Author Topic: Speculation: SFR (mini-BFR) as fully reusable Falcon Heavy replacement  (Read 265333 times)

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
So in Musk's Reddit AMA he mentioned that raptor has an optimal thrust to weight ratio when it is 500,000 pounds of force thrust. Falcon heavy has exactly 9 times that amount of vacuum thrust, 4,500,000 lbs. this seems to provide an opening for something smaller than the BFR but capable of serving falcon heavy payloads while being fully reusable due to Raptor's much higher Isp. A 9-Raptor "SFR" (Small, um, Falcon Rocket). It would have the same sort of configuration as F9, but could take advantage of improvements in mass ratio (which would have to be large in order to also compensate for methane/LOx's lower bulk density) and much better Isp and much higher thrust of Raptor.

This has several infrastructure related advantages over just using falcon heavy or using BFR:

Full, return-to-launch-site reusability for both stages for all FH class payloads (perhaps barge landing would be needed initially for ones at the very peak of FH performance, but I don't think that'd be necessary).
Simple, 2-stage operations (versus 2 boosters, a core, and an upper stage)
Lack of kerosene's coking problems
Cheaper fuel (not significant)
Less hassle due to fewer engines but still plenty for engine-out capability on first stage
Capability of about 100tons IMLEO for full expendable in case anyone wants it (or in between for barge-landing and partial reuse)

Could still use the Falcon Heavy test stand if core logistics were figured out since it has the same thrust.
LC39a, Vandenberg, and Brownsville's launch pads could still be used (modified) since the thrust level is the same as Falcon Heavy.

There is a global market (including US govt) for around 25-30 launches per year that are outside partially-reusable F9R's capability (especially for pure return to launch site). The launch market has been growing at about 5-6% per year for the last decade, so at that growth rate, there should be around 40 launches per year by the time something like this could be ready and test launched a couple times (6-8 years). Around 40 is usually the "is it worth it?" line for full RLVs. So this may be a better first platform for Raptor to be tested on than BFR. It seems to be fairly economically viable as a full RLV without any miracles, and it should be significantly more operationally efficient than the 27 coking engines and 3 flyback/barge-landing cores of Falcon Heavy without requiring totally new and expensive infrastructure to be built. Due to the necessarily wider cores, it'd have to be almost fully reusable right from the start, but given half a decade experience in F9R and Dragon V2, that should be much less of a challenge than it is now. Plus there's so much extra performance for the majority of payloads, you could initially budget an enormous retro-burn for the upper stage if the TPS is under-performing.


Anyway, I wouldn't be surprised if something like this were revealed later this year as a stepping stone to full BFR. And given the simplicity of the design and operational advantages even at the very heaviest (and thus highest margin) of current payloads, it'd be nearly impossible to compete with.
« Last Edit: 01/06/2015 09:47 pm by Robotbeat »
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Space Ghost 1962

  • Senior Member
  • *****
  • Posts: 2780
  • Whatcha gonna do when the Ghost zaps you?
  • Liked: 2925
  • Likes Given: 2247
Maybe a reaction to competitive pressures?

But I doubt that as it doesn't seem to help in "getting to Mars", which seems to be Musk's new focus.

He needs to accelerate BFR to do that.

That's why I think its a "fall back" strategy. Just in case the launch services market stays irrational for an indefinite time, possibly funded by annoyed governments indefinitely, attempting to wait for SpaceX to die.

So, by having a fallback, they can do BFR as it suits Mars, then wait out the political mediocrities until they die off. 

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
I actually think he's being honest in saying it's for technical reasons that Raptor will be lower thrust. There really is a mass penalty if you scale an engine up big enough.

But also, a smaller engine can be built and tested earlier, with less risk of difficult-to-solve thrust instability issues.

But keep in mind that as rocket engines are tweaked and matured, they almost always increase in thrust. Merlin more than doubled in thrust from Merlin 1A to today's Merlin 1D and there's still room for increased thrust. So it's certainly possible that a vehicle which would be capable of just 100 tons initially could grow to be capable of 200tons after a decade of tweaking and upgrading.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Mongo62

  • Full Member
  • ****
  • Posts: 1072
  • Liked: 834
  • Likes Given: 156
So it's certainly possible that a vehicle which would be capable of just 100 tons initially could grow to be capable of 200tons after a decade of tweaking and upgrading.

Oh that wee baby! Only 100 tons to LEO! But you can't expect much from such small rockets. I remember that when Saturn was V, he could only put 130 tons into LEO. And SLS Block I would only lift 75 tons to LEO. Pathetic!

Seriously, I have no doubt that a 9-Raptor, 100 ton to LEO launcher would be entirely feasible from an engineering standpoint. I wonder if it would make sense from a financial standpoint, given the finite SpaceX development budget. Given that everything that a 9-Raptor launcher could carry to LEO could also be carried by a 30-Raptor megalauncher, for (presumably) a little more in operating costs but saving the development cost, I am not sure that the financial case for a 100-ton launcher closes.

Online TrevorMonty

The excess capacity of this mini BFR could be used to supply a LEO fuel depot. Deliver a Dragon to ISS or Bigelow station, then offload surplus fuel (approx 30- 40t) to nearby depot. Given the mission is paid for by the Dragon the fuel delivery would be free.

The fuel could then be used for BLEO missions eg lunar flyby.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
BFR would be too powerful for LC39a. They'd need an entirely new launch complex, most likely.

BFR may just be a step too far, with no near-term market need for that large of a vehicle. Mini-BFR, however, would be like a cost-optimized version of Falcon Heavy and could launch from 3 of their current 4 launch sites (as well as use their Texas test stand) and give Raptor a much nearer-term market, and with gradual improvements in thrust it could (like v1.0 to v1.1) organically grow along with the payloads (assuming larger payloads start appearing).

I'm not saying it's a guarantee, just that it seems to be a credible next step. And it could certainly enable initial Mars settlement as well, as the Mars base is built up. And again, because it's much more optimized for cheap commercial launch than either FH or BFR, it'd also cement SpaceX's domination of the launch industry. (BFR would leave an opening for a better-optimized RLV which could still compete in the high-revenue Ariane5/Proton market.)
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Mongo62

  • Full Member
  • ****
  • Posts: 1072
  • Liked: 834
  • Likes Given: 156
If excess thrust is a problem, perhaps a partly-fueled BFR with only some of the engines lit could work? It would have a lower payload mass fraction to be sure, given the unused engines and tankage, but as long as it can get the payload into the right orbit, who cares? Engine wear and tear would be almost the same in both cases (maybe 10 engines lit instead of 9).
« Last Edit: 01/07/2015 12:08 am by Mongo62 »

Offline GWH

  • Full Member
  • ****
  • Posts: 1741
  • Canada
  • Liked: 1925
  • Likes Given: 1277
Would it at all be feasible to use the BFR to launch multiple satellites destined for similar orbits?  E.g. One yearly launch putting up a dozen satellites in GEO?  Additional launches to access other orbits as required.
Fuel reserves and delta V capabilities on individual satellites could be increased significantly given the massive payloads on a BFR.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
If excess thrust is a problem, perhaps a partly-fueled BFR with only some of the engines lit could work? It would have a lower payload mass fraction to be sure, given the unused engines and tankage, but as long as it can get the payload into the right orbit, who cares? Engine wear and tear would be almost the same in both cases (maybe 10 engines lit instead of 9).
I doubt that'd be a problem since the engines no doubt will be throttleable. Especially for an RLV, you're almost always better off loading up to the top with propellant and using the extra propellant as margin. For an upper stage, this extra propellant is especially helpful because you can use it to do a big retro-burn after payload sep and reduce the wear and tear on your heatshield (especially relevant in this case because we're probably talking about some version of PICA-X).
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline go4mars

  • Senior Member
  • *****
  • Posts: 3748
  • Earth
  • Liked: 158
  • Likes Given: 3463
...because it's much more optimized for cheap commercial launch than either FH or BFR, it'd also cement SpaceX's domination of the launch industry. (BFR would leave an opening for a better-optimized RLV which could still compete in the high-revenue Ariane5/Proton market.)
What do you see as the big differences in op costs (key driver's and guess of their cost difference) between the truly Big FR (aka transorbital railroad), and the smaller rocket you're suggesting? 

As you know, I lean strongly toward a "millions of colonists sooner" approach; whatever form that may take.
Elasmotherium; hurlyburly Doggerlandic Jentilak steeds insouciantly gallop in viridescent taiga, eluding deluginal Burckle's abyssal excavation.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
Would it at all be feasible to use the BFR to launch multiple satellites destined for similar orbits?  E.g. One yearly launch putting up a dozen satellites in GEO?  Additional launches to access other orbits as required.
Fuel reserves and delta V capabilities on individual satellites could be increased significantly given the massive payloads on a BFR.
I don't think that'd make much sense. Remember, BFR and this proposed mini-BFR are full RLVs, so a launch need not cost very much (especially on the mini-BFR, one could imagine launches for just, say, $15 million, if the overall launch rate were high enough), and you kind of need a high launch rate anyway for it to make sense. Instead of launching, say, 4 satellites at once, where you have to wait for each payload to arrive at the launch site and be integrated into a big carrier adapter thing, you could just launch each one as it becomes available. It might come close to saving you money on less costly integration, and it certainly would be worth more to your customers if they could launch right away instead of having to wait for the yearly launch.

It is no doubt still worth doing for smaller payloads which are cheaper and more numerous, but on the multi-hundred-million dollar GSO birds, you'd be better off with a dedicated launch. And besides, if you have such enormous performance, you could actually insert right into GSO, not merely GTO, saving a bunch of propellant and/or time.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
...because it's much more optimized for cheap commercial launch than either FH or BFR, it'd also cement SpaceX's domination of the launch industry. (BFR would leave an opening for a better-optimized RLV which could still compete in the high-revenue Ariane5/Proton market.)
What do you see as the big differences in op costs (key driver's and guess of their cost difference) between the truly Big FR (aka transorbital railroad), and the smaller rocket you're suggesting? 

As you know, I lean strongly toward a "millions of colonists sooner" approach; whatever form that may take.
Well, one of them is launch pad wear and tear. You're dumping a bunch more energy onto the pad with the BFR than a mini-BFR (which would have the same thrust as Falcon Heavy), and there's some suggestion that the relationship between pad damage and thrust is worse than just a linear relationship (sorry, no citation, but if anyone finds anything, let me know). Another is liability, i.e. if the rocket blows up, it's a much bigger explosion with BFR than with mini-BFR.

Also, larger things have lower natural frequency. There are scaling laws that come into effect that makes it disproportionately harder to move an object around for processing and the like as the object becomes larger. This is why larger internal combustion engines tend to operate at much lower RPMs than smaller internal combustion engines (think about it: it will take longer for a piston to go through a longer stroke if it's moving at the same speed, for instance).

A larger stage is going to require a bigger crane, which obviously costs more money to rent or buy and maintain.

Plus, you're going to be consuming more fuel in general.

Plus the raw material costs are greater, of course. SpaceX uses Al-Li to get its excellent mass fractions... 3D printed parts--used a lot for Raptor--have a cost proportional to their mass (and they aren't cheap.... thousands or tens of thousands of dollars per kilogram... and even if you buy a machine, a nice DMLS machine costs around a million dollars, weighs a ton, and will take a year just to print its own mass in 3d printed parts if run around the clock, plus there's a significant amount of labor involved plus powder and maintenance).
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
There's a pretty good chance that for colonization a BFR-sized rocket will be cheaper. But a mini-BFR will be cheaper for at least the next decade.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
Another issue is reliability-related:

Suppose BFR has 30 engines on the first stage alone (FH will have 27). If BFR is developed first, as the very first application of Raptor, that means a huge probability of engine failure occurring at least once on a flight (in fact, I'd probably bet money on it... a little, not a lot) or at very least causing a LOT of aborts during the launch hold-down. You think F9 has a lot of scrubs, just wait until you see the BFR monster trying to spin up 30 new Raptors at once!

Going with a smaller 9-engine vehicle initially, however, allows you to increase the reliability of the engine without being bogged down in endless aborts and nearly guaranteed engine failure during flight (which doesn't, of course, mean launch failure). If SpaceX had went straight to the Falcon Heavy, they probably would've folded before they ironed out all of Merlin's little glitches. (This still may be a problem for FH, of course, but by flying Falcon 9 v1.1 at least a dozen times beforehand, it gives Falcon Heavy's first flight a shot of launching without engine failure or endless aborts.) Also, it allows you to improve the performance of Raptor before integrating it on the larger vehicle where more is at stake.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline docmordrid

  • Senior Member
  • *****
  • Posts: 6333
  • Michigan
  • Liked: 4204
  • Likes Given: 2
Another issue is reliability-related:

Suppose BFR has 30 engines on the first stage alone (FH will have 27). If BFR is developed first, as the very first application of Raptor, that means a huge probability of engine failure occurring at least once on a flight (in fact, I'd probably bet money on it... a little, not a lot) or at very least causing a LOT of aborts during the launch hold-down. You think F9 has a lot of scrubs, just wait until you see the BFR monster trying to spin up 30 new Raptors at once!
>

Unless they divide the thrust chambers up between a more manageable number of pumps, say 3 per. Then you have 1.5 mlbf per engine pump, but 500 klbf per chamber.
« Last Edit: 01/07/2015 01:49 am by docmordrid »
DM

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
...that would require a pretty creative interpretation of Musk's remarks. But I think there are other scaling laws that point away from this. A smaller turbine engine can, at some point, have better power density than a larger one. We shall see.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline GWH

  • Full Member
  • ****
  • Posts: 1741
  • Canada
  • Liked: 1925
  • Likes Given: 1277
I completely agree with all your points however I would perceive this rocket would be in direct competition with the BFR and money to develop it. 
As long as a BFR can deliver payload (in multiples if required)  cheaper than a first stage reusable F9/FH and the competition than it should get used, furthering spacex 's goals.  Until there is a launcher comparable in cost than this big excessive rocket can get utilized. 
Regarding integration I do have my own ideas on how the interior of a MCT would be designed around a shipping container like architecture for delivering cargo to Mars (pre fanned habitats etc).  Definitely it's own topic, however if this same modular structure were applied to integrating satellites, the cost and lead time should be low.  Adapters are simple if mass isn't a major consideration in my opinion.
Really the comes down to how to justify using a massive vehicle commercially to make development of the MCT feasible..  Just my thoughts anyways. 

Offline go4mars

  • Senior Member
  • *****
  • Posts: 3748
  • Earth
  • Liked: 158
  • Likes Given: 3463
Another issue is reliability-related:

Suppose BFR has 30 engines on the first stage alone (FH will have 27). If BFR is developed first, as the very first application of Raptor, that means a huge probability of engine failure occurring at least once on a flight (in fact, I'd probably bet money on it... a little, not a lot) or at very least causing a LOT of aborts during the launch hold-down. You think F9 has a lot of scrubs, just wait until you see the BFR monster trying to spin up 30 new Raptors at once!

Going with a smaller 9-engine vehicle initially, however, allows you to increase the reliability of the engine without being bogged down in endless aborts and nearly guaranteed engine failure during flight (which doesn't, of course, mean launch failure). If SpaceX had went straight to the Falcon Heavy, they probably would've folded before they ironed out all of Merlin's little glitches. (This still may be a problem for FH, of course, but by flying Falcon 9 v1.1 at least a dozen times beforehand, it gives Falcon Heavy's first flight a shot of launching without engine failure or endless aborts.) Also, it allows you to improve the performance of Raptor before integrating it on the larger vehicle where more is at stake.
If there are 30 raptors on a core, the plan could be to assume a few don't make it.  There was a F9 flight where this happenned and the primary mission was still accomplished.  Since only 1 engine failed in flight in all the F9 launches (one in however many merlin's that's been in total), 1/30 or 1/100 (depending how many raptors per BFR) seems less problematic than 1/9 (secondary payload might have hit the right orbit too).  But it's almost certainly better than a FH sidecore engine failure because presumably the opposite core would quickly switch one off to negate thrust imbalance.  So 2/27 out. 

I wonder if the Big Falcon first stage will be cylindrical or something else - like SuperNexus.
« Last Edit: 01/07/2015 02:25 am by go4mars »
Elasmotherium; hurlyburly Doggerlandic Jentilak steeds insouciantly gallop in viridescent taiga, eluding deluginal Burckle's abyssal excavation.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 39270
  • Minnesota
  • Liked: 25223
  • Likes Given: 12114
Remember, N1 also could do engine-out. But when you have that many engines, it becomes really complicated (especially in the Soviets' case since they were rushed and never actually ground-test-fired their rocket before launching it).


Remember, Raptor has NEVER flown before while Merlin has flown (in one form or another) 135 times. Merlin 1D has flown 80 times, thus you can probably safely assume its reliability is around 99.4%, giving the Falcon Heavy a roughly 16% chance of having at least one engine-out (and roughly a 2 or 3% chance of dual-engine-out, which could easily mean loss-of-mission). If another, say, 9 F9s fly with no engine loss before FH flies sometime in 2015 or 2016, then that further improves the odds of Falcon Heavy having an engine-out to a mere 8% (and a dual-engine-out probability of around 0.6%).

But look at the Raptor. You don't know, a priori, what its engine reliability is. The first few Merlins had problems (although F1 flight 3 just had a residual thrust error, so probably isn't relevant for this calculation), so let's say we estimate roughly 98% reliability for Raptor or any such new engine that SpaceX might develop without much heritage. That gives a roughly 17% chance of first stage engine-out on the first mini-BFR (and dual-engine-out of 2.5%). But on the BFR, that gives you a whopping 45% chance of engine-out on the first launch and even a 20% chance of two engine out and 9% chance of losing three engines! Designing your avionics to safely handle 3 engine failures is a lot more complicated than just one engine-out, and as I showed, 3 engine failures has a fairly high probability.

...but instead, suppose we launch mini-BFR 200 times--hey, it's an RLV!--, and it has maybe three engines out in that time (they occur early on, most likely, and their root cause is determined and fixed, thus their reliability improves quite rapidly). You've flown Raptors a total of 2000 times (including the upper stage). Now, your chance of losing a first stage BFR engine on the first BFR flight is more like 4% (so quite unlikely), and your chance of a 3 engine out is a miniscule 0.01% (well, not quite that good, probably, since the odds of failure aren't likely to be completely independent... but that's difficult to calculate :) ).

So you greatly reduce your odds of the first launch of BFR being a dramatic failure if you build and operate the mini-BFR for a while, first. This is the same way SpaceX has operated from the beginning. Falcon 1-> Falcon 9 v1.0->v1.1->Heavy.
« Last Edit: 01/07/2015 03:03 am by Robotbeat »
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline hkultala

  • Full Member
  • ****
  • Posts: 1198
  • Liked: 748
  • Likes Given: 945
They dont need to equal the capasity of Falcon Heavy, there are no payloads that need anything close to the the full capasity of Falcon heavy.

Even with only 5 engines it could lift _any_ commercial or goverment satellite and still be fully reusable.

Though second stage with one raptor would then be way overpowered, but making second stage big and staging early might make reuse of first stage easier/more efficient.

Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
0