10,000psi chamber pressure

[jabowery]

The Raptor engine's nearly 5,000psi is a truly impressive chamber pressure, but compare its complexity to the 10,000psi ultracentrifugal engine at halfwaytoanywhere.


For armchair engineers, critique the Mathematica calculations.

[meberbs]

The first line of the first link says:
" The combustion chamber is nominally 2,500 psi,"

There are mentions of higher pressures further down, but that just leads to the question of which is correct. Chamber pressure isn't nearly as important as other metrics: thrust and Isp. The link claims an Isp of 300 s, though notably lacking in what the nozzle and environment assumptions are. Merlin already beats that at 311 s in vacuum. Raptor expects to reach 380 s in vacuum once they build a vacuum nozzle. Of course fuel plays into this with propane being an unusual choice.

I can't find any clear explanation on the site about how the fuel and oxidizer are supposed to be pumped. Clear diagrams of fuel flow and engine cycle would be helpful. (I don't have anything on my computer at the moment that can do anything useful with Mathematica files, in case there is anything useful in them.)

[jabowery]

The first line of the first link says:
" The combustion chamber is nominally 2,500 psi,"

That is for the first of two proposed prototypes.  The first is 2000lbf thrust and runs a very conservative set of parameters in order to avoid catastrophic failure given the crude numeric modeling.

There are mentions of higher pressures further down, but that just leads to the question of which is correct.

The second of the two proposed prototypes:
"Nominal thrust ...6725.lbf
ChamberPressure ... (11340 psi)"

Even so, it is still running very rich, so as to assure adequate cooling flow.

Chamber pressure isn't nearly as important as other metrics: thrust and Isp.

The thrust scales favorably. 

The link claims an Isp of 300 s

At a very rich mixture.

Raptor expects to reach 380 s in vacuum once they build a vacuum nozzle. Of course fuel plays into this with propane being an unusual choice.

It is hard to beat 380s, even in vacuum and I suspect propane is rather difficult to avoid in this engine due to the shared thermal environment between LOX and the fuel.

I can't find any clear explanation on the site about how the fuel and oxidizer are supposed to be pumped.

It is simply an axial feed for both fuel and oxidizer, going to radial cooling channels that are spinning about the axis and acting as a centrifugal pump.

The patent description may help, although the calculations linked in the original post indicate the number of cooling channels needed to be increased -- particularly if a lower conductivity metal than aluminum is used.

[john smith 19]

So let me see if I understand this correctly.

This is a 20 year old patent that's just about expired for this concept that presumably has never been built before.

Once I'd realized the exhaust nozzles are at the top of the drawing I think I've got it.

This is basically a staged combustion cycle without the staging. IOW there is no preburner (or no after burner if you follow Russian terminology and your PoV).

Logically it follows the thinking of a gas tap off cycle to eliminate one of the combustion chambers but at the pressure levels of a full flow SC design.

So full preburner pressure in the main chamber but compensated for reducing the volume of the main chamber.

But the drive turbine is also acting as the pump impeller.

So you've got multiple fluid connections and multiple bearings inside the main combustion chamber.

While these are conceptually simple their implementation in this environment is going to be challenging.
Likewise the extreme thermal environment using a single rigid drive shaft/fluid connector guarantees lots of interesting design problems.

Personally I'm not that impressed by high chamber pressures, although they look great on study contract proposals. Changing fuels from RP1 to one of the short chain HC's gave several seconds improvement without breaking a sweat (more if you can find a low BP strained molecule like Propyne)

It looks like it could give better T/W than even an SC engine but there's lots of engineering in there and  stuff working ways that are pretty far outside the known range. Gas dynamics inside a conventional thrust chamber is fairly complex, but this has the injection system at right angles to the flow direction you want to ultimately achieve.

If the material properties of Aluminum alloys are as good as you think building a low(ish) pressure version of the design should present few problems. That would put many of the doubts about its complexity to rest. 


 

[jabowery]

So let me see if I understand this correctly.

This is a 20 year old patent that's just about expired for this concept

The patent was expired in the early 2000s due to fees.  I don't know what the "Anticipated expiration" of this coming May (2019-05-14) means but the the patent ceased being enforceable long ago.

that presumably has never been built before.

This prototype was tested circa 1996 but had a hard start:

I was the source of funds for the engine's development and patent fees.  I had to make a hard decision between paying lawyers and paying for testing.  As there was an angel investor who indicated interest ("It seems too good to be true but I can't find anything wrong with the concept.") and conducting engine tests on limited funds is either foolish or heroic depending on one's view, I cut off funds for testing and went with patent coverage so there would be a property in which to invest.  The other guys (Roger Gregory and Don Scott) decided to continue with testing without my support, advice and/or consent.  Don, who made his living doing custom machining for silicon fabrication lines, built the prototype and took off with Roger to Roger's family farm in northern Michigan.  The detonation in the "exploded view" above was the predictable result.  No one was injured but the sheriff took a dim view of these explosions going off in his county and no further testing occurred to the best of my knowledge.

Meanwhile, I managed to get another consulting gig during the DotCon era, which permitted me to successfully file internationally in hopes that the angel investor would weigh in and support adequately safe testing.  It never happened and when the DotCon bubble popped, the resulting over-supply of H1-b guest workers absorbed all of the subsistence positions in my profession long enough that I had to abandon supporting the international patent fees -- particularly given that my wife was suffering from the onset of Huntington's disease.

Roger continued refining the design via modeling the engine as documented at the halfwaytoanywhere.com site.

Perhaps the biggest problem with finding investors in rocket engine development, after the obvious technical, market and  political risks:

is that if even one jurisdiction is outside the patent regime, a company can use your patent disclosure to serve the international launch market from that jurisdiction and pay no royalties.  Investors understandably don't like this kind of risk even if they believe in the technology.  It is a far worse problem than other kinds of patents.  In retrospect, I should not have trusted the angel investor's words and, instead, imputed to him the just-described, rational concern -- realizing that it just doesn't make sense to invest in rocket engine development unless you are in a position to vertically integrate, perhaps keeping trade secrets, as Elon Musk was.

Once I'd realized the exhaust nozzles are at the top of the drawing I think I've got it.

You are referring to the patent's "FIG. 1":

This is basically a staged combustion cycle without the staging. IOW there is no preburner (or no after burner if you follow Russian terminology and your PoV).

That's an interesting way to view it:  The injection points of the combustion chamber (at the outermost diameter) is the small-volume "preburner" where coriolis forces rapidly mix the fuel and oxidizer, and combustion progresses as the mixture simultaneously moves inward toward the larger volume because it decreases in density as it increases in temperature.  You really have to pay attention to the enormous coriolis forces during this entire process to adequately model the combustion dynamics.

But the drive turbine is also acting as the pump impeller.

...and as source of the engine's thrust.

So you've got multiple fluid connections

I'm not sure what you mean by "fluid connections" but in terms of liquid connections, there are 4, as pictured in FIG. 1:  40 to 14, 56 to 14, 44 to 32 and 14 to 34. 
The injectors (26) might be thought of as "fluid connections" (32 to 24 and 34 to 24) between the feeds and the combustion chamber, I suppose.

The critical thing about the liquid connections is to avoid cavitation.  The boundary layer effects of the rotating axial feed pipes (14 and 38) matches the tangential velocity as the liquids progress toward entry to the combustion chamber's impeller/cooling channels. There does need to be a modest tank pressure to but that's all.

and multiple bearings inside the main combustion chamber.

No.  There are 3 bearings, and they're outside the combustion chamber (there is no "main" combustion chamber -- there is only (24) in FIG. 1). Two are designated by 48 and one by 54 in FIG.1.  In the prototype that was built, these were silicon nitride bearings and operated well within their speed tolerance since the tangential velocity of the inner radius is so low.  Bearing reuse should be quite feasible without resorting to more exotic bearings.

While these are conceptually simple their implementation in this environment is going to be challenging.
Likewise the extreme thermal environment using a single rigid drive shaft/fluid connector guarantees lots of interesting design problems.

Agreed.  Our limited budget permitted only TK!Solver modeling (supplemented by Mathematica), so there is good reason to be suspicious of the design.  In particular, our objective of achieving a design that would be workable at small scale may not be feasible because scaling prefers larger engines so as to reduce thermal transfer from the combustion chamber to the cooling channels.  Particularly in the case of propane, the heat capacity is limiting.  If the liquid goes supercritical in the cooling channels/impellers, the liquid density goes down, injector pressure goes down, flow rate goes down and pressure drop across the injector goes down (risking "chugging"). 

Nevertheless, the engine's fabrication simplicity is attractive as evidenced by the fact that Don got the pictured prototypes fabricated in a high performance automotive machine shop on virtually no money.

Personally I'm not that impressed by high chamber pressures, although they look great on study contract proposals.

There are reasons both pro and con for higher pressures.  In our case, we were looking to reduce the size of the engine so as to make it cheap to fabricate.  That benefits from high pressures.  One big benefit of very high pressures is that the expansion ratio achievable in very small nozzles, without a bell, goes way up.  The mass savings also go into payload and/or relieving other launch system constraints.  It does appear that the engine Roger came up with in his Mathematica refinements may even be extended by a central aerospike because of the large number of nozzles.  That would, of course, have the corresponding vacuum benefits without much increase in thrust to weight.

It looks like it could give better T/W than even an SC engine but there's lots of engineering in there and  stuff working ways that are pretty far outside the known range.

Agreed.  I guess one reason I'm dredging up this ancient history is that modern modeling tools, combined with the development budgets now invested by vertically integrated companies, may substantially mitigate risk.  Taken in conjunction with the economy of the engine's fabrication, testing then may become feasible even with a number of destructive hard starts while trying to refine the empirical models.

Gas dynamics inside a conventional thrust chamber is fairly complex, but this has the injection system at right angles to the flow direction you want to ultimately achieve.

I'm not sure what injection vectors have to do with thrust vectors.  Injection momentum is tiny compared to thrust momentum.

If the material properties of Aluminum alloys are as good as you think building a low(ish) pressure version of the design should present few problems. That would put many of the doubts about its complexity to rest. 

When I did thermal modeling aluminum came out as an amazing material due to its conductivity.  That combined with the flow velocities due to coriolis effects within the cooling channels/impellers resulted in a very non-intuitive optimum design.  But, as you point out, modeling these things is challenging.

Thanks for your critique.

[meberbs]

The first line of the first link says:
" The combustion chamber is nominally 2,500 psi,"

That is for the first of two proposed prototypes.  The first is 2000lbf thrust and runs a very conservative set of parameters in order to avoid catastrophic failure given the crude numeric modeling.

From reading that site, I had no idea that it was referring to 2 different prototypes, that at least clarifies some of the apparent contradictions.

Even so, it is still running very rich, so as to assure adequate cooling flow.

You mention this a couple times, but you left out an important word: fuel rich or oxygen rich? I assume fuel rich, because oxygen rich anything in any type of rocket engine is a great recipe for explosions unless you are in possession of some advanced metallurgy.

Chamber pressure isn't nearly as important as other metrics: thrust and Isp.

The thrust scales favorably. 

Scales favorably with what? I know chamber pressure is generally proportional to thrust, but multiple other factors also come into play, so it really isn't comparable across engines expect to describe how far you are pushing the limits to make your engine work. Also for a comparison with staged combustion engines, the main chamber wouldn't necessarily be the highest pressure location, Elon Musk recently said about Raptor's design:

Much above 300 bar main chamber pressure means extreme oxygen preburner pressure of 700 to 800+ bar. Definitely pushing the limit of known physics.

I can't find any clear explanation on the site about how the fuel and oxidizer are supposed to be pumped.

It is simply an axial feed for both fuel and oxidizer, going to radial cooling channels that are spinning about the axis and acting as a centrifugal pump.

The patent description may help, although the calculations linked in the original post indicate the number of cooling channels needed to be increased -- particularly if a lower conductivity metal than aluminum is used.

I saw the patent, but I find patents to be the single most painful format to try to extract information from. Your explanation and response to john smith help some but I am still not sure if I fully follow. I understood that you are using centrifugal pumping, but not where the power was being derived from. It seems that you are getting the power from exhausting at an angle. It seems to me this leaves you with a massive control problem, since changing the angle of the multiple exhaust nozzles would probably be complex, but maintaining a controlled spin rate matched correctly with engine throttle settings seems like it would severely limit something, and possibly make startup or shutdown a challenge.


I am sorry to here about the history of your attempt at development. Often patents aren't as useful or necessary as they sound. Test data is relatively easy to keep proprietary/trade secret, never expires, and means a whole lot more in determining if a concept can actually work. Of course for rocket engines, you need to have the funding resources (and appropriate testing location) to have hardware blow up on you multiple times, especially with any kind of new  or radical design. If nothing blows up, it almost certainly means you did insufficient testing, and your rocket will blow up (it might anyway of course.) Timing wise, it seems like developing your engine 5 years ago, or a bit earlier would have been good. There are today literally over 100 companies worldwide trying to make smallsat launchers, most of these will go nowhere, but clearly a lot of funding got tossed around in this area. If you had been lucky with timing and your idea worked, you could have sold the engine to one of them, or merged with one (since overall vehicle design and engine sizing are closely tied together.) Today seems too late for this concept to go anywhere. The market is over-saturated with vehicle developers, and any that have a chance at becoming real players would already have engine designs.

[jabowery]

Perhaps the attached coloring of the drawing will help.

From reading that site, I had no idea that it was referring to 2 different prototypes, that at least clarifies some of the apparent contradictions.

Yes, the site needs a lot of work to make things clearer.  Both Roger and I have been preoccupied, and once it was apparent we needed vertically integration in a business to justify the patent fees hence angel investment, he lost interest.

Even so, it is still running very rich, so as to assure adequate cooling flow.

You mention this a couple times, but you left out an important word: fuel rich or oxygen rich? I assume fuel rich

Correct. 

Chamber pressure isn't nearly as important as other metrics: thrust and Isp.

The thrust scales favorably. 

Scales favorably with what? I know chamber pressure is generally proportional to thrust, but multiple other factors also come into play, so it really isn't comparable across engines expect to describe how far you are pushing the limits to make your engine work.

Thrust scales favorably with respect to the limiting factor on the engine:  heat capacity of the liquid propane.  As the engine gets larger, the thrust goes up and the combustion chamber's ratio of surface area to volume goes down linearly, hence the flow of liquid propane has better cooling capacity.

Also for a comparison with staged combustion engines, the main chamber wouldn't necessarily be the highest pressure location, Elon Musk recently said about Raptor's design:

Much above 300 bar main chamber pressure means extreme oxygen preburner pressure of 700 to 800+ bar. Definitely pushing the limit of known physics.

Combustion is single-stage.  There is a single combustion chamber illustrated in the colored drawing going from red to yellow to indicate progressive combustion from the injectors swirling inward.

...I understood that you are using centrifugal pumping, but not where the power was being derived from. It seems that you are getting the power from exhausting at an angle.

Correct.

It seems to me this leaves you with a massive control problem, since changing the angle of the multiple exhaust nozzles would probably be complex

The nozzles are fixed to the main rotating body -- no gimbling of the nozzles.

but maintaining a controlled spin rate matched correctly with engine throttle settings seems like it would severely limit something

That's right.  Thrust control is a problem that Roger had only vaguely speculated might be possible by (somehow) limiting the flow from the feed pipes to the impeller/cooling channels.  It isn't something Roger worked on -- at least not while I was helping with the modeling.

and possibly make startup or shutdown a challenge.

Well, yes but transients are a problem with any engine.  This engine may need an externally powered spin-up.

...Timing wise, it seems like developing your engine 5 years ago, or a bit earlier would have been good...Today seems too late for this concept to go anywhere. The market is over-saturated with vehicle developers, and any that have a chance at becoming real players would already have engine designs.

That can cut both ways.  Operations are typically far more expensive than development, so as things ramp there could be proportionately more money shaken loose.  Moreover, we were targeting a NASCAR machine shop mindset with this engine -- very simple design, cheap to fabricate; and it was.  I can easily imagine some Russians developing it for the fun of it.