### Author Topic: ASDS fly-back recovery  (Read 13401 times)

#### joek

• Senior Member
• Posts: 2800
• Liked: 567
• Likes Given: 337
##### ASDS fly-back recovery
« on: 04/09/2016 09:38 PM »
Moved from CRS-8 discussion thread.  Bits and pieces of this discussion are spread over many threads.  Musk indicated that fly-back from ASDS may be in the cards (as opposed to waiting for ASDS to return the stage to port).  That could  contribute to "rapid reuse".  This is the thread to discuss trade-offs and express your opinions.
« Last Edit: 04/09/2016 09:44 PM by joek »

#### joek

• Senior Member
• Posts: 2800
• Liked: 567
• Likes Given: 337
##### Re: ASDS fly-back recovery
« Reply #1 on: 04/09/2016 09:39 PM »
Moved from the CRS-8 discussion thread...

"Rapid reuse" has many possible interpretations:
1. Time to recover a stage (e.g., transport).
2. Time to make a recovered stage ready for re-flight (e.g., refurbish).
3. Time to make a stage available for re-flight (e.g., transport+refurbish).

Those times will be driven primarily by launch rate and inventory costs.  Inventory costs will be driven by how many stages are required in the pipeline to maintain a given launch rate given time (3) to make the stage available for re-flight.

The general figure of merit for these types of calculations is the inventory costs (or the costs of non-revenue-generating assets, which is why, e.g., airlines are so obsessed with turnaround times).  So let's try some BOTE...

A. Near-term optimistic with ASDS barge recovery: Launch rate: once every 168hr (7 days)
1. Time to recover stage = 72hr (barge)
2. Time to refurbish stage - 24hr
3. Total time to make stage available for reflight: 96hr.
= Stage inventory required = 92/168 = 0.547 stages.

B. Long-term optimistic with ASDS barge recovery: Launch rate: once every 4hr
1. Time to recover stage = 72hr (barge)
2. Time to refurbish stage = 2hr
3. Total time to make stage available for reflight: 74hr.
= Stage inventory required = 74/4 = 18.500 stages.

C. Near-term optimistic with ASDS fly-back recovery: Launch rate: once every 168hr (7 days)
1. Time to recover stage = 2hr
2. Time to refurbish stage = 2hr
3. Total time to make stage available for reflight: 4hr.
= Stage inventory required = 4/168 = 0.023 stages.

D. Long-term optimistic with ASDS fly-back recovery: Launch rate: once every 4hr
1. Time to recover stage = 2hr
2. Time to refurbish stage = 2hr
3. Total time to make stage available for reflight: 4hr.
= Stage inventory required = 4/4 = 1.000 stages.

... somewhere in there is an answer, but I don't think it's nearly as simple as most make it out to be.  Specifically, fly-back recovery from ASDS is unlikely to pay off until there is a significant increase in flight rate, or unless the stage inventory cost is very high (another discussion).  And yes, there are many more factors to consider; knock yourself out.

#### sewebster

• Full Member
• Posts: 236
• British Columbia
• Liked: 171
• Likes Given: 85
##### Re: ASDS fly-back recovery
« Reply #2 on: 04/09/2016 09:41 PM »
Well, I guess the legs would have to fold back up.

#### meekGee

• Senior Member
• Posts: 7894
• N. California
• Liked: 4141
• Likes Given: 843
##### Re: ASDS fly-back recovery
« Reply #3 on: 04/09/2016 09:46 PM »
Well, I guess the legs would have to fold back up.

Yup.  And that's true for the other alternative, which is shipping the stage back by fast-boat after horizontalizing it.

There are certainly more tricks coming up with the ASDS, one way or the other.

ABCD - Always Be Counting Down

#### schaban

• Full Member
• Posts: 111
• Liked: 22
• Likes Given: 17
##### Re: ASDS fly-back recovery
« Reply #4 on: 04/09/2016 09:57 PM »
grasshopper didn't need to fold legs for re-fly.

However, keeping pretty big volume of LOX and RP-1 where hot rocket is coming in is a pretty dangerous proposition. Ensured loss of barge in cases, similar to SES9...

Also, safing/checkouts and refueling would probably take a day to perform. so, for typical ship lending, how much time you would actually save?

same saving would be achieved if you make ship go 10 knots instead of current 5...

#### meekGee

• Senior Member
• Posts: 7894
• N. California
• Liked: 4141
• Likes Given: 843
##### Re: ASDS fly-back recovery
« Reply #5 on: 04/09/2016 10:19 PM »
grasshopper didn't need to fold legs for re-fly.

However, keeping pretty big volume of LOX and RP-1 where hot rocket is coming in is a pretty dangerous proposition. Ensured loss of barge in cases, similar to SES9...

Also, safing/checkouts and refueling would probably take a day to perform. so, for typical ship lending, how much time you would actually save?

same saving would be achieved if you make ship go 10 knots instead of current 5...

If Musk talked about it publicly, I think the trivial feasibility aspects can be taken for granted.  It's not like they forgot to consider the refueling logistics.

What's at play is whether it will prove worth while.

I really like the idea of fly-back, since it's a potentially same-day solution, compared to basically a two-week cycle for a barge that is 1000 km downrange.

But using an offloading crane and carrier boats also allows the barge to loiter indefinitely, you can launch as soon as the rocket's offloaded (so it doesn't hurt the launch rate), but you just have potentially several stages en-route.

It's all a function of launch rate.   For 1/week, I'd say an offloading crane is plenty good.   For 1/day, you'll want fly-back.

ABCD - Always Be Counting Down

#### John Alan

• Full Member
• Posts: 525
• Central IL - USA - Earth
• Liked: 261
• Likes Given: 1283
##### Re: ASDS fly-back recovery
« Reply #6 on: 04/09/2016 10:20 PM »
My 2 cents...
I think the process will remain as is for a while... float it back to shore...
I think OCISLY will be busy over the next several years with few modifications...
I hope they start painting stars or something on her for each mission...

BUT..at some point... TBD
Spacex will have to design and invest in a much larger manned platform to act as a catch platform...
What that may look like is up in the air at this time... I don't think they know even as of yet...
The cost to build such a beast has to be recovered by the savings of recycling cores and then some...

But in short... a crew can "shelter in place" under an armored deck as a stage comes in...

Then either (depending on what is decided for a plan and process)
They can then secure it and crane it onto a launch mount on the ship... send it back to land with a hop...
OR...
They secure it and stow it horizontally below deck and begin the refurb process even..

In summary... this may all be pie in the sky crazy ideas that make no sense \$\$\$\$\$ wise...
But... maybe it does...

Either way... I think the current ship design is WAY too small to add relaunch ability to it's bag of tricks...
Replacement is needed... in my opinion... and only if there is \$\$\$ to be made or saved to cover costs...
« Last Edit: 04/09/2016 10:23 PM by John Alan »

#### joek

• Senior Member
• Posts: 2800
• Liked: 567
• Likes Given: 337
##### Re: ASDS fly-back recovery
« Reply #7 on: 04/09/2016 10:23 PM »
grasshopper didn't need to fold legs for re-fly.

Grasshopper is in no way comparable; its legs are completely different than F9v1.2.

Quote
Also, safing/checkouts and refueling would probably take a day to perform. so, for typical ship lending, how much time you would actually save?

same saving would be achieved if you make ship go 10 knots instead of current 5...

Good points...
- How much work can you do on-ship that would othewise be done on land?  Unknown.
- How much recovery time could you save by simply increaseing ship speed?  Unknown.
...and great questions.
It is often these simple optimizations that pay significant dividends.

#### CorvusCorax

• Full Member
• Posts: 358
• Germany
• Liked: 384
• Likes Given: 205
##### Re: ASDS fly-back recovery
« Reply #8 on: 04/09/2016 10:23 PM »
Fuel and Oxygen should not be kept on board the barge to minimize Falcon 9's chances of hit and sink her. But that's not a big problem, they have support vessels nearby on standby that approach the barge anyway to secure and safe the Rocket and would have to do a checkup anyway before refueling.

A big point in favor of flyback is to minimize exposure of the rocket to salt water. Especially on GEO missions, the barge is going to be several hundred miles out at sea and would take several days to come back during which the rocket will be exposed to a very corrosive seawater environment (salt water spray from waves, etc...) especially when the weather is less than ideal. Flying it back would reduce the exposure to this, which might be even more important than a fast turnaround time.

When they say "quick turnaround" what they usually mean is "low effort turnaround" - aka no work extensive overhauls, no dis- and reassembly. If there's enough stages "in the queue" (which there has to be anyway for contingency reasons in case a core is lost) then the actual time needed to recycle the stage is less important than how work intensive the stage is during that time.

As such, the time needed for the stage to transfer home would likely matter a lot less than the corrosive effect of several days of salty seawater spray.

What also might matter if launch cadence gets high is the availability of the drone ship itself. They need a lot more drone ships if each of them needs to travel back to port and back out after each recovery. (Then again having a few spare drone ships wouldn't hurt too much in case one sprouts a hole again or becomes otherwise unavailable for maintenance, overhaul or repair)

#### karki

• Member
• Posts: 30
• Liked: 9
• Likes Given: 51
##### Re: ASDS fly-back recovery
« Reply #9 on: 04/09/2016 10:40 PM »
What also might matter if launch cadence gets high is the availability of the drone ship itself. They need a lot more drone ships if each of them needs to travel back to port and back out after each recovery. (Then again having a few spare drone ships wouldn't hurt too much in case one sprouts a hole again or becomes otherwise unavailable for maintenance, overhaul or repair)

I think this invalidates option B. Would need 19 barges and support ships (assuming 19 flights in a row that all need ASDS landings). Doesn't seem likely.

#### douglas100

• Senior Member
• Posts: 2157
• Liked: 217
• Likes Given: 96
##### Re: ASDS fly-back recovery
« Reply #10 on: 04/09/2016 11:02 PM »

...Yup.  And that's true for the other alternative, which is shipping the stage back by fast-boat after horizontalizing it...

Yes, that would be my solution to increasing the flight rate. Also I think it would be easier (and cheaper) to increase the stage inventory than get involved with complex processing ops off shore.
Douglas Clark

#### meekGee

• Senior Member
• Posts: 7894
• N. California
• Liked: 4141
• Likes Given: 843
##### Re: ASDS fly-back recovery
« Reply #11 on: 04/09/2016 11:06 PM »

...Yup.  And that's true for the other alternative, which is shipping the stage back by fast-boat after horizontalizing it...

Yes, that would be my solution to increasing the flight rate. Also I think it would be easier (and cheaper) to increase the stage inventory than get involved with complex processing ops off shore.

The safe statement is that true gas-and-go is a pre-requisite for fly-back.
ABCD - Always Be Counting Down

#### schaban

• Full Member
• Posts: 111
• Liked: 22
• Likes Given: 17
##### Re: ASDS fly-back recovery
« Reply #12 on: 04/09/2016 11:36 PM »
I really like the idea of fly-back, since it's a potentially same-day solution, compared to basically a two-week cycle for a barge that is 1000 km downrange.

Where is such a difference in estimated time is coming from? 1000km/5kt is basically 5 days.

Remember, after fly-back, stage still need to be transported horizontally to the same hangar. So, difference in ground post fly/ship delivery is, literally, between tracking from port vs. tracking from landing zone; which, in the case of Port Canaveral, is negligible.

From what I see, the difference is really only between time it take for ship to get to the port vs pre-re-flight checks and fuel load. And it actually not only fuel load, you need helium, nitrogen and whatever liquid used for fins. And I'm not sure if everything could be loaded from the stage bottom in current modification.

#### CorvusCorax

• Full Member
• Posts: 358
• Germany
• Liked: 384
• Likes Given: 205
##### Re: ASDS fly-back recovery
« Reply #13 on: 04/09/2016 11:43 PM »
One potential issue with "gas and go" from the drone ship is that you don't have hold down clamps. You also have no payload, so the stage is relatively light - even when fuelled - that means even partial engine thrust will possibly give you >1.0 thrust2weight (depending on how many engines are used)

The last (and for a quick refuel and go maybe only) possible point in time to see a major engine problem is when they ignite and ramp up thrust.  During a regular launch, the computer aborts and the clamps don't release if there's something wrong.

Sitting "loose" on its legs, the rocket has no such option. It also can't ramp up thrust too "slowly" (as to stay below 1:1 thrust2lift for the self check), as that would put a helluva thermal load on the deck. You want to get clear asap once that million-pounds-of-thrust-plasma-torch is lit.

At least with the current drone ship. Of course they might build a bigger one that has an inbuilt flame trench and launch clamps, cranes, etc...

Maybe SpaceX could buy or lease a decommissioned aircraft carrier - or an oil rig - from someone Does the old sea-launch rig still exist?

« Last Edit: 04/09/2016 11:48 PM by CorvusCorax »

#### Rhyshaelkan

• Full Member
• Posts: 260
• Liked: 22
• Likes Given: 33
##### Re: ASDS fly-back recovery
« Reply #14 on: 04/09/2016 11:55 PM »
FH side boosters will probably land close to launch site. While the middle core will likely be caught at sea. However you already have two back at base. One additional "in-stock" core would allow another FH launch relatively quick.

I think two slow boats going out to catch and retrieve would be simpler and more cost effective in a long run. Rather than having a cryo support ship in addition to the barge to refuel. As others have said, having fuel and lox on the barge during recovery is a bad idea.

KISS... just have a barge or 3 to go out and catch. All of this dependent on flight rate.
I am not a professional. Just a rational amateur dreaming of mankind exploiting the universe.

#### dnavas

• Full Member
• Posts: 115
• San Jose
• Liked: 19
• Likes Given: 70
##### Re: ASDS fly-back recovery
« Reply #15 on: 04/09/2016 11:56 PM »
Of course they might build a bigger one that has an inbuilt flame trench and launch clamps, cranes, etc...

At which point, why not not just launch directly from the platform?  In fact, have a few, launch from Texas and land at BFP1, relaunch and land at BFP2, relaunch and land in Florida.  You wind up only needing to "return" a core from a launch in Florida.  Bonus that there are likely a few oil rigger guys used to being stationed out in the Gulf....

#### CorvusCorax

• Full Member
• Posts: 358
• Germany
• Liked: 384
• Likes Given: 205
##### Re: ASDS fly-back recovery
« Reply #16 on: 04/10/2016 12:13 AM »
Launching from sea might actually be a goodviable idea - long term.

The issues with launching from Kwajalein but also for Sea Launch originate - at least partially - from the high cost and time needed to get the launch vehicle there.

Of course if you can build the launch vehicle in Texas, then launch it on a suborbital hop out to the launch platform further south in the Caribbean (on a nice sunny and calm day) mate with 2nd stage and payload there (flown in by helicopter) and then launch into orbit from there.

Of course you still have the issue that high orbit launches cannot land at launch site - regardless whether the launch site is on land or at sea.

Advantages include - you can be in international waters. No FAA licenses required, you can launch at any time on short notice and do all kinds of sh***t - and also you can keep stuff secret if needed. People wouldn't even know a launch took place until your orbital vehicle shows up on an African radar or Telescope.

The launch site might be much more convenient (aka closer to the equator) than anything on mainland US soil.

Replace the term "platform" for "Caribbean volcano island" as needed

• Senior Member
• Posts: 2707
• Boca Chica, Texas
• Liked: 3606
• Likes Given: 243
##### Re: ASDS fly-back recovery
« Reply #17 on: 04/10/2016 12:32 AM »
In the Texas thread, they've been trying to figure out how employees would get to work from South Padre. What better way than to shuttle them with a used F9?

#### meekGee

• Senior Member
• Posts: 7894
• N. California
• Liked: 4141
• Likes Given: 843
##### Re: ASDS fly-back recovery
« Reply #18 on: 04/10/2016 12:51 AM »
I really like the idea of fly-back, since it's a potentially same-day solution, compared to basically a two-week cycle for a barge that is 1000 km downrange.

Where is such a difference in estimated time is coming from? 1000km/5kt is basically 5 days.

Remember, after fly-back, stage still need to be transported horizontally to the same hangar. So, difference in ground post fly/ship delivery is, literally, between tracking from port vs. tracking from landing zone; which, in the case of Port Canaveral, is negligible.

From what I see, the difference is really only between time it take for ship to get to the port vs pre-re-flight checks and fuel load. And it actually not only fuel load, you need helium, nitrogen and whatever liquid used for fins. And I'm not sure if everything could be loaded from the stage bottom in current modification.

So yeah - 5 days each way, plus "ocean induced variability"...

As for the complexity of fly-back - sure.  It's just that since it's been broached by Musk, I'd think that such obvious issues have been addressed at a fundamental level.
ABCD - Always Be Counting Down

#### meekGee

• Senior Member
• Posts: 7894
• N. California
• Liked: 4141
• Likes Given: 843
##### Re: ASDS fly-back recovery
« Reply #19 on: 04/10/2016 12:54 AM »
One potential issue with "gas and go" from the drone ship is that you don't have hold down clamps. You also have no payload, so the stage is relatively light - even when fuelled - that means even partial engine thrust will possibly give you >1.0 thrust2weight (depending on how many engines are used)

The last (and for a quick refuel and go maybe only) possible point in time to see a major engine problem is when they ignite and ramp up thrust.  During a regular launch, the computer aborts and the clamps don't release if there's something wrong.

Sitting "loose" on its legs, the rocket has no such option. It also can't ramp up thrust too "slowly" (as to stay below 1:1 thrust2lift for the self check), as that would put a helluva thermal load on the deck. You want to get clear asap once that million-pounds-of-thrust-plasma-torch is lit.

At least with the current drone ship. Of course they might build a bigger one that has an inbuilt flame trench and launch clamps, cranes, etc...

Maybe SpaceX could buy or lease a decommissioned aircraft carrier - or an oil rig - from someone Does the old sea-launch rig still exist?

If you start following the road that leads from the "fly back to shore" statement, then yes - it gets real interesting real fast.

You have to fold the legs.
The rocket is not at a known location.
You're going to be using 3 engines (at least), and take off is slower than landing
Hold-downs - definitely.
All manners of refueling

So this is not for once-per-week launch rate, as I said.  This is for the case that you want to launch every day.

ABCD - Always Be Counting Down

Tags: