Page 2 of 3

Re: Reaction Engines

Posted: Mon Feb 03, 2014 12:39 am
by ladajo
Why not fly a single engine test vehicle?
A single engine should have enough push to get itself and enough fuel off the ground to then demonstrate a full transition cycle to rocket, then back to jet, and come in for a landing. This is about a full scale engine demonstrating its full flight regime, not full endurance.

You don't need to orbit it. You are only looking to test the engine, and get it back.

It is even simpler if you don't want the engine back.

JMHO

Re: Reaction Engines

Posted: Mon Feb 03, 2014 1:31 am
by GIThruster
My sentiment exactly, save I would want the effort put into such a demonstrator to be able to serve some sort of function. If I were at AFRL, I'd be interested in an X-37 replacement, meaning orbital. Doesn't need to be manned nor have much payload.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 3:22 am
by 93143
ladajo wrote:Why not fly a single engine test vehicle?
A single engine should have enough push to get itself and enough fuel off the ground to then demonstrate a full transition cycle to rocket, then back to jet, and come in for a landing. This is about a full scale engine demonstrating its full flight regime, not full endurance.

You don't need to orbit it. You are only looking to test the engine, and get it back.

It is even simpler if you don't want the engine back.
That's true. But one advantage of the SABRE concept is that the inlet is the only thing that cares what Mach number the vehicle is flying at. Behind the inlet, the engine could (for all it knows) be flying really slowly through very hot air. And behind the precooler, the only difference is the power balance between the precooler and the preburner, both of which heat the helium loop.

It can be ground-tested.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 4:28 am
by GIThruster
93143 wrote:It can be ground-tested.
It can but for it to make TRL5, it needs to be tested across a broad range of environmental conditions. If that can be done someplace like Marshall, then there's an advantage to ground testing. If there is no preexisting thrust stand that can duplicate the expected environment for TRL5, then going straight to flying prototype is likely a better ROI than building such a test apparatus.

I would want to know how much modification would be needed to drop this into something like X-37 and be able to fly it. If there's any existing frame that can provide TRL7, that impacts the decision how to get to TRL5.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 5:50 am
by Aero
GIThruster wrote:
93143 wrote:It can be ground-tested.
It can but for it to make TRL5, it needs to be tested across a broad range of environmental conditions. If that can be done someplace like Marshall, then there's an advantage to ground testing. If there is no preexisting thrust stand that can duplicate the expected environment for TRL5, then going straight to flying prototype is likely a better ROI than building such a test apparatus.

I would want to know how much modification would be needed to drop this into something like X-37 and be able to fly it. If there's any existing frame that can provide TRL7, that impacts the decision how to get to TRL5.
If there's any existing frame ... Well you've found the problem. I was hoping to find an existing airframe to use for Sabre engine testing but it doesn't look hopeful to me. It is purely a matter of size, performance and propellant tanks.

First, to answer your example of the X-37. The X-37 has a 4.5 m wingspan and the Sabre engine is 2-3 meters in diameter. The X-37 thrust is around 6600 lb-f, while the Sabre sea level thrust is 440,000 lb-f and much more in rocket mode. That is a mismatch. It is almost like you would drop the X-37 into the Sabre engine, not the other way around.

The Sabre engine is huge, it masses over 14 metric tonnes.

Sabre provides over 440,000 lb-f thrust so the test airframe must be strong enough to withstand that thrust force. For reference, the total thrust from all 4 engines on a Boeing 747 is 266,000 lb-f, that is, 66,500 lb-f per engine.

Sabre must be flight tested with its own nacelle because the nacelle controls the airflow to the heat exchanger and engine. Some might go so far as to say that the nacelle is part of the engine.

The test bed flight vehicle fuel tanks must be insulated to contain liquid hydrogen and LOX. That alone rules out all existing airframes that I know of.

The test bed flight vehicle must be capable of controlled supersonic and hypersonic flight. There are only a handful of supersonic candidates, no hypersonic candidates that I know of.

Bottom line is, build a Sabre engine test bed flight vehicle or test the Sabre engine on the ground. But, baring serious show stoppers being discovered during ground testing, the Sabre engine testing will need a test bed flight vehicle sooner or later.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 6:19 am
by 93143
Ah... you might want to cut all those numbers in half. Skylon uses two SABREs, and their combined thrust and weight are similar to the figures you provided.

Accurate numbers (for SABRE 2; the latest version is SABRE 4) can be had from the C1 trajectory spreadsheet.

Static sea level thrust: <140,000 lbs per engine (not much more than the current world record for an airbreathing engine)
Peak airbreathing thrust (net): <320,000 lbs per engine (less than an Airbus A380's maximum takeoff thrust)
Vacuum thrust: <340,000 lbs per engine (less than a Space Shuttle Main Engine's sea level thrust)

A detailed mass breakdown of Skylon C1 yields more accurate mass numbers:

Engine mass (less nacelle): 5436 kg
Nacelle mass: 1961 kg

See? Not that bad.

...

If you want to get really technical, a large chunk of the engine is duplicated in each nacelle. Each cycle feeds two chambers. Each precooler feeds two cycles (for four chambers per nacelle). The cutaway shows a single compressor, so I assume that each compressor feeds two cycles too... so the figures above are either accurate or double the true values, depending on how you interpret the situation.

Still too big for an X-37...


EDIT: I goofed. I've been peddling this set of numbers for a while, and I never noticed that the intake captured airflow in the spreadsheet is per nacelle. So the peak airbreathing thrust is actually about 230,000 lbs.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 2:13 pm
by ladajo
The other advantage for a flight testing regime is you get easier access to inlet pressures and densities given alititude.

I think it is not that big a leap to build a single engine test vehicle capable of a 15 minute flight cycle. Size is driven by duration.

I am also thinking that it would neccesitate a two field profile. One for launch, one for recovery. That also simplifies the vehicle.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 4:24 pm
by Aero
Ah... you might want to cut all those numbers in half. Skylon uses two SABREs, and their combined thrust and weight are similar to the figures you provided.

Accurate numbers (for SABRE 2; the latest version is SABRE 4) can be had from the C1 trajectory spreadsheet.
My information is from

http://en.wikipedia.org/wiki/SABRE_(rocket_engine)

Your reference title sounds better than Wikipedia but I don't find such a thing as a C1 trajectory spreadsheet. Would you give some guidance as to where to find that spread sheet beyond, it's on the REL web site? A link if possible.

Until then I really only have the Wikipedia numbers and your word that they've used the Skylon thrust numbers instead of the Sabre thrust numbers. Quite believable but unsupported.

Thanks.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 4:53 pm
by happyjack27
Aero wrote: Until then I really only have the Wikipedia numbers and your word that they've used the Skylon thrust numbers instead of the Sabre thrust numbers. Quite believable but unsupported.
Wikipedia has a strict policy that all information must be sourced. If you click on the raised number after the wikipedia in question, it will bring you to the "References" section, and highlight the source of the information.

Re: Reaction Engines

Posted: Mon Feb 03, 2014 11:47 pm
by GIThruster
Aero wrote:If there's any existing frame ... Well you've found the problem. I was hoping to find an existing airframe to use for Sabre engine testing but it doesn't look hopeful to me. It is purely a matter of size, performance and propellant tanks.
X-37 grew out of "Copper Canyon", NASP and X-30. X-30 is probably the best fit.

Specifications (X-30 as designed)

General characteristics
Length: 160.0 ft (48.8 m)
Wingspan: 74.0 ft (22.6 m)
Gross weight: 300,000 lb (136,078 kg)
Powerplant: 1 × Scramjet
Performance
Propellants: Air/Slush LH2

Seems to me if they're working this design currently they know they have something they can adapt. If the tanks will work, this is the right size. X-30 was cancelled because of the drive to put a 2 man crew on it instead of making it a demonstrator. Given a working SABRE, they might just resurrect this older work and update it with some of the X-37 and X-33 tech. The aluminum tank for X-33 still exists somewhere, though I think it was scaled down from what would work with SABRE. Since X-33 however, Musk developed the aluminum welding techniques needed to drive the cost down on such structures and I think NASA intends to use this for SLS. And Scaled could certainly build a composite tank. They could have built it for X-33 if asked. Who knows, we might see something like this fly for USAF.

Re: Reaction Engines

Posted: Tue Feb 04, 2014 12:46 am
by 93143
Aero wrote:Your reference title sounds better than Wikipedia but I don't find such a thing as a C1 trajectory spreadsheet. Would you give some guidance as to where to find that spread sheet beyond, it's on the REL web site? A link if possible.
C1 trajectory spreadsheet

presentation with C1 mass breakdown (pg. 33)
GIThruster wrote:X-37 grew out of "Copper Canyon", NASP and X-30. X-30 is probably the best fit.
Wait, are you trying to think of something the U.S. Air Force would do with it?

X-30 was never built; it would need substantial work to get it ready. Even if it had been, it was designed as a scramjet-powered waverider with loads of forebody compression, which doesn't really fit with SABRE. I don't think the U.S. has an appropriate airframe it could simply bolt a SABRE to.

REL are waffling on the question of whether to bother flight-testing the inlets before flying the first boilerplate Skylon prototype. I really don't think the rest of the engine is worth developing a whole airplane for, but if you must flight-test it, why not do it on a dumbed-down version of the final target airframe? In other words, from REL's perspective, use the Skylon Y-plane.

If the Air Force has a specific target in mind for the technology, they could get going on that. If they're just interested in the concept in general, there's probably no rush.

...

There was going to be a subscale demonstrator called SCEPTRE. Unfortunately it wouldn't have had a hydrogen turbopump, as doing one that small (about a third of a million RPM) would have cost a couple hundred million by itself. It was to use a highly pressurized tank. And the system wouldn't have been in a useful form factor for flight testing; Bond described it as a "dissected rabbit" with all the parts spread around. In any case, they seem to have gotten more investor interest than expected and have decided to go full scale right away.

Re: Reaction Engines

Posted: Tue Feb 04, 2014 1:17 am
by Skipjack
I think the SR71 might be a good fit.

Re: Reaction Engines

Posted: Tue Feb 04, 2014 1:31 am
by happyjack27
Have they done a full-engine computer simulation? Not sure if the technology is out there to make it economically viable, but it's certainly an option to consider.

Re: Reaction Engines

Posted: Tue Feb 04, 2014 3:47 am
by 93143
Skipjack wrote:I think the SR71 might be a good fit.
Too small. SABRE 2 is twice the mass and between four and nine times the thrust of the J-58. And the SR-71 can't go much past Mach 3.6-3.7 or the windshield comes loose (so I hear).
happyjack27 wrote:Have they done a full-engine computer simulation? Not sure if the technology is out there to make it economically viable, but it's certainly an option to consider.
Both REL and the von Karman Institute have detailed engine models, though the REL model is more up-to-date (or was at the time the ESA assessment report was written).

But I suspect you're talking about doing a full CFD simulation. That's probably beyond the state of the art right now... I mean, you could do it, sort of, but what with the turbulence modelling and the turbulent combustion modelling and the multiphase flow modelling and the expensive and/or inaccurate equations of state and the models and mixing rules for stuff like diffusion coefficients, you'd get at least a few percent and possibly double-digit uncertainties at each stage, and you know what the engineering rule of thumb is about compounding errors: they never subtract. Thermal and mechanical couplings with the engine structure shouldn't be too hard to get right in principle (unless something starts to fail), but their existence could compound inaccuracies in the fluid/thermo side. The one saving grace is that with hydrogen/air you probably won't get a lot of soot, so radiation should be pretty easy to get right, as well as being less important...

You'd have to calibrate the whole thing very carefully to get a quantitatively meaningful result, which means it would be more of a predictive tool than a testing tool.

Re: Reaction Engines

Posted: Tue Feb 04, 2014 4:28 am
by Aero
93143 wrote:
Aero wrote:Your reference title sounds better than Wikipedia but I don't find such a thing as a C1 trajectory spreadsheet. Would you give some guidance as to where to find that spread sheet beyond, it's on the REL web site? A link if possible.
C1 trajectory spreadsheet

presentation with C1 mass breakdown (pg. 33)

Thanks for the links.

Here is the support data for the Wikipedia numbers.
link: [url http://www.reactionengines.co.uk/sabre_howworks.html [/url]
For space access, the thrust during air-breathing ascent is variable but around 200 tonnes
per engine. During rocket ascent this rises to 300 tonnes but is then throttled down
towards the end of the ascent to limit the longitudinal acceleration to 3.0g
From the spread sheet it is hard to see how that could be right but the documents were both from late 2012 so that's not it. One thing is evident though and that is the gross thrust varies a great deal over the atmospheric flight regime and so does the net thrust. The result is that a wide range of thrust values can be justified for this engine but from the spread sheet and calculating the thrust from mass times acceleration given, (not using the gross thrust given) the net thrust per engine does not approach 200 tonnes even in the most favorable flight regime.

And I did find another source for the mass, that being just over seven tonnes per engine. So the Sabre engine is not so huge after all, just very, very large.

Finding an existing airframe that could take those engines supersonic with cryogenic prop tanks is still not likely. I vote for REL to continue on their development plan but eliminate the subscale nacelle step as it is like "a government build" approach but doesn't reduce risk for money spent much at all. If there is a problem with the nacelles, a new design can be retrofitted onto an existing airframe and probably at a much lower cost that the cost of the subscale nacelle program.