Kiwis in Spaaaaace [Rocketlab Electron makes orbit]

New Zealand joins the list of space-faring nations with the successful launch of their Electron rocket:

It’s a small rocket with a 150 kg payload (approximately the mass of a large sheep), but good enough for smaller satellites. Satellites of this size usually share a ride with larger satellites, but they don’t get to launch on their schedule, and don’t necessarily get the orbit they want. So it’s nice to have a cheap launcher with that flexibility.

Rocketlab had a previous launch of this, which was largely successful as a test but didn’t quite make orbit. This launch made orbit and successfully deployed its payload.

This is also the first launcher with electric pumps on the engines. Usually, a small portion of the normal propellant is burned in a turbine to pump the fuel. Here, they use electric motor-driven pumps. It’s not quite as efficient as other techniques, but it’s vastly simpler and probably more reliable. We’ll never see it with large rockets like the Falcon 9, but it’s a great idea for smaller rockets. The rocket even dumps depleted batteries during the launch to save mass (just as burned propellant is dumped overboard and reduces the vehicle mass).

Previous thread on the subject:


(Of course it would be sheep-sized…)

Well, this means I won the pool at work. However, I’m somewhat doubtful about their business case; there is a burgeoning demand for a dedicated small satellite launcher, but even at the base manifesting cost of US$5M there are only a handful of customers in that payload class who can afford a launch, and the rideshare options appear to be pretty limited for anything beyond a CubeSat in a P-POD deployer. If this had flown a couple of years ago they’d probably have payloaders lining up, but a number of prospective satellite operators have either changed their focused for folded up shop. The recently reconstituted Firefly Space Systems is significantly sizing up their Alpha 2 vehicle specifically because of these issues, and of course SpaceX retired the Falcon 1 and got out of the smallsat business because they realized their most optimistic projections showed no profit in it even with essentially unlimited reusability. (The bulk of the cost in a launch is the labor and oversight in integration and launch activities, and below a certain threshold reuse offers little or no savings practical advantage other than a reduction in manufacturing throughput.)

There are a couple of other companies on the horizon planning an inaugural orbital launch demonstration “within the coming year” (which practically speaking means two or three years down the road) but the smallsat industry as a whole is looking weaker rather than stronger despite the number of entrepreneurial efforts and support for smallsats by component OEMs. I’d like to hope that RocketLabs (and Firefly, Blue Origin, IOS, and Virgin Orbit) are successful in making the dedicated small satellite launch industry viable, but given that we’re in the third iteration of commercial space launch boom-and-bust cycles with no one yet showing clear profitability on purely commercial (non-governmental) flight, I remain skeptical.


That makes me miss the old Muppets skits.

Good to see you rooting for the little guy :).

As I alluded to in the previous thread, I’m also skeptical about their business model.

I expect that with their own launchpad, automatic flight termination, and other features, they’re keeping overhead low enough that they’re just barely profitable at $5M. But they still need a market. Big LEO constellations are going to use a large rocket with a dispenser system, as with Iridium NEXT. Little cubesats can’t afford their own launch and will continue to be rideshares.

The best payloads I can think of are for startups, universities, etc. wanting to go past cubesats but not yet at the level of the big guys. HiakaSat was an example here; lost in the Super Strypi launch failure (along with my payload…). However, I have a hard time believing that there are more than a few payloads per year of this type.

The Electron could in principle handle quick replacements if there are losses in a constellation, but they all have hot spares and so I don’t really see a need here.

Maybe they can skim some rideshare profits from the others if they offer better customer service. My experience with cubesats is that SpaceX, etc. only just barely see smallsat rideshares as worth their time, and basically act as if they’re doing you a favor (of course, cubesats always go through an intermediate integrator like Nanoracks, but the principle is the same). Maybe by focusing on smallsats, Rocket Lab can offer a more pleasant experience. You still don’t get to pick a launch day, but maybe they won’t bump their rideshares from flight to flight without explanation.

Well, I dunno. Just brainstorming–I don’t have any real insight into the market. I wish them well in any case. And hey, they made orbit, which is a big deal in anyone’s book.

A while back I argued that it’s getting easier and easier to build a 2-stage kerolox rocket (not easy, but easier). I think that events are bearing this out, though I didn’t anticipate electric pumps at the time. Rocket Lab was very clever in reducing their development risks and costs and deserve a lot of credit for that. It’s one of those ideas that’s obvious in retrospect–pressure-fed engines are a thing (and used by the upper stage of the Falcon 1), but modern batteries have a much greater energy density than compressed gas, so of course battery-powered pumps are viable. It’s a great middle ground between pressure fed and turbopump engines.

So this thread isn’t about intergalactic fruit?

Seriously, that’s where my mind went when I first read the thread title. I thought it was an agricultural experiment.

They have the contract with Moon Express for their Lunar X Prize entry, due to launch later this year. I wish them all well, but … I have my doubts.

I’m always in favor of increased competition (and anything that forces ULA to admit to overpricing their flights is alright by me) but the fundamental problem at the small satellite launch arena is that costs have to be low enough to support a nascent industry and the venture capitalists investing in it, but high enough that a company can make a return on investment and profit in the near term. So far, no one has yet made a profit in commercial spaceflight without subsidy or deep pocket government payloaders to help defray the costs of infrastructure and staff, and while I haven’t seen a specific projection on the number of flights per year that RocketLabs would have to achieve to get to breakeven, my off-the-cuff guestimate would be 20 to 30 per year, or essentially one every two weeks.

There are some particular efficiencies with small satellites that can drive costs down below the traditional thresholds. For one, the payloads are smaller and require less complex equipment and labor to transport and handle them. They tend to be more self-contained and generally don’t require site-load of propellants (a major processing and integration issue), often not requiring even on the ground health and status monitoring, which allows them to be encapsulated at leisure and integrated as a vehicle comes together, which means that payloads could potentially be stored and launched as opportunity arises rather than having to be specifically manifested between other payloads and suffer delays because of integration or technical issues with other missions. Most importantly, small payloads tend to have relatively minor mass modal contribution to vehicle stability and controllability, and thus, there is not a need to perform extensive modal testing and multiple coupled loads analysis cycle analyses peculiar to the payload. (The small size limits the need for displacement-limited testing and the general lack of flexible deployable elements that are not securely locked down during flight often means that analysis alone can demonstrate that payload dynamic response is well away from or above vehicle modes.)

On the down side, however, is the fact that New Zealand is logistically difficult to reach (although again, the compact size and low mass makes this less of a problem in terms of payload transportation), the size of the payload envelope will force design restrictions or eliminate certain types of payloads, unknown reliability is challenging for payloaders who need to meet specific business objectives or timelines to demonstrate viability, and the vehicle design itself appears very limited in terms of upgrading capability, which means if it falls just below a sweet spot in payloader needs it may not be competitive in the market place even against significantly more expensive launchers.

Rideshare does suck for payloaders, and while the situation is better than the days when you had to rely on ‘Big Daddy’ to get you a lift on a ULA rocket, it still means having to compromise science or mission objectives. It is basically suitable for space science experiments or educational projects that just need to get to someplace in LEO, but not good for anyone who needs to achieve specific orbital arguments. And while SpaceX gotten a particularly bad reputation for dumping CubeSats and other payloads at the last minute to make up differences in mass budget or concerns about reliability, this is going to be the case with any commercial space launch provider who has to serve their primary payloader first. (ULA and other government-sponsored vehicles have mandates to carry a certain number of rideshares and so are a little more dilligent about sticking with a secondary payloader unless there is a genuine risk, but if you aren’t on-pad come integration time there will be a mass simulator flying in your place. Nobody waits for hitchhikers.). Dedicated space launch is crucial for commercial smallsat, but the costs are still prohibitive without a solid plan for a large return on investment that is hard to demonstrate.

High grade kerosense (RP-1) and liquid oxygen (LOX) rocket engines may seem “easy” to build, but that is only because we’ve had decades of prior experience learning (through failure as much as success) how to build and operate them, and even then there are a lot of phenomena that are not fully understood in the combustion of such a complex hydrocarbon fuel that requires a lot of trial and error in design. A liquid-to-gas-phase fuel such as methane is actually easier to design for despite the lower specific energy. However, the thermodynamic limitations of combustion-based engines are always going to dictate pretty small payload mass margins just because most of the vehicle has to be propellants and the systems to store and pressurize them. The next evolution in chemical propulsion technology will be the development of continuous wave detonation engines (CWDE), also sometimes known as rotating detonation engines (RDE), which can substantially improve thermodynamic yield and offer greater compactness than conventional combustion chamber and de Laval nozzle engines. CWDEs are particularly desirable for small launch applications because of the compactness, but simluating and controlling detonation phenomena is another challenging problem in propulsion.


Most certainly. Standing on the shoulders of giants and all that.

SpaceX had a presentation on their Raptor simulation efforts a while back. Unlike RP-1, it was actually possible to fully simulate methane combustion: it has “only” around 300 intermediate reaction products, which was actually tractable for their purposes. RP-1 has thousands. I doubt that anyone has really done a full RP-1 combustion simulation at this point (of course it’s always possible to approximate).

I’m certainly looking forward to this, but have seen very little in the way of development so far. Even aerospike engines, which have been in development for decades, and have made it to test stands, are still not being used anywhere (though we’ll see about Firefly). If RDEs follow a similar path, they’re decades off.

Raptor may be the most advanced engine available for quite some time. It seems to be the final refinement as far as staged combustion goes.

Maybe not. They’d have to launch by March 31, since the prize deadline isn’t going to be extended anymore, and their business case without it seems imaginary.

The problem is less one of modeling combustion products (which is largely of concern for external ballistics) is less an issue than specific dynamic phenomena that affect ignition and combustion stability. One of the problems with rocket-grade kerosene is that it is desirable to operate in an oxidizer-rich regime to obtain maximum combustion efficiency and the lightest reaction products, but high oxidizer levels result in greater instability and potential for problems in ignition. Methane, being a lightweight fuel with a high hydrogen load, is far more stable provided you stay away from well-known detonation regimes, and is a much easier fuel to handle and store than cryogenic hydrogen. It is also relatively easy to synthesize which I’m sure appeals to the notion of in-situ propellant production on Mars, although practically manufacturing usable quantities of methane and oxidizer in-situ has yet to be demonstrated and has significant hurdles to overcome.

Aerospike nozzles have been flown (although not on orbital class vehicles as of yet) and while their are practical issues with cooling and control they are fairly well understood. The thing that prevents adoption is largely a lack of necessity; they’re really only necessary for single stage to orbit (SSTO) applications, and conventional solutions using extendable exit cones are better proven and provide better vacuum performance than an aerospike nozzle in upper stage applications.

Research on RDEs is largely confined to stationary power generation applications, but they can produce thrust just like any other combustion application, and allow greater theoretical energy converstion efficiency than any combustion process with comparable propellants. The problems with them come in predicting and controlling detonation phenomena and throttleability as well as building a combustion chamber robust enough to withstand detonation pressure ewaves, but their basic simplicity in design and compactness make them very interesting for future propulsion applications, particularly in using lightweight fuels like methane, propane, or dimethyl ether that are storable and can be readily synthesized despite offering lower performance and or less energy density than ambient pressure liquid hydrocarbon fuels. I have a ‘toy model’ of an orbital space launch vehicle for which one variant uses DME/LO[SUB]2[/SUB] propellants in a two stage heavy lift application that demonstrates overall good performance.

I haven’t seen enough detail on the design of the Raport engine to assess how advanced it may be, but numerous companies have developed staged combustion engines including Rocketdyne (now Aerojet-Rocketdyne), NPO Energomash, and Blue Origin, as well as the engines used on the more recent version of the Chiense Long March vehicles (purportedly). I would infer by “advanced” that you mean the Raptor is a full-flow staged combustion (as opposed to fuel-rich on the RD-25 or oxygen-rich on the NK-33/43 and RD-180 family). However, while there are certain performance advantages to full flow staged combustion, it is not the end all of engine system performance and may not represent the ‘best’ engine design for stability and reliability.

For instance, the RD-25 “Space Shuttle Main Engine” (SSME) was highly regarded for its compactness and high specific performance (which was quite impressive), but this required very high pressures and operating speeds in its staged combustion system which in the original design necessitated a complete rebuild every two or three operating full cycles (which is not so impressive, although later blocks improved operating life substantially). The original engine system design proposed by Rocketdyne (which, BTW, used an aerospike nozzle because the Space Transportation System was essentially an augmented SSTO) actually delivered comparable performance with much lower operating pressures using a much simpler combustion tap-opff cycle, and achieved slightly better overall ground-to-orbit total thrust. In general, I would take any pronouncements by SpaceX or any other company that their engine is somehow “more advanced” with skepticism unless they shows some performance metric which actually distinguishes them from other existing engines.


Right. “Best” can be a lot of different things and depends on the application. A gigantic pressure-fed, low chamber pressure engine was the best choice for the Truax Sea Dragon, because it was suited for the core principle of the Big Dumb Booster concept. But no one would call it particularly advanced.

Full-flow seems to be the final evolution of turbopump rocket engines. Not best, necessarily–that depends on cost, development time, reliability, complexity, risk, reusability, and a host of other things–but if you could handwave away all the difficulties, that’s the design you would want. Well, except for maybe the expander cycle for small LH2 engines.

As far as Raptor goes, full-flow, along with the quite high chamber pressure (though reduced from early estimates) as well as the fuel choice contribute to it being “advanced” as far as I’m concerned. Of course methane has been considered before but in terms of actual use, it’s a modern propellant.

Exactly. The issue is with the intermediate products, not the stuff in the exhaust–combustion is a pretty slow process, all things considered (on the order of a millisecond), and a lot of things can happen between when the propellant is injected and it leaves the nozzle. Diffusion, turbulent mixing, acoustic waves, etc. all play a part and at many different length scales. The reactions themselves are highly nonlinear with pressure and temperature. If you really want to know what’s going on, you have to model all this stuff. Keeping the molecules simple makes this a tractable problem.

Incidentally, here is the video where SpaceX talks about their simulation efforts. I’m happy to see that they use my company’s processors for their simulation :).

Question: I hadn’t heard of dimethyl ether as a propellant before. What are its advantages compared to RP-1, methane, or LH2?

The other problem with LH2, other than it’s dammed cold, is it has an affinity for migrating through the walls of the tanks due to small molecular size. As Jerry Pournelle once pointed out, “It wants OUT”.

Turns out Rocket Lab had a couple of secret payloads on this launch.

One is basically a giant disco ball, which is supposed to produce predictable flashes of reflected sunlight like the Iridium flares.

The second is much more important and interesting, IMO. It’s a previously undisclosed “kick stage”, with a new engine using a new type of monopropellant. It boosted two of the Lemur 2 cubasats from a 300x500 km orbit to a 500x500 km orbit.

By my quick eyeball estimate of the kick stage mass fraction, I’m guessing it won’t be very useful as a 3rd stage for single heavy payloads. However, I think it can provide a lot of flexibility for Rocket Lab to be able to launch cubesats from multiple customers, as in this test launch.

Any ideas about how much the new kick stage will improve the business case for Rocket Labs and their potential customers?

Interesting. I wonder what the monopropellant is. Some possibilities:

Apparently their order book is already full for a couple of years, and they are planning to scale up the frequency of launches quite markedly - possibly up to 50 a year.

All in all a remarkable feat.

I looked to see if the old standby for exotic propellants–Ignition! by John Clark–contained any references to dimethyl ether, but I couldn’t find any.

Of course, in the process I was reminded of some of the whack-a-doodle things they did try or at least look at. Just in the dimethyl category, we have:
dimethyl mercury
dimethyl beryllium
“a dimethylamino group attached to a mercaptan sulfur”

As for this latter one, it is compared unfavorably with another propellant already described as intense, pervasive and penetrating, and resembling the stink of an enraged skunk, but surpassing, by far, the best efforts of the most vigorous specimen of Mephitis mephitis. Clark leaves the odor to the imagination: whose odor can’t, with all the resources of the English language, even be described.

As a point of order, that is one very large sheep.

150kg payload would be 4-6 adult sheep i.e. a modest ovarian harem.

As an avowed non-expert in all matters ariesian, I turned to Wikipedia, which claims:
Ewes typically weigh between 45 and 100 kilograms (100 and 220 lb), and rams between 45 and 160 kilograms (100 and 350 lb)

If correct, this would imply a maximum of three small adult sheep. Or a single large one.

Ovine. An ovarian harem would be something entirely different, and probably harder to fit into a capsule.