Whats wrong with Skylon?

[wumpus]

5 hours ago, sevenperforce said:

Yeah, it's not supersonic. Basically the same concept as the SR-71: takeoff with empty tanks, refuel at modest speed, then burn to orbit from a much more optimal launch location. 

It's basically the same as air-launching, but you can use existing tanker aircraft rather than designing an entirely new mothership or dealing with decoupling. 

Of course, there is the issue of why [US] commercial flights of sufficient distance don't refuel mid-air.  Certainly US-Asia flights could be re-fueled over Hawaii for a considerable savings of weight over the course of the flight.  Certainly nearly all US commercial pilots (especially the ones flying the big jets to and from Asia) have experience in mid-air refueling.  Somehow I don't see the FAA allowing such things for passenger (and presumably Fed-Ex as well, do they do any refueling?).

 

I really have to question the direction this thread has gone.  If a turbojet's efficiency isn't enough to propel a rocket, you have bigger problems than the Isp of the turbojet.  If mach 5 isn't fast enough (the rocket has to carry too much fuel for mach 5-"mach" 25), then you are very unlikely to fix the problem by increasing the Isp of the turbojet's performance <mach 5.

I still wonder if burning liquid oxygen in combination with a scramjet would help at all.  Before playing KSP, I thought it was a much better idea, but it might have promise.  The idea is that while scramjets appear to work above mach 5 (and prototypes have proven so), they are lucky to provide as much thrust as they experience drag.  Assuming (the big if, and why KSP dashed my dreams) is that this number *must* be positive or the whole scheme collapses.  If it is positive, then adding additional thrust (from a liquid oxygen oxidized rocket) would increase efficiency, decrease the time it took to accelerate to maximum airspeed (presumably cutting the length of time it has to withstand brutal temperatures) and hopefully increase the fuel spent to accelerate (kind of like increasing TWR to decrease drag losses, only with drag losses* instead of gravity ones).  If the number is negative, then you are better off setting your gravity turn to accelerate out of the atmosphere.  Also note that the Isp of such a jet-rocket would be miserable compared to the SABRE (and maybe 50-100% better than a hydrogen rocket).  The idea is to make it up by extending the speed you can use air at.

* yes, I know going faster will increase drag losses.  But going faster is the entire point of the exercise.  The point is to get going as fast as possible while still breathing air.

[AngelLestat]

5 hours ago, magnemoe said:

Did an few calculations on Skylon, it can reach 1700 m/s air breathing, isp of engines in rocket mode is 465s and will mostly run in vacuum.
Empty weight is  fully loaded is 325 ton, empty is 53 ton, payload is 15 ton, 
This gives ln(325/68)=1.56
1.56*465*9.8=7109, now add 1700 m/s and you start at the rockets at 30 km attitude, 

Now lets try an two stage version, assuming an total upper stage mass of 110 ton, its the same as falcon 9 upper stage with payload. 
ln(325/(53+110)=0.7, 0.7*465*9.8=3190, we end up 4km/s short with an maximum speed of 5900 m/s 

Assuming we use hydrogen for second stage to and an isp of 450, we will need an mass fuel faction of around 2.6 for upper stage. (ln(2.6)*450*9.8=4200 m/s
110/2.6=42 ton to orbit, assume second stage dry weight is 4 ton or the same as falcon 9 second stage, it can be fragile as most structure stays in the skylon.

So you are taking into account the second stage of falcon9 which is not reusable and it use kerosene, your stage use hydrogen (which increase a lot your drag and it makes harder  to take advantage of the air breathing mode).
Take into account that all rockets pass through the atmosphere drag layer very fast, almost vertically, but an air breathing engine should spent a considerable amount of time in the atmosphere.
The development cost would be much higher than falcon9, and I dont see many advantages.
The best advantage that skylon has over falcon9, is that it lands, refuel and launch again. No much need for planning, operation, asembly, etc.
And you can launch a lot of sats just with 1 skylon (depending its lifetime).
There is when you save most of the money, that is why Elon Musk wants 2 reusable stages instead just one, even if still have assembly cost and some of operation cost that will be higher, but it can save a lot of money and increase a lot its launch rate, something that it can not do with its normal second stage, because it needs to build a new one and spent a lot of time and money in testing.
A SSTO will have less probabilities of fail (1 single part), so the testing is highly reduce. 

[sevenperforce]

3 hours ago, blowfish said:

All the air is ram-compressed.  The question is how much additional turbocompression you apply.

Sure, the incoming air is ram-compressed. The question is, when your thermal load is too great to apply any additional turbocompression, can you make any use of the ram-compression you still have? Not with the SABRE.

3 hours ago, blowfish said:

An afterburner is entirely in front of the nozzle. Combustion takes a nontrivial amount of time.  Afterburners are very long tubes with fuel injection at the front for that reason.  Flow is entirely subsonic over the course of combustion.

That's kind of splitting hairs; the afterburner is obviously upstream of the final expansion nozzle, but it is downstream of the main combustion region, which was all I was trying to say. And sure, there will be an optimal length to allow more complete mixing. But for an engine like the SABRE, there's an advantage in moving the air injection downstream of the initial combustion chamber, because it allows it to accept a greater range of air pressures than if it had to be injected straight into the combustion chamber.

With injection directly into the combustion chamber, you HAVE to use turbocompression; otherwise, your chamber pressure will have to be kept exceedingly low to prevent unstart. Such designs are therefore limited by the thermal capacity of the turbocompressor. Downstream compression, on the other hand, cannot unstart and thus can accept air at speeds greater than the capacity of the turbocompressor.

3 hours ago, blowfish said:

The inlet geometry still doesn't make sense.  You're certainly not going to get any better performance than a shock cone intake.

The shock cone intake is still there. It's just rearranged.

Here's the basic geometry for a shock cone inlet, in a simple ramjet sort of arrangement:

The cone compresses the air into the intake while some air spills around the edges; controlling the position of the cone can adjust how much air enters the intake.

Reducing by symmetry, we're looking at this basic shape:

This shape can be taken through various transformations to produce different engines with the same basic geometry. For example, translation, forming an inlet ramp:

Or rotation about the cone axis, forming a typical shock cone inlet:

Or, by rotation around an alternate axis, which is what I'm proposing:

It's the same shape and the same function. The only difference is where the spill air goes. Supersonic fighters often use the spill air to help generate lift, while in this particular instance, spill air travels through the open center. It could therefore be used to potentially augment thrust, where it would be wasted in the former case.

Edited March 10, 2016 by sevenperforce

[blowfish]

53 minutes ago, sevenperforce said:

Sure, the incoming air is ram-compressed. The question is, when your thermal load is too great to apply any additional turbocompression, can you make any use of the ram-compression you still have? Not with the SABRE.

Overheating will happen at compressor exit - if you want you can just reduce the amount of energy that the helium loop is outputting and it will reduce the compression ratio.

53 minutes ago, sevenperforce said:

That's kind of splitting hairs; the afterburner is obviously upstream of the final expansion nozzle, but it is downstream of the main combustion region, which was all I was trying to say. And sure, there will be an optimal length to allow more complete mixing. But for an engine like the SABRE, there's an advantage in moving the air injection downstream of the initial combustion chamber, because it allows it to accept a greater range of air pressures than if it had to be injected straight into the combustion chamber.

Meh ... everything has to be at the same pressure by the time it enters the nozzle, so it's just a question of where you're adding/removing energy.  Turborockets have a separate combustion chamber because there needs to be something to drive the compressor, not because it provides any benefit to separate the two.  The SABRE doesn't have this problem because the compressor is driven by the helium loop, though it does still have a preburner, presumably because the main chamber exhaust is too hot for a heat exchanger to handle.

53 minutes ago, sevenperforce said:

It's the same shape and the same function. The only difference is where the spill air goes. Supersonic fighters often use the spill air to help generate lift, while in this particular instance, spill air travels through the open center. It could therefore be used to potentially augment thrust, where it would be wasted in the former case.

Actually using that spill air isn't as trivial as you think.  The required capture area varies widely with mach number - in order to accommodate this you would probably have to have a nozzle throat that could vary in area by a factor of several.

[sevenperforce]

1 minute ago, blowfish said:

Overheating will happen at compressor exit - if you want you can just reduce the amount of energy that the helium loop is outputting and it will reduce the compression ratio.

And unstart your engine.

2 minutes ago, blowfish said:

Actually using that spill air isn't as trivial as you think.  The required capture area varies widely with mach number - in order to accommodate this you would probably have to have a nozzle throat that could vary in area by a factor of several.

Or have a bypass area large enough to accommodate any amount of spill.

[blowfish]

1 minute ago, sevenperforce said:

And unstart your engine.

You'd probably have to have a variable area nozzle for this, yes, but that would be the case regardless.

2 minutes ago, sevenperforce said:

Or have a bypass area large enough to accommodate any amount of spill.

The area of the bypass doesn't matter - it only mass limits through the nozzle throat where it chokes.  And that nozzle throat has to be the correct area to accept however much air mass the intake is giving you (or, as you way, it will unstart, or going in the other direction the nozzle will cease to choke and you won't produce useful thrust).

The SABRE's bypass ramjets will need to have variable throat areas in order to accept a variable amount of bypass mass flow.  I don't know if the core will to - it's certainly doable on an expansion-deflection nozzle, but maybe not necessary if the bypass is variable enough.

[magnemoe]

26 minutes ago, AngelLestat said:

So you are taking into account the second stage of falcon9 which is not reusable and it use kerosene, your stage use hydrogen (which increase a lot your drag and it makes harder  to take advantage of the air breathing mode).
Take into account that all rockets pass through the atmosphere drag layer very fast, almost vertically, but an air breathing engine should spent a considerable amount of time in the atmosphere.
The development cost would be much higher than falcon9, and I dont see many advantages.
The best advantage that skylon has over falcon9, is that it lands, refuel and launch again. No much need for planning, operation, asembly, etc.
And you can launch a lot of sats just with 1 skylon (depending its lifetime).
There is when you save most of the money, that is why Elon Musk wants 2 reusable stages instead just one, even if still have assembly cost and some of operation cost that will be higher, but it can save a lot of money and increase a lot its launch rate, something that it can not do with its normal second stage, because it needs to build a new one and spent a lot of time and money in testing.
A SSTO will have less probabilities of fail (1 single part), so the testing is highly reduce. 

First off the spaceplane going at would be on an suborbital trajectory going at 5 km/s more than half the orbital speed. its an release during freefall in space. I do not know the suitable AP and trajectory here but it should be high enough so you can delay the release until you are in vacuum or close enough for it to be irrelevant.  
In short its an separation in space as in freefall and vacuum just as shuttle or skylon cargo, yes cargo is 8x larger and you have time restrain, this is the only difference. As both launches are unmanned you might skip a bit on the safety distance for second stage ignition. 

Main benefit over Falcon 9 is that staging is at 5 km/s not below 2 km/s, note that cargo capacity is more than 3x falcon 9 and pretty close to falcon heavy, 
Another option is to make everything smaller and / or have lower performance requirements. You could use an more dense propellant than hydrogen for second stake and trim airframe size,

This could also let you recover second stage, think it would be to hard to recover the hydrogen tank but you could recover engine and systems.
Recovery is very dependent on payload orbit anyway. 

Atmospheric flight path would be identical, so would start of rocket stage, you might want to adjust the end of first stage trajectory as upper stage separation will take a minute or two and upper stage will have lower twr. 
Again this is also relevant for an full SSTO Skylon for various launches, I liked the idea of returning the tug to skylon for return however it would then have to dock inside the hold. 
For an GTO mission I think aerobraking and separate return is probably better. 

 

 

[sevenperforce]

43 minutes ago, blowfish said:

You'd probably have to have a variable area nozzle for this, yes, but that would be the case regardless.

The area of the bypass doesn't matter - it only mass limits through the nozzle throat where it chokes.  And that nozzle throat has to be the correct area to accept however much air mass the intake is giving you (or, as you way, it will unstart, or going in the other direction the nozzle will cease to choke and you won't produce useful thrust).

The SABRE's bypass ramjets will need to have variable throat areas in order to accept a variable amount of bypass mass flow.

That is part of the point of having spill air bypass through the center, rather than being redirected outside the engine entirely. You don't need a variable area nozzle, because the nozzle opens against the stream of spill air.

[magnemoe]

3 hours ago, wumpus said:

Of course, there is the issue of why [US] commercial flights of sufficient distance don't refuel mid-air.  Certainly US-Asia flights could be re-fueled over Hawaii for a considerable savings of weight over the course of the flight.  Certainly nearly all US commercial pilots (especially the ones flying the big jets to and from Asia) have experience in mid-air refueling.  Somehow I don't see the FAA allowing such things for passenger (and presumably Fed-Ex as well, do they do any refueling?).

 

I really have to question the direction this thread has gone.  If a turbojet's efficiency isn't enough to propel a rocket, you have bigger problems than the Isp of the turbojet.  If mach 5 isn't fast enough (the rocket has to carry too much fuel for mach 5-"mach" 25), then you are very unlikely to fix the problem by increasing the Isp of the turbojet's performance <mach 5.

I still wonder if burning liquid oxygen in combination with a scramjet would help at all.  Before playing KSP, I thought it was a much better idea, but it might have promise.  The idea is that while scramjets appear to work above mach 5 (and prototypes have proven so), they are lucky to provide as much thrust as they experience drag.  Assuming (the big if, and why KSP dashed my dreams) is that this number *must* be positive or the whole scheme collapses.  If it is positive, then adding additional thrust (from a liquid oxygen oxidized rocket) would increase efficiency, decrease the time it took to accelerate to maximum airspeed (presumably cutting the length of time it has to withstand brutal temperatures) and hopefully increase the fuel spent to accelerate (kind of like increasing TWR to decrease drag losses, only with drag losses* instead of gravity ones).  If the number is negative, then you are better off setting your gravity turn to accelerate out of the atmosphere.  Also note that the Isp of such a jet-rocket would be miserable compared to the SABRE (and maybe 50-100% better than a hydrogen rocket).  The idea is to make it up by extending the speed you can use air at.

* yes, I know going faster will increase drag losses.  But going faster is the entire point of the exercise.  The point is to get going as fast as possible while still breathing air.

turbofan planes are very efficient, even for very long flight past 12.000 km the long distance planes has enough fuel capacity. 
Military has other needs it might need an thirsty fighter jet to cross the pacific or keep station for 12 hours. 
Even dirt cheap space launch like an well used Skylon is more expensive than high cost military flights. 

Yes it has been jokes about Ukrainian or Russian tankers as an response to talks of fuel taxes for planes in EU. 
Deep sea fishing boats sometimes bunker in open sea from east block supply ships and trucks often install lots of extra fuel capacity so they can go Norway to Portugal and only refuel in Andorra so it has some realty but more likely you will fly to East Europe and switch plane. 

[blowfish]

3 minutes ago, sevenperforce said:

That is part of the point of having spill air bypass through the center, rather than being redirected outside the engine entirely. You don't need a variable area nozzle, because the nozzle opens against the stream of spill air.

It doesn't quite work like that.  The expanding section of the nozzle can be actuated by pressure (though you probably don't want that on a rocket), but the nozzle throat has to be actively controlled to maintain the correct area.

[fredinno]

11 hours ago, sevenperforce said:

Yeah, it's not supersonic. Basically the same concept as the SR-71: takeoff with empty tanks, refuel at modest speed, then burn to orbit from a much more optimal launch location. 

It's basically the same as air-launching, but you can use existing tanker aircraft rather than designing an entirely new mothership or dealing with decoupling. 

Why? Seems like a lot more complexity for a niche feature, with very little performance increase. The 0 to subsonic part is usually pretty efficient.

10 hours ago, magnemoe said:

Did an few calculations on Skylon, it can reach 1700 m/s air breathing, isp of engines in rocket mode is 465s and will mostly run in vacuum.
Empty weight is  fully loaded is 325 ton, empty is 53 ton, payload is 15 ton, 
This gives ln(325/68)=1.56
1.56*465*9.8=7109, now add 1700 m/s and you start at the rockets at 30 km attitude, 

Now lets try an two stage version, assuming an total upper stage mass of 110 ton, its the same as falcon 9 upper stage with payload. 
ln(325/(53+110)=0.7, 0.7*465*9.8=3190, we end up 4km/s short with an maximum speed of 5900 m/s 

Assuming we use hydrogen for second stage to and an isp of 450, we will need an mass fuel faction of around 2.6 for upper stage. (ln(2.6)*450*9.8=4200 m/s
110/2.6=42 ton to orbit, assume second stage dry weight is 4 ton or the same as falcon 9 second stage, it can be fragile as most structure stays in the skylon.

Umm, I'm pretty sure you need to stage at at least 70 km to make sure the atmosphere does not rip the 2nd stage apart at staging. The 2nd stage is to do the circularisation, ascent above the Karman line, and (plus a third stage, possibly) doing BLEO missions.

[fredinno]

4 hours ago, AngelLestat said:

So you are taking into account the second stage of falcon9 which is not reusable and it use kerosene, your stage use hydrogen (which increase a lot your drag and it makes harder  to take advantage of the air breathing mode).
Take into account that all rockets pass through the atmosphere drag layer very fast, almost vertically, but an air breathing engine should spent a considerable amount of time in the atmosphere.
The development cost would be much higher than falcon9, and I dont see many advantages.
The best advantage that skylon has over falcon9, is that it lands, refuel and launch again. No much need for planning, operation, asembly, etc.
And you can launch a lot of sats just with 1 skylon (depending its lifetime).
There is when you save most of the money, that is why Elon Musk wants 2 reusable stages instead just one, even if still have assembly cost and some of operation cost that will be higher, but it can save a lot of money and increase a lot its launch rate, something that it can not do with its normal second stage, because it needs to build a new one and spent a lot of time and money in testing.
A SSTO will have less probabilities of fail (1 single part), so the testing is highly reduce. 

>implying you won't make a resuable 2nd stage considering the extra performance for a 2STO skylon

And a F9 with a resuable 2nd stage would likely just be a Falcon Heavy (with 5-8 T payloads using the original Falcon 9) due to the performance hit it would take. Skylon has better ISP, but higher development costs- if you have no problem with Saturn V-sized rockets, then the full reuse of 2 stages via rcoket propulsion is fine.

There is a time and place for everything, but not for SSTOs- at least not now.

3 hours ago, magnemoe said:

First off the spaceplane going at would be on an suborbital trajectory going at 5 km/s more than half the orbital speed. its an release during freefall in space. I do not know the suitable AP and trajectory here but it should be high enough so you can delay the release until you are in vacuum or close enough for it to be irrelevant.  
In short its an separation in space as in freefall and vacuum just as shuttle or skylon cargo, yes cargo is 8x larger and you have time restrain, this is the only difference. As both launches are unmanned you might skip a bit on the safety distance for second stage ignition. 

Main benefit over Falcon 9 is that staging is at 5 km/s not below 2 km/s, note that cargo capacity is more than 3x falcon 9 and pretty close to falcon heavy, 
Another option is to make everything smaller and / or have lower performance requirements. You could use an more dense propellant than hydrogen for second stake and trim airframe size,

This could also let you recover second stage, think it would be to hard to recover the hydrogen tank but you could recover engine and systems.
Recovery is very dependent on payload orbit anyway. 

Atmospheric flight path would be identical, so would start of rocket stage, you might want to adjust the end of first stage trajectory as upper stage separation will take a minute or two and upper stage will have lower twr. 
Again this is also relevant for an full SSTO Skylon for various launches, I liked the idea of returning the tug to skylon for return however it would then have to dock inside the hold. 
For an GTO mission I think aerobraking and separate return is probably better. 

 

 

Actually, you might as well take advantage of the extra performance to make a Shuttle 2.0.

Anyways, recovering the 2nd stage is easily possible with a 2STO Skylon, in terms of performance.

[magnemoe]

1 hour ago, fredinno said:

Umm, I'm pretty sure you need to stage at at least 70 km to make sure the atmosphere does not rip the 2nd stage apart at staging. The 2nd stage is to do the circularisation, ascent above the Karman line, and (plus a third stage, possibly) doing BLEO missions.

Ok then we start separation at 75-80, Ap should be at final orbit, second stage is fragile, do not see an need for an stage three outside an deep space probe ion engine. 

[sevenperforce]

4 hours ago, blowfish said:

It doesn't quite work like that.  The expanding section of the nozzle can be actuated by pressure (though you probably don't want that on a rocket), but the nozzle throat has to be actively controlled to maintain the correct area.

This is the part where I wish I had a really good flow modeling software package, because I suspect that the airflow through the center will have some pretty specific effects that can potentially be tuned to do exactly that. Just like the atmosphere functions as the opposite side of the exhaust bell in an aerospike nozzle, the central flow functions as a virtual aerospike to hold the exhaust against the traditional exhaust bell. 

At subsonic speeds near sea level, the low pressure at the inlet caused by the compressor will tend to pull air through and into the bypass, so that it is coming by with a pressure equal to ambient but a rearward velocity relative to the engine. This will cause minor thrust augmentation, though not much. There is no significant benefit to reheat at this stage. 

Once the stable supersonic flight regime is reached, however, the flow through the actual inlet is choked while the spill through the central bypass is unchoked. Spill air is higher pressure than ambient, so it has a net positive thrust (or at least reduces back pressure like the ramjet burners on the SABRE). However, it's necessarily lower pressure than the choked air at the terminal shock in the inlet, so whatever the pressure in the combustor is, it will expand against the spill air stream rather than pushing back against the terminal shock and unstarting.

This at least eliminates the need for an adjustable exhaust nozzle and an adjustable inlet spike. 

[blowfish]

@sevenperforce The nozzle also has to choke, otherwise you won't generate useful thrust.  The diverging section beyond the throat can be pressure actuated, and is in many engines.  However, before the throat, the flow is subsonic and must be constricted to reach mach 1 at the throat so that it can go supersonic in the diverging section.  Note that pressure works the opposite way you want for isentropic flow when it's subsonic - if flow A has greater pressure than flow B, then flow A will expand, but you really want it to contract.

Supersonic jets typically have hydraulic actuators on the nozzle throat and allow the diverging flaps to be actuated by pressure.  Rockets don't have this problem because they are (mostly) constant flow devices, so the entire nozzle geometry is fixed (keeping the throat choked is a major reason why throttling is so hard).

[Octa]

Sorry for interrupting the current discussion about jet engines, i'm definetly no expert regarding the concepts of reusable transports....

Isn't the maintenance between flights the main factor that makes spaceplanes currently too expensive/inefficent? In KSP, we pay funds for some cheap fuel after the spaceplane landed and it can lift of again, while IRL the space shuttle had to go through maintenance to keep reliable.

I have a phrase like "Why would i invest energy to put 50t into orbit when 45 of them would get back to earth anyway?" in my head, most likely heard or read about the space shuttle. I guess this particular point could be solved with progress in fuel effiency or engine development. Then again: why should i put these on a fully reusable spaceplane instead of a "cheap" 1 use rocket? Maybe make some parts of that rocket reusable but not the whole thing....

Are these valid points in this discussion or do i miss some huge aspects? Please enlighten me.

[sevenperforce]

44 minutes ago, Octa said:

Isn't the maintenance between flights the main factor that makes spaceplanes currently too expensive/inefficent? In KSP, we pay funds for some cheap fuel after the spaceplane landed and it can lift of again, while IRL the space shuttle had to go through maintenance to keep reliable.

I have a phrase like "Why would i invest energy to put 50t into orbit when 45 of them would get back to earth anyway?" in my head, most likely heard or read about the space shuttle. I guess this particular point could be solved with progress in fuel effiency or engine development. Then again: why should i put these on a fully reusable spaceplane instead of a "cheap" 1 use rocket? Maybe make some parts of that rocket reusable but not the whole thing....

Are these valid points in this discussion or do i miss some huge aspects? Please enlighten me.

Absolutely valid points. With the Shuttle program, the amount of reconditioning, rebuilding, and refurbishing between each flight was simply astronomical. Extreme sink of money and time. This was partly because it was intended to do so much; it was a launch vehicle that was also a crew ferry that was also a flying lab that was also a Space Station Construction system that was also a cargo carrier system that was also a satellite delivery system that was also a satellite recovery system that was also a plane that was also a launch engine return system. Altogether there were so many systems that refurbishing between flights was an extreme nightmare.

Skylon's engine system is more complex, but as a concept it is much simpler: a plane that flies to orbit, drops a payload, and comes back.

[AngelLestat]

15 hours ago, magnemoe said:

ln(325/(53+110)=0.7, 0.7*465*9.8=3190, we end up 4km/s short with an maximum speed of 5900 m/s 

What you mean with a maximun speed of 5900m/s?? I guess you want to said 4900 m/s

15 hours ago, magnemoe said:

In short its an separation in space as in freefall and vacuum just as shuttle or skylon cargo, yes cargo is 8x larger and you have time restrain, this is the only difference. As both launches are unmanned you might skip a bit on the safety distance for second stage ignition. 

Main benefit over Falcon 9 is that staging is at 5 km/s not below 2 km/s, note that cargo capacity is more than 3x falcon 9 and pretty close to falcon heavy, 
Another option is to make everything smaller and / or have lower performance requirements. You could use an more dense propellant than hydrogen for second stake and trim airframe size,

Yeah but your numbers are very sharp as I said before and they does not have into account drag (those 1700m/s and 30km altitude you get for almost free), neither structural reinforcement for a 2 stage vehicle with a deploy system (long area doors is a safety concern, the vehicle can be broke in half).
I dont know how to explain it in few words, but you will only realize of the trouble you get into if you try to imagine the deploy system at scale.
Make some draws and you will notice that is not a problem that can be skip it so easily. 

Then you need to deal as I said with the reusable second stage... how do you solve that?

15 hours ago, magnemoe said:

This could also let you recover second stage, think it would be to hard to recover the hydrogen tank but you could recover engine and systems.
Recovery is very dependent on payload orbit anyway.

No.. you should recover the second stage, you kill all the launch rate and cost saves of the whole vehicle.
That is the point that it sale most with skylon..  it does not matter if its payload is similar to falcon9, if the cost is even 10 times lower than falcon9 fully reusable.  Companies would design their sats or payloads to fit the skylon.
Again.. the benefit of just refuel and launch again without even need to make huge calculations to achieve trajectory or staging, or operation (take into account all the operations that requires to launch a falcon9, all the systems), all those details + assembly and mostly "TIME", are the ones that saves money.. The amount of payload you can rise to orbit is secondary.
 
So your payload to orbit in reusable mode (heat shield, legs, aerobraking, etc) will reduce your payload to 25 tons with lucky taking into account the extra lost due drag from the first stage.
Other thing.. in your case, if you want to launch your first stage sable from the equator, you need to have enough fuel left to go back, in the skylon case you just need to go back 1 vehicle, no 2, and the wings can save a lot of deltav.
 

15 hours ago, magnemoe said:

Again this is also relevant for an full SSTO Skylon for various launches, I liked the idea of returning the tug to skylon for return however it would then have to dock inside the hold. 
For an GTO mission I think aerobraking and separate return is probably better. 

I dont see problem docking inside the hold, those maneuvers are done at almost 0m/s.
If you want a tug able to return from GTO, it will need a lot of mass to make it reusable and resist the reentry at those speeds, the amount of deltav you save with aerobraking is not enough to enter in such mess.
The gto tugs returns almost emptly with almost no mass, so just a little of fuel left can achieve the orbit change.
If a client wants to push up the GTO payload limit, then he should make its own ion propulsion to achieve gto from LEO.

[fredinno]

16 hours ago, magnemoe said:

Ok then we start separation at 75-80, Ap should be at final orbit, second stage is fragile, do not see an need for an stage three outside an deep space probe ion engine. 

A 3rd stage can be done to offer GEO circulation capacity, increase GTO capacity (like would have been great for SES-3) or increase payload for planetary probes. I was thinking a upgraded STAR motor.

4 hours ago, sevenperforce said:

Absolutely valid points. With the Shuttle program, the amount of reconditioning, rebuilding, and refurbishing between each flight was simply astronomical. Extreme sink of money and time. This was partly because it was intended to do so much; it was a launch vehicle that was also a crew ferry that was also a flying lab that was also a Space Station Construction system that was also a cargo carrier system that was also a satellite delivery system that was also a satellite recovery system that was also a plane that was also a launch engine return system. Altogether there were so many systems that refurbishing between flights was an extreme nightmare.

Skylon's engine system is more complex, but as a concept it is much simpler: a plane that flies to orbit, drops a payload, and comes back.

Yeah, Skylon is also supposed to carry all kinds of cargo- even crew. It suffers most of the same complexity problems as the Shuttle, minus the crew cabin.

3 hours ago, AngelLestat said:

What you mean with a maximun speed of 5900m/s?? I guess you want to said 4900 m/s

Yeah but your numbers are very sharp as I said before and they does not have into account drag (those 1700m/s and 30km altitude you get for almost free), neither structural reinforcement for a 2 stage vehicle with a deploy system (long area doors is a safety concern, the vehicle can be broke in half).
I dont know how to explain it in few words, but you will only realize of the trouble you get into if you try to imagine the deploy system at scale.
Make some draws and you will notice that is not a problem that can be skip it so easily. 

Then you need to deal as I said with the reusable second stage... how do you solve that?

No.. you should recover the second stage, you kill all the launch rate and cost saves of the whole vehicle.
That is the point that it sale most with skylon..  it does not matter if its payload is similar to falcon9, if the cost is even 10 times lower than falcon9 fully reusable.  Companies would design their sats or payloads to fit the skylon.
Again.. the benefit of just refuel and launch again without even need to make huge calculations to achieve trajectory or staging, or operation (take into account all the operations that requires to launch a falcon9, all the systems), all those details + assembly and mostly "TIME", are the ones that saves money.. The amount of payload you can rise to orbit is secondary.
 
So your payload to orbit in reusable mode (heat shield, legs, aerobraking, etc) will reduce your payload to 25 tons with lucky taking into account the extra lost due drag from the first stage.
Other thing.. in your case, if you want to launch your first stage sable from the equator, you need to have enough fuel left to go back, in the skylon case you just need to go back 1 vehicle, no 2, and the wings can save a lot of deltav.
 

I dont see problem docking inside the hold, those maneuvers are done at almost 0m/s.
If you want a tug able to return from GTO, it will need a lot of mass to make it reusable and resist the reentry at those speeds, the amount of deltav you save with aerobraking is not enough to enter in such mess.
The gto tugs returns almost emptly with almost no mass, so just a little of fuel left can achieve the orbit change.
If a client wants to push up the GTO payload limit, then he should make its own ion propulsion to achieve gto from LEO.

Hey, you know that most HLVs only go up to 25T? Considering the Skylon carries ~12T in SSTO mode, that's double, and you have both stages reusable (assuming what you suggested- landing the 2nd stage on legs). And if you launch from the equator, you need to just wait until you pass over your landing site, which is almost certainly near your launching site (Cape Canaveral, the facilities are already there- even if you don't launch there, it'll be easy to build a landing pad- hell, you can put that on a barge ship).

A GTO tug can slowly aerobrake after a small burn at apoapsis. You need more shielding, (unless you slowly aerobrake so that it doe not need shielding (Mars probes do this)) but the thing is most GEO customers expect at least being able to launch on their LV at least to GTO. The customer is priority- using its electric thrusters reduces the lifetime by huge amounts, and not everyone uses them.

[sevenperforce]

4 minutes ago, fredinno said:

4 hours ago, sevenperforce said:

With the Shuttle program, the amount of reconditioning, rebuilding, and refurbishing between each flight was simply astronomical. Extreme sink of money and time. This was partly because it was intended to do so much; it was a launch vehicle that was also a crew ferry that was also a flying lab that was also a Space Station Construction system that was also a cargo carrier system that was also a satellite delivery system that was also a satellite recovery system that was also a plane that was also a launch engine return system. Altogether there were so many systems that refurbishing between flights was an extreme nightmare.

Skylon's engine system is more complex, but as a concept it is much simpler: a plane that flies to orbit, drops a payload, and comes back.

Skylon is also supposed to carry all kinds of cargo- even crew. It suffers most of the same complexity problems as the Shuttle, minus the crew cabin.

The difference is that the Shuttle was expected to do most of those things at the same time. It had a crew cabin, a hab, a cargo bay, and an SRMS. It ran satellite servicing missions and satellite recovery missions and space station assembly missions. Skylon, on the other hand, doesn't need orbital persistence or service capability. It won't be expected to carry a crew and a satellite at the same time, and it won't be expected to return a payload to the surface.

[fredinno]

37 minutes ago, sevenperforce said:

The difference is that the Shuttle was expected to do most of those things at the same time. It had a crew cabin, a hab, a cargo bay, and an SRMS. It ran satellite servicing missions and satellite recovery missions and space station assembly missions. Skylon, on the other hand, doesn't need orbital persistence or service capability. It won't be expected to carry a crew and a satellite at the same time, and it won't be expected to return a payload to the surface.

Actually, it will be expected to return a payload to the surface when necessary, what else are you going to use the massive downmass for? Satellites can be carried down for repair if it is necessary, or for recoverable satellites.

[sevenperforce]

3 minutes ago, fredinno said:

Actually, it will be expected to return a payload to the surface when necessary, what else are you going to use the massive downmass for? Satellites can be carried down for repair if it is necessary, or for recoverable satellites.

Still, it's a much smaller mission requirement set than the Shuttle, which had to fly crewed on every mission.

Shuttle-flying the Shuttle to the ISS was like taxiing a Boeing 747 three blocks down the street to grab lunch.

[wumpus]

40 minutes ago, sevenperforce said:

The difference is that the Shuttle was expected to do most of those things at the same time. It had a crew cabin, a hab, a cargo bay, and an SRMS. It ran satellite servicing missions and satellite recovery missions and space station assembly missions. Skylon, on the other hand, doesn't need orbital persistence or service capability. It won't be expected to carry a crew and a satellite at the same time, and it won't be expected to return a payload to the surface.

Did the shuttle ever run satellite recovery runs?  While it seems to be designed to bring back a spy satellite (pretty much the only reason for half the design silliness), it never took off from Vandenburg, so presumably never entered a polar orbit and never launched or recovered spy satellites (presumably they would mention bringing back non-spy satellites).  It did service the Hubble, but nearly all that available downward cargo was wasted all 100+ trips (the internal space laboratory presumably used the space and did come back down, but that could never justify the downward mass.

[fredinno]

1 minute ago, wumpus said:

Did the shuttle ever run satellite recovery runs?  While it seems to be designed to bring back a spy satellite (pretty much the only reason for half the design silliness), it never took off from Vandenburg, so presumably never entered a polar orbit and never launched or recovered spy satellites (presumably they would mention bringing back non-spy satellites).  It did service the Hubble, but nearly all that available downward cargo was wasted all 100+ trips (the internal space laboratory presumably used the space and did come back down, but that could never justify the downward mass.

yes. https://en.wikipedia.org/wiki/STS-51-A

" Palapa B2 and Westar 6, meanwhile, had been deployed during the STS-41-B mission earlier in the year, but had been placed into improper orbits due to the malfunctioning of their kick motors; they were both safely recovered and returned to Earth during STS-51-A."

[cantab]

On 10/03/2016 at 1:17 AM, sevenperforce said:

If you're going to use an airbreather as your first stage then eschew LOX altogether and just have it accelerate in-atmo on a trajectory that will take it above the Karman line. Open internal bay, release vacuum-optimized second stage+payload, and then re-enter. 

Should be able to beat the Falcon on a very narrow range of flight profiles. 

In-air refueling is by far the simplest way to get an SSTO spaceplane though. 

Assuming you have a working SABRE or similar engine, you will get better results with a more even division of delta-V between the carrier aircraft and the orbit insertion stage.

As for in-air refuelling, I disagree. It's logistically complicated, rather hazardous, and demands extra pilot training. There's a reason that really only the military uses it. In the long run it will be simpler, safer, and cheaper to take off fully-fuelled from the ground.