So, correct me if I’m wrong here: the advantage of a fully reusable SSTO over a fully reusable TSTO is the reduced ground operations to get it back into the air on the next flight? Is that all?
Plus no staging. Plus the ability to travel to anywhere on Earth in less than an hour. Plus the ability to incrementally test every aspect of a flight profile. Those are pretty important…
Another reason for one-stage vs more is that when you design an SSTO, well, you design one vehicle. When you design a two-stage, you actually design and test three: each stage alone and the mated pair. SSTO means you don’t have to recover stages downrange, or find a way to fly them back. Further, as NASA Langley analysis showed thirty years ago, the benefits of staging diminish greatly as the propellant mass fraction increases, so if you can build high PMF stages, the benefits from using more than one are reduced. These are just a few of many reasons; one day I will write a book. 😉
.. and I will buy it. Need an advance?
[O]ne day I will write a book.
Please do!
One interesting idea I/we came up with on Sunday for an X-vehicle was a staging demonstrator, that would simply repeatedly go to Mach something, do a non-explosive separation, then have both fly back, to eliminate one of the more common causes of launch failure, and to demonstrate reliable mechanical (as opposed to pyro, for operability) separation. The cheapest way to do it would be bimese, so you’d only have to develop two vehicles (the twin stages, and the mated configuration). It doesn’t work that well for orbital performance, but bimese might make sense for a very cheap X-vehicle.
With cross-feed that would make a lot of orbital sense, too, as you’d essentially be using high-Mach in-flight refueling for the orbital stage but without the cost of developing a dedicated tanker. I figure with LOX/RP-1 you’d then stage at about Mach 5 (it’s necessarily a little less than a 2 to 1 mass ratio). If you used four mated sections, the first pair would separate at about Mach 5 and the second pair at Mach 10 or so, making it pretty easy to get the one of the four into orbit.
A two-stage idea I’ve had is to nest the upper stage inside the first stage, basically turning strap-on boosters into a complete wrap-around first stage. I’d suggest doing this by having an inverse taper on the upper stage, so it nests into the first stage like a plastic cup, while the first stage starts looking a lot like the DC-X, fat and squat.
The advantages are:
The second stage engines can ground start (and are cross fed from the main tanks).
Having the second stage engines provide much of the lift for the second stage reduces the compressive loaing on the main stage, which no longer has to completely support the second stage mass.
If the two stages use similar fuels, the second stage and the inner surface of the main stage doesn’t need to be insulated, because the second stage LOX tank is nestled next to (essentially inside of) the main stage LOX tank, and ditto for the RP-1 or LH2 tanks.
The second stage is no longer subject to aerodynamic pressures it would encounter either at the front of the rocket or attached to the side of a big booster. Nothing can fall off and hit it, and nothing on it can fall off and hit the main booster, either. It’s shielded inside the first stage.
If staging is high enough, the payload shroud/nosecone can be a reusable part of the first stage, opening for stage separation and then closing again.
Unlike normal tandem matings, the aerodynamic and thrust loads all remain axially symmetric, and wetted area isn’t dramatically increased, along with interference drag.
In the event of an in-flight abort prior to separation, the cross feed could reverse and run the main engines off the second stage fuel tanks, which would also lighten the landing weight.
Obviously there are a few downsides, such as having an inner tank wall on the first stage, and the possibility of stage-to-stage impact during separation, and that you might keep the second stage engines running during separation, impinging on the first stage inner tank.
What I think a lot of people fail to appreciate is that Falcon 9 is essentially bimese. The second stage is shorter, and has less motors, but it’s built on the same hardware as the first stage.
It’s probably more accurate to state the the FH first stage and boosters will be a very close approximation to a true trimese layout.
People wonder if a F9v1.1 first stage could be used as an SSTO – but using a FH without a 2nd stage (full trimese mode) would be very interesting as well.
Looking forward to that book Gary. I don’t know if this will be such a disadvantage for a general reusable TSTO as it is for the Falcon 9 v1.1 , but Elon has said they expect to lose 40%(!) of the payload for the reusable version of the F9 v1.1:
I gather from what was said in that Popular Mechanics interview that that’s primarily because of the loss of efficiency in not letting the first stage get too far down range to return to the launch site.
It is notable as well that the Falcon 9 v1.1 first stage might have nonreusable SSTO capability if it really has the claimed 20 to 1 mass ratio, but just barely. According to SpaceX, the Merlin 1D has a 311 s vacuum Isp. Then the delta v the first stage could reach would be: 311*9.81*ln(20) = 9,140 m/s. The delta-v commonly given for LEO is in the range of 9,100 m/s.
And the Falcon Heavy side boosters almost certainly would have SSTO capability if they really reach the 30 to 1 mass ratio claimed. The delta v would be: 311*9.81*ln(30) = 10,400 m/s.
Of course, the delta v and the payload capability as SSTO would be much higher if the engines were given altitude compensation. 😉
Bob Clark
Well, if the drop some of those Merlins on the way up, like an old Atlas, the Falcon 9 could easily do SSTO. 🙂
But isn’t the “SSTO or bust” mentality just the wrong direction? You are falling into the old trap of optimizing for performance instead of cost.
Taking the reusability aspect of F9v1.1/F9R and pushing it to be an expendable SSTO with virtually no payload doesn’t seem very helpful. It won’t lower the cost of going into space, however nice of a thought experiment it is.
When you take into account that possible 40% loss in payload for a reusable TSTO, a reusable SSTO becomes much more competitive when you also take into account its lowered operational costs.
But having a high Isp would be essential, and the Isp’s of the first stage versions of the Merlins are inadequate. What would work if they were replaced by the NK-33’s used on the Orbital sciences Antares while maintaining the high mass ratio of the Falcon 9.
SpaceX of course would not want to do that. But using altitude compensation on the Merlin’s would also work, and there are already known various methods of accomplishing that.
Bob Clark
Well, my idea for cost reduction is to get rid of the hideously expensive turbopumps and replace them with a thruster blasting into a pipe where fuel is pumped in from the walls. If the thruster’s Ve is 4,000 m/sec (LOX/LH2, ISP 406) and its flow is 1 kg/sec, you pump in 20 kg/sec of fuel and due to conservation of momentum, the mix ends up traveling at 190 m/sec (425 mph) which has a pressure equivalent for LOX of 3000 psi. For 1000 psi your mix ratio can be 38:1 (fuel mass vs burner fuel flow)
For a spreadsheet:
*****
Ve burner exhaust velocity (ISP*9.806)
x mix ratio (fuel flow to burner flow)
Vf final burner-exhaust/fuel velocity
rho fuel density in kg/m^3
Vf = Ve/(x+1)
specific KE = 0.5*Vf^2
equivalent pressure = rho*specific KE (in Pascals, divide by 6894.757 for psi)
*****
LH2/LOX is about 40% efficient at converting chemical energy to exhaust momentum, and the mixing is largely elastic so KE is conserved. If the fuel and oxidizer are chilled well below their boiling points, they’ll stay liquid despite the slight increase in temperature from the thermal energy in the burner exhaust, which with cyrogenics would condense into ice and mix with the fuel or oxidizer, giving the benefits of a staged combustion cycle for free.
It doesn’t seem to quite match the efficiency of a turbopump, but it’s not much worse, and it could be built with pipe, a welder, and a drill (and of course a thruster). Of course all the cleverness has to go into pressure recovery from a spray of high velocity droplets.
If engine cost can be greatly reduced, (along with engine mass), system architectures gain quite a bit more flexibility.
Both Bob Clark and George Turner address the capability of the F9R to make orbit as an SSTO or stage-and-a-half. The drop engine approach would (in my opinion) be preferable to the boost-back approach that SpaceX is employing. But a TAN nozzle, dual-bell, or extendible exit cone to deliver vacuum Isp would make it easy for an F9R equivalent to reach orbit as an SSTO, and with a useful payload. I’d guesstimate that it would be feasible to achieve 10-20K lbm payload for 1M lbm launch mass with 9 Merlin 1Ds.
I’ve wondered if you could make an inflatable bell extension out of some silicone-based ultra-high temperature plastic, perhaps with some woven metal on the inner surface. Inflated tubular support ribs might not even need any cooling. You could tap part of the GOX flow for pressure or just use a burst disk and a tiny tank.
Goodyear designed such an extension for the J-2; the report is on line somewhere. But metal extensions are cheap and well characterized.
So, correct me if I’m wrong here: the advantage of a fully reusable SSTO over a fully reusable TSTO is the reduced ground operations to get it back into the air on the next flight? Is that all?
Plus no staging. Plus the ability to travel to anywhere on Earth in less than an hour. Plus the ability to incrementally test every aspect of a flight profile. Those are pretty important…
Another reason for one-stage vs more is that when you design an SSTO, well, you design one vehicle. When you design a two-stage, you actually design and test three: each stage alone and the mated pair. SSTO means you don’t have to recover stages downrange, or find a way to fly them back. Further, as NASA Langley analysis showed thirty years ago, the benefits of staging diminish greatly as the propellant mass fraction increases, so if you can build high PMF stages, the benefits from using more than one are reduced. These are just a few of many reasons; one day I will write a book. 😉
.. and I will buy it. Need an advance?
Please do!
One interesting idea I/we came up with on Sunday for an X-vehicle was a staging demonstrator, that would simply repeatedly go to Mach something, do a non-explosive separation, then have both fly back, to eliminate one of the more common causes of launch failure, and to demonstrate reliable mechanical (as opposed to pyro, for operability) separation. The cheapest way to do it would be bimese, so you’d only have to develop two vehicles (the twin stages, and the mated configuration). It doesn’t work that well for orbital performance, but bimese might make sense for a very cheap X-vehicle.
With cross-feed that would make a lot of orbital sense, too, as you’d essentially be using high-Mach in-flight refueling for the orbital stage but without the cost of developing a dedicated tanker. I figure with LOX/RP-1 you’d then stage at about Mach 5 (it’s necessarily a little less than a 2 to 1 mass ratio). If you used four mated sections, the first pair would separate at about Mach 5 and the second pair at Mach 10 or so, making it pretty easy to get the one of the four into orbit.
A two-stage idea I’ve had is to nest the upper stage inside the first stage, basically turning strap-on boosters into a complete wrap-around first stage. I’d suggest doing this by having an inverse taper on the upper stage, so it nests into the first stage like a plastic cup, while the first stage starts looking a lot like the DC-X, fat and squat.
The advantages are:
The second stage engines can ground start (and are cross fed from the main tanks).
Having the second stage engines provide much of the lift for the second stage reduces the compressive loaing on the main stage, which no longer has to completely support the second stage mass.
If the two stages use similar fuels, the second stage and the inner surface of the main stage doesn’t need to be insulated, because the second stage LOX tank is nestled next to (essentially inside of) the main stage LOX tank, and ditto for the RP-1 or LH2 tanks.
The second stage is no longer subject to aerodynamic pressures it would encounter either at the front of the rocket or attached to the side of a big booster. Nothing can fall off and hit it, and nothing on it can fall off and hit the main booster, either. It’s shielded inside the first stage.
If staging is high enough, the payload shroud/nosecone can be a reusable part of the first stage, opening for stage separation and then closing again.
Unlike normal tandem matings, the aerodynamic and thrust loads all remain axially symmetric, and wetted area isn’t dramatically increased, along with interference drag.
In the event of an in-flight abort prior to separation, the cross feed could reverse and run the main engines off the second stage fuel tanks, which would also lighten the landing weight.
Obviously there are a few downsides, such as having an inner tank wall on the first stage, and the possibility of stage-to-stage impact during separation, and that you might keep the second stage engines running during separation, impinging on the first stage inner tank.
What I think a lot of people fail to appreciate is that Falcon 9 is essentially bimese. The second stage is shorter, and has less motors, but it’s built on the same hardware as the first stage.
It’s probably more accurate to state the the FH first stage and boosters will be a very close approximation to a true trimese layout.
People wonder if a F9v1.1 first stage could be used as an SSTO – but using a FH without a 2nd stage (full trimese mode) would be very interesting as well.
Looking forward to that book Gary. I don’t know if this will be such a disadvantage for a general reusable TSTO as it is for the Falcon 9 v1.1 , but Elon has said they expect to lose 40%(!) of the payload for the reusable version of the F9 v1.1:
Elon Musk on SpaceX’s Reusable Rocket Plans.
By Rand Simberg
February 7, 2012 6:00 PM
http://www.popularmechanics.com/science/space/rockets/elon-musk-on-spacexs-reusable-rocket-plans-6653023
I gather from what was said in that Popular Mechanics interview that that’s primarily because of the loss of efficiency in not letting the first stage get too far down range to return to the launch site.
It is notable as well that the Falcon 9 v1.1 first stage might have nonreusable SSTO capability if it really has the claimed 20 to 1 mass ratio, but just barely. According to SpaceX, the Merlin 1D has a 311 s vacuum Isp. Then the delta v the first stage could reach would be: 311*9.81*ln(20) = 9,140 m/s. The delta-v commonly given for LEO is in the range of 9,100 m/s.
And the Falcon Heavy side boosters almost certainly would have SSTO capability if they really reach the 30 to 1 mass ratio claimed. The delta v would be: 311*9.81*ln(30) = 10,400 m/s.
Of course, the delta v and the payload capability as SSTO would be much higher if the engines were given altitude compensation. 😉
Bob Clark
Well, if the drop some of those Merlins on the way up, like an old Atlas, the Falcon 9 could easily do SSTO. 🙂
But isn’t the “SSTO or bust” mentality just the wrong direction? You are falling into the old trap of optimizing for performance instead of cost.
Taking the reusability aspect of F9v1.1/F9R and pushing it to be an expendable SSTO with virtually no payload doesn’t seem very helpful. It won’t lower the cost of going into space, however nice of a thought experiment it is.
When you take into account that possible 40% loss in payload for a reusable TSTO, a reusable SSTO becomes much more competitive when you also take into account its lowered operational costs.
But having a high Isp would be essential, and the Isp’s of the first stage versions of the Merlins are inadequate. What would work if they were replaced by the NK-33’s used on the Orbital sciences Antares while maintaining the high mass ratio of the Falcon 9.
SpaceX of course would not want to do that. But using altitude compensation on the Merlin’s would also work, and there are already known various methods of accomplishing that.
Bob Clark
Well, my idea for cost reduction is to get rid of the hideously expensive turbopumps and replace them with a thruster blasting into a pipe where fuel is pumped in from the walls. If the thruster’s Ve is 4,000 m/sec (LOX/LH2, ISP 406) and its flow is 1 kg/sec, you pump in 20 kg/sec of fuel and due to conservation of momentum, the mix ends up traveling at 190 m/sec (425 mph) which has a pressure equivalent for LOX of 3000 psi. For 1000 psi your mix ratio can be 38:1 (fuel mass vs burner fuel flow)
For a spreadsheet:
*****
Ve burner exhaust velocity (ISP*9.806)
x mix ratio (fuel flow to burner flow)
Vf final burner-exhaust/fuel velocity
rho fuel density in kg/m^3
Vf = Ve/(x+1)
specific KE = 0.5*Vf^2
equivalent pressure = rho*specific KE (in Pascals, divide by 6894.757 for psi)
*****
LH2/LOX is about 40% efficient at converting chemical energy to exhaust momentum, and the mixing is largely elastic so KE is conserved. If the fuel and oxidizer are chilled well below their boiling points, they’ll stay liquid despite the slight increase in temperature from the thermal energy in the burner exhaust, which with cyrogenics would condense into ice and mix with the fuel or oxidizer, giving the benefits of a staged combustion cycle for free.
It doesn’t seem to quite match the efficiency of a turbopump, but it’s not much worse, and it could be built with pipe, a welder, and a drill (and of course a thruster). Of course all the cleverness has to go into pressure recovery from a spray of high velocity droplets.
If engine cost can be greatly reduced, (along with engine mass), system architectures gain quite a bit more flexibility.
Both Bob Clark and George Turner address the capability of the F9R to make orbit as an SSTO or stage-and-a-half. The drop engine approach would (in my opinion) be preferable to the boost-back approach that SpaceX is employing. But a TAN nozzle, dual-bell, or extendible exit cone to deliver vacuum Isp would make it easy for an F9R equivalent to reach orbit as an SSTO, and with a useful payload. I’d guesstimate that it would be feasible to achieve 10-20K lbm payload for 1M lbm launch mass with 9 Merlin 1Ds.
I’ve wondered if you could make an inflatable bell extension out of some silicone-based ultra-high temperature plastic, perhaps with some woven metal on the inner surface. Inflated tubular support ribs might not even need any cooling. You could tap part of the GOX flow for pressure or just use a burst disk and a tiny tank.
Goodyear designed such an extension for the J-2; the report is on line somewhere. But metal extensions are cheap and well characterized.