of things deploying in rapid sequence, and a winch. Winch bad. Winch have many fiddly bits. I wonder if they also did a tradeoff study of reliability vs capacity of this concept and a brute-force lander.

Re:Cool Movie - but bad idea!

[Mr2cents]

Indeed, you are right about Spirit and Opportunity.. I almost forgot.

I am fairly confident that NASA has made the correct choice about the EDL method, but I really wonder if they are going to be ready in 2009. I haven't seen any photo's of the MSL construction, so I don't even know if they have already started building it or not. Or are they planning to do construction in just a few months? I know the science instruments are all (nearly) finished, but I hardly hear news about the rest of the hardware.

WTF: Hyper-realistic?

[corsec67]

Almost realistic: the simulation approaches what the same inputs would do to the real system.

Realistic: the simulation behaves the same way as the real system.

Hyper-realistic: the simulation is better at realism that they real system?

What next, über-realistic? Or is profit next?

Re:WTF: Hyper-realistic?

[Anarchitect_in_oz]

I think it's to cover all the storytelling camera movements and focus effects, that go above and beyond the simulation of the concept itself. More fun to watch that way. Like news is more interesting with handheld cameras and fast cuts.

Why not use what works?

[ryanisflyboy]

Why does NASA have to spend money on new untested methods? If the old baloon method worked well for two previous rovers, why not use it again? It is hard to beat a 100% success rate. Does anyone know why they want to use this over other methods?

The sysadmin in me says: The more moving parts, the greater the chance something will break.

Re:Why not use what works?

[ColdWetDog]

Why does NASA have to spend money on new untested methods?

Trebuchet's don't scale very well. AFAIR, neither did the beach ball. This thing is lots bigger and heavier.

Re:Why not use what works?

[corsec67]

True, but how much does a triangular pyramid covered with self-inflating balloons weigh? Plus it had the hinged ramps that would fold down, so the old landing method wasn't completely without complex stuff.

Re:Why not use what works?

[Anonymous Coward]

It has to do with mass, the MSL rover, at 900 kg, is much too heavy to land using the airbag methods that landed the 180 kg Spirit and Opportunity rovers. To give a sense of scale, the MSL rover is the size of a minivan, while Spirit and Opportunity are the size of small riding mowers. See:

http://en.wikipedia.org/wiki/Image:Mars_Science_Laboratory_empty_chassis.jpg

And

[Konster]

We've got to flex some of our engineering muscle in front of our Martian friends so they will less inclined to invade us. In this light, clearly this > bouncing beach ball delivery.

Maybe have an external speaker system that blasts Ride of the Valkyries during descent, too.

Re:And

[memristance]

In that vein, I propose we scrap the whole robo-probe thing and go straight to a manned lander.

Re:Who am I to question NASA's physics...

[zappepcs]

That is the reason for the skycrane. If the air were as dense, a standard parachute landing would be all that is required. The problem is an inability to decelerate to a stop without the sudden stop part. I believe that gliders are also out of the question due to atmospheric conditions. This is a compromise of all known possible options. One that will land the heavier weight, and perhaps give them much greater accuracy in choosing a landing spot... With the speed of the rover hitting the right spot first try can be quite a bit of savings.

Not entirely practical

[Chazerizer]

As cool as this is, we've succesfully landed rovers on mars (and the moon, though not a robotic system), as well as landing non-motile craft on other planets. All used relatively simple delivery systems, and frankly, worked pretty well. The Apollo system (at 40 years old) landed softly enough not to smash human beings, which can be a lot more sensitive than robots. Maybe this type of technology will have a use in the future (though it's not like it's a super-high-tech idea). All of that being said, GOD DAMN that's cool!

Re:Not entirely practical

[DerekLyons]

As cool as this is, we've succesfully landed rovers on mars (and the moon, though not a robotic system), as well as landing non-motile craft on other planets. All used relatively simple delivery systems, and frankly, worked pretty well.

So what? This thing is much larger than anything landed anywhere before, with the exception of the manned LEM.
 

The Apollo system (at 40 years old) landed softly enough not to smash human beings, which can be a lot more sensitive than robots.

The Apollo capsule had an atmosphere much thicker (as in height) and much thicker (as in density) than the Mars landers have available. The Apollo LEM could use rocket braking because of the Moon's low gravity.
 
Mars is a stone cold bitch to land on because the atmosphere is too thin to completely rely on parachutes, and Mars' gravity is too high to rely completely on rockets.

what I don't understand

[the_Bionic_lemming]

We seem to be able to get to mars better, the Russians do land landings better than us.

why not have them design and build the landing mechanism, and we just fly it there?

Re:turtle on its back

[Profane MuthaFucka]

The video showed the lander dropping straight down on the wheels without any side motion at all. This seems unlikely to me. The rover would be swinging like my dick on those cables unless there were some thrusters used to stop any swinging motion.

Other than that, it's a pretty neat idea, ESPECIALLY for a spacecraft which is not a rover. A rover can move out of the landing zone, but a stationary spacecraft cannot. It would be sitting on soil which has traces of hydrazine from the landing rockets - but this system would avoid that problem.

Re:No solar panels?

[Concerned Onlooker]

Not exactly. It is radioisotope power but it is not nuclear power in the sense that there is a reaction going on. The simple decay gives off heat. As I recall it's not that much, either. Something on the order of 110 watts. Still, it's much better than relying on solar panels. Here's a nice page of fact sheets for the mission.

Landing on Mars is hard

[Animats]

The problem with soft-landing heavy objects on Mars is that there's not enough atmosphere for aerobraking and parachutes to do the job, so the approaches used for Earth re-entry won't work. There's too much gravity for landing on rockets. as with lunar landers, without most of the payload being landing fuel. The problem gets harder as the mass goes up. This was realized only about five years ago, to the embarrassment of some within NASA. So there are now various complicated hybrid schemes, like this.

The scheme with the cables does not look promising. Unlike Luna, Mars has winds and weather. This looks like one of the student lander designs from NASA's high school curriculum.

One bad feature of this design is that the actual landing forces have to be taken by the rover's suspension. Previous designs had the rover inside the landing module, not underneath it. That approach uses crushable components (air bags, crushable blocks, collapsible legs, etc.) to cushion the landing. With this "flying crane" approach, the autopilot has to do a really, really smooth landing or the rover will be broken.

Really Poor Concept

[VendettaMF]

The whole operation is horrendously complicated, with dozens of potential failures at each point, and no realistic means of allowing for such failures. Every step would have to function perfectly, or we've just sent another multi-billion paperweight to a dead planet.

Whatever happened to KISS?

The engineer who proposed this really needs to look into alternate fields of employment. I suggest Fecal Matter Relocation.

Re:This is needlessly complicated and HERE is why:

[the_other_chewey]

Even in their most recent plan for this Mars descent, their first mode of descent is to drop the module like a stone, using elaborate and expensive heat shielding to protect the even-more-expensive gear. But maybe -- just maybe -- they could take a lesson from Spaceship One and just take their time getting this thing down to the ground.

Sorry - but you have no clue. Mars' atmosphere at the surface has about 1% of the Earth's density, making something like aerodynamic flying impossible.
There simply isn't any other way than "dropping like a stone" - even on their parachutes, the rovers did exactly that. Those parachutes were supersonic, and their
main purpose was trajectory stabilization (although they did of course contribute to the braking).

Go read this article already linked above for a well written explanation about why
landing on Mars is actually very hard and cannot in any way be compared to landing on Earth.

Why the crane part?

[Somegeek]

I'm OK with everything up until they start winching the lander down under the crane. How is that better than sitting the rover on top of the retro rocket module, hovering, then landing, and having the rover drive off the top of the lander?

The number of failures that could happen to the winching system seems nuts; a line might not lower, or at the wrong speed, or a line could tangle, or a side to side oscillation while descending, or a cable not disconnect, and if any of these go wrong, you have no time to fix it.

Re:Why the crane part?

[ChrisA90278]

I think some one figured out that a winch is ligher weight then a folding ramp. You have to remember that this next rover is the size of a car. Think about how you'd make a folding ramp that would allow a car to drive off a lander. The winch is very compact and light

A ramp carries a risk too. It would need a set of motoers and gears to make it unfold. If there is anything in spacecraft that will fail it is those kinds of moving parts and they depand on batteries that have beenin space for a year.

The winch could be a simple fiction device. Not a motor, just a spool of cable that is pulled open by the weight of the rover and some kind of friction device to reduce th speed that the cable can unroll.

Re:Anyone know? Audio recordings of Mars?

[Somegeek]

Just a little longer...

http://www.space.com/businesstechnology/081001-tw-phoenix-microphone.html