-
#240
by
cambrianera
on 29 Sep, 2013 21:50
-
Recognizing cooked eggs from raw eggs is easy, you spin them on a table; if the egg is cooked the rotating movement easily transmits to the inside, giving some momentum to the rotation, if it's raw doesn't trasmit, and the rotation is short (only the eggshell rotates).
What has this to do with F9?
To get centrifuging of propellant it's not enough to spin the tanks, it's necessary to transmit the movement to the propellant inside.
This can't happen in a few second burn, had to begin before, probably from aero forces.
An egg doesn't have slosh baffles, F9 has.
How many propellant you think is needed for final burn?
It's all in the domes, without baffles.
-
#241
by
clongton
on 29 Sep, 2013 21:52
-
Recognizing cooked eggs from raw eggs is easy, you spin them on a table; if the egg is cooked the rotating movement easily transmits to the inside, giving some momentum to the rotation, if it's raw doesn't trasmit, and the rotation is short (only the eggshell rotates).
What has this to do with F9?
To get centrifuging of propellant it's not enough to spin the tanks, it's necessary to transmit the movement to the propellant inside.
This can't happen in a few second burn, had to begin before, probably from aero forces.
An egg doesn't have slosh baffles, F9 has.
How many propellant you think is needed for final burn?
It's all in the domes, without baffles.
Correct. The domes are the sumps.
-
#242
by
intrepidpursuit
on 29 Sep, 2013 21:56
-
They've already got landing legs to deploy (and a few suggestions that they might be part of a roll control strategy); would it make sense to try to attach control surfaces of some kind to these?
This has been discussed at length in the Grasshopper threads. My assumption is that they have already thought of this and already have a solution in place. Something failed, whether it was mechanical or just an engineering failure. With the data from this flight they can tweak the solution already in place to compensate for whatever unexpected parameter presented itself during this flight.
I think the problem was two fold. First, they did not have legs on this rocket. Elon has stated in the past that the legs would be used to help slow and stabilize the rocket, so without them this stage would have had less margin than it would on a normal flight. Second, somehow they used more RCS gas/required more ACS compensation than anticipated. Calculating how much stabilization is needed for a system that will be doing something no one has ever done before requires some big assumptions for variables. They can likely find some efficiencies in the flight path along the way to increase margin, and then once the legs are on and deploying that margin will be well above where it needs to be. No major redesign (fins, aerodynamic control surfaces, etc) necessary.
-
#243
by
R7
on 29 Sep, 2013 21:56
-
Correct. The domes are the sumps.
Yeah, I hastily generalized anti-vortex baffles as slosh baffles too. My bad.
-
#244
by
JimO
on 29 Sep, 2013 21:57
-
Now that TLEs are available for stage-2, can anybody work out quick&dirty ground visibility locations/times/directions for early revs, to seek out serendipitous visual sightings related to fuel dump and other post-insertion dynamic events.
-
#245
by
corrodedNut
on 29 Sep, 2013 22:02
-
There are two kinds of baffles, anti-slosh baffles, and anti-vortex baffles. The anti-slosh baffles are on the tank walls, the anti-vortex baffles are in the sumps. If you want to see the phenomena in action, go to your toilet and flush it. Pretend the chugging and gulping you hear is the Merlin engine sucking helium. Or fill a bottle half full of water, turn it upside down, give it a twist, and watch all the water cling to the sides of the bottle as it spins out the neck.
-
#246
by
Mongo62
on 29 Sep, 2013 22:05
-
I think the problem was two fold. First, they did not have legs on this rocket. Elon has stated in the past that the legs would be used to help slow and stabilize the rocket, so without them this stage would have had less margin than it would on a normal flight. Second, somehow they used more RCS gas/required more ACS compensation than anticipated. Calculating how much stabilization is needed for a system that will be doing something no one has ever done before requires some big assumptions for variables. They can likely find some efficiencies in the flight path along the way to increase margin, and then once the legs are on and deploying that margin will be well above where it needs to be. No major redesign (fins, aerodynamic control surfaces, etc) necessary.
Might this be why Elon tweeted that the CRS-3 booster (the next attempt at controlled reentry) will have legs attached? The data he saw possibly indicating that without the legs' extra control authority, the RCS thrusters were insufficient to keep the nearly empty booster from spinning too quickly, but with them, it should keep the spinning under control?
-
#247
by
clongton
on 29 Sep, 2013 22:07
-
Launch video is up:
-
#248
by
corrodedNut
on 29 Sep, 2013 22:18
-
Anyone want to speculate what the MerlinDvac nozzle extension is made of? Looks like Niobium to me...
-
#249
by
Robotbeat
on 29 Sep, 2013 22:32
-
Recognizing cooked eggs from raw eggs is easy, you spin them on a table; if the egg is cooked the rotating movement easily transmits to the inside, giving some momentum to the rotation, if it's raw doesn't trasmit, and the rotation is short (only the eggshell rotates).
What has this to do with F9?
To get centrifuging of propellant it's not enough to spin the tanks, it's necessary to transmit the movement to the propellant inside.
This can't happen in a few second burn, had to begin before, probably from aero forces.
An egg doesn't have slosh baffles, F9 has.
How many propellant you think is needed for final burn?
It's all in the domes, without baffles.
Correct. The domes are the sumps.
... And there are baffles at the sumps.
-
#250
by
ugordan
on 29 Sep, 2013 22:43
-
Anyone want to speculate what the MerlinDvac nozzle extension is made of? Looks like Niobium to me...
Steve Jurvetson says it's "niobium halfnium [sic]" :
http://www.flickr.com/photos/jurvetson/10008137326/It does look like the same stuff as on v1.0 to me.
-
#251
by
Kabloona
on 29 Sep, 2013 23:14
-
Anyone want to speculate what the MerlinDvac nozzle extension is made of? Looks like Niobium to me...
Steve Jurvetson says it's "niobium halfnium [sic]" : http://www.flickr.com/photos/jurvetson/10008137326/
It does look like the same stuff as on v1.0 to me.
IIRC, the S2 nozzle extension was on the critical path in the schedule, and SpaceX said it was coming from a vendor. It appears that their in-house development of a rumored carbon-carbon extension to replace the niobium wasn't far enough along to produce flight hardware for this launch, so presumably they fell back on what worked for them on v1.0. (Just my speculation, though.)
Does anyone know if they're still working on the carbon-carbon extension?
-
#252
by
gospacex
on 29 Sep, 2013 23:25
-
My hunch is the roll has accumulated during the long fall after the 3-engine burn which killed horizontal velocity component, and the stage started falling almost vertically.
With engines not running during the fall, only ACS was available to keep proper attitude. The magnitude of roll moment (due to rocket not being perfectly symmetric) was larger than SpaceX expected.
-
#253
by
titanmiller
on 30 Sep, 2013 00:03
-
If you have to counter the aero-torque, then does this require some extra new thruster, or can the existing octoweb/center-engine be tweaked to stop the roll on the way down?
Why not just aerodynamic control surfaces like fins?
-
#254
by
Lars_J
on 30 Sep, 2013 00:08
-
If you have to counter the aero-torque, then does this require some extra new thruster, or can the existing octoweb/center-engine be tweaked to stop the roll on the way down?
Why not just aerodynamic control surfaces like fins?
Because tail fins, once over a certain size, are going to make the stage harder to control as it re-enters the atmosphere tail first. Like throwing a dart in reverse.
-
#255
by
gospacex
on 30 Sep, 2013 00:12
-
Because tail fins, once over a certain size, are going to make the stage harder to control as it re-enters the atmosphere tail first. Like throwing a dart in reverse.
Why said the fins must be on the rear end?
-
#256
by
Avron
on 30 Sep, 2013 00:13
-
If you have to counter the aero-torque, then does this require some extra new thruster, or can the existing octoweb/center-engine be tweaked to stop the roll on the way down?
Why not just aerodynamic control surfaces like fins?
Because tail fins, once over a certain size, are going to make the stage harder to control as it re-enters the atmosphere tail first. Like throwing a dart in reverse.
drogue parachute off set ??
-
#257
by
Jcc
on 30 Sep, 2013 00:15
-
If you have to counter the aero-torque, then does this require some extra new thruster, or can the existing octoweb/center-engine be tweaked to stop the roll on the way down?
Why not just aerodynamic control surfaces like fins?
Because tail fins, once over a certain size, are going to make the stage harder to control as it re-enters the atmosphere tail first. Like throwing a dart in reverse.
Fins on the interstage? Never mind, they will fix the roll issue without fins.
-
#258
by
Lars_J
on 30 Sep, 2013 00:19
-
Launch video is up:
Thanks!
Here are two shots that I didn't catch in the live webcast:
- IR view of M1DVac in interstage
- fairing separation moment
-
#259
by
QuantumG
on 30 Sep, 2013 00:31
-
The legs are going to be aerosurfaces.