Does anyone know why they did that (admittedly cool) hovering maneuver, rather than just landing on long extendable legs or something? In other words, rather than hovering (using up reaction mass) and slowly dropping the rover, why not land, and then slowly drop the rover?
It’s heavy, so it had to use engines to slow itself down, but those engines would have stirred up a great deal of dust and grit, some of which would have settled back into the rover, possibly incapacitating some of the equipment; the sky crane kept the engines away from the rover.
It avoids getting dust everywhere. From the “Seven Minutes of Terror” video (relevant quote starts around the 3:50 mark):
ETA: Looks like I got scooped. Well, go & watch the video anyway; it explains a lot about why they did things the way they did, and how preposterous it is that they actually pulled it off.
Dust doesn’t make sense to me. Would those jets stir up more dust than typical bad weather would? This isn’t the moon, where there’s no weather of any kind. Mars has an atmosphere, and days of good weather and of bad weather. Curiosity MUST have been designed to withstand this sort of thing.
That’s hard to comprehend. They were so worried about dust that they made the landing an order of magnitude harder. I would think it would have been easier to manage the dust cloud than create a rocket sky crane. Why not have the rover in a sealed bay? Wait 30 minutes for the dust to settle then open the doors.
I’m sure they thought of that.
Think about it this way. The rover is already designed to deal with uneven ground. If you have a lander the rover drives off you need to ALSO design a lander that deals with uneven ground. Then you have to have a “dust proof” garage and a ramp for the rover to get off the lander. And with a lander you will have all that mass up high so that makes it more unstable/harder to land on uneven ground.
So, what takes more mass overall? A system to hover for a pretty short period of time or a lander you drive off of? When you run the engineering numbers I suspect it is the latter.
One of the engineers the other day made a good point. He noted that is was very easy to “real world” test lowering the rover on uneven ground. They did it many times an apparently it was boring as heck. So they knew that part would likely work.
Then you had the hovering part. Once you throw the ground part out of the equation, the hovering part is “easy” to simulate on a computer. Well, easier to simulate than the ground part.
So, you had one part easy to test and one part easy to simulate. Though each part was probably pretty hard to do the opposite with.
So, they had pretty good confidence for each part working basically.
Unfortunately that dust would not have settled conveniently back into the large craters created by the engines.
Why does everyone seem to think that hovering and lowering the rover on cables is “an order of magnitude” harder than having telescoping legs or something? It really isn’t. The flight system is the same either way. The rover needs to be able to detach from the flight system on command either way. If you have legs descending from the flight system around the rover, they need to telescope outwards during flight, since the rover’s wheels extend during descent from the folded flight configuration. Telescoping legs that can take the rover’s weight are more complicated and heavier than some cable and a spool. A box that surrounds the entire rover would be much, much heavier than the cable system, as well as being significantly more complex. With the box, not only do you need active mechanisms to open the door and lower the ramps, you also need additional communication and radar antennas, and an additional descent imaging camera, since the ones on the rover aren’t going to be usable when it’s in the descent box.
The sky crane wasn’t just chosen because of dust. It’s also lighter than just about any alternative, and simpler than anything other than just having the rocket pack directly attached to the rover, which would have the problem of deposition dust all over the rover.
ETA: You also seem to be under the impression that the rover hovered stationary in midair while the rover lowered on cables. It didn’t. While it was descending, the flight system reeled out the cables, so the rover was descending slightly faster than the flight system. When the rover touched the ground, the cables were automatically cut, and the flight system flew away.
It’s that it added so many extra moving parts. If there are already engines being used to slow the descent, it seems simpler to use that mechanism with some tweaks to land the system. Instead, they added a whole additional system. So now two systems need to work together perfectly instead of just one. I can understand the benefits mentioned, but from a reliability standpoint I would think there would be more success getting one system to work instead of two.
I’m amazed and astounded that it worked flawlessly. It’s a pinnacle of human engineering.
The sky crane cable was over 7 metres long. Designing a set of telescoping legs that long, that could land in a stable configuration, then lower the rover safely to the ground, is at least as difficult, probably more - not to mention the extra payload mass that way.
I just wanted to thank everyone for the answers. Someone at work mentioned the dust as a possibility, but I couldn’t find any explanation from NASA or anyone else on the web. As usual, the Straight Dope is the place to find, uh, the straight dope.
and to continue the amazing, consider how much dust and grit the engines did stir up. See:
at 49 secs.
along with getting everything right, the lander took video to prove it. You know: “Pics or it didn’t happen!” 
Even well above the surface, there was a lot of dust kicked up by the thrusters.
True, but the only simpler alternative would be to keep the engine pack directly connected up to the moment of landing. A box or landing platform would require a mechanism to extend or lower ramps, as well as landing gear for the platform.
Same here. Honestly I was not expecting the landing to work as flawlessly as it did. I watched the live coverage of the landing and was really quite astonished when everything worked exactly as designed.
They could have attached Curiousity to a Japanese Spider Crab.
How would you have bet?
I think they already have those on Mars, so there seems little point in bringing one with you.
Clear, detailed panoramic photo of Curiosity, showing dust and rubble from landing.
An extraordinary photo. Imagine what is to come.
OK, I want to know how they got that image.
It sure looks like you can see 100% of the top of the lander - so, where’s the pylon holding the camera?
The camera mast is located at the front-right corner of the rover. If you look straight down on the panorama, you can see that the stitching is a bit off around where the mast is supposed to be mounted. Looks like they stitched the images together not quite taking into account that there’s a region of the top of the rover the mast can’t quite see.