SpaceX BFR Discussion.

[tater]

2 minutes ago, Kerbal7 said:

And it's going to fall flat on its face.

They have to work on the next gen vehicle now, dev takes years. BO will be flying NG in that time frame. It will have performance between F9 and FH, and a 7m fairing. Bezos is aiming for operational booster reuse from the start, so it can likely compete with F9 from day one.

It's not like F9 will cease to exist, either. They will operate selling F9 launches the whole time, and the staff used to dev BFR would otherwise be idle.

9 minutes ago, sevenperforce said:

It's a valid concern. The DC-X tests didn't have big tail fins to deal with. The Falcon 9/H boosters have a very low center of mass, so they automatically orient themselves engine-first. But the only way I see the BFS being able to orient itself tail-first is if it does an aerodynamic pitch-up followed by an RCS-controlled stall, which strikes me as rather risky idea.

BFS has 7 engines on the back. Merlins are ~500kg, so call it 4 tons at the back. The cargo version (the only thing I imagine flying, honestly) has a dry mass well under the 85 ton figure. With the reserved fuel and engines at the back, the CM has to be closer to the bottom on return (empty).

[sevenperforce]

10 hours ago, tater said:

The booster they are unconcerned about (except for the composites, it's just a big F9).

BFR is actually a good deal more straightforward than Falcon Heavy. More engines, sure, but they are all mounted on a single thrust plate so there is no differential torque to worry about. 

Jury is still out on "landing on the launch pad".

[RCgothic]

I wonder if the solution may involve robotic launch clamps that "catch" the booster by moving to engage target latches on the booster. With a system like that returning within 5-10m would be good enough.

[tater]

6 minutes ago, sevenperforce said:

Jury is still out on "landing on the launch pad".

This seems absurd to me to start. I'd bet the first iteration has legs.

1 minute ago, RCgothic said:

I wonder if the solution may involve robotic launch clamps that "catch" the booster by moving to engage target latches on the booster. With a system like that returning within 5-10m would be good enough.

I think he said they already land close enough RTLS to use clamps, so maybe I'm wrong.

[sevenperforce]

4 minutes ago, tater said:

BFS has 7 engines on the back. Merlins are ~500kg, so call it 4 tons at the back. The cargo version (the only thing I imagine flying, honestly) has a dry mass well under the 85 ton figure. With the reserved fuel and engines at the back, the CM has to be closer to the bottom on return (empty).

The center of pressure is what worries me. Without a bunch of grid fins or something (which would be a pretty brute-force approach, and wouldn't really work with the biconic heat shield), I don't see the center of pressure going high enough for passive tail-first aerodynamic stability.

And if the center of pressure WAS high enough for passive tail-first aerodynamic stability, I don't see how biconic entry would be possible.

I know KSP without FAR isn't exactly a reliable aerodynamic model, but I've thus-far been unable to build any BFS clone that isn't either a nose-first lawn dart or tailspins on entry. And that's with very cheaty fuel-pumping.

[Gaarst]

6 minutes ago, sevenperforce said:

BFR is actually a good deal more straightforward than Falcon Heavy. More engines, sure, but they are all mounted on a single thrust plate so there is no differential torque to worry about. 

Dunno, last time they tried to stick 30 engines on a single core, the stupidly complex plumbing caused 3 booms out of 4 launches. Falcon Heavy feeds its 3*9 engines from 3 separate fuel tanks.

[RCgothic]

N1 was not a SpaceX rocket. The problems weren't just the number of engines, but the lack of testing. SpaceX has that thoroughly covered.

Edited February 24, 2018 by RCgothic

[tater]

10 minutes ago, sevenperforce said:

The center of pressure is what worries me. Without a bunch of grid fins or something (which would be a pretty brute-force approach, and wouldn't really work with the biconic heat shield), I don't see the center of pressure going high enough for passive tail-first aerodynamic stability.

And if the center of pressure WAS high enough for passive tail-first aerodynamic stability, I don't see how biconic entry would be possible.

I know KSP without FAR isn't exactly a reliable aerodynamic model, but I've thus-far been unable to build any BFS clone that isn't either a nose-first lawn dart or tailspins on entry. And that's with very cheaty fuel-pumping.

Retractable fins as a solution?

[sevenperforce]

10 hours ago, mikegarrison said:

All the cost benefits he claims only work if there is a fleet of these things making hundreds or thousands of flights each. And they are huge. Is there any way the payload demand is going to be that big? Even if the launch price/tonne is low?

Launch price/tonne is low, yes, but that's not the point. His model is that launch price is low, period, regardless of price per tonne. You only need a half-dozen flight articles to corner the market, and that's if it is just a tenth as rapidly-reusable as Musk claims.

9 hours ago, YNM said:

It's said they could somehow be cheaper per launch than F9...

I had no idea overnight mail is cheaper on An-225 !

No, no. Cheaper per launch than Falcon 1.

5 hours ago, NSEP said:

The BFR is going to beat all the other rockets, but i will have to say, i have no clue if this thing will fly to LEO or not.

I have no doubt that SpaceX will eventually have a fully-reusable, TSTO, composite-body, Raptor-based launch vehicle operating in LEO. Beyond that, who knows?

1 hour ago, tater said:

SpaceX has to build a next gen LV anyway. BO is going to hit the market like a freight train.

Is NG going to reuse the upper stage?

1 hour ago, tater said:

SpaceX might be WalMart (Sears?) at the moment, but Bezos is still Amazon. Any people down on SpaceX ( ) need to internalize this. Bezos is serious. They haven't done anything spectacular yet... but they will. NG is almost Saturn sized, and NA might make BFR look like a toy.

I can't imagine that New Armstrong will actually be built. Something that much larger than Saturn V seems like crazy talk, even to me.

1 hour ago, tater said:

While their info says that the BFR booster would be reused 1000 times, and the BFS 100, let's assume that is grossly optimistic, and they both last the same # of flights.

Assume that they set launch costs pretty close to F9, or maybe slightly cheaper. What they'd want is to be much cheaper, however, to compete with NG, etc, but let's talk where they'd otherwise just have F9, but now have BFR, instead.

If they last only 10 flights, the per launch cost of a stack is... 25 M$. Propellant costs several hundred thousand. They are still cheaper to launch than F9, even if they deliver an F9 sized payload, but they could deliver a 150 ton payload for the same cost to them (~26M$). Reuse 5 times? Now BFR is costing 50M$/launch. They'd not lose money sending up a 2 ton sat, but they'd be leaving money on the table at the same launch price as F9, or be competitive with a lower margin.

So take the 10 use case. Launching F9 sized payloads is not super competitive if they want to make more money per launch, though they can charge a little less, and make the same. Comanifesting, or launching large payloads, however, becomes a cash cow. With an SLS launch costing billions (since the customer actually paid the dev costs), SpaceX can charge what a Delta OV Heavy costs, and pocket hundreds of millions, and it's still a steal for the customer (NASA).

I think that a clean sheet design is going to be reused more than 10 times, honestly, that;s the point of the exercise. Assume they learn something from Block 5 operations.

I'm guessing they get F9B5 to five flights without refurb and ten flights before end of life. For their eventual methalox TSTO, I'm guessing 20 or 30 flights before refurb and 50 flights before EOL. But even that changes everything.

23 minutes ago, Gaarst said:

Dunno, last time they tried to stick 30 engines on a single core, the stupidly complex plumbing caused 3 booms out of 4 launches. Falcon Heavy feeds its 3*9 engines from 3 separate fuel tanks.

One static fire will eliminate all the concerns that doomed N1.

15 minutes ago, tater said:

Retractable fins as a solution?

Then you have three times as many LOV failure modes.

Conservatively.

[tater]

9 minutes ago, sevenperforce said:

Is NG going to reuse the upper stage?

Not that I know of. I'm thinking in the case where SpaceX sits on what they have, and develops nothing new past incremental F9 improvements (what some BFR opponents here apparently think is what they should do?). In that case, they get hit by a freight train, THEN they work on a next gen vehicle. Which would be profoundly stupid, IMO. They are right to work on the rocket that makes their current rocket obsolete.

11 minutes ago, sevenperforce said:

I can't imagine that New Armstrong will actually be built. Something that much larger than Saturn V seems like crazy talk, even to me.

Bezos wants to go to the Moon. Himself. I fully expect BO to start on NA dev as soon as they get NG flying. This is his $#@! you money, lol.

13 minutes ago, sevenperforce said:

I'm guessing they get F9B5 to five flights without refurb and ten flights before end of life. For their eventual methalox TSTO, I'm guessing 20 or 30 flights before refurb and 50 flights before EOL. But even that changes everything.

Exactly. I think that once full reuse is a thing (meaning all stages, and almost no refurb, like aircraft use), then even if the number oif uses is somewhat small, then they can simply work on production costs to further reduce costs. Propellant is cheap.

[sevenperforce]

One thing I have always wanted to see is a TSTO with an upper stage that thrusts perpendicular to its long axis. By the time you get to staging, you're far enough out of the atmosphere that it's not going to cause any drag losses to pitch 90 degrees. Then you can have a low-impact biconic entry with supersonic retropropulsion and no need for a flip on landing, and engine-out or landing leg failure is typically survivable because you're already laying down on the ground.

It requires linear aerospikes or some other funky engine arrangement, though. But damn if it wouldn't be a beauty.

7 minutes ago, tater said:

Bezos wants to go to the Moon. Himself. I fully expect BO to start on NA dev as soon as they get NG flying. This is his $#@! you money, lol.

Bezos can go to the Moon on New Glenn. Musk could go to the Moon on Falcon Heavy, if he wanted to, though a Raptor-based lunar ship would be far more efficient for the same development effort.

I know that SpaceX will never mix kerosene and methane, but a reusable methalox upper stage to drop onto F9 or FH would be SO stinking capable.

7 minutes ago, tater said:

Exactly. I think that once full reuse is a thing (meaning all stages, and almost no refurb, like aircraft use), then even if the number oif uses is somewhat small, then they can simply work on production costs to further reduce costs. Propellant is cheap.

I am thinking we will always see more refurb than a 747 or a CRJ. Best case we could hope for would be the sort of refurb that a B-2 or an SR-71 has.

[tater]

Yeah, operational AF flying of SR-71 is a great analogy. Or RB-57F (I know a guy who used to fly those, they shared bases with SR-71 because they needed the same support resources... he ended up a surgeon, and since his retirement, I end up hanging with him talking airplanes at medical conferences he still attends while my wife is at meetings).

[Elthy]

It all boils down to the heatshield. If they can make the heatshield work several times without replacement they will have the BFS we all dream of, the remaining work is just rocket science. I would love to see thermal simulations of the BFS reentering on earth (even just from LEO) and how Pica-X would hold up there. Since it can protect dragon (with way higher ballistic coefficient) it will be good enough for at leats one flight, but how many more remains the big question...

[sevenperforce]

9 minutes ago, Elthy said:

It all boils down to the heatshield. If they can make the heatshield work several times without replacement they will have the BFS we all dream of, the remaining work is just rocket science. I would love to see thermal simulations of the BFS reentering on earth (even just from LEO) and how Pica-X would hold up there. Since it can protect dragon (with way higher ballistic coefficient) it will be good enough for at leats one flight, but how many more remains the big question...

BFS will have a much better ballistic coefficient than even the Shuttle. I expect we see a small amount of ablation on each flight but nothing substantial. It's the reliability of landing that concerns me.

[magnemoe]

1 hour ago, tater said:

SpaceX has to build a next gen LV anyway. BO is going to hit the market like a freight train. SpaceX might be WalMart (Sears?) at the moment, but Bezos is still Amazon. Any people down on SpaceX ( ) need to internalize this. Bezos is serious. They haven't done anything spectacular yet... but they will. NG is almost Saturn sized, and NA might make BFR look like a toy.

So the dev costs, while substantial, have to happen even for a slightly improved (and larger) F9. The point of BFR is not (proximally, at least) to land colonists on Mars, but to deal with the upcoming competition from Blue, I think.

Forget his promises of Mars, etc. I think those are part of the show. How many they need? Dunno.

Based on what they said the costs were for ITS, I think a BFR/BFS stack is around 250 million $.

Assume an F9 launch costs 30 M$, and they pocket 35M$.

While their info says that the BFR booster would be reused 1000 times, and the BFS 100, let's assume that is grossly optimistic, and they both last the same # of flights.

Assume that they set launch costs pretty close to F9, or maybe slightly cheaper. What they'd want is to be much cheaper, however, to compete with NG, etc, but let's talk where they'd otherwise just have F9, but now have BFR, instead.

If they last only 10 flights, the per launch cost of a stack is... 25 M$. Propellant costs several hundred thousand. They are still cheaper to launch than F9, even if they deliver an F9 sized payload, but they could deliver a 150 ton payload for the same cost to them (~26M$). Reuse 5 times? Now BFR is costing 50M$/launch. They'd not lose money sending up a 2 ton sat, but they'd be leaving money on the table at the same launch price as F9, or be competitive with a lower margin.

So take the 10 use case. Launching F9 sized payloads is not super competitive if they want to make more money per launch, though they can charge a little less, and make the same. Comanifesting, or launching large payloads, however, becomes a cash cow. With an SLS launch costing billions (since the customer actually paid the dev costs), SpaceX can charge what a Delta OV Heavy costs, and pocket hundreds of millions, and it's still a steal for the customer (NASA).

I think that a clean sheet design is going to be reused more than 10 times, honestly, that;s the point of the exercise. Assume they learn something from Block 5 operations.

New Glenn is upcoming but not know how far it is, new Shepard is not even operational.
However if New Glenn can take 40 ton with first stage reuse, it should be able to do +20 ton with second stage reuse. 

Yes BFR is larger than it need to be for commercial operations, Musk has his agenda and want an big rocket to fulfill it. 
And yes it opens some possibilities who is totally unrealistic to day. Collect an failed satellite in GEO and return it to earth after putting another in GEO is one. 
Collecting the failed satellite is an secondary objective. 
25 man on Moon for 14 days is another, yes this require lots of refueling but still cheap as science missions go. 
No its not optimized but until you get new Glenn with second stage reuse you don't need to be optimal. Then that happens Glenn will be cheaper for everyday missions.
 

[kerbiloid]

21 minutes ago, Elthy said:

It all boils down to the heatshield.

And here we return to aerospikes with their wide central body.

[tater]

The key is that their math (based on ridiculously optimistic ideas about reusing the BFS 100 times (LEO) and the BFR (booster) 1000 times) shows that the launch cost (not per kg, the cost for the entire thing) is lower than Falcon 1.

Even dropping the reuse by 1-2 orders of magnitude (1 for ship, 2 for the booster), it's still price competitive per launch. It's game-changing when you can pay 50 M$, and launch 6 tons, or 150, whatever. Stuff like Bigelow hotels starts looking far more plausible (or other, sideways ideas we haven't considered).

2 minutes ago, kerbiloid said:

And here we return to aerospikes with their wide central body.

I still wonder why no one has tried to do a Bono-style plug engine for these reusables.

[sevenperforce]

3 minutes ago, tater said:

The key is that their math (based on ridiculously optimistic ideas about reusing the BFS 100 times (LEO) and the BFR (booster) 1000 times) shows that the launch cost (not per kg, the cost for the entire thing) is lower than Falcon 1.

Even dropping the reuse by 1-2 orders of magnitude (1 for ship, 2 for the booster), it's still price competitive per launch. It's game-changing when you can pay 50 M$, and launch 6 tons, or 150, whatever. Stuff like Bigelow hotels starts looking far more plausible (or other, sideways ideas we haven't considered).

Exactly. The way to disrupt a market is not "if you build it, they will come" but rather "meet current demand with added capacity". Fill the market while giving it room to grow.

Big Dumb Boosters don't work because they depend on economies of scale. BFR is solvent on current demand but can also exploit whatever economies of scale arise in the process.

[NSEP]

Combine This with BFR and im completely done fam.

Ok sure thing, they are both powerpoint slides.

[magnemoe]

1 hour ago, RCgothic said:

N1 was not a SpaceX rocket. The problems weren't just the number of engines, but the lack of testing. SpaceX has that thoroughly covered.

Soviet could not ground test the N1, they had to launch it and see that failed. 
Pretty much how spacex learned how to land first stages, however the stages was junk anyway so why not experiment on them. 
On the other hand the BFR would need need launch pads, the saturn 5 pads should work but spacex only have one of them. 
You would need an N1 / buran sized transporter / erector. If you want to land first stage back on pad you need an backup pad as you will get fails. 

8 minutes ago, sevenperforce said:

Exactly. The way to disrupt a market is not "if you build it, they will come" but rather "meet current demand with added capacity". Fill the market while giving it room to grow.

Big Dumb Boosters don't work because they depend on economies of scale. BFR is solvent on current demand but can also exploit whatever economies of scale arise in the process.

One fun idea of me then they get the crew version qualified is to use it for standard satellite launches. You launch deploy an satellite in high inclined orbit while having 50 paying tourists. 
Yes you have to stay up an day or two but if you fly parallel with the satellite that is an bonus as you can grab it if it has problems. 
 

[tater]

I also think that the disruption of excess capacity has not been internalized yet, though we have little to work with right now, other than 1 FH flight demonstration. Still, with FH you could launch something for what is still competitive (90 M$) that is high mass. Doesn't need to be an extra-capable billion dollar sat, it could be cheap, because it is huge. Use less expensive, heavier materials, because who cares? Use off the shelf electronics, but shield them with lead, because who cares, it's just mass, and mass is now sorta free in the launch cost.

BFR (I'm almost exclusively talking about the cargo version when I talk about the spacecraft) is a total game-changer here. Buy what would be a F9 launch now, and you get 150 TONS. Even if you need small sats in different orbits, you can add propulsion, and have them plane change. You can start experimenting with something else, or sublet space... I can't wait to see this happen (fingers crossed it works out).

 

Edited February 24, 2018 by tater

[magnemoe]

9 minutes ago, NSEP said:

Combine This with BFR and im completely done fam.

Ok sure thing, they are both powerpoint slides.

Loved the refueling station  
Prefer to take first stages up to orbit myself. 

[StrandedonEarth]

23 minutes ago, magnemoe said:

Loved the refueling station

Yeah, it looked oddly familiar, but I can't put my finger on why... 

[NSEP]

25 minutes ago, magnemoe said:

Loved the refueling station  
Prefer to take first stages up to orbit myself. 

I just realized the refuelling station is a KSP screenshot, lol, do they know? Is it a hidden joke?

[sevenperforce]

31 minutes ago, tater said:

BFR (I'm almost exclusively talking about the cargo version when I talk about the spacecraft) is a total game-changer here. Buy what would be a F9 launch now, and you get 150 TONS. Even if you need small sats in different orbits, you can add propulsion, and have them plane change. You can start experimenting with something else, or sublet space... I can't wait to see this happen (fingers crossed it works out).

I can also see smoothly-sliding scales for mass costs in a way that adding a couple of COTS SRBs (I'm looking at you, ULA) could never achieve. Every tonne of payload costs about a tonne of propellant on LEO, so SpaceX can offer discounts for lower payloads (or charge a premium for more massive payloads) because any propellant it doesn't use can be offloaded to an LEO tanker for whatever the next big BLEO mission is.

"You want your comsat to go up next week? Great. Up to 30 tonnes to LEO for $15M. If you want it to go to GTO, it's an extra $5M, and if you want direct to GEO it'll be $30M. Whatever margins we don't use, we'll dump into our LEO tanker anyway. Your bird is 12 tonnes? Oh, good; you qualify for a 25% flyweight discount. GEO it is, then. Next!"