SpaceX Falcon 9 v1.1 - Dragon - CRS-5/SpX-5 -Jan. 10, 2015 - DISCUSSION

[rcoppola]

The fact that the grid fins and associated control software seemed to have worked on the first try is an amazing achievement. And although the stage is reported to have "hit hard" that's a relative term. I think Hans mentioned their landing velocity is around 4.5 MPH. They may have exceeded that by only a few percent.

So couple things that cross my mind:

1. Even though the Grid Fins seemed to have worked, there may have been a few small surprises beyond the models, such as possibly influencing the rate of decent towards the end of the return profile. Or some other effect.

2. Even though it "hit hard" I'm curious as to what the angle of attack was. If it hit hard coming down completely vertical or off axis where one or two of the legs hit first, snapping, bending and forcing the piston to puncture the tank and/or essentially beginning to tear the bottom of the stage apart.

3. I'm curious if they received enough data now to (as Elon has said is their intent) redesign the legs (etc.) to allow them to deploy at higher altitudes-velocities for further fuel savings and giving even more command and control authority towards the end of the flight profile, ie. landing burn and landing.

Needless to say, IMO and frankly from observable events, it would seem inevitable now that they will  succeed at returning multiple cores in 2015. I'll even go so far as to say they may even be allowed to RTLS by EOY.

[mikelepage]

I like the fuel hypothesis.

Since we know that nominal landings work well (from GH) then either there was a mundane failure (unlikely IMO) or the failure has to do with the inbound phase.

If the inbound phase put them just beyond the edge of the envelope, the final landing maneuver might have run out of fuel even a few feet high, resulting in a hard landing.

The deceleration on landing is such that if you stop thrusting before the end of the maneuver, you're still coming down pretty fast.

Agreed. 
Also relevant is from the AMA

Previously, you've stated that you estimate a 50% probability of success with the attempted landing on the automated spaceport drone ship tomorrow. Can you discuss the factors that were considered to make that estimation?
In addition, can you talk more about the grid fins that will be flying tomorrow? How do they compare to maneuvering with cold-gas thrusters?
permalink
[–]ElonMuskOfficial 3641 points 4 days ago
I pretty much made that up. I have no idea
The grid fins are super important for landing with precision. The aerodynamic forces are way too strong for the nitrogen thrusters. In particular, achieving pitch trim is hopeless. Our atmosphere is like molasses at Mach 4!

I'm trying to read between the lines of Elon's words to figure out what he was most uncertain about before the launch the part where it comes in through the upper atmosphere is the least tested part of this process, and produces the largest error in the final landing solution - the grid fins have an almost impossible task of keeping the stage going in a straight direction.  Elon mentioned Pitch trim - whether the rocket noses up or down (it's somewhat confusing that the engines are now at the "nose" of the rocket) and it seems that is very difficult to control even with the grid fins. If it's travelling at Mach 4, even nosing up or down a little will push the whole thing outside the envelope where it can still land on the target with fuel to spare.

So because of that I think the problem this time around likely was not having enough fuel for the size of the sideways divert it had to do - because it was pushed off target by aerodynamic forces in the upper atmosphere.  The stage will probably have to do some form of sideways divert every single time it lands, but hopefully they can control it more precisely in future than they did this time. 

It would be educational (but also suck) if they discovered as a result of doing this that the forces in the upper atmosphere are so variable that they can't guarantee that every stage will make it back.  Hopefully they accounted for that when they decided how big the grid fins would be.

[meekGee]

If that was the case, they'll find out in about 3 weeks... 

You gotta love the speed of this program...

[SoulWager]

3. I'm curious if they received enough data now to (as Elon has said is their intent) redesign the legs (etc.) to allow them to deploy at higher altitudes-velocities for further fuel savings and giving even more command and control authority towards the end of the flight profile, ie. landing burn and landing.

I think they'll do more testing at spaceport america before redesigning the legs' aerodynamics, though if they're just redesigning the actuators that was probably already progressing without this test data.

[rcoppola]

3. I'm curious if they received enough data now to (as Elon has said is their intent) redesign the legs (etc.) to allow them to deploy at higher altitudes-velocities for further fuel savings and giving even more command and control authority towards the end of the flight profile, ie. landing burn and landing.

I think they'll do more testing at spaceport america before redesigning the legs' aerodynamics, though if they're just redesigning the actuators that was probably already progressing without this test data.

I think that's correct. I think it will be in parallel. SA and mission testing. There is probably some return data they'll want/need that only mission return profiles may give them. But Elon never said exactly what further data they needed to accomplish the redesign except to say it would be a relatively small effort but just lacking some data to do it.

[robertross]

3) What about the piece of something that floated of at solar deployment... Is this normal? Is this allowed?

My impression was that it was some ice.

To me it looked like a piece of torn silver Mylar, say 3 inch x 4 inch (although that's just a guess on dimensions due to the camera).

Nothing to worry about though.

[mvpel]

Elon mentioned Pitch trim - whether the rocket noses up or down (it's somewhat confusing that the engines are now at the "nose" of the rocket) and it seems that is very difficult to control even with the grid fins. If it's travelling at Mach 4, even nosing up or down a little will push the whole thing outside the envelope where it can still land on the target with fuel to spare.

I think you're reading too much into that. He said it's hopeless to manage pitch trim with nitrogen thrusters at Mach 4, not that it's difficult to do so with grid fins which work quite well at supersonic and hypersonic speeds, right?

ANALYSIS OF GRID FINS AS EFFICIENT CONTROL SURFACE IN COMPARISON TO CONVENTIONAL PLANAR FINS - Salman Munawar, Department of Aerospace Engineering, College of Aeronautical Engineering, National University of Sciences and Technology, Risalpur 24090, Pakistan

[cambrianera]

To enable early opening of the legs SpaceX must design them to keep the falling stage aerodinamically stable.
IMHO this is linked mainly to the shape of the A-frame.

[SoulWager]

3. I'm curious if they received enough data now to (as Elon has said is their intent) redesign the legs (etc.) to allow them to deploy at higher altitudes-velocities for further fuel savings and giving even more command and control authority towards the end of the flight profile, ie. landing burn and landing.

I think they'll do more testing at spaceport america before redesigning the legs' aerodynamics, though if they're just redesigning the actuators that was probably already progressing without this test data.

I think that's correct. I think it will be in parallel. SA and mission testing. There is probably some return data they'll want/need that only mission return profiles may give them. But Elon never said exactly what further data they needed to accomplish the redesign except to say it would be a relatively small effort but just lacking some data to do it.

I think the necessary data would come from high altitude leg deployment, and F9R-dev2 is the vehicle that it makes sense to do such testing with. Doing that testing on missions would probably slow down the process of getting permission to land back at the cape, though they may try it on a stage they cannot attempt recovery on.

[SoulWager]

To enable early opening of the legs SpaceX must design them to keep the falling stage aerodinamically stable.
IMHO this is linked mainly to the shape of the A-frame.

I think it would be pretty hard to make the stage unstable, with the center of mass so close to the engines.

[cambrianera]

2. Even though it "hit hard" I'm curious as to what the angle of attack was. If it hit hard coming down completely vertical or off axis where one or two of the legs hit first, snapping, bending and forcing the piston to puncture the tank and/or essentially beginning to tear the bottom of the stage apart.

This: if the stage comes down inclined, and one leg hit first, once damping travel ends you have a four fold spike in load (two fold if inclination involves two legs).
Obviously only a chance, but a good one.

[cambrianera]

To enable early opening of the legs SpaceX must design them to keep the falling stage aerodinamically stable.
IMHO this is linked mainly to the shape of the A-frame.

I think it would be pretty hard to make the stage unstable, with the center of mass so close to the engines.

Should be 13 m from the bottom of the rocket (rim of the nozzle).
Likely enough to be unstable with actual legs deployed.

[rcoppola]

To enable early opening of the legs SpaceX must design them to keep the falling stage aerodinamically stable.
IMHO this is linked mainly to the shape of the A-frame.

I think it would be pretty hard to make the stage unstable, with the center of mass so close to the engines.

I'm not so sure at those potential velocities. The survivability of legs themselves and the added forces to the primary stage structures need to be considered. But instability can also be introduced through Rate of leg deployment as well as synchronization of leg deployments. If even one leg is a few seconds behind the others in deploying at those velocities, you could induce both structural and flight path instability.

[SoulWager]

To enable early opening of the legs SpaceX must design them to keep the falling stage aerodinamically stable.
IMHO this is linked mainly to the shape of the A-frame.

I think it would be pretty hard to make the stage unstable, with the center of mass so close to the engines.

I'm not so sure at those potential velocities. The survivability of legs themselves and the added forces to the primary stage structures need to be considered. But instability can also be introduced through Rate of leg deployment as well as synchronization of leg deployments. If even one leg is a few seconds behind the others in deploying at those velocities, you could induce both structural and flight path instability.

but the legs lose a lot of surface area, and the side of the stage gains a lot of surface area, as the angle of attack increases. The grid fins also have a significant impact on center of pressure.  I think it may be possible for the aerodynamics to set up some oscillations, but it's never going to cause the stage to flip end over end. I think the main issue is predictability, or being able to control exactly how much drag you get from the legs, in order to use them to slow down without overshooting or undershooting your landing site.

[Robotbeat]

Grid fins seemed to do their job just fine! Good job, SpaceX!

[JamesH]

I'm trying to read between the lines of Elon's words to figure out what he was most uncertain about before the launch the part where it comes in through the upper atmosphere is the least tested part of this process, and produces the largest error in the final landing solution - the grid fins have an almost impossible task of keeping the stage going in a straight direction.  Elon mentioned Pitch trim - whether the rocket noses up or down (it's somewhat confusing that the engines are now at the "nose" of the rocket) and it seems that is very difficult to control even with the grid fins. If it's travelling at Mach 4, even nosing up or down a little will push the whole thing outside the envelope where it can still land on the target with fuel to spare.

So because of that I think the problem this time around likely was not having enough fuel for the size of the sideways divert it had to do - because it was pushed off target by aerodynamic forces in the upper atmosphere.  The stage will probably have to do some form of sideways divert every single time it lands, but hopefully they can control it more precisely in future than they did this time. 

It would be educational (but also suck) if they discovered as a result of doing this that the forces in the upper atmosphere are so variable that they can't guarantee that every stage will make it back.  Hopefully they accounted for that when they decided how big the grid fins would be.

That cannot be right - if it had insufficient fuel to do a sideways divert to the barge, it wouldn't have hit the barge. My understanding is it's a controlled fall, not a powered decent. So there just need to be enough fuel for the hover slam, and it's not used for divert (except for the initial burn at top of flight?). You can aerodynamically divert a lot way just with grid fins from the apogee of the flightpath.

[nadreck]

Grid fins seemed to do their job just fine! Good job, SpaceX!

Actually Elon just tweeted: "Grid fins worked extreemly well from hypersonic velocity to subsonic, but ran out of hydrolic fluid right before landing"


then

"Upcomming flight already has 50% more hydraulic fluid, so should have plenty of margin for landing attempt next month" (Sounds like DSCOVR slipped to February)."

Running out of fine control just before landing probably meant that vertical velocity was within parameters but lateral was not. (My WAG on failure)

EDIT: make that lateral and/or pitch and yaw

[mmeijeri]

How does that work? I didn't know you could run out of hydraulic fluid, unless there's a leak somewhere. If it were pneumatic, I'd understand, but where does the hydraulic fluid go?

[JamesH]

Military aircraft sometimes have total loss hydraulic system - once you run out you crash. (The Tornado for example). I presume it because they run at such high pressure it impossible to seal them well enough to keep enough fluid onboard, so you build for loss.

[kenny008]

Possibly just an accumulator. High pressure runs the fins, dumps to a low pressure tank as it's used.