The Reaction Engines Skylon Master Thread (1)

[FinalFrontier]

OK, semi-seriously, has anyone heard of any interest in Skylon from Richard Branson/Virgin Galactic? For a few hundred million, I bet Bond and company might be willing to rename Skylon 'SpaceShipFour'...

And how would such an investment work, assuming that at a later date the project is likely to be lead by a consortium of industry giants?

I could reasonably see a firm like Boeing, Spacex, or VG making the investment. I think it's simply a matter of time. I think also, that after REL builds a full up SABRE engine and demonstrates it can work as advertised it will be far more likely, the pre-cooler is without a doubt a breakthrough but people still want to see the full size engine work.

[hkultala]

I haven't been keeping up with the mass and volume details of Skylon, but a point I've made off-and-on for the past few years is that the propellant mass fraction of the vehicle is highly ambitious, and I say this as a long-time advocate of SSTO.  Using the data above, I note that if we replaced the 1100m³ of propellant volume with low-average density LOX-LH2, we're looking at a PMF of 0.94. 

.. And here is where you make your critical mistake.
Skylon has much more hydgogen (which is lighter than oxygen) and much less oxygen, so the total weight of the fuel + propellant is much smaller than in your calculations.

And this affects all the later conclusions that you make, making them false.

[HMXHMX]

I haven't been keeping up with the mass and volume details of Skylon, but a point I've made off-and-on for the past few years is that the propellant mass fraction of the vehicle is highly ambitious, and I say this as a long-time advocate of SSTO.  Using the data above, I note that if we replaced the 1100m³ of propellant volume with low-average density LOX-LH2, we're looking at a PMF of 0.94. 

.. And here is where you make your critical mistake.
Skylon has much more hydgogen (which is lighter than oxygen) and much less oxygen, so the total weight of the fuel + propellant is much smaller than in your calculations.

And this affects all the later conclusions that you make, making them false.

No, no mistake.  I am not showing the PMF of Skylon at the air-breather mix ratio, but at the pure rocket LOX-LH2 mix ratio, nominally 6:1.  To a first order, assuming equivalent ullage pressures, the mass of a launch vehicle tank is related only to its volume, and not the density of propellant contained therein.  Mine is a perfectly valid means by which to compare the relative difficulty of building Skylon's tanks, say compared with the ET of the Space Shuttle.  And the message is those tanks will be very hard to build, indeed.  And even harder to make reusable.  Not impossible: but if they can be built at all, I can show fairly easily that a pure rocket vehicle with the same PMF assumptions, deleting the low t/w air-breather engines and substituting conventional pure rocket engines, will place a 2x or greater payload into LEO for the same physical vehicle size (volume).

I've only been doing these calculations for the past 43 years, so I think I may have it right... 

Just keep saying to yourself, the metric by which one measures the feasibility of a vehicle's PMF is based on tank volume, irrespective of propellant type.  I'm strongly tempted to add that I have a marvelous proof of this thesis, but neither the time nor room to write it down...

[HMXHMX]

Reaction Engine's Skylon technical page lists an unladen mass of 41 mT, and a fuel load of 220 mT, with 12 mT payload adding up to basically the cited 273 mT maximum takeoff mass. That's more like 81% fuel at takeoff if you count the payload in with the dry mass, or 85% if you don't.

Note that I said "...using the data above" to perform the calculation.  Redoing it for the data you present from the Skylon technical page results in a pure rocket PMF of about 0.909, almost exactly what the S-IVB stage delivered as an expendable. (Note that you have to go dig up the real empty shipped weight of the S-IVB stage, since too many web references add in the mass of the Saturn Instrument Unit, which badly skews the PMF...)

It's still a hard structure to build, as a reusable.  And I maintain it still would be superior to go pure rocket, rather than rely on air-breather technology.

[Andrew_W]

I blogged a year ago:

Sunday, June 12, 2011
I'm a bit skeptical about Skylon.
After running through the mass numbers for Skylon that Reaction Engines offers, I'm left believing that they've got an engine that will theoretically get them to orbit in one stage, but only by them making unrealistic assumptions about the structural weight of the rest of the vehicle, total dry mass 53 tonnes, engine thrust 270 tonnes, engine T/W 14 therefore mass of engines ~19 tonnes, therefore mass of the rest of the vehicle ~34 tonnes. 34 tonnes for a winged vehicle that's 83 meters long, carries all its landing gear to orbit, it has a propellant volume around 1400^3 meters, it uses cryogenic propellants and it has to endure re-entry. I know they're promoting Skylon as having revolutionary construction materials and methods, but it seems to me they've had to make some excessively optimistic assumptions about the structural weight to get the numbers to come together  so they can continue with their pet project - the engines.

Looking at it another way: The combined propellant tank volume by my math (with a few assumptions on the current LH2:LOX ratio) would have enough volume to hold 500 tonnes of LH2/LOX at a 1:6 ratio, lets allow structural weight growth of 20% for the heavier take-off weight making structural wt 40.8 tonnes, 2 SSME's (or easily maintained equivalent) is + 6.4 tonnes, so total unladen weight is 47.2 tonnes, add a P/L of 15 tonnes and also the 500 tonnes LOX/LH2 and you get a take-off weight of 562.2 tonnes, at engine shut off weight is 62.2. Mo/M1 is 9.03, delta V at Ve 4500 m/s is 9907m/s.

9.2 km/s is about all you need to get into orbit.

[MP99]

I'm strongly tempted to add that I have a marvelous proof of this thesis, but neither the time nor room to write it down...

You'd write it in if only you could find a way to put it in the margin of the page? 

cheers, Martin

[john smith 19]

No, no mistake.  I am not showing the PMF of Skylon at the air-breather mix ratio, but at the pure rocket LOX-LH2 mix ratio, nominally 6:1. 

 Mine is a perfectly valid means by which to compare the relative difficulty of building Skylon's tanks, say compared with the ET of the Space Shuttle.  And the message is those tanks will be very hard to build, indeed.  And even harder to make reusable.  Not impossible: but if they can be built at all, I can show fairly easily that a pure rocket vehicle with the same PMF assumptions, deleting the low t/w air-breather engines and substituting conventional pure rocket engines, will place a 2x or greater payload into LEO for the same physical vehicle size (volume).

I've only been doing these calculations for the past 43 years, so I think I may have it right... 

Just keep saying to yourself, the metric by which one measures the feasibility of a vehicle's PMF is based on tank volume, irrespective of propellant type.  I'm strongly tempted to add that I have a marvelous proof of this thesis, but neither the time nor room to write it down...

There are a few things yo may not have factored in.

REL seem to have started to quote propellant mass as a single figure. IIRC the split is 70/150 but I can't recall if that's O2/H2 or vice versa. I'll note SABRE is designed specifically to reduce LO2 consumption and has substantial excess LH2, which it dumps to the "spill ramjet," however in the film Alan Bond states this amount was halved compared to the original RR design and for the SABRE 4 iteration has been halved again, which required a re-think on how the CG shift was handled in the D1 re-design as not enough LH2 was being used to counteract the CP shift (this issue has been resolved according to the IAF paper).

REL have stated they plan to use sub cooled propellants in unvented tanks. I recall 18k rather than the usual 20k for LH2 (although NASA seem to have looked at 16K) and I think 82k for LO2. This would definitely reduce tank size over NBP designs

They have also stated they have an interest in low ullage pressure tanks to lower the amount of either He or vaporized propellants needed to pressurize them.

The shift in propellant mass ratios alone would seem to make a big difference.

[HMXHMX]

No, no mistake.  I am not showing the PMF of Skylon at the air-breather mix ratio, but at the pure rocket LOX-LH2 mix ratio, nominally 6:1. 

 Mine is a perfectly valid means by which to compare the relative difficulty of building Skylon's tanks, say compared with the ET of the Space Shuttle.  And the message is those tanks will be very hard to build, indeed.  And even harder to make reusable.  Not impossible: but if they can be built at all, I can show fairly easily that a pure rocket vehicle with the same PMF assumptions, deleting the low t/w air-breather engines and substituting conventional pure rocket engines, will place a 2x or greater payload into LEO for the same physical vehicle size (volume).

I've only been doing these calculations for the past 43 years, so I think I may have it right... 

Just keep saying to yourself, the metric by which one measures the feasibility of a vehicle's PMF is based on tank volume, irrespective of propellant type.  I'm strongly tempted to add that I have a marvelous proof of this thesis, but neither the time nor room to write it down...

There are a few things yo may not have factored in.

REL seem to have started to quote propellant mass as a single figure. IIRC the split is 70/150 but I can't recall if that's O2/H2 or vice versa. I'll note SABRE is designed specifically to reduce LO2 consumption and has substantial excess LH2, which it dumps to the "spill ramjet," however in the film Alan Bond states this amount was halved compared to the original RR design and for the SABRE 4 iteration has been halved again, which required a re-think on how the CG shift was handled in the D1 re-design as not enough LH2 was being used to counteract the CP shift (this issue has been resolved according to the IAF paper).

REL have stated they plan to use sub cooled propellants in unvented tanks. I recall 18k rather than the usual 20k for LH2 (although NASA seem to have looked at 16K) and I think 82k for LO2. This would definitely reduce tank size over NBP designs

They have also stated they have an interest in low ullage pressure tanks to lower the amount of either He or vaporized propellants needed to pressurize them.

The shift in propellant mass ratios alone would seem to make a big difference.

Volume is volume.  And if Skylon gets to use sub-cooled propellant and low ullage pressures, the same applies to the pure rocket so the relative output of the calculation remains unchanged. I have variously used both tricks over the years in different design concepts, and much favor boost pumps for lower ullage pressure over sub-cooling, which increases propellant price and infrastructure expense.

Shifting to more LH2, stored at NBP or sub-cooled, only makes the comparison more favorable to the rocket, since it has an density five times lower than the average propellant density of LOX-LH2 at nominal mix ratios.

[HMXHMX]

I'm strongly tempted to add that I have a marvelous proof of this thesis, but neither the time nor room to write it down...

You'd write it in if only you could find a way to put it in the margin of the page? 

cheers, Martin

Aye.  Nice talking to perceptive people; that's NASAspaceflight for you!

[mmeijeri]

And I maintain it still would be superior to go pure rocket, rather than rely on air-breather technology.

Do you feel the same way about the utility of airbreathing on a reusable first stage? And what if you only look at it from a "deliberately sunk" cost perspective?

[HMXHMX]

I blogged a year ago:

Sunday, June 12, 2011
I'm a bit skeptical about Skylon.
After running through the mass numbers for Skylon that Reaction Engines offers, I'm left believing that they've got an engine that will theoretically get them to orbit in one stage, but only by them making unrealistic assumptions about the structural weight of the rest of the vehicle, total dry mass 53 tonnes, engine thrust 270 tonnes, engine T/W 14 therefore mass of engines ~19 tonnes, therefore mass of the rest of the vehicle ~34 tonnes. 34 tonnes for a winged vehicle that's 83 meters long, carries all its landing gear to orbit, it has a propellant volume around 1400^3 meters, it uses cryogenic propellants and it has to endure re-entry. I know they're promoting Skylon as having revolutionary construction materials and methods, but it seems to me they've had to make some excessively optimistic assumptions about the structural weight to get the numbers to come together  so they can continue with their pet project - the engines.

Looking at it another way: The combined propellant tank volume by my math (with a few assumptions on the current LH2:LOX ratio) would have enough volume to hold 500 tonnes of LH2/LOX at a 1:6 ratio, lets allow structural weight growth of 20% for the heavier take-off weight making structural wt 40.8 tonnes, 2 SSME's (or easily maintained equivalent) is + 6.4 tonnes, so total unladen weight is 47.2 tonnes, add a P/L of 15 tonnes and also the 500 tonnes LOX/LH2 and you get a take-off weight of 562.2 tonnes, at engine shut off weight is 62.2. Mo/M1 is 9.03, delta V at Ve 4500 m/s is 9907m/s.

9.2 km/s is about all you need to get into orbit.

And if you use denser propellants than LOX-LH2, you can do even better.  Making a few very BOE assumptions I conclude that – based on volume alone without adjusting engine mass – a VTHL LOX-kerosene equivalent vehicle would carry about 28 tons to LEO (same physical size, using 9 NK33-equivalent engines).  That slightly more than double what their very best payload estimate is for Skylon (12 tons).  With no propulsion development required...

[HMXHMX]

And I maintain it still would be superior to go pure rocket, rather than rely on air-breather technology.

Do you feel the same way about the utility of airbreathing on a reusable first stage? And what if you only look at it from a "deliberately sunk" cost perspective?

I see very little advantage in using high Isp, low t/w air-breathing powerplants which only operate in the early part of a flight where Isp is largely unimportant and engine t/w is very important, in place of high t/w rockets which are required anyway for the vast bulk of the delta-V required to get to orbit.  Even if the air-breathing powerplants are "free" along with the oxygen (because they are not).
 
Using a carrier a/c to get to launch altitude, or towing to get there – those can be shown to make some sense. For example, they solve the launch noise problem.  Even a RASV-like HTHL can be shown to work.  But my bottom line is don't chase Isp, and don't send an air-breather to perform an accelerator mission; give it to a rocket and be done with it.

[john smith 19]

And if you use denser propellants than LOX-LH2, you can do even better.  Making a few very BOE assumptions I conclude that – based on volume alone without adjusting engine mass – a VTHL LOX-kerosene equivalent vehicle would carry about 28 tons to LEO

Except it would not *be* an equivalent vehicle.

A true equivalent would be a HTHL winged LOX/Kero system, which might an interesting vehicle in its own right, but is OT for this thread.

I've long been a fan of your proposals and the results you have achieved and it's only recently I've come to realize one of Skylon's hidden strengths is it's aimed to give a TSTO payload fraction in an SSTO vehicle.

AFAIK previous SSTO attempts have countered this by either saying (basically) "Build it 4x bigger" or "launch 4x more frequently." Both are viable *engineering* solutions but (I think) hit major issues with the people who would write the checks for development.

There are many possible solutions to the problem "Get X Kg of payload to orbit Y Km in altitude at Z degrees inclination." REL have chosen their approach and have stuck with it for nearly 25 years.

I cannot say if Skylon is the best (depending on what metric you're using) but I can say it's the one that seems to be getting funded, *despite* the fact that  any US investors would face potential criminal charges if they invested in REL (Mark Hempsell, Space Show interview IIRC) and hence are not an option to them. 

I'd still be interested in finding out how changing to a 70/150 mt O:F ratio (or vice versa) affects your calculations. I'd consider the tank reduction caused by sub cooling H2 (I think 5% but I'm using m1v1/t1=m2v2/t2, not exactly cutting edge math) should be substantial.

[HMXHMX]

And if you use denser propellants than LOX-LH2, you can do even better.  Making a few very BOE assumptions I conclude that – based on volume alone without adjusting engine mass – a VTHL LOX-kerosene equivalent vehicle would carry about 28 tons to LEO

Except it would not *be* an equivalent vehicle.

A true equivalent would be a HTHL winged LOX/Kero system, which might an interesting vehicle in its own right, but is OT for this thread.

I've long been a fan of your proposals and the results you have achieved and it's only recently I've come to realize one of Skylon's hidden strengths is it's aimed to give a TSTO payload fraction in an SSTO vehicle.

AFAIK previous SSTO attempts have countered this by either saying (basically) "Build it 4x bigger" or "launch 4x more frequently." Both are viable *engineering* solutions but (I think) hit major issues with the people who would write the checks for development.

There are many possible solutions to the problem "Get X Kg of payload to orbit Y Km in altitude at Z degrees inclination." REL have chosen their approach and have stuck with it for nearly 25 years.

I cannot say if Skylon is the best (depending on what metric you're using) but I can say it's the one that seems to be getting funded, *despite* the fact that  any US investors would face potential criminal charges if they invested in REL (Mark Hempsell, Space Show interview IIRC) and hence are not an option to them. 

I'd still be interested in finding out how changing to a 70/150 mt O:F ratio (or vice versa) affects your calculations. I'd consider the tank reduction caused by sub cooling H2 (I think 5% but I'm using m1v1/t1=m2v2/t2, not exactly cutting edge math) should be substantial.

I won't belabor my point since I don't wish to hijack the thread.  But equivalency is in the eye of the beholder.  To a first order vehicles with comparable dry mass will cost the same to develop and individually manufacture.  (Except when totally new technology is required for a subsystem such as propulsion, which is the case for Skylon.  But that exception doesn't favor new technology.)  My main point is that – for a fixed volume – a pure rocket reusable will always be superior to a mixed rocket/air-breather-powered vehicle.  That is largely independent of propellant mixture ratios, Isp or anything else, and is largely dependent upon tank mass, which is a direct function of tank volume.

A note about possible criminal charges for U.S. investors.  I think that is incorrect.  I can't help them design or build their vehicle, but I can certainly legally fund it if I had the money. ITAR doesn't restrict investment from the US to foreign firms.  It can be a problem in reverse, because foreign investors in a US firm would have ownership interests in US technology and would presumably want to know the details of that technology in order to make the investment in the first instance.

I'd also add a personal note. I've known Alan Bond since about 1972, when he led the Daedalus study for the BIS, and have high respect for him.  Regrettably we haven't had a recent face-to-face opportunity to debate the Skylon issues I have raised, but we both want the same thing: routine, reliable and inexpensive access to orbit.  I wish the Skylon team every success.

[93143]

It seems to me that you are neglecting the effect of increased weight loading on structural requirements.

A Skylon loaded with LOX/LH2 at 6:1 would collapse its landing gear.  Strengthen the landing gear and the spaceframe would fail.  Strengthen the spaceframe and the wings would either tear off or be inadequate to get the thing off the runway at all.

You could switch to vertical takeoff, but then your thrust has to increase significantly, requiring extra engine mass (doubling the vehicle mass and increasing T/W by 50% means triple the thrust, which means that an engine T/W of 60 gives you 70% of the mass of the SABRE engines on the original vehicle), and the increase in axial loading will at least partially offset the (likely) reduction in design bending moment vs. the horizontal takeoff configuration.

Either way you're dealing with close to twice the total weight of the original Skylon.  A 20% increase in structural mass (cf. Andrew_W) is not going to cover this.

Also, 4500 N·s/kg mission-average (ibid.) is rather ambitious for something ground-started...

[HMXHMX]

It seems to me that you are neglecting the effect of increased weight loading on structural requirements.

A Skylon loaded with LOX/LH2 at 6:1 would collapse its landing gear.  Strengthen the landing gear and the spaceframe would fail.  Strengthen the spaceframe and the wings would either tear off or be inadequate to get the thing off the runway at all.

You could switch to vertical takeoff, but then your thrust has to increase significantly (requiring extra engine mass - doubling the vehicle mass and increasing T/W by 50% means triple the thrust, which means that an engine T/W of 60 gives you 70% of the mass of the SABRE engines on the original vehicle), and the increase in axial loading will at least partially offset the (likely) reduction in design bending moment vs. the horizontal takeoff configuration.

Either way you're dealing with close to twice the total weight of the original Skylon.  A 20% increase in structural mass (cf. Andrew_W) is not going to cover this.

Also, 4500 N·s/kg mission-average (ibid.) is rather ambitious for something ground-launched...

I'm not.  First off, naturally the vehicle using kerosene I described would be VTHL; I so stated.  Next, the mass of the Skylon SABRE powerplants total 10.8 MT according to the most recent (2003) info I could find by searching for 60 seconds (see SABRE engine, Wikipedia, first reference).  Add 3.9 MT for the nacelles.  Nine NK33s weigh 12 tons.  Since they would be aft body mounted and not at the tips of the wings, you'd save wing mass.  In addition, I'm neglecting much lighter main and nose gear mass that come from using the gear only for landing empty.

Increasing the mass of a VTHL has no bearing on the matter.  Fuel is cheap. LOX is the cheapest propellant you can buy.  As a wise man once said: "When LOX is 3 cents  per pound it doesn't make much sense to want to get it from the air for free."

(FYI, I don't know where your 4500 number comes from.  The vacuum Isp of an NK33 at 27:1 is 331 seconds.  That's what I used.)

[93143]

I stated I was referencing Andrew_W's earlier post, in which he specified both 20% extra structural mass and 4500 N·s/kg.

I was not responding only to your latest kerolox idea, but to the entire string of posts about how Skylon would work so much better with rockets instead of SABREs.  Maybe it's true, but I'm not remotely convinced.

SSTO is easy if you don't need to reuse the vehicle - just go with balloon tanks and kerolox.  Since this is a reusable vehicle we're trying to build, you need to either have load-bearing tanks with TPS and mounting structure outside them or go with something like Skylon, which has non-structural tanks and a load-bearing spaceframe suspending them within the TPS.

If you go with the spaceframe option, the mass of the spaceframe should (to first order) scale linearly with the load it is expected to support.  If you don't, you can't assume similar structural mass to Skylon.

My engine mass calculation was based simply on the idea that SABRE has a T/W of 14 (which is an oversimplification, but good enough for this sort of handwaving).  The comparison was with a hydrolox rocket.  If you have to triple the thrust due to the extra LOX mass and vertical takeoff, a T/W of 60 (not sure what I was thinking there) only saves you 30% in engine mass.  A T/W of 70 (~SSME) saves 40%.

Also, putting the engines on the back is what took down HOTOL and VentureStar.  Admittedly a conventional rocket engine cluster would be lighter than either an RB545 or an RS-2200...

[simonbp]

Also, putting the engines on the back is what took down HOTOL and VentureStar.  Admittedly a conventional rocket engine cluster would be lighter than either an RB545 or an RS-2200...

At least for HOTOL the issue was that with most of the dry mass at the back, the center of mass shifted radically while the center of pressure remained near-constant. This is a wonderful recipe for an unstable aircraft, which isn't so much a problem for a rocket that just blasts up on a gravity turn, but is a real problem for an airbreather that has to fly horizontally through the atmosphere for much longer.

Using a denser propellant would alleviate this somewhat, but hydrogen is so light compared to a hydrocarbon that I'm finding it hard to believe it makes much difference.

[MP99]

In addition, I'm neglecting much lighter main and nose gear mass that come from using the gear only for landing empty.

Would it still be feasible to abort to a landing? Would you / could you dump the remaining prop first?

cheers, Martin

[Ric Capucho]

I'd have thought that one of the advantages of Skylon would be (re)startable engines on approach, hence go around capability. Hence I'd expect a modest amount of fuel onboard upon landing. Yer airliner is carefully fuel planned for an entire trip, including reserves for diversions and go arounds, so you can bet the same for a Skylon mission.

I'm sure Mr Bond or someone close to REL will disillusion me anytime soon...

"Top her up to the brim, Mr Pumpy. She's off on a jaunt up yonder, and we don't know when she'll be back."

"Aye, Mr Bond."

Ric