That would be amazing. I wonder how difficult it will be for them to get permission to fly a first stage back to the launch site.
It’d be even more efficient if they could get permission to land downrange, e.g. launch in Texas and land the first stage in Florida, and thereby use less propellant.
It shouldn’t be that hard, given the lower speeds and reduced propellant load on return. But they’ll have to do a lot of flight testing to demonstrate their ability to control it.
Jim,
You’d only save some of the boostback propellant because you need to be able to slow down from the current staging speed to something the stage can survive reentry with. All their previous failed reuse attempts were due to the state structurally failing during reentry due to aerodynamic loads. That said yeah, it would help performance, but only along a single launch azimuth…
Of all the RTLS approaches I’ve seen my favorites are Glideback for nanosat and microsat payloads, air-launched glide-forward for medium payloads, and boostback for heavier stuff.
~Jon
If they can just soft-land it on the ocean it should at least give them
lot of good data from the forensics when they pull it apart.
Is hitting a landing zone on a barge at all reasonable? (After a whirl or two of practicing control.)
Well I would think you’d have to have a gyro stabilized barge to take out the rolling and pitching of the ocean. There will be limits to what sea state the barge could handle and still provide a stable landing platform. If the limits were exceeded your booster doesn’t have the fuel to return to the launch site as the profile of the mission assumed barge landing. If you then decided to always allow enough gas for a flyback then you lose a lot of the point of the barge.
And there might weather situations where the launch site is ok if we launch now but the weather at the barge is questionable.
Lastly, what is the legal status of a booster long beyond the territorial limit? I haven’t a clue. I assume the Shuttle’s solids fell outside territorial waters, and no one interfered with us retrieving them. Were they still owned by the US? Or were they up for salvage by anyone and no one bothered us just to be considerate?
If barge stability is an issue, a small water area multihull design has merit.
As for legal status of a stage landed in international waters, my impression is that so long as it isn’t abandoned ownership isn’t in question, and would not be open for salvage. I’ve heard that naval salvage even in the case of abandoned ships, the ships owner can retrieve the salvaged ship from the retriever for a salvage fee.
Dr. No was on TV a few weeks back; at the end Bond and Ryder are rescued by a Skyhook (slow moving airplane collects a cable attached to a balloon). Some of the CIA recon satellites photos were recovered by a similar mechanism. I suppose someone, somewhere, has proposed aerial recovery of first stages. It doesn’t appear particularly practical.
I was reading an old Boeing magazine from the 60s recently which contained an article about Saturn SIC recovery studies. That mentioned they’d considered aerial recovery before settling on parachuting it into the sea.
Seeing a plane catch an SIC stage and bring it back to the ground would have been about as awesome as the SIB parawing recovery plan.
No, Bezos patented it
I don’t suppose landing them on a barge would be practical?
OK, so I should refresh the page before commenting.
They’re getting in some good practice with Grasshopper. I don’t see why they couldn’t land on a large barge. Accounting for payload capacity to fly back to launch site, the cost of a barge with a small crew may be a net win.
Would it be advantageous to try and integrate a Merlin 1-D vacuum engine into the center position engine on stage 1? You would lose performance during a portion of the initial ascent, but the F9 v1.1 has the engine out performance margin SpaceX claims, a reduced efficiency vacuum rated engine may not be that big of a penalty. At a certain point in the climb, the vacuum engine will be more efficient than the standard Merlin 1D’s surrounding it, and overall vehicle performance may recover from the initial losses. The larger advantage with the 1D vacuum engine is in the efficiency that engine can deliver a deltaV to slow down the first stage prior to reentry into the atmosphere. That burn requires a significant amount of fuel. The higher efficiency during the retro burn should have significant mass savings on fuel. Not sure how the 1D-Vac would perform on landing though. May need to do something exciting with a specialized nozzle design for the last few thousands of meters to touchdown.
I doubt it’s worth it. Even if the propulsive delta V profile of the first stage is perfectly symmetrical it’s not as horrendous fuel use wise as it may seem at first blush. Because on the one hand you have to boost the entire stack off the launch pad and up to speed and on the other hand you merely have to slow down a mostly empty first stage, which is orders of magnitude lighter. Also, the situation is not symmetrical, because there will be gravity and aerodynamic losses on the way up and also any vertical velocity will be easily reversed merely by letting gravity do its work.
I appreciate the response, and still am trying to put some numbers to it. I. Don’t know how to figure the performance losses resulting from a suboptimal center engine firing from sea level to maybe 40,000 ft, where it’s performance should equalize out with the standard 1D’s. However working the landing sequence backwards gives encouraging results when I plug in some numbers to model the descent profile. Say the empty mass of the first stage is 13mT (as speculated by some folks, I don’t know if that’s a good number ) and you need 2000kg of fuel for the “hoverslam” landing, that means you will have a mass of 15mT after your deceleration burns in the vacuum of space. No matter what direction they choose to apply the thrust from these burns, it ultimately can be expressed as a change in Delta V, yes? So if you speculate to he needed change in delta V for the booster to survive re entry, you can compare the fuel requirements of the Merlin 1D to the 1D Vac. From what I can find, the 1d has Alan ISP of 311s, and the 1D Vac is speculated to be at 343s. If you decided that you needed to remove 3000 m/s of Delta V, QuantumG’s calculator says it will take 25,139 kg of fuel with the 1D, and 21,617 kg of fuel with a 1D Vac engine. That’s 3522 kg’s, which is more than the speculated fuel need for the entire descent. The whole concept is on the razors edge of being doable according to SpaceX, every drop will count. Not only for the descent through the atmosphere, but in slowing down the booster so it can survive reentry.
Has anybody given any consideration of the feelings of the poor naysayers? This pace of development has got to be too fast for them. They hardly have any time to say it can’t be done before it already is. This is obviously an intolerable situation.
OK, who wants to make book on who will be first with a reusable first stage, XCOR or SpaceX?
That’d be a hard bet to make.. there’s so much negotiating when one tries to compare apples to oranges.
Here’s the best part: the first stage is the easiest part to re-use (due to the milder speeds and re-entry environment) and also the most expensive part of the launcher due to its use of 9 engines and larger size. There’s still a lot of work left to do but they’ve already covered a significant amount of ground.
The impetus behind the Big Dumb Booster is that the first stage is not the most expensive part.
Historical examples are offered. Thor-Agena was what launched Discoverer-Corona, and the Agena upper stage was the expensive part.
I’m familiar with BDB designs and while they have merit I think they aren’t that practical. The problem is that propellant is the easy part, the hard parts are making high-thrust engines that don’t melt or blow up, rockets that stay on course through supersonic and hypersonic speeds, etc. In a way the Falcon 9 is the best effort at a BDB-style design that’s been done. It’s a very straightforward and simple design that has heritage all the way back to the 1950s, and it’s certainly the lowest cost rocket of its class out there. And the obvious way to add more payload is to stretch the tanks, then build bigger engines and scale up in every dimension, and so on. Which is precisely what SpaceX is doing (with the Falcon 9 v1.1, the Falcon Heavy, and the planned MCT).
Perhaps it could be done even cheaper, but I doubt it could be achievable without increasing failure rates a lot.
The more I see the numbers on multi-stage reusability the more it seems like the right way forward. The big thing that RLVs enable is flight rate. Which is valuable all on its own even if they turned out to be significantly more expensive than BDBs at first. Being able to fly hardware multiple times means you can explore the flight envelope and increase the reliability and especially confidence in the vehicle design. More so, it vastly lowers the costs of iterative development, because you can try out hardware modifications and flight profile changes using real hardware and get incredibly valuable data without having to destroy tens of millions of dollars of equipment every time.
I don’t think people fully appreciate how fast things are going to develop once reusability starts to become a reality. It’s going to lower the cost of flights, which will translate to an even larger increase in launch volume and a total increase in revenue for space launch providers. More so, it’s going to make launcher R&D vastly cheaper because of the ability to run a hell of a lot more test flights at less cost. This is a positive feedback loop.
Testing it out over the ocean is a strike of genius.
Assume a transoceanic return profile rather than RTL. What, I wonder, is the ballistic range of a reusable F9 first stage? If it’s long enough, such a stage (suitably modified) could form the basis of a Heinlein-style ballistic transport.
If I were Musk, I’d fair the Grasshopper’s landing legs in with fins and paint the stage in a red and white Tintin checkerboard pattern. Looks like a ’50s SF rocket, flies like a ’50s SF rocket…
That one happened to be nuclear. And awesome.
Build a little, test a little…
And who’s to say Elon Musk doesn’t have an atom or two hidden up his sleeve? If the Mouse can use a little Di$ney Magic to obtain legal license to “Own, operate, and maintain electric power plants… for the generation and transmission of power through nuclear fission” [1] in heavily-populated central Florida, might nor Musk use his PayPal powers to get the right to do something similar in sparsely-populated Cameron, Texas?
Hmmmmm…
So, take a Dragon Rider capsule and put it on top of a reusable Falcon 9 first stage. It boosts up to Mach 2 or 3 or so and goes exo-atmospheric, coasts on a ballistic trajectory some distance then turns around in space and feathers the atmospheric re-entry a bit to make it a little easier, falls down through the atmosphere and then comes in for a powered landing at the destination. A mobile gantry comes out and fetches the passengers then takes them off to the terminal where they can be on their way. I’d have to do the math but it might work out.
Also, if they use a full Dragon capsule instead of just integrating one into the stage then it would have its own escape and re-entry system in case anything went wrong.
You’ve got it.
Add wings to extend re-entry x-range and you have Heinlein’s transport rocket.
Flight time to Sydney: 45 minutes
Time for compulsory TSA cavity search: 90 minutes
Strange fact I just made up: The impetus for re-using the first stage actually comes from the second and third stage folks who are sick of sweating bullets as their own stages fire up, then looking over to see the first stage people flopping back in their chairs, popping open beers, lighting cigarettes and high-fiving each other. With a re-usable first stage everyone will have to sweat for the whole 15 minutes. It’s all about fairness.
That’s awesome.
When SpaceX announced last year that they were considering a launch site near Brownsville, Texas, I did some rough calculations. The distance from Brownsville to a reasonable location near the Florida Gulf Coast is roughly 1550 km. I chose Cape Coral, Florida as a possible landing site. Using a ballistic range calculator and selecting 40 km for the burnout altitude, you’d need a burnout velocity of about 3.7 km/sec to reach Florida. Allowing a bit more for atmospherics, you’re looking at closer to 4.2 km/sec. It seems to me that it’d be easier to stage and then fly the first stage to Florida than to return to the launch site in Texas. This would be more suitable for missions going to a low inclination orbit like GTO than to the ISS but I think that’s where most of their launch manifest is heading, anyway.
That would be amazing. I wonder how difficult it will be for them to get permission to fly a first stage back to the launch site.
It’d be even more efficient if they could get permission to land downrange, e.g. launch in Texas and land the first stage in Florida, and thereby use less propellant.
It shouldn’t be that hard, given the lower speeds and reduced propellant load on return. But they’ll have to do a lot of flight testing to demonstrate their ability to control it.
Jim,
You’d only save some of the boostback propellant because you need to be able to slow down from the current staging speed to something the stage can survive reentry with. All their previous failed reuse attempts were due to the state structurally failing during reentry due to aerodynamic loads. That said yeah, it would help performance, but only along a single launch azimuth…
Of all the RTLS approaches I’ve seen my favorites are Glideback for nanosat and microsat payloads, air-launched glide-forward for medium payloads, and boostback for heavier stuff.
~Jon
If they can just soft-land it on the ocean it should at least give them
lot of good data from the forensics when they pull it apart.
Is hitting a landing zone on a barge at all reasonable? (After a whirl or two of practicing control.)
Well I would think you’d have to have a gyro stabilized barge to take out the rolling and pitching of the ocean. There will be limits to what sea state the barge could handle and still provide a stable landing platform. If the limits were exceeded your booster doesn’t have the fuel to return to the launch site as the profile of the mission assumed barge landing. If you then decided to always allow enough gas for a flyback then you lose a lot of the point of the barge.
And there might weather situations where the launch site is ok if we launch now but the weather at the barge is questionable.
Lastly, what is the legal status of a booster long beyond the territorial limit? I haven’t a clue. I assume the Shuttle’s solids fell outside territorial waters, and no one interfered with us retrieving them. Were they still owned by the US? Or were they up for salvage by anyone and no one bothered us just to be considerate?
If barge stability is an issue, a small water area multihull design has merit.
As for legal status of a stage landed in international waters, my impression is that so long as it isn’t abandoned ownership isn’t in question, and would not be open for salvage. I’ve heard that naval salvage even in the case of abandoned ships, the ships owner can retrieve the salvaged ship from the retriever for a salvage fee.
Dr. No was on TV a few weeks back; at the end Bond and Ryder are rescued by a Skyhook (slow moving airplane collects a cable attached to a balloon). Some of the CIA recon satellites photos were recovered by a similar mechanism. I suppose someone, somewhere, has proposed aerial recovery of first stages. It doesn’t appear particularly practical.
I was reading an old Boeing magazine from the 60s recently which contained an article about Saturn SIC recovery studies. That mentioned they’d considered aerial recovery before settling on parachuting it into the sea.
Seeing a plane catch an SIC stage and bring it back to the ground would have been about as awesome as the SIB parawing recovery plan.
No, Bezos patented it
I don’t suppose landing them on a barge would be practical?
OK, so I should refresh the page before commenting.
They’re getting in some good practice with Grasshopper. I don’t see why they couldn’t land on a large barge. Accounting for payload capacity to fly back to launch site, the cost of a barge with a small crew may be a net win.
Would it be advantageous to try and integrate a Merlin 1-D vacuum engine into the center position engine on stage 1? You would lose performance during a portion of the initial ascent, but the F9 v1.1 has the engine out performance margin SpaceX claims, a reduced efficiency vacuum rated engine may not be that big of a penalty. At a certain point in the climb, the vacuum engine will be more efficient than the standard Merlin 1D’s surrounding it, and overall vehicle performance may recover from the initial losses. The larger advantage with the 1D vacuum engine is in the efficiency that engine can deliver a deltaV to slow down the first stage prior to reentry into the atmosphere. That burn requires a significant amount of fuel. The higher efficiency during the retro burn should have significant mass savings on fuel. Not sure how the 1D-Vac would perform on landing though. May need to do something exciting with a specialized nozzle design for the last few thousands of meters to touchdown.
I doubt it’s worth it. Even if the propulsive delta V profile of the first stage is perfectly symmetrical it’s not as horrendous fuel use wise as it may seem at first blush. Because on the one hand you have to boost the entire stack off the launch pad and up to speed and on the other hand you merely have to slow down a mostly empty first stage, which is orders of magnitude lighter. Also, the situation is not symmetrical, because there will be gravity and aerodynamic losses on the way up and also any vertical velocity will be easily reversed merely by letting gravity do its work.
I appreciate the response, and still am trying to put some numbers to it. I. Don’t know how to figure the performance losses resulting from a suboptimal center engine firing from sea level to maybe 40,000 ft, where it’s performance should equalize out with the standard 1D’s. However working the landing sequence backwards gives encouraging results when I plug in some numbers to model the descent profile. Say the empty mass of the first stage is 13mT (as speculated by some folks, I don’t know if that’s a good number ) and you need 2000kg of fuel for the “hoverslam” landing, that means you will have a mass of 15mT after your deceleration burns in the vacuum of space. No matter what direction they choose to apply the thrust from these burns, it ultimately can be expressed as a change in Delta V, yes? So if you speculate to he needed change in delta V for the booster to survive re entry, you can compare the fuel requirements of the Merlin 1D to the 1D Vac. From what I can find, the 1d has Alan ISP of 311s, and the 1D Vac is speculated to be at 343s. If you decided that you needed to remove 3000 m/s of Delta V, QuantumG’s calculator says it will take 25,139 kg of fuel with the 1D, and 21,617 kg of fuel with a 1D Vac engine. That’s 3522 kg’s, which is more than the speculated fuel need for the entire descent. The whole concept is on the razors edge of being doable according to SpaceX, every drop will count. Not only for the descent through the atmosphere, but in slowing down the booster so it can survive reentry.
Has anybody given any consideration of the feelings of the poor naysayers? This pace of development has got to be too fast for them. They hardly have any time to say it can’t be done before it already is. This is obviously an intolerable situation.
OK, who wants to make book on who will be first with a reusable first stage, XCOR or SpaceX?
That’d be a hard bet to make.. there’s so much negotiating when one tries to compare apples to oranges.
Here’s the best part: the first stage is the easiest part to re-use (due to the milder speeds and re-entry environment) and also the most expensive part of the launcher due to its use of 9 engines and larger size. There’s still a lot of work left to do but they’ve already covered a significant amount of ground.
The impetus behind the Big Dumb Booster is that the first stage is not the most expensive part.
Historical examples are offered. Thor-Agena was what launched Discoverer-Corona, and the Agena upper stage was the expensive part.
I’m familiar with BDB designs and while they have merit I think they aren’t that practical. The problem is that propellant is the easy part, the hard parts are making high-thrust engines that don’t melt or blow up, rockets that stay on course through supersonic and hypersonic speeds, etc. In a way the Falcon 9 is the best effort at a BDB-style design that’s been done. It’s a very straightforward and simple design that has heritage all the way back to the 1950s, and it’s certainly the lowest cost rocket of its class out there. And the obvious way to add more payload is to stretch the tanks, then build bigger engines and scale up in every dimension, and so on. Which is precisely what SpaceX is doing (with the Falcon 9 v1.1, the Falcon Heavy, and the planned MCT).
Perhaps it could be done even cheaper, but I doubt it could be achievable without increasing failure rates a lot.
The more I see the numbers on multi-stage reusability the more it seems like the right way forward. The big thing that RLVs enable is flight rate. Which is valuable all on its own even if they turned out to be significantly more expensive than BDBs at first. Being able to fly hardware multiple times means you can explore the flight envelope and increase the reliability and especially confidence in the vehicle design. More so, it vastly lowers the costs of iterative development, because you can try out hardware modifications and flight profile changes using real hardware and get incredibly valuable data without having to destroy tens of millions of dollars of equipment every time.
I don’t think people fully appreciate how fast things are going to develop once reusability starts to become a reality. It’s going to lower the cost of flights, which will translate to an even larger increase in launch volume and a total increase in revenue for space launch providers. More so, it’s going to make launcher R&D vastly cheaper because of the ability to run a hell of a lot more test flights at less cost. This is a positive feedback loop.
Testing it out over the ocean is a strike of genius.
Assume a transoceanic return profile rather than RTL. What, I wonder, is the ballistic range of a reusable F9 first stage? If it’s long enough, such a stage (suitably modified) could form the basis of a Heinlein-style ballistic transport.
If I were Musk, I’d fair the Grasshopper’s landing legs in with fins and paint the stage in a red and white Tintin checkerboard pattern. Looks like a ’50s SF rocket, flies like a ’50s SF rocket…
That one happened to be nuclear. And awesome.
Build a little, test a little…
And who’s to say Elon Musk doesn’t have an atom or two hidden up his sleeve? If the Mouse can use a little Di$ney Magic to obtain legal license to “Own, operate, and maintain electric power plants… for the generation and transmission of power through nuclear fission” [1] in heavily-populated central Florida, might nor Musk use his PayPal powers to get the right to do something similar in sparsely-populated Cameron, Texas?
Hmmmmm…
So, take a Dragon Rider capsule and put it on top of a reusable Falcon 9 first stage. It boosts up to Mach 2 or 3 or so and goes exo-atmospheric, coasts on a ballistic trajectory some distance then turns around in space and feathers the atmospheric re-entry a bit to make it a little easier, falls down through the atmosphere and then comes in for a powered landing at the destination. A mobile gantry comes out and fetches the passengers then takes them off to the terminal where they can be on their way. I’d have to do the math but it might work out.
Also, if they use a full Dragon capsule instead of just integrating one into the stage then it would have its own escape and re-entry system in case anything went wrong.
You’ve got it.
Add wings to extend re-entry x-range and you have Heinlein’s transport rocket.
Flight time to Sydney: 45 minutes
Time for compulsory TSA cavity search: 90 minutes
Strange fact I just made up: The impetus for re-using the first stage actually comes from the second and third stage folks who are sick of sweating bullets as their own stages fire up, then looking over to see the first stage people flopping back in their chairs, popping open beers, lighting cigarettes and high-fiving each other. With a re-usable first stage everyone will have to sweat for the whole 15 minutes. It’s all about fairness.
That’s awesome.
When SpaceX announced last year that they were considering a launch site near Brownsville, Texas, I did some rough calculations. The distance from Brownsville to a reasonable location near the Florida Gulf Coast is roughly 1550 km. I chose Cape Coral, Florida as a possible landing site. Using a ballistic range calculator and selecting 40 km for the burnout altitude, you’d need a burnout velocity of about 3.7 km/sec to reach Florida. Allowing a bit more for atmospherics, you’re looking at closer to 4.2 km/sec. It seems to me that it’d be easier to stage and then fly the first stage to Florida than to return to the launch site in Texas. This would be more suitable for missions going to a low inclination orbit like GTO than to the ISS but I think that’s where most of their launch manifest is heading, anyway.