Author Topic: CRS-3 Falcon 9 first stage to sport legs and attempt soft splashdown  (Read 219948 times)

Offline cambrianera

  • Full Member
  • ****
  • Posts: 1435
  • Liked: 313
  • Likes Given: 258
I've simulated the situation for launch of about 9.8 kg to LEO with 42 tonnes of prop reserved or recovery.

The stage 1 state 10 seconds after MECO is
V - z              V - x      velocity   height   downrange
1,004.88   1,535.43      1,835.03   77,577   70,256

Boosting back at 6 g's takes 26 seconds to bring the horizontal velocity to zero, and it takes 140 seconds for gravity to bring the stage back to the altitude it had achieved after those 26 seconds. It then takes 60 seconds for the stage to descend to the stratopause at 50 km altitude. That means, in order for the stage to reach the stratopause it must return from the boost back burn at about 260 m/s. That horizontal velocity bleeds off as the stage approaches sea level, due to drag, but the stage still moves something like 7 km horizontally within the atmosphere. The prop used for the boost back burn (stop and return) is about 39.6 tonnes in this case.

Of course the longer the boost back burn is delayed after MECO, the less time the stage will have to travel the ever increasing distance back to the landing point. I estimate it to be more likely 20 seconds after MECO to complete 180 degrees rotation and ignition for the boost back burn. Maybe we'll find out some better numbers.

Interesting, some comments.
I believe you are talking about 9800 kg payload to LEO (not 9.8 ).
Moreover for 1800 m/s dv the rocket equation gives me mini/mfin = 1.8 assuming ISP = 311
This means that you assumed a dry mass higher than 40 ton?

Oh to be young again. . .

Offline Jamie T

  • Member
  • Posts: 26
  • Earth
  • Liked: 21
  • Likes Given: 1
Hi.. first post ever here.. 

I find you guys all go pretty crazy on the details.. certainly calculating the buoyancy mechanics of a floating empty stage in the middle of the ocean is neat.. but rather moot considering that once SpaceX demonstrates a fixed point 'soft' landing I doubt they'll ever do this again.  Once they have the Boca Chica site operational I'd expect them to land the stages in Florida.. I know that's what I'd do.

As for the legs falling off in accent I find this kinda low risk? I mean don't they attach solid rocket motors and other stuff to the outside of rockets all the time? and NASA be frakked that they didn't pay for testing of this on their launches?  I thought this is exactly what NASA paid for and wants.

As for deployment.. each leg are operated by a single compressed gas strut just like you find in your car? a little more fancy albeit but certainly very elegant and simple.

So first re entry burn occurs and the explosive bolts securing the legs are fired.. the legs start extending but can't extend any more than the aerodynamic pressures on the feet/legs allow.. so the rocket takes a 'shuttlecock' configuration.. which I think is a nice stable way of entering into the atmosphere and slowing into subsonic speeds.

As the rocket slows down and the pressures decrease the legs slowly continue to swing outward and down.. and on the final burn before landing the legs if still not fully deployed are snapped open from the de acceleration G's and are deployed fully.

touch down.. party time!

J
« Last Edit: 02/27/2014 09:59 pm by Jamie T »

Offline QuantumG

  • Senior Member
  • *****
  • Posts: 8830
  • Australia
  • Liked: 3832
  • Likes Given: 917
Welcome to the forum!
Human spaceflight is basically just LARPing now.

Offline dorkmo

  • Full Member
  • ****
  • Posts: 707
  • Liked: 336
  • Likes Given: 845
Has anyone mentioned that the legs were built by Swift Engineering?

Please disregard if this is old news.

interesting! apparently switch engineering is owned by the grandson of panasonic founder. (wikipedia) funny since panasonic is telsa's battery supplier.

Offline Rocket Science

  • Senior Member
  • *****
  • Posts: 9446
  • NASA Educator Astronaut Candidate Applicant 2002
  • Liked: 3296
  • Likes Given: 9189
Has anyone mentioned that the legs were built by Swift Engineering?

Please disregard if this is old news.
Constructor of Formula Ford Racers! :)
"The laws of physics are unforgiving"
~Rob: Physics instructor, Aviator, Vintage auto racer

Offline corrodedNut

  • Full Member
  • ****
  • Posts: 1543
  • Liked: 213
  • Likes Given: 130
Has anyone mentioned that the legs were built by Swift Engineering?

Please disregard if this is old news.
Constructor of Formula Ford Racers! :)

I thought QG was just making a bad pun

Offline QuantumG

  • Senior Member
  • *****
  • Posts: 8830
  • Australia
  • Liked: 3832
  • Likes Given: 917
Has anyone mentioned that the legs were built by Swift Engineering?

Please disregard if this is old news.
Constructor of Formula Ford Racers! :)

I thought QG was just making a bad pun

Ha! I found this on the google:

http://globenewswire.com/news-release/2012/06/18/479610/259461/en/AC-A-is-Pleased-to-Announce-the-Appointment-of-the-Former-Director-of-Composites-at-SpaceX-Steve-Smith-to-Manager-of-Composite-Manufacturing-at-AC-A.html

Quote
Prior to SpaceX, Smith served as the Composite Shop Manager at Swift Engineering out of San Clemente, California where he over-saw the composite production of experimental aircraft, Indy cars UAV's and structures.

It's funny how these things work, isn't it?
Human spaceflight is basically just LARPing now.

Online Comga

  • Senior Member
  • *****
  • Posts: 4613
  • Liked: 1882
  • Likes Given: 1551
Hi.. first post ever here.. 
Welcome to the forum

Quote
As for deployment.. each leg are operated by a single compressed gas strut just like you find in your car? a little more fancy albeit but certainly very elegant and simple.
It's never really that simple.  Also, look for the discusion of the pivot points, and why the pistons can't start the motion.

Quote
So first re entry burn occurs and the explosive bolts securing the legs are fired..
SpaceX has eschewed explosive bolts.  They use motorized latches.
We also don't know if there will be a re-entry burn.  SpaceX may be trying to minimize or eliminate it.

Quote
the legs start extending but can't extend any more than the aerodynamic pressures on the feet/legs allow.. so the rocket takes a 'shuttlecock' configuration.. which I think is a nice stable way of entering into the atmosphere and slowing into subsonic speeds

As the rocket slows down and the pressures decrease the legs slowly continue to swing outward and down.. and on the final burn before landing the legs if still not fully deployed are snapped open from the de acceleration G's and are deployed fully..
We have no evidence that this is how it will be configured and a strong indication that it won't be.  See Jim's post above.
My impression is that the last statent is part of the plan.  The upward acceleration of the stage will aid the deployment of the legs.
What kind of wastrels would dump a perfectly good booster in the ocean after just one use?

Offline Jamie T

  • Member
  • Posts: 26
  • Earth
  • Liked: 21
  • Likes Given: 1
Cheers Comga for the clarification..

Happy to be part of the forums.. although I'm hardly anything other than a armchair enthusiast..   I can contribute 3D models and animation.. that I'm good at!

thx

J

Offline aero

  • Senior Member
  • *****
  • Posts: 3050
  • 92129
  • Liked: 827
  • Likes Given: 294
I've simulated the situation for launch of about 9.8 kg to LEO with 42 tonnes of prop reserved or recovery.

The stage 1 state 10 seconds after MECO is
V - z              V - x      velocity   height   downrange
1,004.88   1,535.43      1,835.03   77,577   70,256

Boosting back at 6 g's takes 26 seconds to bring the horizontal velocity to zero, and it takes 140 seconds for gravity to bring the stage back to the altitude it had achieved after those 26 seconds. It then takes 60 seconds for the stage to descend to the stratopause at 50 km altitude. That means, in order for the stage to reach the stratopause it must return from the boost back burn at about 260 m/s. That horizontal velocity bleeds off as the stage approaches sea level, due to drag, but the stage still moves something like 7 km horizontally within the atmosphere. The prop used for the boost back burn (stop and return) is about 39.6 tonnes in this case.

Of course the longer the boost back burn is delayed after MECO, the less time the stage will have to travel the ever increasing distance back to the landing point. I estimate it to be more likely 20 seconds after MECO to complete 180 degrees rotation and ignition for the boost back burn. Maybe we'll find out some better numbers.

Interesting, some comments.
I believe you are talking about 9800 kg payload to LEO (not 9.8 ).
Moreover for 1800 m/s dv the rocket equation gives me mini/mfin = 1.8 assuming ISP = 311
This means that you assumed a dry mass higher than 40 ton?

Yes, 9.8 tonnes.

My mass ratio is 1.801092624 as is yours. Dry mass 20,000.00 kg, wet mass 61,528.41 kg at the start of the BB burn. My numerical simulation on launch stages when the remaining prop is lower than the limit. Since stage 1 burns about 2 tonnes/second of prop, stopping the simulated burn at any precise  prop mass value is not practical. Hence 42 tonnes fuel reserved for landing but only 41,528 kg remained at simulated MECO.
Retired, working interesting problems

Offline meadows.st

  • Full Member
  • *
  • Posts: 152
  • Vancouver BC, Canada
  • Liked: 88
  • Likes Given: 3998
meadows.st:  piece-wise integration of your velocity curves yields around 140 meters.
That is about twice the height of the first stage.
Perhaps you have a factor of two in a wrong place?

Very good point, I get about 100m rather than 140 and the plot is for the tip which will scribe an arc of around 78m (50 * 2 * pi/4) so there is obviously a problem but it doesn't look like a factor of two but something is definitely wrong.  Thanks for the suggestion.

edit:
Duh... forest for the trees issue... I have been maintaining a constant moment of inertia even though the pivot point changes. 

No. DOH on my part!
It wasn't a factor of 2.  It was Pi/2 like you suggested.
My bad.
The changing moment of inertia is a smaller issue, although I do see your point.

Despite one comment regarding the futility of this exercise (in my own defense, I am a nerd :P and I also play sports which IMHO are equally futile, and equally enjoyable) I started this exercise to answer the question, will there be a chance to salvage and study the stage or will it split open and therefore sink?

For those that don't care about the details but are interested in the results, I now believe that the system will come to rest in between 6.5 and 8 seconds and the tip of the interstage will have slowed down to a relatively gentle 1 to 4 m/s before contacting the surface of the water (after reaching a peak velocity of between 18 and 25 m/s). (see pic below) The stage will come to rest essentially horizontal and I believe it will be in one piece although there are some some significant torques generated during the tipping sequence so hopefully the pressurized tanks will be strong enough to prevent buckling.

I have appreciated the feedback and input most especially from those of you who are much smarter than I am as I attempt to refine my (admittedly oversimplified) model. Thanks especially to @Comga for pointing me in the right direction after spotting a rookie mistake that caused me to display an entirely impossible physical. I have found that the problems in my original calculations were due to two things (both factors of trying to do things too simply first then not bubbling the refinements through all of my equations): 1) I calculated the Moment of Inertia as if the entire system pivoted around a fixed point (where the "downhill" side of the body intersected the water - this was the starting pivot point of the rotation but certainly not the final pivot point); and 2) I was only calculating the delta theta based on the gravity and buoyancy moments and a constant Moment of Inertia (since I introduced drag vectors due to water (mostly from legs) and air resistance (of the falling body) later in my process - I did not carry these back to the seminal angle) which caused the dilation of distance traveled by the tip.

My original hypothesis (based only on an intuitive guess) was that the system would sink to about 1/4 to 1/3 of its height and relatively slowly settle onto its side. It seems pretty clear (now) that I have no intuitive understanding for how light the F9S1 dry mass (no pun intended) really is - it is closer to a cork and only submerges to around 10% of its length at the point of alightment.  The ranges of dry mass (including legs, residual propellant, engines, I/S, etc) estimated by this group and a number of other sources ranges between ~20mt (per @aero and others) and ~30mt (spacelaunchreport) but I found that the model was more sensitive to the assumed non-buoyant length at the base of the rocket (which I call "open at base" in the simulation results picture) than to the entire range of potential masses that I simulated in the model.

-Steve
« Last Edit: 02/28/2014 02:13 am by meadows.st »
A little rudder far from the rocks is a lot better than a lot of rudder close to the rocks. L. David Marquet

Offline aero

  • Senior Member
  • *****
  • Posts: 3050
  • 92129
  • Liked: 827
  • Likes Given: 294
Did you consider the buoyancy of the helium cylinders? I guess them to be about 10 inches in diameter, about .25 meters. Cross sectional area =~ .05 meters, buoyancy per meter ~ .05*1000 = 50 kg/meter. Not a big force but a force none the less acting through a lever arm that diminishes as the stage tips over.
Retired, working interesting problems

Offline meadows.st

  • Full Member
  • *
  • Posts: 152
  • Vancouver BC, Canada
  • Liked: 88
  • Likes Given: 3998
Did you consider the buoyancy of the helium cylinders? I guess them to be about 10 inches in diameter, about .25 meters. Cross sectional area =~ .05 meters, buoyancy per meter ~ .05*1000 = 50 kg/meter. Not a big force but a force none the less acting through a lever arm that diminishes as the stage tips over.

Short answer, No. Longer answer is: I know both the legs and pneumatic cylinders have some buoyant effect but I didn't take the time to include them in the model. (It is on my list of things to do, as listed on the last slide of the slide deck from my previous post, but I don't know how much additional time I want to spend on it now).
A little rudder far from the rocks is a lot better than a lot of rudder close to the rocks. L. David Marquet

Offline MichaelRBrumm

  • Member
  • Posts: 16
  • Liked: 10
  • Likes Given: 16
NASA knew it was possible this vehicle would have legs. They had no problem with that. It was unlikely as SpaceX would press ahead with having legs on the vehicle until they knew they would integrate on the vehicle, which came late in the flow. They very nearly didn't go with legs on this F9.

Now they have got legs on it. NASA is still fine with that.

I would venture to say that NASA is more than fine with legs being added. Contrary to what some people here seem to believe, NASA actually welcomes advances in space technology. "NASA's mission is to pioneer the future in space exploration, scientific discovery, and aeronautics research." (from NASA's own websites)

If SpaceX weren't being so proactive about doing it themselves, they could actually propose the addition of legs as a separate reusability research program, which NASA would partially fund. So, NASA is more than willing to risk loosing some relatively inexpensive cargo for a major advance like this, as long as SpaceX can demonstrate that they have taken the usual risk evaluations and mitigations.

Loosing the cargo is not going to cause a major issues for NASA. They expect (in the form of risk assessment probabilities) to loose cargo on new launch vehicles, and plan accordingly. SpaceX is taking all the risks here with the costs of the launch.

Now, if it were a manned mission... That would be a completely different story.

Offline cambrianera

  • Full Member
  • ****
  • Posts: 1435
  • Liked: 313
  • Likes Given: 258
I've simulated the situation for launch of about 9.8 kg to LEO with 42 tonnes of prop reserved or recovery.

The stage 1 state 10 seconds after MECO is
V - z              V - x      velocity   height   downrange
1,004.88   1,535.43      1,835.03   77,577   70,256

Boosting back at 6 g's takes 26 seconds to bring the horizontal velocity to zero, and it takes 140 seconds for gravity to bring the stage back to the altitude it had achieved after those 26 seconds. It then takes 60 seconds for the stage to descend to the stratopause at 50 km altitude. That means, in order for the stage to reach the stratopause it must return from the boost back burn at about 260 m/s. That horizontal velocity bleeds off as the stage approaches sea level, due to drag, but the stage still moves something like 7 km horizontally within the atmosphere. The prop used for the boost back burn (stop and return) is about 39.6 tonnes in this case.

Of course the longer the boost back burn is delayed after MECO, the less time the stage will have to travel the ever increasing distance back to the landing point. I estimate it to be more likely 20 seconds after MECO to complete 180 degrees rotation and ignition for the boost back burn. Maybe we'll find out some better numbers.

Interesting, some comments.
I believe you are talking about 9800 kg payload to LEO (not 9.8 ).
Moreover for 1800 m/s dv the rocket equation gives me mini/mfin = 1.8 assuming ISP = 311
This means that you assumed a dry mass higher than 40 ton?

Yes, 9.8 tonnes.

My mass ratio is 1.801092624 as is yours. Dry mass 20,000.00 kg, wet mass 61,528.41 kg at the start of the BB burn. My numerical simulation on launch stages when the remaining prop is lower than the limit. Since stage 1 burns about 2 tonnes/second of prop, stopping the simulated burn at any precise  prop mass value is not practical. Hence 42 tonnes fuel reserved for landing but only 41,528 kg remained at simulated MECO.

Thanks of the answer (and not looking at third or fourth digits  ;) ) but something still missing.
You say The prop used for the boost back burn (stop and return) is about 39.6 tonnes but  61500/(61500-39600)=2.8 (not 1.8 ).
Moreover each engine burns about 230 kg/s.
Three engines * 26 sec is 230*3*26=17940 kg.

The fuel for the brake-RTLS burn should be about half of 39.6 ton, therefore lower performance hit.
Oh to be young again. . .

Offline cambrianera

  • Full Member
  • ****
  • Posts: 1435
  • Liked: 313
  • Likes Given: 258
Did you consider the buoyancy of the helium cylinders? I guess them to be about 10 inches in diameter, about .25 meters. Cross sectional area =~ .05 meters, buoyancy per meter ~ .05*1000 = 50 kg/meter. Not a big force but a force none the less acting through a lever arm that diminishes as the stage tips over.

Short answer, No. Longer answer is: I know both the legs and pneumatic cylinders have some buoyant effect but I didn't take the time to include them in the model. (It is on my list of things to do, as listed on the last slide of the slide deck from my previous post, but I don't know how much additional time I want to spend on it now).

Thanks meadows.st,
I think your work is good enough to show that the stage has 99% probability to survive toppling.
The fact you were conservative on some issues (Cd and legs/cylinders buoyancy) shows how much margin there is against the possibility of damage due to toppling.

Much appreciated  :D
Oh to be young again. . .

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32650
  • Cape Canaveral Spaceport
  • Liked: 11550
  • Likes Given: 344
So, NASA is more than willing to risk loosing some relatively inexpensive cargo for a major advance like this, as long as SpaceX can demonstrate that they have taken the usual risk evaluations and mitigations.

Loosing the cargo is not going to cause a major issues for NASA.

Not applicable in this case.  The ISS needs space suits.

Online LouScheffer

  • Full Member
  • ****
  • Posts: 1999
  • Liked: 2789
  • Likes Given: 299
So, NASA is more than willing to risk loosing some relatively inexpensive cargo for a major advance like this, as long as SpaceX can demonstrate that they have taken the usual risk evaluations and mitigations.

Loosing the cargo is not going to cause a major issues for NASA.

Not applicable in this case.  The ISS needs space suits.
In that case, shouldn't they put one on this Dragon, then another on the next Cygnus?  That way it avoids the most serious problem should either of the two (both new) providers fail.

Offline Rocket Science

  • Senior Member
  • *****
  • Posts: 9446
  • NASA Educator Astronaut Candidate Applicant 2002
  • Liked: 3296
  • Likes Given: 9189
Has anyone mentioned that the legs were built by Swift Engineering?

Please disregard if this is old news.
Constructor of Formula Ford Racers! :)

I thought QG was just making a bad pun

Ha! I found this on the google:

http://globenewswire.com/news-release/2012/06/18/479610/259461/en/AC-A-is-Pleased-to-Announce-the-Appointment-of-the-Former-Director-of-Composites-at-SpaceX-Steve-Smith-to-Manager-of-Composite-Manufacturing-at-AC-A.html

Quote
Prior to SpaceX, Smith served as the Composite Shop Manager at Swift Engineering out of San Clemente, California where he over-saw the composite production of experimental aircraft, Indy cars UAV's and structures.

It's funny how these things work, isn't it?
Yes, they did... A lot of synergy between aerospace and motorsports engineering, for example; Bob Riley worked Apollo IIRC and builds Grand Am racers. Alas my early racing days never got me into an Indy car but one of my on track peers from those days went on to a 2nd place finish at the 500 and another went on to a class win at LeMans for Team Corvette! :)
« Last Edit: 03/01/2014 12:57 am by Rocket Science »
"The laws of physics are unforgiving"
~Rob: Physics instructor, Aviator, Vintage auto racer

Online yg1968

  • Senior Member
  • *****
  • Posts: 9822
  • Liked: 1502
  • Likes Given: 904
On the safety issue, one of the advantages of Soyuz (and of Falcon 9 eventually) over Shuttle is that they have an uncrewed version and a crewed version. They can take more risks on the uncrewed version for new components or features. Obviously, anytime you add something new, you are adding some risks. But having a design that never evolves is not a good idea for an agency that is expected to be innovative.
« Last Edit: 02/28/2014 02:30 pm by yg1968 »

Tags: