Separate names with a comma.
Discussion in 'SpaceX' started by doug, Apr 15, 2015.
CRS-6 First Stage Landing - YouTube
As it fell over I thought "oh almost - maybe they can recover something from the dented carcass - oh, nevermind..."
I know this stuff is hard, but I wonder why the control software allowed the rocket to get so far out of vertical in the first place, that such a massive last-minute correction was needed?
yep, undamped control oscillations.... Time to tweak the control loops or check latency on sensor input.
Elon tweeted yesterday exactly what they believe the problem was. Valve stiction...
He said in a tweet that it's an easy fix
the drone ship looks a little crispy. Ouch. I hope they can refurb it quickly.
or control system output(s)... Missed that one.
And that answers my question about damage... I guess it was another RUD.
Best thing is they probably got a lot of data that they can use in the simulation/modeling - this will give them the info to perfect this provided the hardware is working. They should be able to figure out if it was algorithm/coding or a real hardware performance problem from the data- They certainly learned a lot more about the dynamics and how the 1st stage responds to it's controls. This is amazing. They came so close. and this is very hard rocket science. Too bad the news media is playing this as a failure- it was a learning experience, and not an unexpected result... I suspect with a lot of modeling/simulation they may nail this on the next try. Great Job SpaceX team.
Watching this, my reaction was the same as watching a gymnast almost, but not quite, stick a landing..."hold it, hold it, ... crap." From that explosion, it looks like they should have used up more fuel or have some procedure for dumping fuel just before landing.
That means fuel is not the issue, but throttling is the challenge here.
Kind of hard to understand why any rocket engine can't produce a fine-tuned lower level of thrust.
There are many reasons... only one of which is that, for a given size necessary to support the maximum thrust a turbopump must be designed for, there's a minimum amount fuel necessary to drive that turbopump at the lowest RPM it can still support combustion at.
In other words, a turbopump that can pump 5.5 tons of fuel per second at max-throttle, may be capable of throttling down by a factor of 10X. That's still 11,000lbs/sec of fuel... which may generate significantly more thrust that the now-empty rocket weighs.
Wouldn't sending more payload mass to orbit or ISS (e.g. espresso packets) be a better usage of the fuel? Making a controlled descent burns a lot of propellant, plus it looks like it would be a high probability, high consequence on the failure risk chart.
Engineering is a great profession, and this is the fun part--the inverted pendulum problem in controls theory.
It looks like the lateral thrusters ran out of gas shortly after touchdown, but doubt they had enough control authority for the tilt.
Watch it in full screen, 1080p in slo motion and you'll see that the thrusters at the top of the rocket failed to fire early enough .5 seconds before touchdown thereby failing to prevent the lateral motion that doomed the landing. The video in slomo shows just how close they came. Amazing feat.
I'm afraid I'm with the Control Freaks - that is, the control system should never have allowed it to get that far out of column and particularly to over correct like that. The valve sticktion explanation might be it, but frankly I'm doubtful that's the only issue.
You must be new to SpaceX and what it is trying to do. I'll do my best to explain but be aware there are plenty of videos of Elon explaining this better. The idea is that fuel is only about 2% of the cost of a rocket. The current rocket philosophy is the one you mention, maximize your payload. Elon likes to describe it as building a 747 passenger jet flying it across the country dropping off the passengers then destroying the jet. It would make it outrageously expensive to fly. It is outrageously expensive to get stuff into orbit. By reusing the booster and capsule the cost would drop significantly. That is why SpaceX is working very hard on reusability.
Here is a great video of Elon explaining things: Elon Musk on SpaceX - YouTube
0:03: landing legs come on and the nose is tilted around say 2 degrees to the left
0:04: perfectly aligned to the bulls eye, nice and vertical and no tilt.
0:045: the lower end starts moving the other way, while the nose is still right where it should be
0:05 to 0:07: the nose thrusters don't fire and the tilt gets worse.
0:07: the lower end is now 25 ft above the ground and directly above the left edge of the platform, while the nose is right above the bulls eye. So i am guessing the tilt is around 5 degrees
0:07: nose thrusters come on, and the lower end starts moving towards the center of the platform, while the nose is still above the center.
0:08: nose thrusters shut off; lower end continues to move towards the center of the platform
0:09: the rocket is exactly vertical; touch down close to bulls eye; engine shuts off
So far so good.
0:10 till 0:14: the lower portion continues to slide to the right away from the center towards the other edge of the platform, possibly because of momentum. Nose thrusters come on again as the nose is falling to the left and they struggle for nearly 4 seconds to push the nose back, but it too late because the legs are already on the ground and that means it is much harder to straighten the rocket.
Those thrusters should have come atleast two seconds earlier when the lower section was moving away from the center. Also it is interesting to see the nose cone pretty much right above the platform throughout the 10 seconds of this video
landing - vidme
Wow! Amazing video, thanks for posting the link. It more fully illustrates just how close they got to a perfect landing.
Nice deck-level footage!
I notice they have vertical thrusters at the top firing straight down trying to plant the rocket on to the deck...
All these rocket scientists and none of them working for SpaceX. :wink: