The Reaction Engines Skylon Master Thread (1)

[Citizen Wolf]

One thing I'm curious about (one thing amongst many) is the potential problem of particulates clogging the fine mesh of the heat exchangers. Do airborne particulates pose a different challenge than the icing problem? We saw all the airlines being grounded over Europe when the Icelandic volcanoes erupted a few years ago. Does the heat exchanger on the sabre engine make it more sensitive to such challenges than a conventional jet engine?

[RobLynn]

One thing I'm curious about (one thing amongst many) is the potential problem of particulates clogging the fine mesh of the heat exchangers. Do airborne particulates pose a different challenge than the icing problem? We saw all the airlines being grounded over Europe when the Icelandic volcanoes erupted a few years ago. Does the heat exchanger on the sabre engine make it more sensitive to such challenges than a conventional jet engine?

This will not be a problem.  Firstly there is very little particulate matter in air unless flying through a dust storm, which no sane operator would ever do. Secondly the heat exchangers are just not that fine - approximately 1mm diameter tubes and gaps that I would guess are >0.3mm.  They wouldn't get clogged by airborne dust but could be eroded if there was enough fine particulate matter over a long time (unlikely given ~30 hour operational life of engine).

Dusty air also erodes and destroys conventional engines - helicopters operating is dusty places such as doing repeated landings in windy deserts can destroy their engines in a couple of hours if they are not fitted with inlet air filters.

Flying through ash plumes is an extremely rare occurrence - and something that a space plane operator could always avoid.  But interestingly it would be less of a problem for Sabre than for normal Gas Turbines because the air is not hot enough to melt the ash until it gets to the combustion chamber - where it would get blown out the nozzle and wouldn't be a problem, and there are no hot turbine or stator blades blades for it to stick to or block up the cooling holes of.

[john smith 19]

?? if you agree we can predict the thrust of course we can predict the Isp. One determines the other. We have had no problems with Isp modelling in the past.

That's an interesting question. It's OT but I've always wondered about the in house computer models used. Sutton's too simplistic (and at least in one case his loss levels have been known to be wrong, at least for O2/H2 engines, since the late 60's) and I know the CPIA has gradually extended what they think is important but I wondered if in house models just used carefully selected fudge factors based on experience or is the first principles understanding good enough that results will not be "optimistic"?

[john smith 19]

I have not found any reference to these kinds of problems/solutions on the REL website. This could be a show-stopper for the reusability of the SABRE engine. Mark Hempsell, if you're reading this, has REL a plan for dealing with this problem?

check this page.
http://www.reactionengines.co.uk/sabre_howworks.html
click on the 3rd picture.

It's complicated because you're looking at the hardware for 2 different engine modes superimposed on each other. The bit you're worried about is in the bottom left hand corner labelled "pre-burner" like the bit in the SSME whose output drives the main pump turbines.

SABRE does not drive any turbine directly with H2 rich superheated steam. All are driven either by Helium or what they are pumping (at different physical conditions to the pumped fluid).

This has 2 effects. a) Parts exposed to the H2 rich super heated steam don't experience the multiplying effect of high mechanical stress. Everything happens through heat exchangers (which is why it has so many of them). Keep in mind these are REL's low pressure loss/high heat transfer heat exchangers so the difference between the pre-burner pressure and the main thrust chamber can be less that in the SSME.
b)There are no "Criticality 1" seals like the ones on the SSME between the preburner driving the HP LOX pump and the LOX flow. Poor seal design here required He purge gas to be 4x it's design flow, partly contributing to the 270lb GHe tank on each engine.

Eliminating such seals, de-coupling fluid heating from turbine drive and using fluid bearings should eliminate most of the major reasons the SSME was taken apart after every flight (although that got better as time went on NASA could have installed many more improvements).
Designing in IVHM sensors to both the engine and the structure (with engine data collection as part of the primary test goals from day 1 on the test stand to drive the health models)

"SSME The first 10 years" teaches a lot of lessons if read carefully.

[Citizen Wolf]

@RobLynn
Thanks, interesting points.

[grondilu]

An other public question about RE asked to Elon Musk:



at about t=49'

« With respect to air-breathing, hybrid stages, I have not seen how the physics of that makes sense.  There may be some assumptions that I have that are incorrect, but really  for an orbital rocket you try to get out of the atmosphere as soon as possible, because the atmosphere is just as thick as soup, when you try to go fast.  And it's not helped by the fact that the atmosphere is mostly not oxygen. »

[QuantumG]

Full transcript here: http://shitelonsays.com/transcript/elon-musk-the-future-of-energy-and-transport-2012-11-14#quote_-746403779

[grondilu]

Someone has to be wrong.  I'm no rocket engineer, but I'll try a very basic calculation.

Skylon can go to Mach 5 while breathing air:

Mach 5 = 340 m/s * 5 = 1700 m/s.

Let's say the target speed for orbiting is 10 km/s (I have no idea of the exact value but it's probably about this speed).

That's a speed ratio of 10e3/1700 = 100/17 ~ 5.9

In terms of energy, that's a ratio of 5.9**2 = 34.  So, the kinetic energy in the air-breathing phase is about 34 times smaller than the energy for orbiting.   (Well, I guess it's a bit more complex since the higher the aircraft, the lighter since it has dropped lots of its propellant.  But I try to keep it simple here first)

Now, kinetic energy is proportional to mass, oxygen is height times heavier than hydrogen (I mean, in the reaction producing water), and above in this thread someone told me that H2-O2 engines do not operate at stoichiometric ratios, but rather at 1:6.

So I may be a big Skylon fan, but I have some doubts now.  I very much wonder if the gain in oxygen weight is worth air-breathing.  But there is also the fact that Skylon has wings and that they are very usefull for reentry and landing.  They save a lot of the energy that would be necessary for breaking with a reusable "cylindrical" rocket.

So, what is the main point of air-breathing?  Is it about saving weight or is it about horizontal take-off and landing?

[MikeAtkinson]

Someone has to be wrong.  I'm no rocket engineer, but I'll try a very basic calculation.

You need to use the Rocket Equation (http://en.wikipedia.org/wiki/Tsiolkovsky_rocket_equation)

Then consider Skylon as a 2 stage rocket, the first stage being airbreathing has high Isp and high drag losses. The second stage being from where it goes to pure rocket mode.

[QuantumG]

Skylon can go to Mach 5 while breathing air:

Mach 5 = 340 m/s * 5 = 1700 m/s.

Let's say the target speed for orbiting is 10 km/s (I have no idea of the exact value but it's probably about this speed).

Actually, it's 9 to 10km/s if you get out of the atmosphere quickly. If you try to fly through the atmosphere you have even more drag and gravity losses.

That's a speed ratio of 10e3/1700 = 100/17 ~ 5.9

In terms of energy, that's a ratio of 5.9**2 = 34.  So, the kinetic energy in the air-breathing phase is about 34 times smaller than the energy for orbiting.

Yes, exactly. The kinetic energy to achieve 1e4 m/s is 1e8 J. The kinetic energy to achieve 1.7e3 is 2.89e6 J. That's a difference of 9.711e7 J. In other words, 97.11% of the required energy to get to orbit happens after Mach 5.

[MikeAtkinson]

That's a speed ratio of 10e3/1700 = 100/17 ~ 5.9

In terms of energy, that's a ratio of 5.9**2 = 34.  So, the kinetic energy in the air-breathing phase is about 34 times smaller than the energy for orbiting.

Yes, exactly. The kinetic energy to achieve 1e4 m/s is 1e8 J. The kinetic energy to achieve 1.7e3 is 2.89e6 J. That's a difference of 9.711e7 J. In other words, 97.11% of the required energy to get to orbit happens after Mach 5.

Wrong! Kinetic energy is 1/2 m v^2 

The mass at the end of the air breathing stage is not the mass when Skylon reaches orbit.

[grondilu]

That's a speed ratio of 10e3/1700 = 100/17 ~ 5.9

In terms of energy, that's a ratio of 5.9**2 = 34.  So, the kinetic energy in the air-breathing phase is about 34 times smaller than the energy for orbiting.

Yes, exactly. The kinetic energy to achieve 1e4 m/s is 1e8 J. The kinetic energy to achieve 1.7e3 is 2.89e6 J. That's a difference of 9.711e7 J. In other words, 97.11% of the required energy to get to orbit happens after Mach 5.

Wrong! Kinetic energy is 1/2 m v^2 

The mass at the end of the air breathing stage is not the mass when Skylon reaches orbit.

Indeed.  I guess there is no way to avoid the rocket equation.

[QuantumG]

Wrong! Kinetic energy is 1/2 m v^2 

The mass at the end of the air breathing stage is not the mass when Skylon reaches orbit.

Yes, good point. Thanks.

[93143]

Let's say the target speed for orbiting is 10 km/s (I have no idea of the exact value but it's probably about this speed).

...so you're willing to articulate such a serious doubt on the strength of a number you pulled out of thin air?

Orbital speed (relative to the equator) is about 7.5 km/s, not 10 km/s.  Now, you generally need more than 9 km/s to get to orbit, even for vertical-takeoff rockets that don't spend much time in the atmosphere, but a lot of that is losses - and the bulk of the losses happen during the early part of the ascent, when the vehicle is fighting gravity and drag and suboptimal nozzle expansion.

Skylon takes the vehicle up to 25 km and Mach 5.5 using airbreathing engines and wings, which sucks up the bulk of the losses and leaves a little over 6 km/s to be handled by the rocket mode.

The kinetic energy to achieve 1e4 m/s is 1e8 J. The kinetic energy to achieve 1.7e3 is 2.89e6 J. That's a difference of 9.711e7 J. In other words, 97.11% of the required energy to get to orbit happens after Mach 5.

No.  Partly because:

The mass at the end of the air breathing stage is not the mass when Skylon reaches orbit.

Skylon C1 is 232 tonnes at the switchover and less than 57 tonnes at MECO.  That's eight and a half times the energy, not thirty-four and a half.  Recall that most of the losses happen during the airbreathing phase...

But that's not all.  If you insist on using kinetic energy with respect to the ground, the efficiency of the engines increases dramatically as the vehicle accelerates, because more of the kinetic energy ends up in the vehicle and less in the exhaust.  In fact, once the vehicle exceeds half its exhaust velocity (rocket mode), it gets more energy from the propellant than the propellant had in it (after inefficiencies of course), because the kinetic energy of the propellant after expulsion is lower than it was in the tanks.

So no, you can't do this with 1/2mv², even by fiddling around with masses.  You need the rocket equation.

...

The rocket equation is logarithmic.  If the required delta-V is significantly larger than the exhaust velocity, a small savings in delta-V produces a large increase in payload.  In this case, the high Isp of the airbreathing mode means the first 5.5 km/s or so is almost free (mass ratio of less than 1.2), and only around 6.3 km/s is left for the rocket phase.  The rocket mode has a predicted Isp of ~4500 N·s/kg, or 459 seconds - this leads to a mass ratio of about 4.  The ascent mass ratio for Skylon C1, from start of takeoff roll to MECO, is 4.85, which compares with roughly 9 for a hydrogen-fueled all-rocket SSTO.

Another bonus, which follows from the same math, is that Skylon as a design is much less sensitive to engine underperformance than an all-rocket SSTO concept.  It doesn't drive the logarithm nearly as far into the region of diminishing returns, so the Isp of either mode can fall by a significant amount without eliminating the payload.

...

Now, if you're comparing Skylon to a TSTO, that's a bit of a different story.  But it can be argued that there are significant operational advantages to the SSTO approach, if you can pull off the technical end.  So it's not really apples to apples.

Besides, Skylon's wet payload fraction beats the Delta IV (an expendable TSTO using the same fuel) by roughly 50%, and even its dry payload fraction is about a match for the Delta IV.  Plainly airbreathing can be worth it, if you can get the performance without taking a huge maintenance penalty or something...

[tlesinsk]

Let's say we need 9.5 km/s total delta-v, nad that our LOX/LH2 engine has a 4.5 km/s exhaust velocity. With air-breathing propulsion to Mach 5 (burning very light hydrogen) the LOX/LH2 delta-v is reduced to 7.8 km/s. the corresponding engine+structure+payload mass fraction is 17.7%. Starting LOX/LH2 propulsion from the ground gives you only 12.1% E+S+P.

REL seems to think this makes the difference between closing a design or not, even given the low-density propellant (accomodated in light tanks suspended in a lightweight spaceframe structure that reduces the ballistic coefficient and reduces needed thermal protection), wings, and increased engine mass (which doesn't increase that much due to lower needed thrust/weight ratio for horizontal take-off).

It's complicated, but it all computes. Hell, read the thread, it's been discussed ad nauseam already

[93143]

Let's say we need 9.5 km/s total delta-v, nad that our LOX/LH2 engine has a 4.5 km/s exhaust velocity. With air-breathing propulsion to Mach 5 (burning very light hydrogen) the LOX/LH2 delta-v is reduced to 7.8 km/s.

Nope.  Read my post above.  You've lumped all the losses into the rocket phase, when they mostly belong in the airbreathing phase.

[QuantumG]

Let's say we need 9.5 km/s total delta-v, nad that our LOX/LH2 engine has a 4.5 km/s exhaust velocity. With air-breathing propulsion to Mach 5 (burning very light hydrogen) the LOX/LH2 delta-v is reduced to 7.8 km/s. the corresponding engine+structure+payload mass fraction is 17.7%. Starting LOX/LH2 propulsion from the ground gives you only 12.1% E+S+P.

Or you could just have two stages...

[john smith 19]

So, what is the main point of air-breathing?  Is it about saving weight or is it about horizontal take-off and landing?

Depends who you ask. Air breathing does not have to be horizontal. In Skylon's specific case it is to build an SSTO vehicle with the payload fraction of a good ELV.

Elon Musk has stated that Spacex's goal is an ELV with with a payload fraction 4% of GTOW weight which will allow them an RLV with 2% of it's GTOW, using 2 stages. Skylons goal is SSTO with an ELV payload fraction. Most historical SSTO concepts have been VTOL and their supports have accepted they would deliver 1% of GTOW, so either 4 flights or an SSTO is 4x bigger than an equivalent ELV.

IOW Skylon gives customers what they want (full size ELV payload fraction) and funding suppliers what they want (a vehicle the same size as an ELV and therefor roughly the same cost, when run through their cost models).

[jded]

Wasn't reusable Falcon supposed to stage at Mach 6?

If airbreathing to Mach 5,5 wasn't worth it, would it be worth to have whole first stage that goes only to Mach 6?

[tlesinsk]

Let's say we need 9.5 km/s total delta-v, nad that our LOX/LH2 engine has a 4.5 km/s exhaust velocity. With air-breathing propulsion to Mach 5 (burning very light hydrogen) the LOX/LH2 delta-v is reduced to 7.8 km/s.

Nope.  Read my post above.  You've lumped all the losses into the rocket phase, when they mostly belong in the airbreathing phase.

Indeed ! So, the delta-v for the LOX/LH2 phase is even lower...

Or you could just have two stages...

...which significantly increases both development and operation costs. Why use staging when you have free oxidizer ?