Reaction Engines Sabre/Scimitar

[mike-wsm]

The faqs (frequently asked questions) on Sabre/Scimitar include:

11. How does your frost control technology work?
Very well, thank you! (We aren't giving everything away!)

This appears to be a fudge. It is icing that is the biggest reason why Sabre/Scimitar will not in my considered opinion work and they simply dodge the issue by claiming to have a "secret" solution.

I don't believe they will get working heat exchangers and I don't believe they have any way of dealing with the icing problem.

Edit - The above opinion was based on the originally cited -150 deg C precooling and does not apply for 665 deg C precooling.

[Jane-DoH]

mike-wsm

Heat Exchanger, Scimitar Engine (Not-Classified)

Page 2 discusses the temperature changes encountered. From what it seems, the pre-cooler takes the temperature from 1,250 K down to 665 K (Truthfully those temperatures seem a touch high for Mach 5 flight as skin temperatures usually just come out to 600 C or 873.15 K), not down to -150 C/123.15 K regardless of mach number (I think the SABRE did operate this way though). This would probably eliminate the turbine-inlet temperature issue I was mentioning earlier (I don't know what the solution for the SABRE was, but it would explain why at around Mach 6, they switched over to rocket propulsion -- that fuel would be awfully hot)

[mike-wsm]

A demonstration precooler will be run in front of a Viper jet engine at REL’s B9 test facility in 2011. This precooler will incorporate full frost control and be built from full size SABRE engine modules. The facility will incorporate a high pressure helium loop that rejects the absorbed heat to a bath of liquid nitrogen.

So they didn't make one yet.

Originally Posted by Robyn

Paul Czysz

Paul Czysz Hypersonic Interview

[Jane-DoH]

mike-wsm

Paul Czysz Hypersonic Interview

I've been looking for this for a long time! Thanks

Out of curiosity, do you remember anything about the Japanese ATREX engine, which had a pre-cooler and worked on a hydrogen-expander cycle? As I understand it, the design was cancelled in favor of a pre-cooled turbojet similar to the Scimitar, but it's pre-cooler seemed a lot smaller, lighter and simpler than the version used by Reaction Engines.

[Jane-DoH]

mike-wsm

Yesterday, I stumbled upon another link which better illustrates both the SABRE and Scimitar Engines as well as the aircraft designs involved

Voila!

[mike-wsm]

development study

[Jane-DoH]

mike-wsm

Fascinating design, though puzzling if you ask me. Firstly the inlet that it has seems to have a tiny throat with what must be enormous duct pressures, I don't know why you'd have so much air go through the porous part of the duct into the pre-cooler (though I understand why you'd have the air from the inlet going through). The pre-cooler looks smaller than the SABRE and Scimitar's though (though this could be an optical illusion).

Why do you have the air going through the pre-cooler from the sides and in a spiral-trajectory rather than just straight through? I don't really get it.


BTW: In the link about SABRE/Skylon and Scimitar/LAPCAT A2 -- there's a mathematica formula listed: T = 220 (1 + 0.2M^2N)

I don't really know what this formula is or what units the 220 is in (I assume Kelvin though I'm not sure); I don't really understand what the function of the N is. It looks almost like an exponent, but it's a subscript instead of a superscript. I don't know that math function.

[mike-wsm]

Jane-DoH

I'll try to find some answers. May take a little time, I need to download and print the references so I can study them more fully, there is a limit to what my tiny human brain can do on-screen.

You do raise some interesting questions!

[Machinbird]

Jane-Doh
First, I'd like to express my appreciation for making us old guys (at least this one) catch up on the more recent developments on the cutting edge of technology.

The formula you posted is not quite correct, but that is caused by the problems with cutting and pasting mathematical formulas with limited editing tools. The original formula read: T= 220(1+ 0.2*Machnumber squared)

This is the stagnation temperature formula expressed in terms of Mach number (Mn). In this case, the 220 is the assumed initial (Freestream) temperature Kelvin. Check out the Wiki on stagnation temp here:
Stagnation temperature - Wikipedia, the free encyclopedia
Using the formula for stagnation temperature T0, they have assumed a starting free stream temperature, and simplified into terms of Mach number. (Good enough for back of the envelope work until you start to get disassociation of the working fluid)

[Jane-DoH]

Machinbird

First, I'd like to express my appreciation for making us old guys (at least this one) catch up on the more recent developments on the cutting edge of technology.

Glad to be of help.

The original formula read: T= 220(1+ 0.2*Machnumber squared)

Okay the equation would go like this
1.) T0 = 220 (1 + 0.2(M)^2)
2.) T0 = 220 (1 + 0.2(5)^2)
3.) T0 = 220 (1 + 0.2(25))
4.) T0 = 220 (1 + 5)
5.) T0 = 220 (1 + 5)
6.) T0 = 220 (6)
7.) T0 = 1320 K

In this case, the 220 is the assumed initial (Freestream) temperature Kelvin.

Which I'm guessing is the outside air temperature they're going to be flying Mach 5 at?

[mike-wsm]

Jane-DoH

This is close enough to the figures you quoted above 1250K to 665K and removes any niggles from me about icing.

[Jane-DoH]

mike-wsm

This is close enough to the figures you quoted above 1250K to 665K and removes any niggles from me about icing.

Yeah, it's hot enough...

BTW: Dunno about the SABRE or Scimitar, but the ATREX, as I understand it to combat frost were using some kind of methanol injection system

[mike-wsm]

Jane-DoH

Methanol

Thin-walled high operating temperature heat exchanger and now methanol injection. This is one incredibly complex engine. What is the reliability over how many re-usage cycles?

Why do you have the air going through the pre-cooler from the sides and in a spiral-trajectory rather than just straight through? I don't really get it.

I've had a chance to examine the references you gave and the spiral construction of the pre-cooler appears to be an attempt to make it behave like a counterflow heat exchanger. A conventional counterflow heat exchanger is continuous in that one fluid is gradually heated along its length as the other fluid travelling in the opposite direction is gradually cooled, At all points along the heat exchanger the temperature difference between fluids is the same, yielding the best performance. In the Sabre pre-cooler, the flow temperatures are stepped but tend towards the counterflow ideal. See spiralflow/crossflow in the wiki link above.

The formula does look ambiguous, but if you look carefully is Mn^2.

Check out the admiring video on jetblast/hamsterwheel, sung by heart-throb Christofer Drew / Never Shout Never.

[Jane-DoH]

mike-wsm

Thin-walled high operating temperature heat exchanger and now methanol injection. This is one incredibly complex engine. What is the reliability over how many re-usage cycles?

Well, the methanol injection was ATREX, not SABRE or Scimitar. Regardless, yes, it is a very complicated engine.

I've had a chance to examine the references you gave and the spiral construction of the pre-cooler appears to be an attempt to make it behave like a counterflow heat exchanger. A conventional counterflow heat exchanger is continuous in that one fluid is gradually heated along its length as the other fluid travelling in the opposite direction is gradually cooled, At all points along the heat exchanger the temperature difference between fluids is the same, yielding the best performance.

So, the air goes one way, and the coolant goes the opposite way?

[mike-wsm]

So, the air goes one way, and the coolant goes the opposite way?

Yes, in the diagram above the air is traveling radially inward and the coolant is traveling outward.

[Jane-DoH]

I don't know exactly how this relates to this but I remember reading something Paul Czysz mentioned about a pre-cooling system which entailed cooling the airflow into the engine above Mach 1.8 to avoid an excessive rise in temperature and optimize the benefits of ram compression or something and simultaneously increasing the engine RPM while doing this and likened it to Skylon. Maybe I got the exact explanation wrong, but do you know what I'm talking about?

[mike-wsm]

Yes, something like that. It's all down to physics, if you want to go that fast and do it by burning fuel with air you have to do what is necessary.

Happily nobody has found a way of using air as both fuel and oxidant, by burning the nitrogen and oxygen, that would leave a real nasty NOx trail. Oops, why did I say that? Now they'll start doing it . . .

[Jane-DoH]

mike-wsm

Yes, something like that.

How exactly does that work, increase density, then increase the RPM due to the density change, or to allow it to pull in more air, or something else?

[Jane-DoH]

Why does the LAPCAT A2 fly at such a low altitude? While 85,000 feet is high up, for Mach 5 it seems relatively low...

[MG23]

Originally Posted by balsa model

I don't have to get the cost where you want it, to win this argument.

The only known mass market for spaceflight requires launch costs below $100 a kilo. At that point tourism starts to become viable enough to sustain a huge increase in payload to orbit, and your flight rate can be high enough to make an SSTO more profitable than an expendable.

Saying 'I can prove SSTOs don't make sense so long as I define the question so that SSTOs don't make sense' seems rather silly. In the medium term they're the only things that make sense if your goal is to build a sizable human population in space (longer than that and something like a space elevator may be viable and cheaper).

The big problem with launcher development right now is that there's no proven large market between communication satellites and tourists. If you're launching a billion-dollar comsat which will produce billions in revenue over its lifetime, then the difference between paying $200 million or $20 million to launch it is small, whereas if you're in the launch business, the difference between charging $200 million or $20 million a launch is enormous if it won't result in any significant increase in the number of launches.

So there is little economic incentive for developing new technologies to dramatically reduce launch costs until you can get them down to levels where ordinary mortals can afford them. Skylon, if I remember correctly, was estimated to cost around $500 per kilo back in the 90s, which puts it squarely into the 'who do you think your customer is?' range. Put a $10,000,000,000 price tag on developing a launcher with no known profitable use, and no-one in their right mind is going to fund it... particularly when you consider that the development cost is almost certainly optimistic.

Which is presumably why RE are now pushing for work on a hypersonic airliner instead of an SSTO. The problem is that there's not much more of a proven market for hypersonic airliners than for spaceflight at $500 a kilo.

I just have to point out that all reusable options have shown to be more expensive than using the throw-away variants. By a lot.

The only operational 'reusable' launcher is the shuttle, which is barely reusable and has to carry a human crew with all the extra cost that entails yet still costs less per flight than a comparable expendable launcher. The problem with the shuttle is the huge fixed costs of the facilities and standing army required to support it; the variable cost of flying an extra shuttle launch is around $200,000,000 while the average cost over the entire program is closer to ten times that amount because the fixed costs have to be spread over a small number of launches.

That is why both reusability and high flight rate are important if you want to dramatically cut costs. So, for that matter, is not running your own 'spaceport' that adds a couple of billion dollars a year to your fixed costs when you could just buy launch slots elsewhere when you need them.

Our propulsion technology hasn't seen any revolutionary invention since the 60's. That's our limit.

No, that's as far as we've had any incentive to go. When the only real markets for your services are happy to pay $200,000,000+ a launch with a 98% success rate, there's simply no incentive to do any better.

Absolutely! If it cost less to produce a new expandable one than to recover and overhaul a reusable one that carries the same payload, then Yes!

And even the shuttle, which is at best refurbishable and throws away around $100,000,000 in hardware each flight, costs less to overhaul than the cost of building an equivalent expendable launcher.

Should I even need to add that SpaceX are aiming to recover and reuse both stages of their Falcon launchers because they believe it will reduce costs compared building new ones each time?