F9 Third Stage

[arachnitect]

If it's ever been talked about, I've missed it. Something like Blok D or Transtage. Why not?

Right now, F9 has no way of turning its good LEO performance into higher energy. More Stages is how everyone else solves this problem.

Alternative is a whole new second stage (probably with a higher energy propellant and engine cycle) or using FH for payloads that really shouldn't need it.

Would be simple compared to many of spacex' other development projects.

[gospacex]

I believe three stage version is called Falcon Heavy.

[guckyfan]

Right now, F9 has no way of turning its good LEO performance into higher energy. More Stages is how everyone else solves this problem.

Alternative is a whole new second stage (probably with a higher energy propellant and engine cycle) or using FH for payloads that really shouldn't need it.

Would be simple compared to many of spacex' other development projects.

SpaceX is certainly not "everyone else".

A new second stage with higher energy propellant? An expensive detour, no way they would do that. Though a while back I thought they might do a methane upper stage just for training before they build the new large launch vehicle. We know now this is not happening.

Simple would be a third stage with hypergolics. But Falcon Heavy with reusable boosters is easier once a Heavy has been built.

[IRobot]

Right now, F9 has no way of turning its good LEO performance into higher energy. More Stages is how everyone else solves this problem.

Musk had said before that keeping stages to minimum is a goal, to reduce costs and improve reliability.

They don't need more performance if they are cheap and reliable.

[M129K]

Simple would be a third stage with hypergolics. But Falcon Heavy with reusable boosters is easier once a Heavy has been built.

A Falcon 9 with a third stage would be able to bridge the massive gap between F9 and FH, and a hypergolic third stage wouldn't have to be that expensive compared to a higher energy upper stage, yet could increase F9 GTO performance to about 7 tons.

[Tomness]

Super draco derived third stage would be awesome, but i doubt they want or need it.

[inventodoc]

Super draco derived third stage would be awesome, but i doubt they want or need it.

The Superdraco would be really cool and I can see why that would seem appealing.  I thought that a cluster could even power a whole rocket stage (or enable new classes of VSTOL aircraft) but the problem with Superdraco is the crummy Isp - I think it's only around 235 which is fairly poor efficiency.   Merlin gets in the low 300's and the Raptor is aiming for 365ish.   I think that hydrolox rockets get around 430.

One of the first posters mentioned 'Falcon heavy'.  I guess they are right that Falcon Heavy is truly a three stage rocket.   The side boosters are stage one (with cross propellant flow to the center),  the still mostly full center booster is part of stage 1 and all of stage two, and the upper stage is stage 3, so there you have it.

[A_M_Swallow]

Try a reusable space tug.  A Raptor engine with appropriate sized fuel tanks can fly several tons of cargo to lunar orbit and the Lagrange points.  It can pick up 13 tonnes of cargo from a Falcon 9 in LEO and lots of fuel from the Falcon Heavy.

[rst]

A Falcon 9 with a third stage would be able to bridge the massive gap between F9 and FH, and a hypergolic third stage wouldn't have to be that expensive compared to a higher energy upper stage, yet could increase F9 GTO performance to about 7 tons.

They could also bridge that gap with fully-reusable, RTLS Falcon Heavy (eating the heavy performance penalty for RTLS on the core stage).  More generally, the reason that you don't see people launching small payloads on big rockets is that you're throwing away a lot more rocket than you have to.  The long-term goal for SpaceX is to get the rocket back; if they can do that, and keep refurbishment costs to a minimum, that changes the economics.

[sdsds]

I'm guessing SpaceX won't add a third stage to F9 because those few missions that could use it could instead incorporate a solid motor into the payload.

[guckyfan]

I'm guessing SpaceX won't add a third stage to F9 because those few missions that could use it could instead incorporate a solid motor into the payload.

That would require development on many different satellites by many suppliers. A SuperDraco third stage would be a solution for all purposes. The ISP of SuperDraco gets a lot better with a vacuum extension of the bell. It would beat the Ariane 6 before Ariane construction has even begun.

That's not saying I believe it will happen, I don't.

[darkenfast]

Why Super Dracos?  Why not use a Kestral?   Two spherical tanks, helium tanks and a Kestral, contained inside the fairing.  There's not a lot you can do, as this is all eating into payload and space, but is there any niche that a Kestral 3rd stag would fill?  Perhaps certain space probes?  I don't know, I'm just wondering if there would be a business case for this.

[CuddlyRocket]

A Falcon 9 with a third stage would be able to bridge the massive gap between F9 and FH, and a hypergolic third stage wouldn't have to be that expensive compared to a higher energy upper stage, yet could increase F9 GTO performance to about 7 tons.

They could also bridge that gap with fully-reusable, RTLS Falcon Heavy (eating the heavy performance penalty for RTLS on the core stage).  More generally, the reason that you don't see people launching small payloads on big rockets is that you're throwing away a lot more rocket than you have to.  The long-term goal for SpaceX is to get the rocket back; if they can do that, and keep refurbishment costs to a minimum, that changes the economics.

This is probably the lower cost alternative and SpaceX optimises for cost, not performance.

[arachnitect]

Simple would be a third stage with hypergolics. But Falcon Heavy with reusable boosters is easier once a Heavy has been built.

A Falcon 9 with a third stage would be able to bridge the massive gap between F9 and FH, and a hypergolic third stage wouldn't have to be that expensive compared to a higher energy upper stage, yet could increase F9 GTO performance to about 7 tons.

How do you figure 7t to GTO? I'm playing around with a spreadsheet and don't really see that much performance increase for heavier payloads. I'm more interested in what kind of dV you can get for 1800kg payload.

If I'm doing this right (not likely I admit), a surprisingly good option is a Falcon 1 second stage (Kestrel). Also really good: Castor 30B.

Storable propellant stages get screwed twice. First with low Isp, then with bad mass fraction.


As for cost of RTLS FH boosters vs. a 3rd stage, even if you get the boosters back, it's not free to fly them. I have no idea how to even begin estimating the costs though so we'll just have to wait and see I guess.

[macpacheco]

If it's ever been talked about, I've missed it. Something like Blok D or Transtage. Why not?

Right now, F9 has no way of turning its good LEO performance into higher energy. More Stages is how everyone else solves this problem.

Alternative is a whole new second stage (probably with a higher energy propellant and engine cycle) or using FH for payloads that really shouldn't need it.

Would be simple compared to many of spacex' other development projects.

Short term solution is Falcon Heavy, long term solution is scaled up Falcon 9 using Raptors. The later rocket would have twice the payload of Falcon Heavy to most missions, making it possible to accomplish every Falcon Heavy mission in expendable mode with full reuse.
Using a true third stage make little sense. The FH solution is very smart, and scaling up the F9 with Raptor should provide a very affordable launch system (cheap fuel, full reuse).
Remember, SpaceX is about cost effectiveness, not making the technically perfect rocket that is way too expensive. Without economics the Mars Colony won't be viable. Every launch system should be designed for maximum flexibility with minimum cost.
The main issue of F9 and FH efficiency wise is it's low ISP in Vacuum mode compared with LH rockets. Using methane should bridge that gap very well, maintaining a single engine for first and second stage.

[Zed_Noir]

SpaceX might developed a Super-Draco based upper stage if it could also be used as Dragon service module component. Solid motors and nonstorable liquid motors are not applicable for SM usage.

[Elmar Moelzer]

Do we have any official numbers for the Isp for the SuperDraco. According to Musk, a lot of development effort went into that engine. So I would assume that it has better than average performance numbers.

[guckyfan]

Do we have any official numbers for the Isp for the SuperDraco. According to Musk, a lot of development effort went into that engine. So I would assume that it has better than average performance numbers.

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

[Elmar Moelzer]

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

Hmm, wonder what the limits given an optimal engine bell size would be.

[arachnitect]

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

Hmm, wonder what the limits given an optimal engine bell size would be.

Chamber pressure doesn't make much difference once you're in vacuum, expansion ratio is more important.

AJ-10 gets ~320s. I would consider anything above 330s absurdly good. Pump fed has gotten to 340s.

Mass fraction for pressure fed engine is an issue, high chamber pressure becomes a liability.

[chamann]

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

Hmm, wonder what the limits given an optimal engine bell size would be.

Chamber pressure doesn't make much difference once you're in vacuum, expansion ratio is more important.

AJ-10 gets ~320s. I would consider anything above 330s absurdly good. Pump fed has gotten to 340s.

Mass fraction for pressure fed engine is an issue, high chamber pressure becomes a liability.

Would a higher chamber pressure engine with a smaller expansion ratio get the same isp as a lower chamber pressure engine with a larger expansion ratio?  If so, would that mean a smaller nozzle is needed for the higher chamber pressure engine, compensating for the increased weight of the engine?

[arachnitect]

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

Hmm, wonder what the limits given an optimal engine bell size would be.

Chamber pressure doesn't make much difference once you're in vacuum, expansion ratio is more important.

AJ-10 gets ~320s. I would consider anything above 330s absurdly good. Pump fed has gotten to 340s.

Mass fraction for pressure fed engine is an issue, high chamber pressure becomes a liability.

Would a higher chamber pressure engine with a smaller expansion ratio get the same isp as a lower chamber pressure engine with a larger expansion ratio?  If so, would that mean a smaller nozzle is needed for the higher chamber pressure engine, compensating for the increased weight of the engine?

My understanding is that (in vacuum) expansion is orders of magnitude more important than chamber pressure. Also, when you increase the chamber pressure, that means more material not only on the engine, but the entire propellant system as well.

Examples of pressure fed vacuum engines I have found run at ~100-150 psi.

[baldusi]

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

Hmm, wonder what the limits given an optimal engine bell size would be.

Chamber pressure doesn't make much difference once you're in vacuum, expansion ratio is more important.

AJ-10 gets ~320s. I would consider anything above 330s absurdly good. Pump fed has gotten to 340s.

Mass fraction for pressure fed engine is an issue, high chamber pressure becomes a liability.

Would a higher chamber pressure engine with a smaller expansion ratio get the same isp as a lower chamber pressure engine with a larger expansion ratio?  If so, would that mean a smaller nozzle is needed for the higher chamber pressure engine, compensating for the increased weight of the engine?

My understanding is that (in vacuum) expansion is orders of magnitude more important than chamber pressure. Also, when you increase the chamber pressure, that means more material not only on the engine, but the entire propellant system as well.

Examples of pressure fed vacuum engines I have found run at ~100-150 psi.

Please remember that higher chamber pressure means more expansion for the same size of nozzle. Or smaller (and lighter) nozzle for equal expansion. Also, the engine itself is not heavier, since it can be done smaller. But, on pressure fed, you're still bounded by tank mass and pressurizing mass. In other words, for 1000psi chamber you'd need ten times the He or N to pressurize the tanks.
If you do gas generator or expander, higher pressure can actually lower mass.

[hkultala]

Do we have any official numbers for the Isp for the SuperDraco. According to Musk, a lot of development effort went into that engine. So I would assume that it has better than average performance numbers.

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

It's not high pressure, it's "high pressure for a pressure-fed", but all pressure-fed engines are low pressure.

Superdraco is optimized for high thrust-to-weight ratio, and cheap manufacturing, not to high isp. That's where the "lot of development effort" went to. And I'm quite sure the "lot of development effort" was still much cheaper/less  development effort than many other engines of different companies.



The ISP would still be much worse than the ISP of merlin 1dvac, and the third stage could not be reused.


So the third stage just adding much more complexity with very little benefits. Very against spacex philosophy.

[guckyfan]

Do we have any official numbers for the Isp for the SuperDraco. According to Musk, a lot of development effort went into that engine. So I would assume that it has better than average performance numbers.

It is very limited due to its engine bell size. It should be easy though to add a vacuum nozzle and significantly enhance ISP, considered its high pressure. Easy as rocket science goes.

It's not high pressure, it's "high pressure for a pressure-fed", but all pressure-fed engines are low pressure.

Yes, of course.

Superdraco is optimized for high thrust-to-weight ratio, and cheap manufacturing, not to high isp. That's where the "lot of development effort" went to. And I'm quite sure the "lot of development effort" was still much cheaper/less  development effort than many other engines of different companies.

Yes, of course.

The ISP would still be much worse than the ISP of merlin 1dvac, and the third stage could not be reused.

Yes, it is a hypergolic so less ISP, but cheap and reliable, good for a third stage.

So the third stage just adding much more complexity with very little benefits. Very against spacex philosophy.

I agree it is quite unlikely and against SpaceX philosophy. But very little benefit? Poster M129K mentioned upthread it would get Falcon 9 about 7 ton capability to GTO for small cost and small developoment effort. If true this is something that could scare Arianespace into a coma with their Ariane 6 concept.

[aero]

Does anyone happen to know or have an educated guess of the dimensions of the Super Draco engine? Several members saw the engine on display so I wonder if anyone can estimate the Super Draco:

Nozzle diameter,
Throat diameter,
Nozzle length?

I've looked at Wikipedia and some numbers from there are: Isp - 235 s, vacuum thrust - 73,000 N, burn time - 25 s, prop - 1388 kg. (per engine?) These numbers aren't consistent though because they give mdot of 30 or 50 kg/s. Anyone have better numbers for mass flow or know what these mean?

[baldusi]

Well, first consider that isp is most probably sea level. Second, 25s are the 12.5s + 12.5s of the DragonFly demonstration, which does throttle. Third, that's the total fuel stated in the EIS for the capsule, which probably includes margin. Fourth, Dracos also use it. And fifth, stock thrust is less than full thrust, since this engines are clearly under-expanded, isp might actually increase by throttling.

[aero]

Well, first consider that isp is most probably sea level. Second, 25s are the 12.5s + 12.5s of the DragonFly demonstration, which does throttle. Third, that's the total fuel stated in the EIS for the capsule, which probably includes margin. Fourth, Dracos also use it. And fifth, stock thrust is less than full thrust, since this engines are clearly under-expanded, isp might actually increase by throttling.

Thanks, good points all. Especially SL Isp. And I am confident that you're right about the prop load. I'm still looking for mdot to go with 73,000 N vacuum thrust though. Sure would help if I had some nozzle dimensions.

What I'm thinking is that, given the data, I can slap a vacuum nozzle on my engine model to estimate thrust for a 3-rd stage Super Draco engine configuration. Or, a lunar or Mars lander configuration.

[Lobo]

I'm guessing SpaceX won't add a third stage to F9 because those few missions that could use it could instead incorporate a solid motor into the payload.

That would require development on many different satellites by many suppliers. A SuperDraco third stage would be a solution for all purposes. The ISP of SuperDraco gets a lot better with a vacuum extension of the bell. It would beat the Ariane 6 before Ariane construction has even begun.

That's not saying I believe it will happen, I don't.

As I understand, most payload will already have a hypergolic propulsion system for performing various orbital tweaks and corrections during it's operational life.  So it's not much increase in cost to put larger propellant tanks on a payload that will have the propulsion system anyway.  Larger tanks are cheap.

I also think most payloads going to GEO are actually delivered to a GTO orbit, and then perform the remaining dV themselves to their actual orbit.  I think an actual 3rd stage as part of the LV is usually more useful if the payload is going through escape.  Then the payloads propulsion does course corrections during the flight to wherever it's going, as well as whatever propulsive needs when it gets there.

So I doubt you'd see a 3rd stage on any Falcon rocket unless it starts sending payloads through escape.  Which you might if Elon decides to drop a Dragon on the Mars surface or something.  But I think a partially expendable FH could send Dragon to Mars with the need for a 3rd stage in any event.

[baldusi]

I don't  believe that SD would be the ideal choice. It's optimized for thrust, fast start up and deep throttling. Nothing of that is needed for a third stage engine. Elon stated that SD runs at 1000psi, which is ridiculously high for a pressure fed third stage engine. They could leverage some of the technology, but it should be a completely different engine.
To put things in perspective Fregat's engine has a thrust around 20kN and isp 332s. You'd want something a bit bigger for the F9, like ~30kN and 8tonnes of mass for the whole stage. But remeber that once you're orbital your thrust can be extremely low.

[edkyle99]

Yes, a third stage would allow Falcon 9 to increase its GTO payloads substantially. 

Using a Falcon 1 second stage with a pressure fed Kestrel (and offloading a bit of stage 2 propellant) would increase GTO x 28.5 deg payload to 6.4 tonnes and GEO-1,500 m/s payload to 5.5 tonnes.  Values of 7 and 6 tonnes, respectively, might be possible with a custom built Kestrel-powered third stage.

That's almost Proton territory (and almost reusable Falcon Heavy territory).  It might match the most powerful Delta 4 Medium+ version.  I'm sure the company will be tempted.

 - Ed Kyle

[baldusi]

What cost would that add? They would need a new tooling and to encase everything with a fairing extension. But if it could be done for 20% extra cost it would, basically, eat the whole market. Ironically, a 4m fairing might get them some extra performance, too.

[TrueBlueWitt]

Yes, a third stage would allow Falcon 9 to increase its GTO payloads substantially. 

Using a Falcon 1 second stage with a pressure fed Kestrel (and offloading a bit of stage 2 propellant) would increase GTO x 28.5 deg payload to 6.4 tonnes and GEO-1,500 m/s payload to 5.5 tonnes.  Values of 7 and 6 tonnes, respectively, might be possible with a custom built Kestrel-powered third stage.

That's almost Proton territory (and almost reusable Falcon Heavy territory).  It might match the most powerful Delta 4 Medium+ version.  I'm sure the company will be tempted.

 - Ed Kyle

How much more of the DoD range of payloads would SpaceX be able to cover?

Wasn't there talk, before the move to F9, of an improved Falcon 1 Second stage with Lighter alloy and a Kestrel 2 with tighter tolerances and higher ISP?  Could SpaceX 3D print a better version of Kestrel for such a purpose?

[Elmar Moelzer]

I like the idea of a low cost F9 3rd stage to increase GTO capability. I am not sure reviving the Kestrel is the right answer though. From a pure cost perspective, I would think that it is better to use something that SpaceX has to build in larger quantities anyway, like the SuperDracos.
But then, maybe SpaceX can just increase the performance of the 2 stage F9 via extensive optimization and almost achieve the same thing.

[GalacticIntruder]

IMO, I don't think SpaceX cares about nor needs more GTO performance, for either F9 or FH. What they do need is better BLEO performance, for both rockets, especially FH, since it will most likely be tasked for that mission, more than GTO.

[GreenShrike]

Yes, a third stage would allow Falcon 9 to increase its GTO payloads substantially. 

Using a Falcon 1 second stage with a pressure fed Kestrel (and offloading a bit of stage 2 propellant) would increase GTO x 28.5 deg payload to 6.4 tonnes and GEO-1,500 m/s payload to 5.5 tonnes.  Values of 7 and 6 tonnes, respectively, might be possible with a custom built Kestrel-powered third stage.

Or SpaceX could keep the same 4-5t limit to GTO and gain a couple more tons for Stage 2 reusablity hardware. I have to imagine that expending a small (read: cheap) Kestrel- or SuperDraco-powered third stage would cost less than expending an entire Falcon9 second stage.

I don't imagine a third stage would help with LEO at all, but if a third stage can increase F9's GTO payload capacity, then it could also and/or alternatively provide more margin for S2 reuse during GTO missions.

[MP99]

Please remember that higher chamber pressure means more expansion for the same size of nozzle. Or smaller (and lighter) nozzle for equal expansion.

It's obvious how that applies at SL, but I'm having trouble picturing how that works in vacuum?

If you ran the chamber at 1 PSI, it will have a certain mass flow rate for a certain throat diameter.

But, shouldn't the Isp & thrust scale with the expansion ratio, right up to the point where the exhaust gets cold enough to hit a phase change? Or is that the issue you're discussing?

cheers, Martin

[deruch]

Any discussion on the technical difficulties of adding another separation with a 3rd stage? 

[edkyle99]

Any discussion on the technical difficulties of adding another separation with a 3rd stage? 

It seems apparent that a three-stage rocket should be slightly less reliable than a two-stage rocket, all other things being equal.  However, this hypothetical three-stage Falcon 9 could be compared with a Delta 4M+5,4, which has two core stages and four solid boosters, or with an Atlas 5-421 with its two strap on solids, or a Proton which uses four stages, and so on.  Falcon 9 v1.1 and Atlas 5-401 are the most capable two stage rockets to GTO on earth.  Anything more capable is going to require more propulsion stages or motors - or bigger rockets with more engines.

 - Ed Kyle