Introducing Firefly Space Systems

[jongoff]

I'm skeptical that aero spikes are worth the hastle, but it sure will look awesome.

Performance is improved, though by exactly how much depends on the design. I'm not sure if Firefly has said how much of a performance boost is gained with the aerospike compared to if the aerospike were removed from the design.

I'll ask on Twitter, who knows if they'll respond, though.

I've always wondered about this. Because in order to do an aerospike you pretty much have to use an open cycle of some form (gas generator, etc) because you need bleed gas to let you truncate the spike. But that bleed gas is costing you performance vs going with a closed cycle like expander cycle or staged combustion. You do get an altitude compensation benefit, but you're starting off at a disadvantage in Isp at any given altitude due to the open cycle losses.

Personally, I'd like to see a staged combustion or expander cycle engine using thrust augmented nozzles to get both the T/W ratio boost, and some level of altitude compensation.

But I'm glad someone may actually fly an aerospike, so we can get real data on how it performs in practice vs theory.

~Jon

[baldusi]

I sure hope they are successful and do a second vehicle. If they stick to aerospike that would tell something.

[Robotbeat]

I'm skeptical that aero spikes are worth the hastle, but it sure will look awesome.

Performance is improved, though by exactly how much depends on the design. I'm not sure if Firefly has said how much of a performance boost is gained with the aerospike compared to if the aerospike were removed from the design.

I'll ask on Twitter, who knows if they'll respond, though.

I've always wondered about this. Because in order to do an aerospike you pretty much have to use an open cycle of some form (gas generator, etc) because you need bleed gas to let you truncate the spike. But that bleed gas is costing you performance vs going with a closed cycle like expander cycle or staged combustion. You do get an altitude compensation benefit, but you're starting off at a disadvantage in Isp at any given altitude due to the open cycle losses.

Personally, I'd like to see a staged combustion or expander cycle engine using thrust augmented nozzles to get both the T/W ratio boost, and some level of altitude compensation.

But I'm glad someone may actually fly an aerospike, so we can get real data on how it performs in practice vs theory.

~Jon

You don't necessarily need bleed air. You can just accept the performance disadvantage of a truncated spike or don't truncate the spike as much.

[TrevorMonty]

Firefly are not going down a dead end path, if aerospike configuration doesn't work as expected on test stand, they can still build conventional multi engine LV using engines they have.
Performance may not be great but at least it will be creating revenue which is critical thing at this stage Firefly development.

[whitelancer64]

Firefly are not going down a dead end path, if aerospike configuration doesn't work as expected on test stand, they can still build conventional multi engine LV using engines they have.
Performance may not be great but at least it will be creating revenue which is critical thing at this stage Firefly development.

Firefly Alpha is pretty small. 76.5 ft. (23.3 m) tall and 6 ft. (1.8 m) in diameter. If the aerospike doesn't work the way they think it will, it probably won't be able to make orbit. But aerospikes in general are pretty well understood, and they have done years of work on it already. I doubt that is something they are worried about.

[Elmar Moelzer]

I always thought that aerospikes would be more relevant to SSTOs than to TSTOs.

[Davidthefat]

I posted this on Reddit, but I'll post here anyways. It looks like they are film cooling their combustor. Notice the bleed off from the main fuel inlet line? They probably won't be doing regen cooling as the pressure drop would be something to avoid; and most of the fuel is going straight to the injector manifold. They seem to be film cooling at both just downstream the injector face and at the converging section of the engine.

What get's me is the flexible pipe interconnects they use. AFAIK, the electromechanical linear actuators for TVC are no longer being used on the first stage. In the renders shown on the website, only 4 of the 12 had them, so they might potentially be still using them. However, the pipe interconnects provide an unnecessary pressure drop of the propellants, especially without gimballing the combustors. Perhaps this is used during testing for finding the optimal angle of the combustors for the best performance?

And they actually achieved stable combustion for a relatively long time as the plume shown is fully developed unlike the last engine fire (notice the big poof of exhaust in the old one? That means that the picture was taken just after choked flow was established)

It's also much more underexpanded than the old engine. Which makes sense for usage in an aerospike engine.

Looks like they are still going with the 90 degree bend for the LOX manifold inlet. Probably for mounting volume constraints, but I'd personally go for a straight inlet to minimize the turbulence and pressure losses. That that probably causes mounting issues.

[Proponent]

Personally, I'd like to see a staged combustion or expander cycle engine using thrust augmented nozzles to get both the T/W ratio boost, and some level of altitude compensation.

Are there any particular factors that lead you away from pressure feed?  Tom Markusic knows plenty about turbopumps, so he must have had a good reason for going pressure-fed.  I wonder whether something like pressured-fed TAN might have been in the trade space (maybe licensing is too much of a PITA?).

[Proponent]

I always thought that aerospikes would be more relevant to SSTOs than to TSTOs.

I think the key thing that makes an aerospike potentially attractive in this application is the use of pressure feed.  To keep the mass of the prop tanks down, the engine operates at modest pressure.  That means a conventional nozzle would be particularly vulnerable to pressure losses and, hence, altitude compensation is particularly important.

SSTO designs these days, on the other hand, are usually based on high-pressure engines, for which altitude compensation is less important.

[yg1968]

It looks like Firefly raised $19M this month:

https://www.crunchbase.com/organization/firefly-space-systems#/entity

http://www.bizjournals.com/austin/blog/techflash/2016/06/cedar-park-rocket-maker-raises-19-million-funding.html

[gongora]

Tweet from Jeff Foust:

Emerson Gardner, chair of Firefly Space Systems advisory board: 1st launch planned for March 2018; 4 launches that year. CDR this fall.

[Steven Pietrobon]

That's also the launch month for the NASA VCLS satellites! I wouldn't want my cubesat riding on the first launch of a new vehicle.

[Proponent]

The Orbital Mechanics podcast recently did its second interview with Firefly.  There was one geeky detail that I found particularly interesting.  It turns out that one of the reasons for dropping lox/methane in favor of lox/RP-1 relates to cooling.  As the methane enters the cooling channels it is supercritical.  Eventually, though, the pressure drops below the critical pressure, at which point both the liquid and gaseous phases are present.  Determining the heat-transfer characteristics of the two-phase fluid is tricky -- you don't necessarily know how much of each phase is in contact with the walls of the cooling channels.  There are ways around this, like the use of turbulators, but it was looking like a whole research project all by itself.

Apparently this is not a problem for pump-fed methane engines, because the pressures are much higher.

[Davidthefat]

The Orbital Mechanics podcast recently did its second interview with Firefly.  There was one geeky detail that I found particularly interesting.  It turns out that one of the reasons for dropping lox/methane in favor of lox/RP-1 relates to cooling.  As the methane enters the cooling channels it is supercritical.  Eventually, though, the pressure drops below the critical pressure, at which point both the liquid and gaseous phases are present.  Determining the heat-transfer characteristics of the two-phase fluid is tricky -- you don't necessarily know how much of each phase is in contact with the walls of the cooling channels.  There are ways around this, like the use of turbulators, but it was looking like a whole research project all by itself.

Apparently this is not a problem for pump-fed methane engines, because the pressures are much higher.

Well, if you know fluid mechanics, the flow rate of a fluid is determined by the pressure differential. Higher the differential, the faster it is. The faster it is in the cooling channel, a specific parcel of fluid is there for a shorter period of time. Meaning to a specific parcel of fluid, less energy is transferred. So that parcel of fluid does not have time to heat up to start nucleating.

This was an issue with the Space Shuttle Main Engine (RS-25). If you throttle too low (meaning lower pressure) the stay time of the cryogenic hydrogen was too long and would cause boiling of the propellants and lead to burn throughs of the chamber wall. Well, that wasn't the only issue of deep throttling the RS-25, but was one of the limiting factors.

[TruthIsStranger]

fascinating stuff, Proponent and Davidthefat. Thanks

[strangequark]

I've always wondered about this. Because in order to do an aerospike you pretty much have to use an open cycle of some form (gas generator, etc) because you need bleed gas to let you truncate the spike. But that bleed gas is costing you performance vs going with a closed cycle like expander cycle or staged combustion. You do get an altitude compensation benefit, but you're starting off at a disadvantage in Isp at any given altitude due to the open cycle losses.

Personally, I'd like to see a staged combustion or expander cycle engine using thrust augmented nozzles to get both the T/W ratio boost, and some level of altitude compensation.

But I'm glad someone may actually fly an aerospike, so we can get real data on how it performs in practice vs theory.

~Jon

Not to go too far down this rabbit hole, but this isn't strictly true. Remember that closed expander are severely pressure limited. Open expanders actually deliver improved performance, because you can crank them to much higher chamber pressures, due to the lower pump power requirement, and the high delta-P turbine.

[ringsider]

From Reddit:-

https://www.reddit.com/r/space/comments/4utclo/virgin_galactic_vs_firefly_ceo_law_suit_getting/

[docmordrid]

If the court accepts the notion Markusic's work at Firefly is derived from his work at VG, then why couldn't the argument be made that his VG work on Newton was derived from knowledge gleaned during his days at McGregor? Does SpaceX then own a piece of VG & Newton?

[ringsider]

I think the issues are much deeper.

I actually downloaded the documents mentioned in that post last night, and the complaint of VG is that he:-

- started his - competing - company while at VG (documented);
- used VG company resources to do so (documented);
- tried to hire away his colleagues for his startup while still an employee (documented);
- used his access as an employee of VG to their customers to bankroll his startup (documented);
- and finally used IP he was working on for VG as the core of his competing startup (also documented - according to VG, he actually emailed VG's internal counsel to try to get his aerospike work removed from his IPR agreement while he was still an employee and they refused).

All the emails, slide decks and other docs like FF business plan are in the lawsuit filings.

With that background honestly I wouldn't be surprised if there was not a very bad outcome for Firefly... just looking at the lawsuit(s), this document highlights the situation and the cost of these actions, with VG claiming for their lawyer fees in just one of the actions:-

https://www.scribd.com/doc/314360301/Firefly-Systems-v-Markusic-v-Galactic-Virgin-Galactic-opposed-motion-to-dismiss-pdf

[Davidthefat]

I'm curious, FFSS says that the information they have on designing their systems is public information and not trade secret. How does one in court confirm that fact in court? Wouldn't both VG and FFSS have to show all their designs and have the judge determine that they are indeed trade secret or public information? Wouldn't the judge need to know what is even constituted as public information?

Similarities in design itself isn't implying it's trade secret; it might as well both be public information that was used to design them.