[EthanHeilman]

Elon didn't directly address the advantage of the Skylon (he said he hadn't really looked into it, so fair enough). The main advantage is not that you don't need to bring oxygen for a portion of our trip (which as Mr. Musk points out wouldn't really save very much), but rather that you don't need to bring reaction mass.

Jet engines are just really really fast propellers. They don't need to carry their own reaction mass like a rocket. This gaves them a completely different efficiency profile (no reaction mass, means less weight, which means less fuel, which means less weight, rinse/repeat to the limit). This is the reason that 747s use jet engines rather than rockets. Being able to use both technologies fluidly within their optimal atmospheric envelopes would be a major breakthrough and would completely change access to space.

tl;dr Skylon can use a small, efficient jet engine to get high and fast enough that air resistance stops mattering and while slowly pivoting to a conventional rocket (which is optimal when the air gets thin).

[FrojoS]

Thanks for the description. So for a conventional rocket, like a Flacon 9, how much of the overall energy used is due to air friction? I would have thought, its just a small fraction.

[EthanHeilman]

According to this link (http://www.g2mil.com/high.htm), air friction is not completely trivial. Additionally rocket engines are more efficient in less dense atmosphere "because the thinner air allows a better plume". They estimate that launching from 30,000 feet provides 9.3% greater thrust.

Note that much of the energy spent for orbital flight is not spend getting height, but spent getting speed. Efficiency is greatly increased if you can reach high speeds without having to carry the reaction mass to reach those speeds.

LEO requires about 7.8 km/s, Skylon's jet engines can go around 1.7 km/s. You are reaching 20% of your orbital velocity without reaction mass.

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

The big savings, like spacex's grasshopper project, is of course a reusable space vehicle. The advantage to spacex's approach is that Skylon is a single stage to orbit space plane (you don't need to land a bunch of stages using retro-rockets).

If I was placing a bet, I would bet on spacex since they have a proven track record (rockets are easy, organizing/funding rocket companies is hard). Skylon is a great idea though and it is the general direction that aircraft engine design is headed (a peak at the future).

[bathat]

>LEO requires about 7.8 km/s, Skylon's jet engines can go around 1.7 km/s. You are reaching 20% of your orbital velocity without reaction mass.

Unfortunately, kinetic energy scales as speed squared, so 20% of your orbital speed represents less than 5% of your orbital kinetic energy. To put this in perspective, the difference in gravitational potential energy between LEO and the earth's surface represents about 15% of your total on-orbit energy.

Now, it's true that an air-breathing engine doesn't need to carry reaction mass (and, sometimes, oxidizer), but the air engines add considerable complexity (which is always a bad thing) as well as weight (because you still need to carry a conventional rocket to finish orbit insertion). So what you need to do is ask how the weight penalty of the air engine compares to the weight penalty of carrying extra fuel in a conventional rocket (bearing in mind, of course, that there a pernicious positive feedback loops when scaling a booster).

[lmm]

>Unfortunately, kinetic energy scales as speed squared, so 20% of your orbital speed represents less than 5% of your orbital kinetic energy.

Fortunately this is counterbalanced by the Oberth effect. Getting to 20% of your orbital velocity requires expending 20% of your rocket's delta-V. And since delta-V is logarithmic in your propellant mass (rocket equation) that could easily translate to needing half as much fuel.

[bathat]

>Fortunately this is counterbalanced by the Oberth effect. Getting to 20% of your orbital velocity requires expending 20% of your rocket's delta-V. And since delta-V is logarithmic in your propellant mass (rocket equation) that could easily translate to needing half as much fuel.

Well, that doesn't really address my point, which was that you need to compare the weight of the hybrid engine to the weight of the extra fuel. The first problem is that an air-breathing engine is going to be something like 3 to 5% of the initial mass, and you have to carry it with you to orbit[1]. The second problem is how the fuel scales:

v_hybrid = 0.8 * v_conventional

Assume both have similar engines:

ln(m_hybrid-initial/ m_hybrid-final) = ln((m_conventonal-initial / m_conventional-final)^0.8)

m_hybrid-intial = (m_conventional-intial / m_conventional-final)^0.8 * m_hybrid-final

let delta equal the expression in parenthesis

m_h-i = delta^0.8 * (m_payload + m_engine) = delta^0.8 * (m_payload + m_h-i * 0.05)

m_h-i * (1 - 0.05 * delta^0.8) = delta^0.8 * m_payload

So the fuel load in a hybrid is going to be:

m_final = delta^(1-0.2) / (1 - 0.05delta^(1-0.2)) m_payload,

The factor in the denominator is what really kills you, and the hybrid is only going to give you a net benefit for deltas less than about 15. So, not only is there not a factor of 2 fuel savings, there isn't any fuel savings at all! Even if you assume an engine weight of only 3% of initial mass, the benefit is only for delta < 35, which is better than just about all actual (as opposed to paper) launchers. By the way, the Shuttle had a delta of about 85-90 for LEO precisely because its designers made the decision to bring wings (which we neglected above) along for the ride to orbit. That also contributed to the 1 in 50 accident rate of that launch system.

And none of this addresses the fact that you are optimizing the f*ck out of one of the least expensive components of launch cost by introducing all sorts of unnecessary complexity.

[1] Ok, I suppose you don't, but then you have to have some way of recovering it, and that adds an enormous amount of complexity to the system.

[EthanHeilman]

>And none of this addresses the fact that you are optimizing the fck out of one of the least expensive components of launch cost by introducing all sorts of unnecessary complexity.

As I understand it, construction of the rocket is the most expensive part of a launch system. The point of skylon is to create a reusable single stage to orbit space plane. Shouldn't skylon's reusability make it "optimize the fck" out of one of the most expensive components?