My latest piece, based on a conversation with the proprietor of SpaceX, is up.
30 thoughts on “Elon’s Reusable Rocket”
What will be reusable on the FH? I would think the central core will re-enter at too high a velocity to recover.
Presumably all of it eventually.
Using FH’s price tag to determine ticket price seems odd to me. Has SpaceX implied at anytime that they would design a “Super Dragon” for more than 7 passengers?
Well if that all pans out expect a Moon base with err.. 8 years.
Rand, in the last paragraph, is the $20,000/ticket figure yours or his? Including the capsule I’d be surprised if it was more than two passengers/ton, which would be $110,000/ticket.
I have to admit being skeptical about the second stage. That’s a full reentry, and it’s got to be a controlled reentry (otherwise you need to protect it all from reentry heat.)
The only viable entry attitude I can see is forward end first, with a heat shield roughly the size of Dragon’s at the forward end. A mostly empty second stage dry masses what… can’t seem to find that, but I do see a ref that the Merlin engine is over half the dry weight. If a Merlin is 1380 pounds, say 2760 lbs for a mostly dry second stage, round it up to 3000 for a bit of fuel for landing, add at least another thousand for landing legs, added attitude thrusters and fuel, etc. So, 4000 for a SWAG. Hrmm, that’s lighter than a Dragon, which if I remember right is about 5500 pounds, so the heat shield should work… keeping stable during re-entry? Hrmm.. problematic… maybe trail a streamer of carbon fiber?
I’m less skeptical on the second stage than I was when I began this post, though still skeptical. I hope I’m wrong.
I do however think recovery of the first stage (or stages, on the heavy) would go a very long way to cutting costs, and looks somewhat less difficult.
I very much enjoyed the article, Rand. Thank you.
Yeah, unfortunately only SpaceX has the numbers and they’re (rightly) keeping them close to the vest.
“If a Merlin is 1380 pounds, say 2760 lbs for a mostly dry second stage, round it up to 3000 for a bit of fuel for landing, add at least another thousand for landing legs, added attitude thrusters and fuel, etc. So, 4000 for a SWAG.”
The Merlin 1-D has better numbers than that, though. From Wikipedia:
At the 2011 AIAA Propulsion Conference SpaceX’s Tom Mueller[24] revealed that the engine would have a vacuum thrust of 690 kN (155,000 lbf), a vacuum specific impulse (Isp) of 310s, an increased expansion ratio of 16, vs. the previous 14.5 of the Merlin 1C, and chamber pressure in the “sweet spot” of 9.7 MPa (1,410 psi).
“The engine’s 160:1 thrust-to-weight ratio would be the highest ever achieved for a rocket engine.”
This gives us a weight of about 970 lbs. That would give a second stage weight of about 1,940lbs. with 300lbs. added propellant for landing that’s 2,240lbs. IIRC, the usual percentage of dry weight for landing gear is about 10-15 percent of mass, which would add about another 300lbs., getting us to 2,540lbs. Add another 15 percent for the weight of the shield itself. This gets us a total of 2921lbs. Since this is less than 32 percent the weight of a Dragon capsule without payload (with the 9,260 weight of Dragon, again from its Wikipedia article at http://en.wikipedia.org/wiki/SpaceX_Dragon), it looks a good bit better.
It is a good article, and we can hope there will be good reason to write more like it in the next 12 months.
It’s interesting how so many different rocket innovators have settled on the VTVL design. Blue Origin, Armadillo, Masten, and now SpaceX have all chosen VTVL rockets and all of them have bent metal and flown test flights, including prolonged controlled hovering. Four companies are pursuing a design that NASA tossed on the junk pile over 15 years ago. It makes you wonder where we’d be today if NASA hadn’t let DC-XA fall by the wayside.
On the other hand, none of the aforementioned companies is pursuing SSTO.
Or where we would be today if NASA hadn’t taken the DC-X from DoD.
None of the players in the game today appears to be pursing SSTO. That proves nothing, but it makes me a little less wistful about DC-XA, which was a long way from orbital capability.
Or they could use one ”Tanker Rocket” to put fuel on Orbit on the cheap, so that the Dragon capsule could get the fuel for powered landing without payload penalty ?
The fuel that the Dragon capsule will use for powered landing is the same it carried for an abort. If you don’t abort, you can do a powered landing. If you do abort, you’d have to use a parachute and land whereever, presumably in water.
“…about $100 per pound. For the average male, that means about 20,000 bucks…”
Surely this figure is accurate only if you are prepared to go into space protected only by your skivvies, and hold your breath for a good long time ?
The weight of the spacecraft and consumables should be included. Dragon: about 5-6 tonnes for 7 people; Apollo CM: 6 tonnes for 3 people; Gemini 4 tonnes for 2 people, Gulfstream G-550 (loaded, no fuel) 25 tonnes for 21 people. So 1 tonne per person seems a reasonable estimate, suggesting a price per seat of $200,000.
And the $10/lb number for the ultimate price seems a little off. Let’s suppose a reusable Falcon Heavy can deliver 20,000 kg (a bit less than 40% of its stated payload) daily. At a 2.56 oxidizer fuel ratio, we’re needing maybe 350,000 kg of RP-1 (about twice an A380) and 900,000 kg of lox. I’m guessing RP-1 costs about the same as Jet-A which is running $1000/mt. and say $250/ton of LOx. That’s still $28.40/lb of payload. If they can get personnel and capital costs down to 1/3 each, we’re still looking at $85/lb minimum. I’d love to see the numbers and assumptions behind the $10/lb and $100/lb.
I think you mean $28.50 per KG, not lb
Also, I read the article to mean a 40% penalty in payload, not 40% OF the payload, so 30 tonnes, not 20 tonnes.
Cost is then about $20 per kg, a figure that seems to come up a lot.
Yes, I got $28.50 a kg, $13.07 a pound.
However, Jet-A and Kerosene prices are running about twice what they were ten years ago, and are far above their historical average. If we rerun Sam’s calculation with RP-1 at $500 a metric ton, the cost comes to $20/kg and $9.09 a pound.
At that point Sam, Elon can consider switching to Methane. Thanks to recent developments, it is getting pretty Frakking cheap!
OK, that makes sense. At $13/lb we still need to pay for capital and personnel which we might assume to be 1/3 and 1/3 of total cost like the airlines (if we don’t switch to methane–good point). I’m still interested in the $100/lb number to see if that includes capital and labor.
Yes Rand, I believe you are getting excited (I thought that was my job?) as noted above by most all bulk $/mass is not quite the same as ticket price. However, the big story…
With superdraco operational today and FH next year all the puzzle pieces are coming together for a single billionaire to put together a mars mission.
A 6 crew, 25k kg class ship, fully assembled on the launch pad (BA330+Merlin… 3 engines for redundancy?) can be put in orbit for less than $250m. Fuel, supplies, and crew would take half a dozen more launches over two years for a total cost of around $750m. The dragon taking the crew up is the same dragon to land on mars. So for around one billion you take six to the surface of mars. Before that…
A FH can send supply dragons to the surface of mars for about $150m. Commit to sending a couple each launch window until people arrive. Each becomes an emergency shelter with 18 man-years of supplies (3×6) and an electric/methane hybrid car (If only Elon had access to some electric car company?)
$150m a year starting next year or so in preparation for the first manned landing. That’s very doable for a wealthy visionary.
We send two ships for two billion (need investors for that.) They tether together for gravity and remaining fuel provides radiation shielding.
They transfer everything to one ship and let the other free return as they get close to mars.
We let NASA pay for the supply dragon test flights to the surface of mars, but the actual mars mission is a totally private venture. The government has no claims on what they accomplish or do.
All could easily be done in less than ten years from today. We are better prepared now than when we landed on the moon.
$150m a year starting next year or so in preparation for the first manned landing. That’s very doable for a wealthy visionary.”
You mean like Elon Musk? 😉
I expect the first billionaire mission won’t be to Mars. It will be India’s Lakshmi Mittal sending a mining operation to an asteroid. He will expect to make the money back. Colonization is inherently unprofitable in the short term. It will take Elon (or Bigelow) a while to make enough money they can afford to just throw away (rather than pour into business operations like launch or station).
——————-
Rand, great article, even with the bodyweight/dry mass goof. Also, to get FH prices and daily flight rates they first need to design a 30 passenger Dragon Bus (or put a regular Dragon on top of a satellite cargo) and have enough orbital destinations to attract that kind of traffic on a daily basis.
Colonization is inherently unprofitable in the short term
That’s not a given. It doesn’t have to be so. Everyone involved could be turning a profit from day one. That’s the magic of accrual accounting. It just requires a method.
In the late seventies people paid $10k for microcomputers that wouldn’t make good doorstops today. Would we have these $500 miracles if they hadn’t?
I don’t see how we get to $200k tickets to mars Zubrin and Musk talk about with any flight rate. The best I can come up with is $50m to $60m per person. If we want a solar system economy we are going to have to deal with it.
$4m to LEO (requires that 30 passenger super dragon.) $35m to mars orbit. $21m to mars surface. w/ 3 yrs of supplies extended to 12 yrs with water ISRU.
So you don’t like my one sq. km. per person idea? Make it 4 or 9. With 75% utilization that give you 1200 or 2700 quarter hectare plots for resale from each persons claim. That means as little as $22k profit over development cost to break even. Expect about $100k over cost to provide profit and income to both the bank and the property owner even before the economy diversifies which it will rapidly do to the needs of the colonists.
Thanks you accrual accounting that means both the bank and the borrower make a profit the moment the ink is dry.
Give them 30 yrs to pay off the loan (with insurance to cover deaths) and colonization is profitable from day one for everyone involved.
Second generation martians are going to wonder how terrans can live in such a chaotic environment in small boxes. They will live in large underground mansions in controlled environments and light gravity.
Martians will not have to deal with the idiotic regulations we do here on earth. Children will be building nuclear reactors. Mars will need lots of power. It really will be too cheap to meter there.
Not everyone will be a farmer, but there will be a thriving community farmers market where people trade what they’ve produced. Not just food, but building materials and other things.
Note that Bigelow and Musk make a profit, but not by colonizing… they just provide transportation at a ticket price that includes their profits… again, from day one.
I just noticed: $60m x 28 ppl is just under that $1.7b NASA prize budget we talked about in an earlier post. Makes ya go hmmmm….
I don’t think the upper stage turnaround time is an issue, whether it’s 24 hours in orbit or a week or a month. You just have to have more upper stages than lower stages so you’ve always got an upper stage ready to mate and launch. The ground crew doesn’t have to work on the extra upper stages in parallel (which would linearly increase the size of the ground crew and processing facilities), it’s just that there would be upper stages queued up in orbit waiting for their re-entry window, with an added cost of next to nothing.
My other thought is that a bit of cleverness could dispense with the landing gear entirely, depending on how accurately you can control a returning stage for a precision touchdown. If you can get it to within a football field you could lay out an X-Y gantry that would latch on to the rocket at what would’ve been its landing-gear attachement points. Or you could grab it from the side, or you could just have the stage tip itself over into a net.
If you’ve got that kind of control, you could use the hook and wire approach proven on the X-13A Vertijet back in the late 1950s. All you’d need for the stage would be a hook.
That should work. You could mount a large robot arm on a tower, with freedom to move only in the horizontal plane, and just reach out and hook the rocket on the wire if it gets within reach. The landing gear is only good if you’re coming down on a pad or a runway, and it wouldn’t take much more precision to eliminate it entirely. It would, however, increase the cost of alternate landing sites by a significant amount.
Or you could just land it in a swimming pool, as long as everything is water proofed, which would mean half the houses in LA would become potential landing sites. 🙂
Another thought on Dragon thrusters for abort or landing is that they’re mounted above the heatshield, as I understand it, so they’re probably directed somewhat out instead of straight down.
There might be a simple fix for that by having a refractory metal coverplate on a 4-bar linkage that swings up and out into the engine’s exhaust stream, redirecting it downwards. If you got fancier with the linkage you could also use it for steering. It would give you the benefits of a thruster aligned on the desired resultant thrust vector, without having to put the actual rocket motor and associated plumping outside the edge of the heat shield or through the heat shield.
Since it would be at the bottom end of the rocket and flip open only when the craft needs to expel propulsion gases, I’m naming my idea “the butt flap”.
Ok all you hep cats… the latest dance craze guys and gals… The dragon butt flap!
Or alter the shape of the underside of the rocket stage with a conical depression and have the rocket stage set itself down onto a cone affixed to the ground. Sort of like the Gauld space ship landing on top of the pyramid in ‘Stargate’.
What will be reusable on the FH? I would think the central core will re-enter at too high a velocity to recover.
Presumably all of it eventually.
Using FH’s price tag to determine ticket price seems odd to me. Has SpaceX implied at anytime that they would design a “Super Dragon” for more than 7 passengers?
Well if that all pans out expect a Moon base with err.. 8 years.
Rand, in the last paragraph, is the $20,000/ticket figure yours or his? Including the capsule I’d be surprised if it was more than two passengers/ton, which would be $110,000/ticket.
I have to admit being skeptical about the second stage. That’s a full reentry, and it’s got to be a controlled reentry (otherwise you need to protect it all from reentry heat.)
The only viable entry attitude I can see is forward end first, with a heat shield roughly the size of Dragon’s at the forward end. A mostly empty second stage dry masses what… can’t seem to find that, but I do see a ref that the Merlin engine is over half the dry weight. If a Merlin is 1380 pounds, say 2760 lbs for a mostly dry second stage, round it up to 3000 for a bit of fuel for landing, add at least another thousand for landing legs, added attitude thrusters and fuel, etc. So, 4000 for a SWAG. Hrmm, that’s lighter than a Dragon, which if I remember right is about 5500 pounds, so the heat shield should work… keeping stable during re-entry? Hrmm.. problematic… maybe trail a streamer of carbon fiber?
I’m less skeptical on the second stage than I was when I began this post, though still skeptical. I hope I’m wrong.
I do however think recovery of the first stage (or stages, on the heavy) would go a very long way to cutting costs, and looks somewhat less difficult.
I very much enjoyed the article, Rand. Thank you.
Yeah, unfortunately only SpaceX has the numbers and they’re (rightly) keeping them close to the vest.
“If a Merlin is 1380 pounds, say 2760 lbs for a mostly dry second stage, round it up to 3000 for a bit of fuel for landing, add at least another thousand for landing legs, added attitude thrusters and fuel, etc. So, 4000 for a SWAG.”
The Merlin 1-D has better numbers than that, though. From Wikipedia:
At the 2011 AIAA Propulsion Conference SpaceX’s Tom Mueller[24] revealed that the engine would have a vacuum thrust of 690 kN (155,000 lbf), a vacuum specific impulse (Isp) of 310s, an increased expansion ratio of 16, vs. the previous 14.5 of the Merlin 1C, and chamber pressure in the “sweet spot” of 9.7 MPa (1,410 psi).
“The engine’s 160:1 thrust-to-weight ratio would be the highest ever achieved for a rocket engine.”
This gives us a weight of about 970 lbs. That would give a second stage weight of about 1,940lbs. with 300lbs. added propellant for landing that’s 2,240lbs. IIRC, the usual percentage of dry weight for landing gear is about 10-15 percent of mass, which would add about another 300lbs., getting us to 2,540lbs. Add another 15 percent for the weight of the shield itself. This gets us a total of 2921lbs. Since this is less than 32 percent the weight of a Dragon capsule without payload (with the 9,260 weight of Dragon, again from its Wikipedia article at http://en.wikipedia.org/wiki/SpaceX_Dragon), it looks a good bit better.
It is a good article, and we can hope there will be good reason to write more like it in the next 12 months.
It’s interesting how so many different rocket innovators have settled on the VTVL design. Blue Origin, Armadillo, Masten, and now SpaceX have all chosen VTVL rockets and all of them have bent metal and flown test flights, including prolonged controlled hovering. Four companies are pursuing a design that NASA tossed on the junk pile over 15 years ago. It makes you wonder where we’d be today if NASA hadn’t let DC-XA fall by the wayside.
On the other hand, none of the aforementioned companies is pursuing SSTO.
Or where we would be today if NASA hadn’t taken the DC-X from DoD.
None of the players in the game today appears to be pursing SSTO. That proves nothing, but it makes me a little less wistful about DC-XA, which was a long way from orbital capability.
Or they could use one ”Tanker Rocket” to put fuel on Orbit on the cheap, so that the Dragon capsule could get the fuel for powered landing without payload penalty ?
The fuel that the Dragon capsule will use for powered landing is the same it carried for an abort. If you don’t abort, you can do a powered landing. If you do abort, you’d have to use a parachute and land whereever, presumably in water.
“…about $100 per pound. For the average male, that means about 20,000 bucks…”
Surely this figure is accurate only if you are prepared to go into space protected only by your skivvies, and hold your breath for a good long time ?
The weight of the spacecraft and consumables should be included. Dragon: about 5-6 tonnes for 7 people; Apollo CM: 6 tonnes for 3 people; Gemini 4 tonnes for 2 people, Gulfstream G-550 (loaded, no fuel) 25 tonnes for 21 people. So 1 tonne per person seems a reasonable estimate, suggesting a price per seat of $200,000.
And the $10/lb number for the ultimate price seems a little off. Let’s suppose a reusable Falcon Heavy can deliver 20,000 kg (a bit less than 40% of its stated payload) daily. At a 2.56 oxidizer fuel ratio, we’re needing maybe 350,000 kg of RP-1 (about twice an A380) and 900,000 kg of lox. I’m guessing RP-1 costs about the same as Jet-A which is running $1000/mt. and say $250/ton of LOx. That’s still $28.40/lb of payload. If they can get personnel and capital costs down to 1/3 each, we’re still looking at $85/lb minimum. I’d love to see the numbers and assumptions behind the $10/lb and $100/lb.
I think you mean $28.50 per KG, not lb
Also, I read the article to mean a 40% penalty in payload, not 40% OF the payload, so 30 tonnes, not 20 tonnes.
Cost is then about $20 per kg, a figure that seems to come up a lot.
Yes, I got $28.50 a kg, $13.07 a pound.
However, Jet-A and Kerosene prices are running about twice what they were ten years ago, and are far above their historical average. If we rerun Sam’s calculation with RP-1 at $500 a metric ton, the cost comes to $20/kg and $9.09 a pound.
At that point Sam, Elon can consider switching to Methane. Thanks to recent developments, it is getting pretty Frakking cheap!
OK, that makes sense. At $13/lb we still need to pay for capital and personnel which we might assume to be 1/3 and 1/3 of total cost like the airlines (if we don’t switch to methane–good point). I’m still interested in the $100/lb number to see if that includes capital and labor.
Yes Rand, I believe you are getting excited (I thought that was my job?) as noted above by most all bulk $/mass is not quite the same as ticket price. However, the big story…
With superdraco operational today and FH next year all the puzzle pieces are coming together for a single billionaire to put together a mars mission.
A 6 crew, 25k kg class ship, fully assembled on the launch pad (BA330+Merlin… 3 engines for redundancy?) can be put in orbit for less than $250m. Fuel, supplies, and crew would take half a dozen more launches over two years for a total cost of around $750m. The dragon taking the crew up is the same dragon to land on mars. So for around one billion you take six to the surface of mars. Before that…
A FH can send supply dragons to the surface of mars for about $150m. Commit to sending a couple each launch window until people arrive. Each becomes an emergency shelter with 18 man-years of supplies (3×6) and an electric/methane hybrid car (If only Elon had access to some electric car company?)
$150m a year starting next year or so in preparation for the first manned landing. That’s very doable for a wealthy visionary.
We send two ships for two billion (need investors for that.) They tether together for gravity and remaining fuel provides radiation shielding.
They transfer everything to one ship and let the other free return as they get close to mars.
We let NASA pay for the supply dragon test flights to the surface of mars, but the actual mars mission is a totally private venture. The government has no claims on what they accomplish or do.
All could easily be done in less than ten years from today. We are better prepared now than when we landed on the moon.
You mean like Elon Musk? 😉
I expect the first billionaire mission won’t be to Mars. It will be India’s Lakshmi Mittal sending a mining operation to an asteroid. He will expect to make the money back. Colonization is inherently unprofitable in the short term. It will take Elon (or Bigelow) a while to make enough money they can afford to just throw away (rather than pour into business operations like launch or station).
——————-
Rand, great article, even with the bodyweight/dry mass goof. Also, to get FH prices and daily flight rates they first need to design a 30 passenger Dragon Bus (or put a regular Dragon on top of a satellite cargo) and have enough orbital destinations to attract that kind of traffic on a daily basis.
Colonization is inherently unprofitable in the short term
That’s not a given. It doesn’t have to be so. Everyone involved could be turning a profit from day one. That’s the magic of accrual accounting. It just requires a method.
In the late seventies people paid $10k for microcomputers that wouldn’t make good doorstops today. Would we have these $500 miracles if they hadn’t?
I don’t see how we get to $200k tickets to mars Zubrin and Musk talk about with any flight rate. The best I can come up with is $50m to $60m per person. If we want a solar system economy we are going to have to deal with it.
$4m to LEO (requires that 30 passenger super dragon.) $35m to mars orbit. $21m to mars surface. w/ 3 yrs of supplies extended to 12 yrs with water ISRU.
So you don’t like my one sq. km. per person idea? Make it 4 or 9. With 75% utilization that give you 1200 or 2700 quarter hectare plots for resale from each persons claim. That means as little as $22k profit over development cost to break even. Expect about $100k over cost to provide profit and income to both the bank and the property owner even before the economy diversifies which it will rapidly do to the needs of the colonists.
Thanks you accrual accounting that means both the bank and the borrower make a profit the moment the ink is dry.
Give them 30 yrs to pay off the loan (with insurance to cover deaths) and colonization is profitable from day one for everyone involved.
Second generation martians are going to wonder how terrans can live in such a chaotic environment in small boxes. They will live in large underground mansions in controlled environments and light gravity.
Martians will not have to deal with the idiotic regulations we do here on earth. Children will be building nuclear reactors. Mars will need lots of power. It really will be too cheap to meter there.
Not everyone will be a farmer, but there will be a thriving community farmers market where people trade what they’ve produced. Not just food, but building materials and other things.
Note that Bigelow and Musk make a profit, but not by colonizing… they just provide transportation at a ticket price that includes their profits… again, from day one.
I just noticed: $60m x 28 ppl is just under that $1.7b NASA prize budget we talked about in an earlier post. Makes ya go hmmmm….
I don’t think the upper stage turnaround time is an issue, whether it’s 24 hours in orbit or a week or a month. You just have to have more upper stages than lower stages so you’ve always got an upper stage ready to mate and launch. The ground crew doesn’t have to work on the extra upper stages in parallel (which would linearly increase the size of the ground crew and processing facilities), it’s just that there would be upper stages queued up in orbit waiting for their re-entry window, with an added cost of next to nothing.
My other thought is that a bit of cleverness could dispense with the landing gear entirely, depending on how accurately you can control a returning stage for a precision touchdown. If you can get it to within a football field you could lay out an X-Y gantry that would latch on to the rocket at what would’ve been its landing-gear attachement points. Or you could grab it from the side, or you could just have the stage tip itself over into a net.
If you’ve got that kind of control, you could use the hook and wire approach proven on the X-13A Vertijet back in the late 1950s. All you’d need for the stage would be a hook.
That should work. You could mount a large robot arm on a tower, with freedom to move only in the horizontal plane, and just reach out and hook the rocket on the wire if it gets within reach. The landing gear is only good if you’re coming down on a pad or a runway, and it wouldn’t take much more precision to eliminate it entirely. It would, however, increase the cost of alternate landing sites by a significant amount.
Or you could just land it in a swimming pool, as long as everything is water proofed, which would mean half the houses in LA would become potential landing sites. 🙂
Another thought on Dragon thrusters for abort or landing is that they’re mounted above the heatshield, as I understand it, so they’re probably directed somewhat out instead of straight down.
There might be a simple fix for that by having a refractory metal coverplate on a 4-bar linkage that swings up and out into the engine’s exhaust stream, redirecting it downwards. If you got fancier with the linkage you could also use it for steering. It would give you the benefits of a thruster aligned on the desired resultant thrust vector, without having to put the actual rocket motor and associated plumping outside the edge of the heat shield or through the heat shield.
Since it would be at the bottom end of the rocket and flip open only when the craft needs to expel propulsion gases, I’m naming my idea “the butt flap”.
Ok all you hep cats… the latest dance craze guys and gals… The dragon butt flap!
Or alter the shape of the underside of the rocket stage with a conical depression and have the rocket stage set itself down onto a cone affixed to the ground. Sort of like the Gauld space ship landing on top of the pyramid in ‘Stargate’.