SpaceX Falcon 9 v1.1 - SpX-6/CRS-6 DRAGON - Discussion Thread

[Fr4nK]

Here is what i would add to JRTI ship to help with tipping rocket. It would be folded down until the engine cut off. then spring up to act as a fence for the rocket. Sry for my poor drawing skills

Not needed.  Too  complex.  The rocket has just to land right

And if the rocket is on the edge, those things will knock it over

It start like this, then when it goes up, you make it so it stop where the top of the rocket would be if the legs would be right on the side of the barge. Of course you can add some brain to it so you adjust depending on where it lands. Just need it to be strong and fast. (To prevent tipping)

You realize that as drawn those arms would be like 100' long...

well, im using paint, it is not to scale. And it would be light weight. But strong, It is just to get the idea. SpaceX could figure out how to make it works. 

With those 5 beers I had since it launched, Im still celebrating launch success   And to *land/hit* on the barge twice in a row is pretty awesome!!

[mme]

My two cents:

Barring additional information, the way to fix the problem is to land with less lateral motion.  It either had the lateral motion because it had not killed off enough horizontal speed or because it it overcorrected/corrected too late.  The former is fixed by adjusting the arc that the stage "flies."  The latter is fixed by refining the terminal landing algorithm.

That's it.  No changes to the barge, no need to anchor the barge, no new two-way communication required.  We are well into the "fine tune" phase of the design/experiments.  Wash, rinse, repeat.

[Kabloona]

What's the Air Force looking for, before it deems an overflight/return as safe?

The exact criteria haven't been spelled out in public, but Elon has said that "repeated precision landings" , or words to that effect, will be required (on the barge).

[StuffOfInterest]

I somewhat wonder if heligrid (mentioned a few months ago) could have saved the day, assuming all four legs were on the deck.  The real proof will be when we get to see the full video.  It could be that there is no way this landing was going to be saved.  Still a great learning experience, however.

Considering how good they are on accuracy, I hope this will allow SpaceX to attempt to land at LC-13 sooner rather than later.  Having a stable (and larger) target can't hurt their chances.

[Fr4nK]

My two cents:

Barring additional information, the way to fix the problem is to land with less lateral motion.  It either had the lateral motion because it had not killed off enough horizontal speed or because it it overcorrected/corrected too late.  The former is fixed by adjusting the arc that the stage "flies."  The latter is fixed by refining the terminal landing algorithm.

That's it.  No changes to the barge, no need to anchor the barge, no new two-way communication required.  We are well into the "fine tune" phase of the design/experiments.  Wash, rinse, repeat.

It sure is the way to go. But some fence to save $$$ recovering engine that are worths millions of dollars might be a good investment in the meantime.

[Kabloona]

My two cents:

Barring additional information, the way to fix the problem is to land with less lateral motion.  It either had the lateral motion because it had not killed off enough horizontal speed or because it it overcorrected/corrected too late.  The former is fixed by adjusting the arc that the stage "flies."  The latter is fixed by refining the terminal landing algorithm.

That's it.  No changes to the barge, no need to anchor the barge, no new two-way communication required.  We are well into the "fine tune" phase of the design/experiments.  Wash, rinse, repeat.

It does seem like the problem is that the horizontal velocity that has to be killed off during the landing burn, which means the stage has to remain tilted off vertical in order for the thrust vector to have some horizontal component. This is unlike the Grashopper/F9R Dev flights that have had very little horizontal velocity to kill.

So the stage has to remain tilted both to kill horizontal velocity and then to do any final position correction, and if it undershoots or overshoots on the horizontal velocity, the result is the last two crashes.

Last time looked like it overshot the barge and tried to come back. This time it may not have overshot the barge, but still had too much horizontal velocity when it touched down.

[sanman]

How about giant airbags? They inflate from the barge deck as soon as the stage touches down, in order to trap it upright.

Of course, car airbags can inflate in microseconds because they're not so huge -- whereas multi-storey airbags might take slightly longer. 

[saliva_sweet]

Landing photos!

https://twitter.com/elonmusk/status/588082574183903232

Distinctness of the soot stripes in these pictures is interesting.

[Radical_Ignorant]

Not part of discussion what SX should do, but nice, well balanced article about today event:
uk.businessinsider.com/watch-the-spacex-launch-livestream-2015-4?r=US

[kdhilliard]

...,, landing on target, oriented vertically, and with near zero vertical velocity as you reach the deck is not enough if your lateral velocity sends you skidding off the deck or toppling over.

It's not something I'm happy to have presaged, but it has always struck me as a difficult job to stick all these conditions at once when working with a T/W >>  1.  Grasshopper videos some impressive low level maneuvering, but at nearly a hover.

I hope they stick the next one, but if they don't I wouldn't be surprised were they to outfit a new core as a F9R to conduct a series of increasingly difficult test landings.  There is a huge difference between T/W ~ 1 and T/W >> 1.

[dorkmo]

do we believe that the tank ruptured and shot off similar to last time?

[Optimist]

Wasnt sure where to post this.

Saw this on Twitter , havent seen it posted here.

Tweets & replies

Elon Musk @elonmusk  ·  33m 33 minutes ago
@teknotus There are nitrogen thrusters at top of rocket. Either not enough thrust to stabilize or a leg was damaged. Data review needed.

[jaufgang]

I was writing something related to first stage landing technical hurdles that SpaceX has already retired, but clearly something is missing.

What technical hurdles are left to solve?

They've never done a landing on a solid surface from high altitude free-fall at terminal velocity.  All grasshopper and F9R-Dev tests had the engine firing continuously and a slow controlled descent.   A F9R-Dev test involving engine cutoff and a separate last moment landing burn has never been attempted, let alone succeeded.   

But what difference does it make? Once the F9R-Dev is under power on the landing burn, it's analogous to the slowly decelerating Grasshopper, isn't it (aside from the differences between a test article and flight hardware)? The conditions at sea are perhaps quite different to McGregor though. Not to mention the vertically unstable (despite all the mention of the stablising ability of the) barge.

I think there are some very significant differences.   For one thing the F9R-Dev tests have not had to contend with the sort of high lateral velocities that were implicated in today's imperfect landing.  Also the F9R-Dev was heavily ballasted to allow hovering and very slow descent.  The high velocity approach, and rapid deceleration  of a real landing would leave significantly less room for error and less time to make corrections. 

I had very much expected that by the time they got to this point attempting to land actual flight hardware, they would have had several F9R-Dev tests under their belt where the test rocket had been launched in the desert to stratospheric altitude, cut it's engine, fell back down under grid-fin guidance and then performed a landing burn similar to what would happen in a real launch.  Had they done that successfully, the claim could be made that their landing system has been definitively proven to work, even before the successful recovery of a production stage at sea.

Without that, those final few moments of bringing the stage from terminal velocity to a stationary upright landing remain a "technical hurdle left to solve".

[sanman]

Wasnt sure where to post this.

Saw this on Twitter , havent seen it posted here.

Tweets & replies

Elon Musk @elonmusk  ·  33m 33 minutes ago
@teknotus There are nitrogen thrusters at top of rocket. Either not enough thrust to stabilize or a leg was damaged. Data review needed.

Just for context on what Musk was replying to:

Daniel P Johnson ‏@teknotus  1h1 hour ago
@elonmusk does the first stage have thrusters at the top to halt wobble or is it entirely controlled by Merlin engines?
6:25 PM - 14 Apr 2015


Elon Musk ‏@elonmusk  39m39 minutes ago
@teknotus There are nitrogen thrusters at top of rocket. Either not enough thrust to stabilize or a leg was damaged. Data review needed.

So it sounds like nitrogen thrusters are intended to stop the stage from tipping over. Are these nitrogen thrusters the same as the RCS, or are they something more specialized for the landing?

[JazzFan]

Wasnt sure where to post this.

Saw this on Twitter , havent seen it posted here.

Tweets & replies

Elon Musk @elonmusk  ·  33m 33 minutes ago
@teknotus There are nitrogen thrusters at top of rocket. Either not enough thrust to stabilize or a leg was damaged. Data review needed.

How much effect do nitrogen thrusters have in the lower atmosphere at sea level?

[Danderman]

An old engineering saying is that 90% of system costs are incurred in the first 10% of the project.

In this case, if the Falcon first stage were intended to perform pinpoint landings on rocking barges, the stage would not be designed to be relatively top heavy; ie it would have a broad base.

So, SpaceX is going to spend some extra cash on trying to get a pencil shaped object to balance on a narrow pad area. Not that this is impossible, only that the lack of systems engineering at the outset (meaning a review of ALL system requirements, not just those for launch) is going to cost more time and money, and probably have a high recurring cost of first stages that are lost during recovery.

On the other hand, this could all be a giant head fake for ULA to sink their time and money into recovering first stage engines.

[sanman]

Musk's reply at least seems to imply that the nitrogen thrusters would be triggered to rebalance the tipping stage if necessary.

[iamlucky13]

Well, oil rigs don't use thrustmasters, and are instead anchored to the ocean floor. If this stage had been landing on something that was more like an oil rig than the current barge, then would this touchdown have had greater chances of success?

Yes, a barge may be cheaper than an oil rig, but if you're going for reusability you may not want to skimp on this cost.

Deepwater oil rigs do and have used dynamic positioning since the 1960's, with steadily improving accuracy over that time. In fact, beyond a certain depth, dynamic position actually results in greater stability than anchoring, at least according to the dynamic position supplier literature. Thrustmaster seems to indicate 5m position keeping is relatively easy:
https://www.thrustmaster.net/portable-dps-shallow-water-applications/

How about giant airbags? They inflate from the barge deck as soon as the stage touches down, in order to trap it upright.

Of course, car airbags can inflate in microseconds because they're not so huge -- whereas multi-storey airbags might take slightly longer. 

Or they could just use the rocket's control system to null out the lateral velocity by improving the control system. They've been making really good progress in that direction.

Besides, airbags would push unevenly on a rocket not centered on the pattern, possibly tipping over an otherwise stable rocket, and nobody makes multi-storey airbags, especially not rated for exposure to rocket exhaust.

[Jim]

It sure is the way to go. But some fence to save $$$ recovering engine that are worths millions of dollars might be a good investment in the meantime.

No, it isn't and you have nothing to back up your claim.  The barge is only temporary.  The fence is a kludge and not worth the effort

[mme]

...,, landing on target, oriented vertically, and with near zero vertical velocity as you reach the deck is not enough if your lateral velocity sends you skidding off the deck or toppling over.

It's not something I'm happy to have presaged, but it has always struck me as a difficult job to stick all these conditions at once when working with a T/W >>  1.  Grasshopper videos some impressive low level maneuvering, but at nearly a hover.

I hope they stick the next one, but if they don't I wouldn't be surprised were they to outfit a new core as a F9R to conduct a series of increasingly difficult test landings.  There is a huge difference between T/W ~ 1 and T/W >> 1.

You can kill horizontal velocity by changing the arc the stage "flies."  You can also kill it earlier in the landing burn.  The landing burn itself is over 30 seconds long, there is a lot of time to make adjustments.

These barge landings are giving them more real-world data for landing that F9R-Dev can.