Jet engine stage

[RobLynn]

You only need the thrust for a minute or two so Isp is far less important than high T/W and low cost.   So how about just throw away the very expensive gas turbine and stick a tip rocket on a fan.
http://forum.nasaspaceflight.com/index.php?topic=26225.msg786083#msg786083

A very durable, safe and damage resistant 500kg GE90 fan can produce about 500kN thrust with about 100MW input.  At 400m/s tip speed that needs tip rockets producing about 250kN, (about 11kN for each of 22 blades).  Using NOBFX rockets with 320s Isp would only need about 80kg/s of fuel, giving an effective Isp of about 600s.  And rocket power does not drop off with altitude (though fan thrust obviously does).

Increase the tip speed of the arm that the thruster is mounted on by 50% and it uses ~50% less fuel for the same fan thrust.

60 seconds worth of max thrust would be about 4800kg fuel - about the same weight as the expensive gas turbine you have eliminated.

Such fans could probably be built for <$1 million each, being subject to mass production of the blades and with lightweight pressure fed tip rockets (integral pumping) and probably a fuel cut off/control valve at the root of each blade.  Probably design as a transonic unducted fan to minimise weight and cost.

T/W probably in range 50-100:1, 600s Isp, with relatively low development costs due to modularity.  Should also be able to be designed to withstand brief exposure to 2nd stage exhaust.  High T/W and low cost means the fans can be sized to deliver sufficient thrust at higher altitudes and run at lower speed at low altitudes.

[Carreidas 160]

Taking this thread to its logical conclusion: what if we used a SABRE as a booster? Sounds awful...

Let me rephrase: what if Skylon was a 2-stage vertical takeoff vehicle with an RL-10 in the middle? (sounds even worse...) How much payload would it get to orbit?

Here's the idea: 3 cores: middle core is LOX/LH2 and goes all the way up to orbit. Outer cores are SABRE powered boosters and drop away when running out of air. Perhaps have only LH2 tank in middle core and crossfeed to boosters, or have 3 standalone LH2 tanks, one on each core. Not sure where to place the SABRE air inlet in the latter case.

(we could add 2 more SABREs for some extra punch)

SABRE has slightly higher T/W than F100/F135 and can operate at much higher machs/altitude. Also, it packs lots more punch. Instead of switching to internal LOX as on Skylon, just drop the boosters at Mach 5.5 and pray they can be recovered

[Lars_J]

How about this concept - Let's add some propeller engines to a 747 or F-22. That should improve takeoff performance, right?

If you thought that sounded stupid beyond belief - Guess what - that is how stupid the idea of adding a jet engine first stage to a rocket is.

[hkultala]

[offtopic]

How about this concept - Let's add some propeller engines to a 747 or F-22. That should improve takeoff performance, right?

If you thought that sounded stupid beyond belief - Guess what - that is how stupid the idea of adding a jet engine first stage to a rocket is.

Reminds me about B-36 - it had 6 heavy piston engines turning propellers, for takeof and cruise, and 4 jet engines, for takeoff and high speed flying.

It did not do very well and was replaced with B-52 as soon as B-52 was ready.

[/offtopic]

[DMeader]

Early models of the B-36 didn't have the jets. They were original equipment for the later models and retrofitted to the earlier ones when weight and range came to be issues.

[Andrew_W]

How about this concept - Let's add some propeller engines to a 747 or F-22. That should improve takeoff performance, right?

If you thought that sounded stupid beyond belief - Guess what - that is how stupid the idea of adding a jet engine first stage to a rocket is.

Well they've added a propeller to the F-35B to improve takeoff performance.

As Quantum G said earlier in the thread, people need to actually do some math to get a measure of the viability of a jet first stage or boosters, hunches mean nothing.

[Danderman]

Yeah, I would love someone to calculate the performance of a Centaur stage lifted to altitude by a bunch of jet engine boosters strapped on.

[simonbp]

Probably something similar to a stack of solids with a 3x turbofan first stage... 

[e of pi]

Well, in theory, a Centaur starting from about 30,000 ft and Mach 1.7 might need about 8600 m/s to hit orbit. With these numbers from Astronautix (I know, I know, always a rsiky beginning to sentence but hing with me), Centaur has a dry mass of 2,026kg, and 20,800 kg of fuel with ISp 451s. So the mass ratio for 8600 m/s can be:

exp(8600/(9.81*452))=6.99

The burnout mass is equal to fuel divided by (R-1):

20800/(6.99-1)=3472 kg

Of this mass, 2026 is the Centaur stage itself, so the amount remaining for payload is about 1470 kg. The first stage would need to be roughly two F135s, given a thrust under afterburners of 191 kN each and a total stack mass of anything between 16 and 32 tons. Given the Centaur and payload will mass about 24.3 tons, I think that's a safe estimate.

Of course, the real question isn't technical viability--it's commercial viability. Say that the two F135 come cheap (P&W promised $10m each), and the remainder of the stage costs only 50% as much as the engines, for a $30m stage. Let's be extremely generous and say this can be re-used 100 times, so about $300,000 per flight. Thus, if you can get Centaur down to under $14m, you can compete with Falcon 1 on cost. Given that RL-10 alone costs $38 million, though...

[simonbp]

Pegasus can launch 443 kg to LEO, and costs of order $20 million. I doubt you can beat that with an expendable LH2 stage...

Plus, you have to account for the mass of the wings* and payload faring, as well as the gravity losses from the Centaur's low thrust/weight ratio. Those will considerably cut into the payload number you list.

* wings reduce gravity losses, but not completely

[QuantumG]

As QuantumG said earlier in the thread, people need to actually do some math to get a measure of the viability of a jet first stage or boosters, hunches mean nothing.

Oh, you mean way back at the start of the thread when I suggested to the OP that he go design a stage that we can talk about, otherwise the thread will devolve into pointless arguments with no substance? ... Which he didn't do?

Yeah.

[RobLynn]

Well, in theory, a Centaur starting from about 30,000 ft and Mach 1.7 might need about 8600 m/s to hit orbit. With these numbers from Astronautix (I know, I know, always a rsiky beginning to sentence but hing with me), Centaur has a dry mass of 2,026kg, and 20,800 kg of fuel with ISp 451s. So the mass ratio for 8600 m/s can be:

exp(8600/(9.81*452))=6.99

The burnout mass is equal to fuel divided by (R-1):

20800/(6.99-1)=3472 kg

Of this mass, 2026 is the Centaur stage itself, so the amount remaining for payload is about 1470 kg. The first stage would need to be roughly two F135s, given a thrust under afterburners of 191 kN each and a total stack mass of anything between 16 and 32 tons. Given the Centaur and payload will mass about 24.3 tons, I think that's a safe estimate.

Of course, the real question isn't technical viability--it's commercial viability. Say that the two F135 come cheap (P&W promised $10m each), and the remainder of the stage costs only 50% as much as the engines, for a $30m stage. Let's be extremely generous and say this can be re-used 100 times, so about $300,000 per flight. Thus, if you can get Centaur down to under $14m, you can compete with Falcon 1 on cost. Given that RL-10 alone costs $38 million, though...

Wow that RL-10 price is insane, how does a 50 year old (yes I know it has been substantially updated) expander cycle engine get so expensive?  It's 277kg so $140,000 per kg?  Nearly 3x times the price of gold!  If I recall correctly back in DC-X days there was suggestion of RL-10 manufacturing cost as low as $0.5 million in large numbers.  Even the SSME is only about $25,000/kg (last number I saw was about $80 million per engine)

The White Knight II has a payload of 17,000kg to 15000m.  It would almost be able to lift Centaur, though the 2x RL-10 Centaur would be a bit underpowered for air launch.

Seems like there is potentially room in the market for a competing upper stage engine and XCOR has a slow-burn project with ULA to develop one.  Why did none of the ML60, RL-60 and RL-50 ever get completed?  At one point seemed they were well along in their development and testing.  Was it just a case of established players being happy with the quasi-monopoly RL-10?

[Jim]

Wow that RL-10 price is insane, how does a 50 year old (yes I know it has been substantially updated) expander cycle engine get so expensive? 

With a low and inefficient production rate

[hkultala]

Well, in theory, a Centaur starting from about 30,000 ft and Mach 1.7 might need about 8600 m/s to hit orbit.

 With these numbers from Astronautix (I know, I know, always a rsiky beginning to sentence but hing with me), Centaur has a dry mass of 2,026kg, and 20,800 kg of fuel with ISp 451s. So the mass ratio for 8600 m/s can be:

exp(8600/(9.81*452))=6.99

The burnout mass is equal to fuel divided by (R-1):

20800/(6.99-1)=3472 kg

Of this mass, 2026 is the Centaur stage itself, so the amount remaining for payload is about 1470 kg. The first stage would need to be roughly two F135s, given a thrust under afterburners of 191 kN each and a total stack mass of anything between 16 and 32 tons. Given the Centaur and payload will mass about 24.3 tons, I think that's a safe estimate.

Your estimate has one big fault:

Centaur lacks the thrust it needs to accelerate itself when staging so low.
RL-10 has thrust of 99 kN. This means it can lift 10 tons against earth gravity. But it weighs 23t so it has T/W ratio of under 0.45

The jet-first stage can only put the craft into ballistic trajectory that has apoapsis at 21 km. So the centaur would have to contribute quite big part of the upwards acceleration (which is cannot do very well because of the bad T/W)

This rise on free ballistic trajectory would take about 48 seconds.
But if the centaur is firing downwards(at T/W of 0.5) it would rise up to 33km,
and rise would take 96 seconds. (of course the T/W changes slightly , but even after ~100 seconds the centaur would still be way below T/W of 1).

So, we get a nice suborbital flight that goes only to 33 km. Then we start falling, with a fuel tank that has lots of fuel left.
Just before we hit the ground we might reach T/W of over 1 and start slowing down our fall.

So practically 2-engined, 170kg heavier centaur is needed. And it still won't have T/W of 1 on staging. But as the T/W will be close to 1, it can consume enough fuel to get T/W over 1 before the apoapsis, so can actually really accelerate itself upwards. And after the upwards acceleration, also horizontally, to get orbit speed.

But I claim it will need more than the 8600 m/s delta-v to reach orbit due the terrible gravity losses, and as it will be 170 kg heavier, payload will be smaller too.

8.9 km/s delta-v and 2 RL-10's would mean payload of about 1030 kg,
but I think even the 8.9 km/s is too optimistic.


Centaur really needs a first stage that can stage at much higher speed and/or much higher.

[Danderman]

I appreciate the data and the analysis.

I would suggest more and cheaper jet engines. I don't know where the requirement for F135s came from, but since the jet engines are ground launched, there is no problem with using more engines, and NOT focusing on thrust-to-weight ratios.