Single stage to orbit, Atlas
Jan Philips
read that with the current (at that time) fuels and minimum weights
for tanks and engines, a single stage to Earth orbit rocket was
impossible. But the Atlas came close.
Now it is clear that dropping off a stage of empty tanks (and their
engines) helps. But the Atlas only dropped off the two outer engines,
right, (no tanks). Empty tanks are clearly dead weight, but not
engines. So intuitively it seems that nothing would be gained by
dropping off two thrust-producing engines, but no dead-weight tanks.
Can someone explain why this "intuition" is wrong?
Thanks, as always.
Ned Pike
news:8o5qmvko5hquhuqkr...@4ax.com...
Because the tanks that fed the dropped engines were the same tanks that fed
the sustainer engine. Atlas was a stressed-skin design, where the tanks
were part of the main load-bearing structure. If the tanks lost pressure,
the whole stack lost its structural integrity. There were no tanks to
jettison on the way up.
Jan Philips
<cap...@nospam.adelphia.net> wrote:
>Because the tanks that fed the dropped engines were the same tanks that fed
>the sustainer engine. Atlas was a stressed-skin design, where the tanks
>were part of the main load-bearing structure. If the tanks lost pressure,
>the whole stack lost its structural integrity. There were no tanks to
>jettison on the way up.
I'm sorry, I didn't make myself clear. I know the original Atlas had
only one set of tanks, and thus none to drop. My question is why drop
the two outer engines? Why not let them continue to provide thrust?
In other words, why is it that the Atlas could not put a payload into
Earth orbit if it didn't drop the two outer engines?
Ned Pike
If it kept the outer engines, Altas wouldn't have enough delta-v to make
orbit. Atlas was designed the way it was. The propellant tanks didn't have
the capacity to keep all 3 engines burning into orbit.
Actually, Atlas is considered a design paradigm. Dropping the outer 2
engines and keeping the tanking did wonders for the booster's mass fraction.
Look at all the other staged boosters we've built. All except Atlas
(handwave) have dropped tanks as well as engines. The entire Atlas design
concept was for low booster dry weight, achieved mainly through the
"stage-and-a-half" design.
JNICHOLS
The thrust of three engines were only needed for launch (much like the
boosters on the shuttle). Once through the denser layers of the atmosphere,
and the tank now lightened by the burning of fuel, the 1/2 stage becomes
somewhat of dead weight. At some point between launch and orbit the rocket
becomes more efficient with a single engine, than with three engines.
If the engines have the same thrust and use the same amount of fuel per
second, you have a rocket with three times the thrust, but with only a third
of the burn time. This three engine rocket is heavier also (dead weight).
With one engine the rocket will only have a third the thrust, but three
times the burn time. This rocket is lighter. This rocket is more efficient
(higher top speed or more payload).
BTW I think it was "Project Score" that put the "main" stage in orbit.
Pat Flannery
JNICHOLS wrote:
> BTW I think it was "Project Score" that put the "main" stage in orbit.
>
>
>
>
>
>
Yes, that was the one- and its orbiting came as quite a surprise to the
ground tracking crew, who had not been told that that was the intention
of the mission. Only those who had a "need to know were" clued in; and
some fancy and secretive work was done on the guidance system
surreptitiously.
Pat
Doug...
jud...@bellsouth.net says...
You've gotten several good answers, but none of them address the central
fact that you're carrying a limited amount of fuel and oxidizer.
Think of it this way -- you're getting where you're going by burning
kerosene and liquid oxygen, right? The total amount of energy in your
rocket is determined by the amount of fuel it carries, not by the number
of engines burning it. It takes more thrust to accelerate at the
beginning of the flight, when you're lifting all the fuel that you're
going to burn later. But there comes a point, after a fair amount of
your fuel and oxidizer are burned and gone, where you will accelerate
just fine using only one engine.
The tank itself was the lightest part of the rocket. By putting all the
fuel and oxidizer in one big tank each, they saved weight -- two tanks
weigh less than four, or eight. On the other hand, the engines were the
heaviest part of the rocket. Dumping two of the three engines probably
reduced the weight of the rocket structure by something like a quarter to
a half.
If you don't drop those two engines, you're using the remaining energy in
your rocket (the remaining fuel and oxidizer) to accelerate a
significantly larger mass. That means your ability to change its
velocity (i.e., delta-V) is reduced. In the case of the Atlas, the fuel
it carried didn't provide enough energy to put the whole structure, with
all three engines, into orbit. But it did have enough energy to get the
structure with only one engine attached into orbit.
--
Do not meddle in the affairs of dragons, for | Doug Van Dorn
thou art crunchy and taste good with ketchup | dvan...@mn.rr.com
JNICHOLS
"Pat Flannery" <fla...@daktel.com> wrote in message
news:3F6D4EAF...@daktel.com...
>
>
> Yes, that was the one- and its orbiting came as quite a surprise to the
> ground tracking crew, who had not been told that that was the intention
> of the mission. Only those who had a "need to know were" clued in; and
> some fancy and secretive work was done on the guidance system
> surreptitiously.
>
> Pat
>
I will have to reread a chapter of John Chapman's "ATLAS The Story of a
Missile", but I think that it had a dummy warhead that was replaced with a
lighter instrument package, just the night before launch.
JNICHOLS
"Doug..." <dvan...@mn.rr.com> wrote in message
news:MPG.19d72665a...@news-server.mn.rr.com...
> fact that you're carrying a limited amount of fuel and oxidizer.
>
> Think of it this way -- you're getting where you're going by burning
> kerosene and liquid oxygen, right? The total amount of energy in your
> rocket is determined by the amount of fuel it carries, not by the number
> of engines burning it. It takes more thrust to accelerate at the
> beginning of the flight, when you're lifting all the fuel that you're
> going to burn later. But there comes a point, after a fair amount of
> your fuel and oxidizer are burned and gone, where you will accelerate
> just fine using only one engine.
>
> The tank itself was the lightest part of the rocket. By putting all the
> fuel and oxidizer in one big tank each, they saved weight -- two tanks
> weigh less than four, or eight. On the other hand, the engines were the
> heaviest part of the rocket. Dumping two of the three engines probably
> reduced the weight of the rocket structure by something like a quarter to
> a half.
>
> If you don't drop those two engines, you're using the remaining energy in
> your rocket (the remaining fuel and oxidizer) to accelerate a
> significantly larger mass. That means your ability to change its
> velocity (i.e., delta-V) is reduced. In the case of the Atlas, the fuel
> it carried didn't provide enough energy to put the whole structure, with
> all three engines, into orbit. But it did have enough energy to get the
> structure with only one engine attached into orbit.
That's what I said, but Doug has a slightly larger vocabulary.
--
People are more violently opposed to fur than leather because it's safer to
harass rich women than motorcycle gangs.
-Unknown
Jan Philips
wrote:
> The thrust of three engines were only needed for launch (much like the
>boosters on the shuttle). Once through the denser layers of the atmosphere,
>and the tank now lightened by the burning of fuel, the 1/2 stage becomes
>somewhat of dead weight. At some point between launch and orbit the rocket
>becomes more efficient with a single engine, than with three engines.
> If the engines have the same thrust and use the same amount of fuel per
>second, you have a rocket with three times the thrust, but with only a third
>of the burn time. This three engine rocket is heavier also (dead weight).
> With one engine the rocket will only have a third the thrust, but three
>times the burn time. This rocket is lighter.
OK, I understand now. Thanks, everyone.
Doug...
jnic...@sport.rr.com says...
LOL -- I know it's what you said, and I credited you and others in my
first statement. But you didn't emphasize the point that the rocket only
had 'x' energy available, in the form of 'x' amount of fuel and oxidizer,
for the whole flight.
I think the problem some people have is equating thrust with energy. If
you have one engine, you have a certain amount of energy available, but
if you have three engines, you ought to have twice again that amount
available, right? Wrong. You only have twice again the amount of thrust
-- your total energy available is still the same. That's where the
"common sense" approach fails.
Pat Flannery
JNICHOLS wrote:
> I will have to reread a chapter of John Chapman's "ATLAS The Story of a
>Missile",
>
You've got that one too!?
> but I think that it had a dummy warhead that was replaced with a
>lighter instrument package, just the night before launch.
>
A receiver/tape recorder/transmitter that broadcast President
Eisenhower's Christmas message to the world:- now, if instead of: "This
is the President of the United States speaking. Through the marvels of
scientific advance, my voice is coming to you from a satellite circling
in outer space. My message is a simple one. Through this unique means I
convey to you and all mankind America's wish for peace on earth and good
will to men everywhere."....someone had sent up a message that went more
like this....
"Hello world, this is President Eisenhower coming to you from outer
space via an Atlas rocket...I hope you all have a merry Christmas, and
that all the boys and girls of the world have been good, and don't get
coal instead of nice presents in your Christmas stockings.
But unfortunately not everybody in this world has been a good person
during the last year...for instance...Vice President Nixon has been a
sneaky conniving little son-of-a-bitch, and deserves to get Checker's
dog poop in his Christmas stocking... this vicious little bastard could
inflict infinite harm on the American Dream if he were ever to rise to
the position of President; and even though he says that he hates the
Communist Chinese as much as I do, I'm willing to bet my five stars he
would kiss their red asses if he ever thought he could make a little
political hay from it, so great and reprehensible is his desire for
personal glory at all costs! So don't let him do that to our world, boys
and girls! And watch out for that military-industrial complex that made
this God-damned overpriced rocket! They, like Nixon, are out to screw
you in the brain, and in the pocketbook! Merry Christmas!"
Ah, if only...
Pat
Henry Spencer
Jan Philips <jud...@bellsouth.net> wrote:
>When I was a kid in the 60s, reading books about space and rockets, I
>read that with the current (at that time) fuels and minimum weights
>for tanks and engines, a single stage to Earth orbit rocket was
>impossible.
That was the common wisdom, but people who'd looked closely at the problem
knew it was wrong even then. Several 60s rocket stages had performance
and mass ratios adequate for SSTO, if you didn't want much payload and
didn't insist on reusability. (The Titan II first stage is the most
spectacular example -- not even balloon tanks, and it could put a few tons
of payload into LEO, if you could somehow reduce [!] its engine thrust.)
>But the Atlas came close.
It would have been possible to put an Atlas variant into orbit without the
engine jettison, if you were willing to accept a very low orbit and zero
payload. The same could have been done with a Delta variant, or an S-IC
variant. Nobody was willing to take the trouble for what would have been
basically a stunt.
>Now it is clear that dropping off a stage of empty tanks (and their
>engines) helps. But the Atlas only dropped off the two outer engines,
>right, (no tanks).
Yes and no. Atlas didn't drop any tanks, no, but the Atlas tankage was so
lightweight that dropping tanks wouldn't have made much difference. For
practical purposes, Atlas was a two-stage rocket; it dropped almost all of
the mass of a complete stage.
(Incidentally, some other things went with the engines. Notably, the tank
pressurization system did; after that, Atlas just relied on the pressurant
gas left in the tanks, augmented by hydrostatic head from acceleration.)
>Empty tanks are clearly dead weight, but not
>engines. So intuitively it seems that nothing would be gained by
>dropping off two thrust-producing engines, but no dead-weight tanks.
Engines do become dead weight eventually. By the time (fairly late in
flight) when Atlas staging occurred, Atlas had far more thrust than it
needed. Much of that thrust was required primarily to get the fully
loaded rocket off the ground and clear of the atmosphere quickly; at very
high altitude, with the tanks largely empty, the need for haste was past
and the extra engines were no longer contributing enough to make them
worthwhile. There was also some optimization of the outer engines for low
altitude and the center engine for high altitude; they were not identical.
--
MOST launched 1015 EDT 30 June, separated 1046, | Henry Spencer
first ground-station pass 1651, all nominal! | he...@spsystems.net
Henry Spencer
Ned Pike <cap...@nospam.adelphia.net> wrote:
>...The entire Atlas design
>concept was for low booster dry weight, achieved mainly through the
>"stage-and-a-half" design.
That, and the "balloon" tanks, which consistently performed better than
even their proponents' fondest hopes.
Atlas also had a requirement, by the way, to have all engines firing at
liftoff. At the time, there was some concern about the reliability of
high-altitude engine ignition. (It was difficult enough to get reliable
ignition at sea level!)
Henry Spencer
JNICHOLS <jnic...@sport.rr.com> wrote:
>> [Project SCORE]
>
> I will have to reread a chapter of John Chapman's "ATLAS The Story of a
>Missile", but I think that it had a dummy warhead that was replaced with a
>lighter instrument package, just the night before launch.
Not quite exactly...
On Atlas, things like instrumentation did *not* ride in the nose. The two
flat fairings along the sides held all the electronics and such, and it
was there that some last-minute substitutions were done, both to include
the tape-recorder package and to lighten ship a bit.
The nose of the SCORE Atlas was just a sheet-metal cone, not a warhead.
This *wasn't* something that could be done at the last minute -- the cone
was welded on -- but some of the test Atlases also had the cone nose, so
it wasn't itself obviously special.
(By the way, when I say "a sheet-metal cone", I mean that quite literally.
No supports, no structure -- just a cone of thin stainless-steel sheet
welded on the front. There wasn't even any deliberate pressurization; the
welding made it airtight, and 1atm of air pressurized it well enough.)
Probably the best account of Project SCORE is the paper in the July/Aug
1999 JBIS, a detailed personal account by one of the guys who did it.
Michael Walsh
Jan Philips wrote:
The main problem with your intuition is that the Atlas dropped not
only two engines but a sizable thrust structure. The first stage engines
were optimized for efficiency at lower levels of the atmosphere. The
sustainer engine that was kept was lower thrust and quite a bit
over-expanded at sea-level but was much more efficient under
vacuum conditions.
You need the high thrust engines at liftoff just to get the missile
off the ground. As the fuel burns off and when the heavy first stage
thrust structure is dropped less thrust is required so the more propellant
efficient single remaining engine is more optimum.
The Atlas tanks were very low weight.
On early Atlas vehicles the weight of the thrust structure and engines
that was dropped was about the same as the dry weight of the
remaining vehicle.
Mike Walsh
Jan Philips
wrote:
>That was the common wisdom, but people who'd looked closely at the problem
>knew it was wrong even then. Several 60s rocket stages had performance
>and mass ratios adequate for SSTO, if you didn't want much payload and
>didn't insist on reusability. ...
Thank you, I was completely unaware of that.
>Engines do become dead weight eventually. By the time (fairly late in
>flight) when Atlas staging occurred, Atlas had far more thrust than it
>needed. Much of that thrust was required primarily to get the fully
>loaded rocket off the ground and clear of the atmosphere quickly; at very
>high altitude, with the tanks largely empty, the need for haste was past
>and the extra engines were no longer contributing enough to make them
>worthwhile.
Thanks, I understand it now.
Jan Philips
<mp1w...@Adelphia.net> wrote:
>The main problem with your intuition is that the Atlas dropped not
>only two engines but a sizable thrust structure. ...
Thanks.
>The Atlas tanks were very low weight.
Did that contribute to some of the early failures? ISTR that they
thickened the walls after a while.
Reed Snellenberger
news:HLL27...@spsystems.net:
> payload and didn't insist on reusability. (The Titan II first stage
> is the most spectacular example -- not even balloon tanks, and it
> could put a few tons of payload into LEO, if you could somehow reduce
> [!] its engine thrust.)
>
Just had a mental image of a set of thrust reversers latched to the rear
end of a Titan II... would have been impressive to watch, anyways...
--
Reed
Jan Philips
>... On the other hand, the engines were the
>heaviest part of the rocket. Dumping two of the three engines probably
>reduced the weight of the rocket structure by something like a quarter to
>a half. ...
Actually it was more than that (of the empty weight). Astronautix
gives the mass of the Atlas D half stage as 3050 kg and the empty mass
of the rest of the rocket as 2347 kg, so the half stage was 56.5% of
the total empty weight! (Of course there was still fuel in the tanks
the I'm not taking into account.) The initial total mass was over
117,000 kg, but I don't know how much of that 112,000 kg of fuel is
left at staging.
Jan Philips
wrote:
> BTW I think it was "Project Score" that put the "main" stage in orbit.
I think I read that even before Sputnik, an Atlas could have gone into
orbit. They were testing re-entry (for warheads) and they put ballast
in them to keep them from going into orbit, so as to not have
something "inadvertently" in orbit (saving that for a planned
satellite).
Henry Spencer
Reed Snellenberger <rsnellenberger@-omitthis-houston.rr.com> wrote:
>> ... (The Titan II first stage
>> is the most spectacular example -- not even balloon tanks, and it
>> could put a few tons of payload into LEO, if you could somehow reduce
>> [!] its engine thrust.)
>
>Just had a mental image of a set of thrust reversers latched to the rear
>end of a Titan II...
Heh. Alas, that wouldn't work -- you need to reduce the fuel consumption
along with the thrust. There's a minor problem with the high thrust
giving brutally high acceleration just before cutoff, but the big problem
is that the burn time is too short. The bulk of the acceleration has to
be done horizontally at high altitude, since horizontal velocity is what
you need to reach orbit, but the Titan II first stage goes through its
fuel too quickly for that -- a lot of the fuel would be expended fairly
inefficiently, before the stage clears the atmosphere.
Now mind you, there ought to be no fundamental problem in building a
*less* powerful engine system with the same mass. :-) You really want
it to be throttlable, too, for reducing acceleration loads. The engine
it's got doesn't lend itself to such changes, I'm told, but doing it
with a new engine is not a big deal.
For bonus points, contract with the Russians for a throttlable RD-253 --
that's the Proton first-stage engine, and it's about the right thrust and
is substantially lighter. Moreover, it has about 30s higher Isp (!) than
the Titan engines, which gives you quite a bit of delta-V margin.
Pat Flannery
Jan Philips wrote:
>I think I read that even before Sputnik, an Atlas could have gone into
>orbit. They were testing re-entry (for warheads) and they put ballast
>in them to keep them from going into orbit, so as to not have
>something "inadvertently" in orbit (saving that for a planned
>satellite).
>
>
I don't think that that would be the case; the original Atlas test
version was the Atlas A; and it had only the two outer booster motors
for propulsion- which didn't give the ability to reach orbit. The
orbital capability only emerged with Atlas B which had all three motors-
this first flew on July 9th, 1958.
There was speculation that von Braun's Army team might "accidentally"
launch a satellite with their Jupiter-C test vehicle to upstage the
Soviets (and even more importantly the Navy), but it was made clear to
them that such a launch would get them in _big_ trouble.
Pat
Henry Spencer
Jan Philips <jud...@bellsouth.net> wrote:
>>The Atlas tanks were very low weight.
>
>Did that contribute to some of the early failures? ISTR that they
>thickened the walls after a while.
The walls got thicker as the hardware sitting on top of the Atlas got
heavier over time and hence tank pressures had to be raised, but not in
response to actual failures. The balloon tanks were far tougher than
anyone had expected.
(There was some suspicion that the loss of an unmanned Mercury test had
been a tank failure, and NASA insisted on adding tank reinforcements...
but the reinforcements were removed later, after flight data confirmed the
Atlas guys' belief that the failure had been in the spacecraft adapter,
not the rocket tanks.)
Convair/GD was convinced that the tanks had much greater margins than
expected, but had a hard time proving it. They convinced NASA to let them
instrument a couple of production space-launch Atlases to get real-life
load data... and by sheer chance, the second one had a worst-case failure,
a booster engine out near maximum dynamic pressure. The *instrumentation*
hit its limits and stopped reporting as the loads passed 150% of design
ultimate strength, with the tanks still intact. The tanks never did fail;
the nose fairing broke off, and Range Safety said "enough!" and pushed the
destruct button.
Henry Spencer
Pat Flannery <fla...@daktel.com> wrote:
>There was speculation that von Braun's Army team might "accidentally"
>launch a satellite with their Jupiter-C test vehicle to upstage the
The Soviets didn't enter the picture, as far as US planning went. The
Navy, yes. :-)
>but it was made clear to
>them that such a launch would get them in _big_ trouble.
And even so, the Vanguard guys were half-expecting it to happen.
ed kyle
>
> Now it is clear that dropping off a stage of empty tanks (and their
> engines) helps. But the Atlas only dropped off the two outer engines,
> right, (no tanks). Empty tanks are clearly dead weight, but not
> engines. So intuitively it seems that nothing would be gained by
> dropping off two thrust-producing engines, but no dead-weight tanks.
>
> Can someone explain why this "intuition" is wrong?
>
Dropping the 3.05 metric ton booster package (43% of the total
vehicle dry mass) provided two benefits. First, it improved
the mass fraction, providing more delta V. Second, shutting
down the powerful booster engines was needed to limit the
acceleration (G-loading) experienced by the vehicle.
According to astronautix.com, the Atlas C (the one that launched
itself into orbit for Project SCORE), mass budget was:
Core Sustainer Stage: 3.98 metric tons empty
107.61 tons propellants
Booster Package: 3.05 tons
and the propulsion system numbers were:
Booster Engine: 176.8 tons thrust (SL) 191.77 tons thrust (vac)
ISP 248 sec (SL), 282 sec (vac)
Sustainer Engine: 31.26 tons thrust (SL), 37.04 tons thrust (vac)
ISP 309 sec (vac)
If the booster package would have stayed attached during the
entire flight, the total no-payload ideal delta-V would have
been (assuming an average 282 sec ISP) about 7590 meters per
second. With booster package jettison, the vehicle could
achieve an ideal no-payload delta-v of about 9135 meters per
second (4642 m/s during boost phase and 4493 m/s during
sustainer phase).
By the end of its 135 second boost phase, an Atlas C would
have burned about 90 tons of propellant. Its mass at that
point would have been about 20.7 tons while its thrust would
have been 191.77 tons, providing more than 9 G of rapidly
increasing acceleration. Shutting down the booster engines
and dropping the booster package changed the vehicle into a
17.54 ton machine powered by 37.04 tons of thrust (only 2 G).
That acceleration would have increased back to only a bit
more than 9 G again by the time the sustainer engine shut
down four minutes after liftoff.
BTW, an Altas V common core booster first stage could do
*almost* as well as an Atlas C (about 8700 m/s ideal
delta-v) without dropping any engine hardware. The
Delta IV common core booster could probably provide about
the same result.
- Ed Kyle
Michael Walsh
Jan Philips wrote:
No, and I don't believe they thickened the walls, unless some of the
later versions had higher tank pressures.
I note that I am speaking about flight failures. The thin tanks that
required pressurization to keep them from collapsing suffered a number
of failures on the ground. Generally because of either loss of pressure
or overpressurizing the LO2 tank and inverting the hemispherical part
of the lower fuel tank that formed the intersection between the LO2
and RP-1 tanks.
In order to depressurize the Atlas tanks it was necessary to put them
in "stretch" where both ends of the tank were rigidly physically
attached
to prevent the thin tank skin from folding in on itself.
Mike Walsh
Jochem Huhmann
> In order to depressurize the Atlas tanks it was necessary to put them
> in "stretch" where both ends of the tank were rigidly physically
> attached to prevent the thin tank skin from folding in on itself.
Do we agree that this would be not practical for a reusable craft (and
SSTO in an ELV would be a quite useless stunt)?
Jochem
--
"A designer knows he has arrived at perfection not when there is no
longer anything to add, but when there is no longer anything to take
away." - Antoine de Saint-Exupery
Michael Walsh
Jochem Huhmann wrote:
> Michael Walsh <mp1w...@Adelphia.net> writes:
>
> > In order to depressurize the Atlas tanks it was necessary to put them
> > in "stretch" where both ends of the tank were rigidly physically
> > attached to prevent the thin tank skin from folding in on itself.
>
> Do we agree that this would be not practical for a reusable craft (and
> SSTO in an ELV would be a quite useless stunt)?
>
> Jochem
Well. the new Atlas 5 no longer has the "balloon" tanks.
However, it still doesn't make it reusable.
Mike Walsh
Peter Stickney
Score's orbiting came as a big surprise to the launch crew, too. In
order to maintain the secrecy of the project (Is not just the Premeire
who like surprises Comrade Patrik Mikhailovith!), and to disguise just
how much stuff they'd stripped out, the USAF built a duplicate to the
LCC. which mirrored the status consoles, and had the ability to feed
the appropriate signals to the real LCC. The Launch Director was also
in on the deal, of course, and kept the count going even though
certain sensors & such that should have (Or should've) stopped teh
whole party. According to a guy I know who was a Cape Canaveral
Telemetry Guy and Range Rat through the 1950s, the Range Safety
Officer also wasn't in on the plot. As the missile started deviating
from the "planned" course, he'd already lifted the cover on the
destruct key when the Air Force Project Officer restrained him.
--
Pete Stickney
A strong conviction that something must be done is the parent of many
bad measures. -- Daniel Webster
Peter Stickney
It's not quite outside of the realm of possibility. I think the
system that you're thinking of, though, was the Redstone Arsenal's
Jupiter C, which was developed from the Army's Redstone MRBM to test
the new ablative reentry vehicles for the Jupiter IRBM. It was well
known that with a slight change of trajectory, the Jupiter C could
place its 4th stage in orbit. In fact, after the officially
sanctioned Vanguard suddenly developed a case of blowing up on the pad
(Vanguard development up to that point had been remarkable smooth),
Werner von Braun and General Medaris, the COmmander of the Army
Ballistic Missile Agency, received official santion to dust off the
mothballed Jupiter Cs, and use them as satellite launchers. 90 days
later, Hey Presto! Explorer I.
Jan Philips
Stickney) wrote:
>It's not quite outside of the realm of possibility. I think the
>system that you're thinking of, though, was the Redstone Arsenal's
>Jupiter C, which was developed from the Army's Redstone MRBM to test
>the new ablative reentry vehicles for the Jupiter IRBM.
Yes, after thinking about it, I think it was the Redstone/Jupiter C.
>sanctioned Vanguard suddenly developed a case of blowing up on the pad
I thought only one Vanguard blew up before the successful Explorer 1.
> 90 days later, Hey Presto! Explorer I.
IIRC, they said they could do it in 90 days - they actually did it in
63.
Jan Philips
<judmccr...@bellsouth.net> wrote:
>I thought only one Vanguard blew up before the successful Explorer 1.
>
>> 90 days later, Hey Presto! Explorer I.
>
>IIRC, they said they could do it in 90 days - they actually did it in
>63.
The first orbital attempt of Vanguard blew up Dec 6, 1957. The
successful Explorer 1 was on Jan 31, 1958 - 56 days after Vanguard, so
I don't know where I got the 63 days. Another Vanguard blew up Feb 5,
1958.
Henry Spencer
Jochem Huhmann <j...@gmx.net> wrote:
>> In order to depressurize the Atlas tanks it was necessary to put them
>> in "stretch" where both ends of the tank were rigidly physically
>> attached to prevent the thin tank skin from folding in on itself.
>
No, we don't. :-) As with an expendable, it is a complication that a
designer might not wish to incur -- ground handling is more complex and
the range of ground mishaps that can lead to structural damage is rather
wider -- but there is nothing inherently ridiculous about it.
(Ground handling is really the only issue. Nearly all launchers rely on
tank pressurization for at least part of their structural strength in
flight.)
>(and SSTO in an ELV would be a quite useless stunt)?
Not at all. If I were asked to design an expendable, I'd think very hard
about making it a single stage. The price is a need for lightweight
structure and engines, either multiple (>2) engines or deep throttling,
and iron-fisted control of the mass budget (plus possibly some difficulty
convincing skeptical investors that it can be done). The payoff is a much
simpler vehicle with many fewer failure modes, and the disappearance of
the "where do the spent stages fall?" problem.
(By the way, you don't need balloon tanks to do an expendable SSTO,
although they might help.)
Brett Buck
> In article <871xu8m...@nova.revier.com>,
> Jochem Huhmann <j...@gmx.net> wrote:
>
>>(and SSTO in an ELV would be a quite useless stunt)?
>
>
> Not at all. If I were asked to design an expendable, I'd think very hard
> about making it a single stage. The price is a need for lightweight
> structure and engines, either multiple (>2) engines or deep throttling,
> and iron-fisted control of the mass budget (plus possibly some difficulty
> convincing skeptical investors that it can be done).
I think you would find that the extra cost would be prohibitive. As
I am sure you know, the last 5% of performance costs 95% of the money
and effort. You just wind up with huge and expensive, rocket with a
comparatively tiny payload and razor-thin margins in a lot of places.
I think (and current successful programs bear this out) that you
wind up better off in most ways by not designing it to require heroic
measures, sticking with something relatively easy to produce, and taking
your hit on the part count. SSTO *could* have been done for the last 30
years - but no one does, and Russians (deified by many Internet Space
Buffs, for right or wrong) took pretty much the diametrically opposite
approach.
Brett
Henry Spencer
Peter Stickney <pe...@adelphia.net> wrote:
>...In fact, after the officially
>sanctioned Vanguard suddenly developed a case of blowing up on the pad
Except that it hadn't really progressed all that far. The big pad blowup
was the very first complete Vanguard, including the first flight of the
second stage, and most of the people involved thought it grossly premature
to pin great hopes on it. They weren't given a choice...
>Werner von Braun and General Medaris, the COmmander of the Army
>Ballistic Missile Agency, received official santion to dust off the
>mothballed Jupiter Cs, and use them as satellite launchers.
And in fact, they had very quietly started preparations in anticipation
of being given official approval.
Henry Spencer
Jan Philips <judmccr...@bellsouth.net> wrote:
>>IIRC, they said they could do it in 90 days - they actually did it in
>>63.
>
>The first orbital attempt of Vanguard blew up Dec 6, 1957. The
However, the go-ahead for Explorer 1 was given on 7 Nov 1957 (a few days
after Sputnik 2 was launched). It took 85 days. Mind you, much of that
was for the satellite rather than the launcher -- Washington stipulated
that the satellite must carry at least one IGY-related science instrument,
so the Huntsville/JPL crew couldn't just launch a beeping transmitter.
Jan Philips
wrote:
>>The first orbital attempt of Vanguard blew up Dec 6, 1957. The
>>successful Explorer 1 was on Jan 31, 1958 - 56 days after Vanguard...
>
>However, the go-ahead for Explorer 1 was given on 7 Nov 1957 (a few days
>after Sputnik 2 was launched).
OK, I thought the go-ahead on Explorer was given after the Vanguard
failure.
Henry Spencer
Brett Buck <buc...@pacbell.net> wrote:
>>>(and SSTO in an ELV would be a quite useless stunt)?
>> Not at all. If I were asked to design an expendable, I'd think very hard
>> about making it a single stage. The price is a need for lightweight
>> structure and engines, either multiple (>2) engines or deep throttling,
>> and iron-fisted control of the mass budget (plus possibly some difficulty
>> convincing skeptical investors that it can be done).
>
> I think you would find that the extra cost would be prohibitive.
I don't think so, except possibly that last item (convincing investors).
Contrary to popular belief, this isn't some leading-edge-of-technology
stunt; all it takes is good design. The right sort of engine performance
and mass ratio were achieved in several 1960s US rocket stages, and we
have better engines (from the Russians), better materials, and better
and lighter control systems.
>...You just wind up with huge and expensive, rocket with a
>comparatively tiny payload and razor-thin margins in a lot of places.
Put six SSMEs under a (pre-aluminum-lithium!) shuttle ET, and it can reach
orbit with 60klb of payload. How does that qualify as "comparatively tiny
payload" and "razor-thin margins"?
I wouldn't use that approach for a real design -- the SSME is too
expensive and NASA controls the production lines for that hardware --
but it illustrates what's possible with 1970s hardware.
(By the way, concerning "comparatively tiny payload", note that Russian
launchers have about half the payload of comparably-sized US launchers,
and are nevertheless the cheapest way to launch things. What matters is
how much the rocket costs, not how big it is. The two are only tenuously
related.)
> I think (and current successful programs bear this out) that you
>wind up better off in most ways by not designing it to require heroic
>measures, sticking with something relatively easy to produce...
Yes, that's what I'm talking about. It is a *MYTH* that SSTO requires
"heroic measures" -- sheer superstition, totally unsupported by facts.
Expendable SSTO simply isn't that hard. The only reason it wasn't done
decades ago is that nobody has tried.
>and taking your hit on the part count.
That's where the cost soars and the reliability plummets. "Current
successful programs" mostly have nowhere near the concern for either
issue that I would think appropriate.
>SSTO *could* have been done for the last 30
>years - but no one does...
Quite so. I'm not interested in designing another 1950s missile
derivative. To get a competitive edge, you have to do something
different.
Henry Spencer
Jan Philips <judmccr...@bellsouth.net> wrote:
>>However, the go-ahead for Explorer 1 was given on 7 Nov 1957 (a few days
>>after Sputnik 2 was launched).
>
>OK, I thought the go-ahead on Explorer was given after the Vanguard
>failure.
Confidence in Vanguard was not high, and with US embarrassment mounting
steadily, having a backup was seen as reasonable. The Vanguard guys had
hopes of beating the Army even so, since they had one launch attempt
imminent and would probably get a second try before the Army was ready.
Vanguard's first attempt was the spectacular failure. The second ended up
slipping past the Army's successful first attempt, due to a combination of
weather problems and technical snags, and then it failed too. They did
manage orbit on the third try.
(Both launchers actually were fairly unreliable, based on the later record.
Vanguard was unlucky, and the Army guys were lucky.)
Jan Philips
wrote:
>(Both launchers actually were fairly unreliable, based on the later record.
>Vanguard was unlucky, and the Army guys were lucky.)
Yes, Vanguard 4 successes out of 12; Jupiter C: 4 out of 9.
Kevin Willoughby
> The balloon tanks were far tougher than anyone had expected.
Then why are balloon tanks so rare today?
--
Kevin Willoughby ke...@scispace.org.invalid
Imagine that, a FROG ON-OFF switch, hardly the work
for test pilots. -- Mike Collins
Henry Spencer
Jan Philips <judmccr...@bellsouth.net> wrote:
>>(Both launchers actually were fairly unreliable, based on the later record.
>>Vanguard was unlucky, and the Army guys were lucky.)
>
>Yes, Vanguard 4 successes out of 12; Jupiter C: 4 out of 9.
Not quite right, unless I've missed something...
Vanguards TV-0 through TV-2 were not orbital launch attempts and were not
full Vanguards. TV-O was a Viking, used to check out some electronics and
range facilities. TV-1 was also a Viking -- the very last -- but with a
Vanguard third stage on top, to test the third stage and its spin table.
TV-2 had a live Vanguard first stage but everything else was dummies; a
number of the engineers thought this was a complete waste and tried to get
live second and third stages onto it, unsuccessfully.
TV-3, TV-3BU, TV-4, and TV-5 were full Vanguards and orbit attempts, as
were SLV-1 through SLV-7, for a total of eleven, of which three (TV-4,
SLV-4, and SLV-7) succeeded.
Similarly, you have to be careful about what you call a Jupiter C. There
were only six launches of the full four-stage Jupiter C (aka Juno I), all
of them orbit attempts, three of them successful.
Henry Spencer
Kevin Willoughby <ke...@scispace.org.invalid> wrote:
>> The balloon tanks were far tougher than anyone had expected.
>
>Then why are balloon tanks so rare today?
Partly because they *are* more hassle to handle, and more vulnerable to
ground-handling damage due to procedural errors.
Partly because they are a bit less versatile. LockMart switched to
non-balloon tanks for Atlas V mostly to make it possible to bundle three
of the core stages together for Atlas V Heavy.
And to a very large extent because of sheer superstition and prejudice.
Convair (later General Dynamics) had an uphill battle all the way to get
them accepted...
The Titan program originally came about largely because influential people
in the USAF thought Atlas's balloon tanks were never going to work, and
hence a backup was required.
For Atlas (and no other large rocket) the USAF insisted on a "burst test",
putting a bullet into the fuel tank of a pressurized Atlas prototype to
find out what would happen. (In fact, nothing much -- the tank vented,
the intertank bulkhead inverted and ruptured, but the Atlas didn't even
collapse, much less explode! It wrinkled a bit but stayed upright. To
make it actually burst, they ended up using explosives to rip the tanks
apart.)
As I noted, balloon tanks instantly got the blame when the MA-2 Mercury
flight test failed, even though the problem turned out to be in the
spacecraft adapter.
Both the Saturn program and the Shuttle program ground-ruled out balloon
tanks, simply because NASA did not trust them, no reason offered.
Pat Flannery
Kevin Willoughby wrote:
>In article <HLLMn...@spsystems.net>, he...@spsystems.net says...
>
>
>>The balloon tanks were far tougher than anyone had expected.
>>
>>
>
>Then why are balloon tanks so rare today?
>
>
The big problem was that you had to keep them pressurized at all times,
or the rocket would collapse under its own weight. In the case of Atlas,
this was first done with nitrogen; then later with helium to make the
stage lighter during transport- but the whole thing was a pain in the
ass from an operational point of view.
Pat
Encyclopedia Astronautica
someone could have demonstrated it long ago (think of Bono's SASSTO,
an incremental improvement of the Saturn S-IVB). Once you had
something demonstrated, you could start tweeking it -- extendable
nozzle to boost Isp, understand where you could safely cut weight,
introduce incrementally new technologies and eventually reusability.
This is the kind of risk reduction NASA of USAF should have done with
relatively little cost (not great leaps forward like the shuttle or
X-33). Then investors or Congress might have been willing by now to
put up the cash for a commercial reusable SSTO.
The aircraft designer Ed Heinemann (who was certainly the greatest
aircraft engineer when it came to weight concerns -- vis. A-4, A-3)
once said that no one had ever looked at reducing launch vehicle
weight with a clean sheet of paper. The only instance I can think of
was Skylon (a hardback single spine with a suspended tank and very
light aeroshell). Most others just seem to hope that 'new materials'
will solve the problem -- but the record of composites and Al/Li (not
to mention many others touted in the 1980's but not heard of again) do
not bear this out -- the dramatic weight reductions hoped for are
often lost due to other material limitations once the actual article
is fabricated.
IMHO the materials aspect has always been over-emphasized by NASA and
the aerospace industry. The Atlas, Navaho, XB-70 and MiG-25 were built
primarily of steel. If you look at the comparison with all-aluminum
versions, the difference is actually very little. Real savings are to
be found in a fundamentally different design approach.
As noted, an Atlas V CCB with conventional structure cannot make it to
orbit (don't forget you have to include propellant residuals in the
calculation). But an early model Atlas, with, for example, a cluster
of 4 LR-105's, should have been able to accomplish the feat easily. It
would have also been interesting to see the results of re-entry of a
lightweight steel balloon tank with a deployable badminton-type skirt
-- perhaps it could even have lead to an alternate re-entry approach
not requiring exotic materials (as studied but rejected for Man in
Space Soonest)....
End of rant.
Mark Wade
http://www.astronautix.com/
he...@spsystems.net (Henry Spencer) wrote in message news:<HLoI1...@spsystems.net>...
> Expendable SSTO simply isn't that hard. The only reason it wasn't done
> decades ago is that nobody has tried.
>
ed kyle
>
> This is the real tragedy of SSTO. For a relatively modest cost,
>
I can think of two reasons that SSTO ELVs haven't happened.
First, SSTO is really only possible for low earth orbits
(LEO), and most space missions go to higher energy orbits
than that. Only 16 of the 41 orbital missions conducted so
far this year were LEO missions, for example. These missions
need at least one upper stage, and I'm not sure that the
two or three-stage design would necessarily be a good match
to a lightweight SSTO first stage.
Second, SSTO ELVs would produce a nasty orbital debris
problem. Each mission would leave a big chunk of hardware
in LEO that, if not deorbited, would pose a reentry hazard
within a few months.
> IMHO the materials aspect has always been over-emphasized by NASA and
> the aerospace industry. The Atlas, Navaho, XB-70 and MiG-25 were built
> primarily of steel. ...
>
I've been told that the Navaho XSM-64 winged missile used
titanium alloys. I'm not as certain about the booster, but
I think it was mostly aluminum construction - integral tank
with ring stiffeners.
- Ed Kyle
Henry Spencer
Encyclopedia Astronautica <eastro...@hotmail.com> wrote:
>IMHO the materials aspect has always been over-emphasized by NASA and
>the aerospace industry. The Atlas, Navaho, XB-70 and MiG-25 were built
>primarily of steel. If you look at the comparison with all-aluminum
The issue for most or all of those was heat resistance, not strength.
Atlas ascent heating, especially on a depressed trajectory, was too high
for aluminum (although it was used in the MX-774 test vehicle that proved
the Atlas concept), and Karel Bossart didn't think he could afford the
mass for external insulation. The aircraft, and I think Navaho as well,
were meant to cruise at aerothermal conditions where aluminum has no
useful strength.
(This is not to say that there wasn't a certain amount of overenthusiasm
for the more exotic materials, but these particular cases really did seem
to need something more heat-resistant than aluminum.)
>As noted, an Atlas V CCB with conventional structure cannot make it to
>orbit (don't forget you have to include propellant residuals in the
>calculation)...
One small reservation: conventional residuals allowances are far higher
than they need to be. 1% is traditional... but the Titan II first stage
was specced at 0.5%, and the S-IV at 0.25% (and it usually beat that
substantially).
Jan Philips
wrote:
>> Vanguard 4 successes out of 12; Jupiter C: 4 out of 9.
>
>Not quite right, unless I've missed something...
I got it from the following web pages, but maybe they count it
differently:
http://www.astronautix.com/lvs/jupiterc.htm
http://www.astronautix.com/lvs/vanguard.htm
ed kyle
> Brett Buck <buc...@pacbell.net> wrote:
> >>>(and SSTO in an ELV would be a quite useless stunt)?
> >> Not at all. If I were asked to design an expendable, I'd think very hard
> >> about making it a single stage. The price is a need for lightweight
> >> structure and engines, either multiple (>2) engines or deep throttling,
> >> and iron-fisted control of the mass budget (plus possibly some difficulty
> >> convincing skeptical investors that it can be done).
> >
> > I think you would find that the extra cost would be prohibitive.
>
> I don't think so, except possibly that last item (convincing investors).
> Contrary to popular belief, this isn't some leading-edge-of-technology
> stunt; all it takes is good design. The right sort of engine performance
> and mass ratio were achieved in several 1960s US rocket stages, and we
> have better engines (from the Russians), better materials, and better
> and lighter control systems.
>
> >...You just wind up with huge and expensive, rocket with a
> >comparatively tiny payload and razor-thin margins in a lot of places.
>
> Put six SSMEs under a (pre-aluminum-lithium!) shuttle ET, and it can reach
> orbit with 60klb of payload. How does that qualify as "comparatively tiny
> payload" and "razor-thin margins"?
>
> I wouldn't use that approach for a real design -- the SSME is too
> expensive and NASA controls the production lines for that hardware --
> but it illustrates what's possible with 1970s hardware.
>
How about an ET and four, supposedly less-expensive RS-68s?
With that setup you could probably get 18-25 metric tons
(40klb-55klb) of payload to LEO. (Though less expensive,
RS-68 is heavier and has a lower vacuum ISP than SSME.)
The RD-68s would have to run at their lowest throttle
setting for much of the flight, they would have to burn
longer than for a Delta IV mission, and you would probably
need to shut some of them down near the end.
- Ed Kyle
Rick DeNatale
> In the case of Atlas,
> this was first done with nitrogen; then later with helium to make the
> stage lighter during transport
And for those on display in museum rocket parks with an air compressor!
Derek Lyons
IIRC at least a few of them have been filled with foam to avoid the
need for an air compressor.
D.
--
The STS-107 Columbia Loss FAQ can be found
at the following URLs:
Text-Only Version:
http://www.io.com/~o_m/columbia_loss_faq.html
Enhanced HTML Version:
http://www.io.com/~o_m/columbia_loss_faq_x.html
Corrections, comments, and additions should be
e-mailed to o...@io.com, as well as posted to
sci.space.history and sci.space.shuttle for
discussion.
Henry Spencer
Jan Philips <judmccr...@bellsouth.net> wrote:
>>> Vanguard 4 successes out of 12; Jupiter C: 4 out of 9.
>>Not quite right, unless I've missed something...
>
>I got it from the following web pages, but maybe they count it
>differently:
>http://www.astronautix.com/lvs/jupiterc.htm
>http://www.astronautix.com/lvs/vanguard.htm
For Vanguard, he's counting TV-2 but not TV-1, which strikes me as
inconsistent (each tested one and only one Vanguard stage). That gives
him one more attempt and one more success.
For Jupiter C, he's counting the three-stage reentry tests as well as the
four-stage space launches, which is defensible, giving nine launches.
But he's counting the first three-stager as a failure, despite his own
commentary saying it was successful.
Henry Spencer
>>...(don't forget you have to include propellant residuals in the
>>calculation)...
>
>One small reservation: conventional residuals allowances are far higher
And of course, if you really get aggressive, with the right combination of
circumstances, you *don't* have to include propellant residuals because
there aren't any. The shutdown sequence for the proposed J-2S upgrade of
the S-IVB included running in a low-power mode to burn the residuals, and
then a gas-fed mode to burn the pressurants!
Henry Spencer
Don't be too sure of that. Re-engine the Titan II first stage with a
throttlable RD-253 variant (or equivalent), add a telescoping nozzle
extension, perhaps re-do some of the structure in composite, and really
pull out all the stops otherwise, and I wouldn't be surprised if you could
do SSTO to GTO or even escape. (The re-engining alone makes an enormous
difference, since the RD-253's Isp is more than 10% higher than that of
the stock Titan engines.)
If you disregard that possibility -- I think it possible but (unlike SSTO
to LEO) it might not be practical for routine operations -- then you need
staging for high-energy launches. But LEO is actually a pretty good place
to do staging, and it opens up options for very-high-performance upper
stages (since staging and upper-stage burn are no longer time-critical,
you can use smaller, lighter engines, or even consider a non-chemical
upper stage).
>...need at least one upper stage, and I'm not sure that the
>two or three-stage design would necessarily be a good match
>to a lightweight SSTO first stage.
Why not? It worked pretty well on the SSTO-class (or near-SSTO-class)
lightweight first stages of Titan II, Atlas-Centaur, Saturn V...
>Second, SSTO ELVs would produce a nasty orbital debris
>problem. Each mission would leave a big chunk of hardware
>in LEO that, if not deorbited, would pose a reentry hazard
>within a few months.
This is actually only slightly worse than non-SSTOs, which also mostly
leave big (although not quite *so* big) chunks of hardware in LEO. The
solution is the same for either type: deorbit the final stage.
Henry Spencer
>> expensive and NASA controls the production lines for that hardware --
>> but it illustrates what's possible with 1970s hardware.
>
>How about an ET and four, supposedly less-expensive RS-68s?
That lowers the cost, and may increase reliability, although four engines
are not as convenient as six for doing deep throttling by shutting some
down.
It still leaves the problem of NASA-controlled hardware (the ET). Of
course, one would assume that equivalent tankage could be built in other
facilities...
Pat Flannery
ed kyle wrote:
>I've been told that the Navaho XSM-64 winged missile used
>titanium alloys. I'm not as certain about the booster, but
>I think it was mostly aluminum construction - integral tank
>with ring stiffeners.
>
>
>
The missile used titanium in high heat areas, such as the nose and
engine, according to "The Navaho Missile Project"- the booster was made
from 20-24 ST aluminum that was chem-milled in one of the earliest uses
of that process.
Pat
Pat
ed kyle
> In article <88d21cfd.03092...@posting.google.com>,
> ed kyle <edky...@hotmail.com> wrote:
> >I can think of two reasons that SSTO ELVs haven't happened.
> >First, SSTO is really only possible for low earth orbits (LEO)...
> >two or three-stage design would necessarily be a good match
> >to a lightweight SSTO first stage.
>
> Why not? It worked pretty well on the SSTO-class (or near-SSTO-class)
> lightweight first stages of Titan II, Atlas-Centaur, Saturn V...
>
Consider the Delta IV-M/RS-68 example. A straight Delta IV-M
with one RS-68 and one RL10B-2 can put about 4 metric tons
into GTO. By the time the second stage completes its first
burn to enter low earth orbit, its mass is about 9.6 tons.
Now, consider a new single-stage to LEO expendable that uses
RS-68 engines. If you wanted to use this vehicle to put 4 tons
into GTO, it would need to be able to put itself and about 13.1
tons into LEO (assuming the upper stage structure has been
downsized to hold the reduced amount of propellant now needed).
As I show below, you wouldn't be able to do this unless you used
more than one, and maybe even more than two, RS-68 engines.
Assuming the following for an RS-68,
ISP average = 393 sec
Engine mass = 6.6 tons
Maximum liftoff mass per RS-68 = 258 tons,
and an ideal LEO delta-v = 9200 meters/sec, I find that the
initial mass needed to orbit 13.1 tons is equal to the following
expression.
Mi = (142.66 + 71.874*N)/(1 - 10.89*PFSMF)
where N = number of RS-68 engines
and PFSMF = the propulsion free (no engine) structural mass
fraction of the vehicle (engineless dry mass/inital mass).
This expression gives the following results.
N PFSMF max Mi
--------------------------------
1 0.015 256.4 tons
2 0.041 512.4 tons
3 0.049 768.2 tons
4 0.054 1030.6 tons
--------------------------------
Based on the best available historical examples, it appears
impossible to get a SSTO with only one RS-68, and very difficult
to do it with two or three. The best PFSMF numbers for LH2
vehicles to date have been 0.057 for Saturn V S-II, 0.073 for
S-IVB, 0.078 for Ariane 5 core, etc. The STS External Tank
has a PFFSMF of 0.04, but it does not have an interstage or
a propulsion section, etc.
Net result, the SSTO expendable first stage concept ends up
needing more total engines and tank structure (and therefore
costing more) than the optimized staged expendable Delta IVM.
- Ed Kyle
Henry Spencer
ed kyle <edky...@hotmail.com> wrote:
>Consider the Delta IV-M/RS-68 example. A straight Delta IV-M
>with one RS-68 and one RL10B-2 can put about 4 metric tons
>Now, consider a new single-stage to LEO expendable that uses
>RS-68 engines. If you wanted to use this vehicle to put 4 tons
>into GTO, it would need to be able to put itself and about 13.1
I haven't got time to check your calculations right now, but this is
broadly plausible. One thing staging does do is that it reduces total
vehicle mass; the fallacy is to believe that total vehicle mass is a
useful figure of merit.
>Based on the best available historical examples, it appears
>impossible to get a SSTO with only one RS-68, and very difficult
>to do it with two or three.
This might suggest that you need a better (and perhaps larger) engine,
or a better fuel. The RS-68 was not optimized for performance -- its
designers knew it was meant for a multi-stage rocket -- and LH2 is a poor
choice for high-delta-V stages*.
(* Contrary to popular belief, it is harder to get high delta-V with
LH2 than with kerosene. When you look at *vehicle* performance rather
than *engine* performance, LH2's high Isp is not enough to make up for
its large, heavy insulated tanks and its poor engine T/W. LH2 stages
have low *mass* but not particularly high delta-V.)
If you must use LH2, at least use the RD-0120 instead of the RS-68.
>...The STS External Tank
>has a PFFSMF of 0.04, but it does not have an interstage or
>a propulsion section, etc.
It does, however, have quite heavy carry-through structure for SRB loads,
not to mention having a massive intertank ring instead of a common
bulkhead, and some other superfluous mass like an insulated LOX tank
(needed mostly because the old "just let the ice fall off at launch"
approach was unacceptable with a fragile orbiter slung on the side).
Pat Flannery
Henry Spencer wrote:
>
>(* Contrary to popular belief, it is harder to get high delta-V with
>LH2 than with kerosene. When you look at *vehicle* performance rather
>than *engine* performance, LH2's high Isp is not enough to make up for
>its large, heavy insulated tanks and its poor engine T/W. LH2 stages
>have low *mass* but not particularly high delta-V.)
>
Doesn't this argument fall short- in the fact that it implies that the
insulation has to be carried during the _entire_ ascent to final
velocity? The Centaur Standard Shroud may well show a method of
jettisoning the weight of the insulation at such time that possible ice
build-up has been passed during the vehicle's ascent- once past around
200,000 feet- jettison the insulation (and the shroud that it is
attached to) from the vehicle; thereby losing its detrimental weight,
and the now unnecessary structural strength it supplied against
aerodynamic stresses during ascent.
Pat
Henry Spencer
Pat Flannery <fla...@daktel.com> wrote:
>>...LH2's high Isp is not enough to make up for
>>its large, heavy insulated tanks and its poor engine T/W...
>
>Doesn't this argument fall short- in the fact that it implies that the
>insulation has to be carried during the _entire_ ascent to final
>velocity? The Centaur Standard Shroud may well show a method of
>jettisoning the weight of the insulation at such time that possible ice
You still suffer somewhat from insulation requirements in places where
it's less easily unloaded, and from poor engine T/W.
The other problem with this theory is that when Centaur switched from its
old jettisonable insulation to permanently-bonded foam, net payload went
*up* slightly.
ed kyle
>
> (* Contrary to popular belief, it is harder to get high delta-V with
> LH2 than with kerosene. When you look at *vehicle* performance rather
> than *engine* performance, LH2's high Isp is not enough to make up for
> its large, heavy insulated tanks and its poor engine T/W. LH2 stages
> have low *mass* but not particularly high delta-V.)
>
I've yet to be convinced that this is true. Consider the following
side-by-side comparison of the world's most powerful liquid hydrogen
engine and the world's most advanced kerosene engine.
RS-68 (LH2/LOX) vs. RD-180 (RP/LOX)
RS-68: Max liftoff mass = 258 tonnes
Dry engine mass = 6.6 tonnes
Avg ISP = 393 sec
For LEO, assume ideal delta V = 9200 m/s
Minitial/Mfinal = e^(9200/(393*9.805)) = 10.89
Mfinal = Mengine + Mstructure + Mpayload
Mfinal = 6.6 + Minitial*PFSMF + Mpayload
where PFSMF = propulsion free structural mass fraction
Also,
Mfinal = Minitial/10.89 = 258/10.89 = 23.69 tonnes
so that Mpayload = 17.09 - 258*PFSMF
To reach LEO, PFSMF < 0.066
If PFSMF = 0.055 (roughly equal to the S-II stage)
Mpayload = 2.9 tonnes
If PFSMF = 0.04 (roughly equal to External Tank)
Mpayload = 6.8 tonnes.
Note that the Delta IV and Ariane 5 core stages
have PFSMF = roughly 0.078
RD-180: Max liftoff mass = 325 tonnes
Dry engine mass = 5.4 tonnes
Avg ISP = 325 sec
For LEO, assume ideal delta V = 9100 m/s
Minitial/Mfinal = e^(9100/(325*9.805)) = 17.39
Mfinal = Mengine + Mstructure + Mpayload
Mfinal = 5.4 + Minitial*PFSMF + Mpayload
where PFSMF = propulsion free structural mass fraction
Also,
Mfinal = Minitial/17.39 = 325/17.39 = 18.69 tonnes
so that Mpayload = 13.29 - 325*PFSMF
To reach LEO, PFSMF < 0.041
If PFSMF = 0.04 (roughly equal to the S-IC stage)
Mpayload = 0.29 tonnes
If PFSMF = 0.025 (roughly equal to the Atlas sustainer stage)
Mpayload = 5.16 tonnes.
Note that the Atlas V CCB and Zenit 3 first stages have
PFSMF = roughly 0.06.
To me, it seems that while both engines offer expendable SSTO
possibilities, the hydrogen engine provides a bit more
"wiggle-room" in the dry mass department. RS-68 could put
4 tonnes into LEO using a 12.9 tonne stage structure. With
RD-180, a 4 tonne payload would require a stage structure mass
of less than 9.3 tonnes. This despite the fact that RD-180 is
the more powerful engine.
- Ed Kyle
Jim Davis
> To me, it seems that while both engines offer expendable SSTO
> possibilities, the hydrogen engine provides a bit more
> "wiggle-room" in the dry mass department. RS-68 could put
> 4 tonnes into LEO using a 12.9 tonne stage structure. With
> RD-180, a 4 tonne payload would require a stage structure mass
> of less than 9.3 tonnes. This despite the fact that RD-180 is
> the more powerful engine.
I draw the opposite conclusion from your numbers, Ed.
RS-68: Propellant Volume/Stage structure mass = .051 m3/kg
RD-180: Propellant Volume/Stage structure mass = .033 m3/kg
The RD-180 vehicle should be much easier to build for the given 4
tonne payload.
Jim Davis
ed kyle
Both should be possible, but the RD-180 vehicle structure would
have to be as efficient as the still-extraordinary Atlas sustainer
stage, which used thin stainless steel balloon tanks. Note that
the Atlas III single-stage-Atlas first stage used balloon tanks
too, but its PFSMF was about 0.04, meaning that the RD-180
powered Atlas III first stage might barely make orbit, but would
not be able to carry any payload. The question for the Atlas III
designers would have to be, "can you cut 4 tonnes of dry mass out
of your stage and still make it work?"
On the other hand, the RS-68 powered SSTO would have to be a
bit more efficient than the Saturn V S-II stage, and not nearly
as efficient as the STS External Tank. Both of these vehicles
used self-supporting tank structures, had insulated tanks, etc.
- Ed Kyle
Jim Davis
> On the other hand, the RS-68 powered SSTO would have to be a
> bit more efficient than the Saturn V S-II stage, and not nearly
> as efficient as the STS External Tank. Both of these vehicles
> used self-supporting tank structures, had insulated tanks, etc.
You're missing my point, Ed. Take the shuttle ET. Because of the
differences in the densities of the two propellant combinations
you can put nearly 3 times the amount of LO2/RP-1 in the ET as
you can LO2/LH2 (although the fractions allocated to fuel and
oxidizer would have to be adjusted). The possible objection that
the ET cannot support 3 times the mass of propellant is not valid
because propellant tanks are essentially pressure vessels and
their masses are determined by the internal pressure maintained
in the tank not by the hydrostatic pressure.
Any possible expendable LO2/LH2 SSTO can always be converted into
an LO2/RP-1 SSTO by this procedure. And once this is done you
will find that either:
a.) The RP-1/LO2 fuelled vehicle has a greater delta-v with the
same payload or
b.) The RP-1/LO2 fuelled vehicle has a greater payload with the
same delta-v or
c.) The RP-1/LO2 fuelled vehicle has the same payload and same
delta-v with heavier (i.e., easier and cheaper to design and
build) propellant tanks.
Your comparisons of historic tank weights are misleading for
various reasons. Details really matter. The S-II and ET are much
larger than the Atlas. The Atlas has a greater fineness ratio
than the S-II and ET. The S-II and ET are essentially upper
stages where much greater effort goes into weight savings because
the payoff is so much greater there.
You have to start with a clean sheet of paper to get a valid
basis of comparison.
Jim Davis
Derek Lyons
>You're missing my point, Ed. Take the shuttle ET. Because of the
>differences in the densities of the two propellant combinations
>you can put nearly 3 times the amount of LO2/RP-1 in the ET as
>you can LO2/LH2 (although the fractions allocated to fuel and
>oxidizer would have to be adjusted). The possible objection that
>the ET cannot support 3 times the mass of propellant is not valid
Um, no. The ET is designed to take a given amount of weight and
distribution of the same in it's tanks. You cannot change that
without redesigning the structure.
>because propellant tanks are essentially pressure vessels and
>their masses are determined by the internal pressure maintained
>in the tank not by the hydrostatic pressure.
Um, no. The mass of a tank is determined by the amount of structure
in the tank. The mass a tank can hold is determined by the amount of
structure in the tank. The pressure a tank can hold is determined by
the amount of structure in the tank. You cannot arbitrarily increase
the amount of fuel or oxidizer in a stage without redesigning, or at
least requalifying, the structural design.
Furthermore, extremely large vertical pressure vessels (Like the ET
LH2 tank) certainly do experience significant hydrostatic pressures
when loaded, and those pressures increase under acceleration.
Jim Davis
> The ET is designed to take a given amount of weight and
> distribution of the same in it's tanks. You cannot change that
> without redesigning the structure.
Of course not. You would definitely have to redesign an ET if it
was going to be filled with LO2/RP-1 rather than LO2/LH2. I'm
merely pointing out that the structural mass of an ET designed to
hold LO2/RP-1 would not be any greater than one designed to hold
LO2/LH2 if the volumes were equal.
> The mass of a tank is determined by the amount of
> structure in the tank.
No kidding, Derek.
> The mass a tank can hold is determined
> by the amount of structure in the tank.
No. The mass a tank can hold is determined by its volume, the
density of the fluid filling it, and the loads the tank is designed
to resist. In the case of rocket propellant tanks the internal
loads the tanks are designed to resist are not driven by the
hydrostatic loads. There probably is a size where hydrostatic
forces are the design driver but we've yet to see it.
> The pressure a tank can
> hold is determined by the amount of structure in the tank. You
> cannot arbitrarily increase the amount of fuel or oxidizer in a
> stage without redesigning, or at least requalifying, the
> structural design.
No doubt about it. But there is no reason to believe that after the
redesign/recertification the tank's mass will be greater.
> Furthermore, extremely large vertical pressure vessels (Like the
> ET LH2 tank) certainly do experience significant hydrostatic
> pressures when loaded,
Of course they are significant. It's just that they do not drive
the design. The design is driven by the need for internal
pressurization to resist bending loads (and all tanks require this
internal pressure, not just ballon tanks) and/or provide net
positive suction head to the propellant pumps.
> and those pressures increase under
> acceleration.
Are you certain? Hydrostatic loads are functions of the level of
acceleration and the amount of the remaining propellant. One is
increasing and the other decreasing. Are you certain that one
effect outweighs the other?
Jim Davis
Scott Hedrick
news:3f7b24e6...@supernews.seanet.com...
> You cannot arbitrarily increase
> the amount of fuel or oxidizer in a stage without redesigning, or at
> least requalifying, the structural design.
Or *type* of fuel, as well.
--
If you have had problems with Illinois Student Assistance Commission (ISAC),
please contact shredder at bellsouth dot net. There may be a class-action
lawsuit
in the works.
Derek Lyons
>Derek Lyons wrote:
>
>> The ET is designed to take a given amount of weight and
>> distribution of the same in it's tanks. You cannot change that
>> without redesigning the structure.
>
>Of course not. You would definitely have to redesign an ET if it
>was going to be filled with LO2/RP-1 rather than LO2/LH2. I'm
>merely pointing out that the structural mass of an ET designed to
>hold LO2/RP-1 would not be any greater than one designed to hold
>LO2/LH2 if the volumes were equal.
You were not at all clear in expressing that point.
ed kyle
> ed kyle wrote:
>
> > On the other hand, the RS-68 powered SSTO would have to be a
> > bit more efficient than the Saturn V S-II stage, and not nearly
> > as efficient as the STS External Tank. Both of these vehicles
> > used self-supporting tank structures, had insulated tanks, etc.
>
> You're missing my point, Ed. Take the shuttle ET. Because of the
> differences in the densities of the two propellant combinations
> you can put nearly 3 times the amount of LO2/RP-1 in the ET as
> you can LO2/LH2 (although the fractions allocated to fuel and
> oxidizer would have to be adjusted).
I understand your point that an RP fueled vehicle can produce
more delta-v for a given amount of dry mass. But an equivalent
volume of heavier RP fuel requires more thrust (engines) to get
airborne, which requires more engine (dry) mass (and money),
etc. When I compare RS-68 against RD-180, I don't find this
theoretical, often-discussed hydrocarbon advantage.
For example, an LH2 ET-based SSTO would only need three RS-68s to
lift itself off the ground. Your kerosene-filled ET SSTO would
need seven RD-180s (and would thus have a higher dry mass). The
LH2 ET+3xRS68s would produce a an ideal no-payload delta-v of
11,607 m/s while the RP1 ET+7xRD180 delta-v would be 11,676 m/s.
The RP vehicle would be only a bit more powerful, but would surely
cost quite a bit more than the LH2 vehicle.
- Ed Kyle
Jim Davis
> For example, an LH2 ET-based SSTO would only need three RS-68s
> to lift itself off the ground. Your kerosene-filled ET SSTO
> would need seven RD-180s (and would thus have a higher dry
> mass). The LH2 ET+3xRS68s would produce a an ideal no-payload
> delta-v of 11,607 m/s while the RP1 ET+7xRD180 delta-v would be
> 11,676 m/s.The RP vehicle would be only a bit more
> powerful, but would surely cost quite a bit more than the LH2
> vehicle.
Now add payload to each vehicle until it can acheive the delta-v
required for LEO. You'll find that the RP1 ET+7xRD180 vehicle has
almost twice the payload of the LH2 ET+3xRS68 vehicle. The RP-1
vehicle may cost more (I'm not sure about the relative costs of RP-1
and LH2 engines) but it costs less per kg of payload.
As you've discovered, if you increase the delta-v requirement to
beyond that required for LEO, LH2 becomes increasingly more
competitive, but the case for some form of staging becomes
increasingly compelling. For upper stages the argument for LH2 is
much more solid.
Jim Davis
ed kyle
> ed kyle wrote:
>
> > For example, an LH2 ET-based SSTO would only need three RS-68s
> > to lift itself off the ground. Your kerosene-filled ET SSTO
> > would need seven RD-180s (and would thus have a higher dry
> > mass). The LH2 ET+3xRS68s would produce a an ideal no-payload
> > delta-v of 11,607 m/s while the RP1 ET+7xRD180 delta-v would be
> > 11,676 m/s.The RP vehicle would be only a bit more
> > powerful, but would surely cost quite a bit more than the LH2
> > vehicle.
>
> Now add payload to each vehicle until it can acheive the delta-v
> required for LEO. You'll find that the RP1 ET+7xRD180 vehicle has
> almost twice the payload of the LH2 ET+3xRS68 vehicle. The RP-1
> vehicle may cost more (I'm not sure about the relative costs of RP-1
> and LH2 engines) but it costs less per kg of payload.
>
This would also be true if you compared a vehicle that had NxRD180
engines against a vehicle with 2xNxRS68 engines. The RS68 machine
would lift more payload so that the launch would cost less per kg
of payload.
I've found no references regarding costs on these engines. My
guess is that they are about equal. RS-68 has a simplier gas
generator cycle, but it uses the more exotic fuel and is assembled
in a higher-wage country. RD-180 costs, on the other hand, will
rise as its assembly begins to transfer to the U.S. (news of the
start of this process finally came out a week or two ago.).
> As you've discovered, if you increase the delta-v requirement to
> beyond that required for LEO, LH2 becomes increasingly more
> competitive, but the case for some form of staging becomes
> increasingly compelling. For upper stages the argument for LH2 is
> much more solid.
>
This seems vitally important, because only 4 in 10 of this year's
space launches went to LEO. The rest went to higher delta-V orbits
or escape trajectories.
I suppose that the SSTO comparison is going to be impossible
to clarify until someone actually builds one of these things,
because success or failure in the SSTO realm depends so much on
structural details.
- Ed Kyle
Jim Davis
> This would also be true if you compared a vehicle that had
> NxRD180 engines against a vehicle with 2xNxRS68 engines. The
> RS68 machine would lift more payload so that the launch would
> cost less per kg of payload.
But in this case the 2xNxRS68 vehicle would be much larger than the
NxRD180 vehicle so I fail to see how the launch would cost less per
payload. Can you provide numbers to back this up?
Jim Davis
James Steven York
wrote:
>convincing skeptical investors that it can be done). The payoff is a much
>simpler vehicle with many fewer failure modes, and the disappearance of
>the "where do the spent stages fall?" problem.
>
>(By the way, you don't need balloon tanks to do an expendable SSTO,
>although they might help.)
Wouldn't balloon tanks also be ideal as drop tanks for an otherwise
usable system (like the old ROMBUS plug-nozzle concept)? Given no
other major systems attached to them, it seems likely that they'd
shred on the way down and not drop any large masses of metal to the
ground, that they'd be light, and produced in quantity, cheap.
ed kyle
Compare a 3xRD180 SSTO expendable with a 6xRS68 vehicle.
For the 3xRD180:
Max Liftoff Mass = 975 tonnes
Engine Mass = 16.2 tonnes
To get a realistic mass budget for structure and
propellant, consider the following historical data.
Mp = Propellant Mass (tonnes)
Ms = Structural & Residual Mass (No Engines) (tonnes)
Stage Mp Ms Ms/Mp
Atlas V CCB 284.1 18.90 0.0665
Zenit 3 Stg 1 318.8 21.95 0.0689
S-IC (Sat V) 2078.9 88.40 0.0425
Atlas III SSA 183.2 8.33 0.0454
Delta II 96.1 4.68 0.0487
Let's assume we can improve on past efforts to make
Ms/Mp = 0.04
Now Mtotal = Ms + Mp + Mengines + Mpayload
Mtotal = Mp(Ms/Mp + 1) + Mengines + Mpayload
Mp = (Mtotal - Mengines - Mpayload)/(Ms/Mp + 1)
Mp = (975 - 16.2 - Mpayload)/(1.04)
Mp = (958.8 - Mpayload)/1.04
and Delta-v = 9200 = 325*9.805*ln(975/(975-Mp))
solving:
Mp = 920.67 tonnes
Ms = 36.82 tonnes
Mpayload = 1.3 tonnes
For the 6xRS68 vehicle:
Max Liftoff Mass = 1,548 tonnes
Engine Mass = 39.6 tonnes
To get a realistic mass budget for structure and
propellant, consider the following historical data.
Mp = Propellant Mass (tonnes)
Ms = Structural & Residual Mass (No Engines) (tonnes)
Stage Mp Ms Ms/Mp
Delta IV CBC 204.1 17.9 0.0877
Ariane 5 Core 170.0 10.35 0.0609
S-II (Sat V) 450.3 8.37 0.0611
LWET 727.3 30.30 0.0417
Here let's be conservative and select a real structure-
propellant mass fraction number (say 0.061) that does not
assume any improvement on existing technology.
Now Mtotal = Ms + Mp + Mengines + Mpayload
Mtotal = Mp(Ms/Mp + 1) + Mengines + Mpayload
Mp = (Mtotal - Mengines - Mpayload)/(Ms/Mp + 1)
Mp = (1548 - 39.6 - Mpayload)/(1.061)
Mp = (1508.4 - Mpayload)/1.061
and Delta-v = 9300 = 393*9.805*ln(1548/(1548-Mp))
solving:
Mp = 1409.47 tonnes
Ms = 85.98 tonnes
Mpayload = 12.95 tonnes
Comparing 6xRS68 with 3xRD180,
Mpayload/Ms = 0.151 (6xRS68)
Mpayload/Ms = 0.035 (3xRD180)
Mpayload per engine = 2.16 tonnes (6xRS68)
Mpayload per engine = 0.43 tonnes (3xRD180)
- Ed Kyle
Jim Davis
> Compare a 3xRD180 SSTO expendable with a 6xRS68 vehicle.
>
> For the 3xRD180:
>
> Max Liftoff Mass = 975 tonnes
> Engine Mass = 16.2 tonnes
So far I'm with you.
> To get a realistic mass budget for structure and
> propellant, consider the following historical data.
>
> Mp = Propellant Mass (tonnes)
> Ms = Structural & Residual Mass (No Engines) (tonnes)
>
> Stage Mp Ms Ms/Mp
>
> Atlas V CCB 284.1 18.90 0.0665
> Zenit 3 Stg 1 318.8 21.95 0.0689
> S-IC (Sat V) 2078.9 88.40 0.0425
> Atlas III SSA 183.2 8.33 0.0454
> Delta II 96.1 4.68 0.0487
>
> Let's assume we can improve on past efforts to make
> Ms/Mp = 0.04
Here's where you lose me. Certainly we can improve on past efforts.
All we have to do is use exactly the same techniques on LO2/RP-1
tanking as we've done on LO2/LH2 tanking. Later in your post you
select the real Ms/Mp of .061 for your LO2/LH2 vehicle. If we build
an LO2/RP-1 tank to the this same efficiency what would the Ms/Mp
be? The density of LO2/LH2 at a mixture ratio of 6 has a density of
358 kg/m3. The density of LO2/RP-1 at a mixture ratio of 2.72 has a
density of 1011 kg/m3. So a realistic Ms/Mp for LO2/RP-1 is .061*
358/1011 or .022, far better than your estimate of .04.
Using these numbers we can calculate:
Mtotal = Ms + Mp + Mengines + Mpayload
Mtotal = Mp(Ms/Mp + 1) + Mengines + Mpayload
Mp = (Mtotal - Mengines - Mpayload)/(Ms/Mp + 1)
Mp = (975 - 16.2 - Mpayload)/(1.022)
Mp = (958.8 - Mpayload)/1.022
and Delta-v = 9200 = 325*9.805*ln(975/(975-Mp))
solving:
Mp = 920.65 tonnes
Ms = 19.91 tonnes
Mpayload = 18.23 tonnes
Mpayload/Ms = 0.035 (3xRD180) Ms/Mp = .04
Mpayload/Ms = 0.92 (3xRD180) Ms/Mp = .022
> Mpayload per engine = 2.16 tonnes (6xRS68)
Mpayload per engine = 0.43 tonnes (3xRD180) Ms/Mp = .04
Mpayload per engine = 6.09 tonnes (3xRD180) Ms/Mp = .022
Note that the RP-1 vehicle has much better economics than the LH2
vehicle even though they are both built to exactly the same level
of structural technology.
Here are my raw numbers:
RS-68 RD-180
dV 9300 9200 m/s
GLOW 1548000 975000 kg
T/W 1.14 1.20
Mp 1409448 920654 kg
Mpl 12975 18259
# engines 6 3
F, ea 2890000 3827000 N
F, total 17340000 11481000 N
Isp 393 325 s
Ms 85976 19886 kg
T/W, eng 44.66 72.28
Me, ea 6600 5400 kg
Ms/Mp 0.061 0.022
Mpl/Ms 0.15 0.92
Mpl/engine 2163 6086 kg
Jim Davis
ed kyle
Ms/Mp = 0.022 is about two times better than the best efforts that
have been demonstrated on booster-class RP-fueled stages to date.
(There must be a reason that this is true.) If you want to use
that mass fraction for comparison purposes, than I'll have to
insist that you assume the same advances for the LH2 vehicle and
use Ms/Mp = 0.031! A 6xRS68 LH2 vehicle with Ms/Mp = 0.031 would
be able to boost 51 tonnes into LEO, 2.8 times more than the
Ms/Mp = 0.022, 3xRD180 machine.
- Ed Kyle
Jim Davis
> Ms/Mp = 0.022 is about two times better than the best efforts
> that have been demonstrated on booster-class RP-fueled stages to
> date.
Yes.
> (There must be a reason that this is true.)
Indeed, there is. It's simply that the heroic weight saving
measures that go into LH2-fuelled stages haven't been necessary
for RP-fuelled stages. RP stages are typically first stages, like
all the ones you list. Shaving weight off the first stage isn't
nearly as effective (and isn't nearly as expensive) as simply
making the first stage bigger to increase payload. Indeed, the
Ms/Mp of the original Thor and Atlas stages got progressively
worse as these stages were lengthened in order for them to cope
with increases in bending loads. LH2 stages on the other hand are
typically upper stages (even if they are ignited at liftoff they
still have to be accelerated all the way to near orbital
velocities). Any mass shaved off these stages typically translate
directly into mass to LEO. So it is no suprise that strenuous and
expensive efforts to save weight go into LH2 upper stages and not
into RP-1 first stages. The experience with the S-IC and S-II
stages are particularly relevant here.
The Delta IV core is in a class by itself as the only genuine LH2
first stage and not suprisingly it has the worst Ms/Mp of any LH2
stage you listed even though it is of recent origin. There is
relatively little payoff in shaving mass from the first stage.
Finally, looking at individual *tanks* instead of complete stages
is instructive. The Shuttle ET LH2 tank has an inert mass of 26640
lbs and holds 227641 lbs of LH2 (numbers from Jenkins) for a Ms/Mp
of 0.117. The Shuttle ET LO2 tank, on the other hand, has an inert
mass of 12000 lbs and holds 1361936 lbs of LO2 for a Ms/Mp of
0.0088, over 13 times better than the LH2 tank. It is not a
coincidence that LO2 is 16 times denser than LH2. Clearly, the
*volumes* of the tanks determine the inert weight of the tanks.
Anyway, it is to be hoped that subsequent events will demonstrate
whether LH2 or RP-1 proponents are correct. All I wanted to do was
demonstrate that the claim made by Henry Spencer (where did he get
off to, anyway?) that RP-1 is superior to LH2 for SSTO
applications is logical if one accepts the premise behind it. You
apparently don't and we can agree to disagree but you now know
where dense fuel proponents are coming from.
Jim Davis
jeff findley
> Finally, looking at individual *tanks* instead of complete stages
> is instructive. The Shuttle ET LH2 tank has an inert mass of 26640
> lbs and holds 227641 lbs of LH2 (numbers from Jenkins) for a Ms/Mp
> of 0.117. The Shuttle ET LO2 tank, on the other hand, has an inert
> mass of 12000 lbs and holds 1361936 lbs of LO2 for a Ms/Mp of
> 0.0088, over 13 times better than the LH2 tank. It is not a
> coincidence that LO2 is 16 times denser than LH2. Clearly, the
> *volumes* of the tanks determine the inert weight of the tanks.
According to NASA's web site, venting occurs when the ullage pressure
of liquid hydrogen tank reaches 38 psig or the ullage pressure of the
liquid oxygen tank reaches 25 psig. This relatively small difference
in pressure (only a factor of 1.5) certainly cannot account for such a
large difference in the Ms/Mp of the tanks.
Comparing the volumes of the tanks, 53,518 cubic feet for the H2 tank
and 19,563 cubic feet for the O2 tank gives a ratio of 2.74, which is
roughly the ratios of the tank weights (2.22). Since the design
pressures of the tanks are roughly similar, the weight of the tanks
appears to scale with the size of the tank, not with the mass of
propellant it carries. This is a distinct advantage for dense
propellants like LOX, and is even more so for kerosene.
I'd also think (based on what little pressure tank design I've done)
that the complex, pointy shape of the O2 tank would make it even more
challenging to keep its Ms/Mp so low when compared to the much more
conventional shape of the H2 tank. We've not even considered that
issue.
Jeff
--
Remove "no" and "spam" from email address to reply.
If it says "This is not spam!", it's surely a lie.
ed kyle
> ed kyle wrote:
>
> The Delta IV core is in a class by itself as the only genuine LH2
> first stage and not suprisingly it has the worst Ms/Mp of any LH2
> stage you listed even though it is of recent origin.
>
The fact that the CBC stage has to handle strap-on booster
loads (especially in the Heavy configuration) may have
something to do with it. (The Atlas V CCB stage is proportionally
heavier in this manner too when compared to Atlas III Single Stage
Atlas.) In addition, the CBC mass includes a lengthy interstage
that must support the second stage and payload.
>
> Finally, looking at individual *tanks* instead of complete stages
> is instructive. The Shuttle ET LH2 tank has an inert mass of 26640
> lbs and holds 227641 lbs of LH2 (numbers from Jenkins) for a Ms/Mp
> of 0.117. The Shuttle ET LO2 tank, on the other hand, has an inert
> mass of 12000 lbs and holds 1361936 lbs of LO2 for a Ms/Mp of
> 0.0088, over 13 times better than the LH2 tank. It is not a
> coincidence that LO2 is 16 times denser than LH2. Clearly, the
> *volumes* of the tanks determine the inert weight of the tanks.
>
Agreed, but consider also that, although the tanks account for
17.53 tonnes of dry mass, non-tank ET structure (interstage, etc.)
accounts for an additional 12.77 tonnes of dry mass (42% of the
total dry mass), resulting in Ms/Mp = 0.04. Nontank structure,
which is a significant part of any vehicle's mass, is less
dependant on tank volume than tank mass.
> Anyway, it is to be hoped that subsequent events will demonstrate
> whether LH2 or RP-1 proponents are correct. All I wanted to do was
> demonstrate that the claim made by Henry Spencer (where did he get
> off to, anyway?) that RP-1 is superior to LH2 for SSTO
> applications is logical if one accepts the premise behind it. You
> apparently don't and we can agree to disagree but you now know
> where dense fuel proponents are coming from.
>
I've yet to convince myself that it is anything but a paper-premise
in that I don't see RP advantages when I compare real, existing
hardware. The premise could prove to be correct in the future, I
suppose, if specialized hardware is designed to fully exploit the
fuel density advantage.
It may take awhile for such a machine to appear. In the mean time,
thanks for helping contemplate the problem here!
- Ed Kyle
jeff findley
> Agreed, but consider also that, although the tanks account for
> 17.53 tonnes of dry mass, non-tank ET structure (interstage, etc.)
> accounts for an additional 12.77 tonnes of dry mass (42% of the
> total dry mass), resulting in Ms/Mp = 0.04. Nontank structure,
> which is a significant part of any vehicle's mass, is less
> dependant on tank volume than tank mass.
Also consider that much of that non-tank structure is there because of
the SRB's strapped to the side of the ET. Still more of this non-tank
structure is due to strapping the shuttle to the side of the tank.
Some of this non-tank structure would "go away" if you were to convert
it into a more conventional (e.g. expendable SSTO) rocket stage.
Damon Hill
news:yz9vfqy...@sgipd572.net.plm.eds.com:
> edky...@hotmail.com (ed kyle) writes:
>> Agreed, but consider also that, although the tanks account for
>> 17.53 tonnes of dry mass, non-tank ET structure (interstage, etc.)
>> accounts for an additional 12.77 tonnes of dry mass (42% of the
>> total dry mass), resulting in Ms/Mp = 0.04. Nontank structure,
>> which is a significant part of any vehicle's mass, is less
>> dependant on tank volume than tank mass.
>
> Also consider that much of that non-tank structure is there because of
> the SRB's strapped to the side of the ET. Still more of this non-tank
> structure is due to strapping the shuttle to the side of the tank.
> Some of this non-tank structure would "go away" if you were to convert
> it into a more conventional (e.g. expendable SSTO) rocket stage.
Wonder what a 'demonstrator' SSTO based on a modified ET could
do if the following assumptions are made:
1. Stage-and-half using a jettionsable ring of four RS-68s and
a SSME sustainer (possibly with extendable nozzle). I don't
recall this being mentioned in the discussion in this thread
and it seems like this would be most helpful in mass reduction
to orbit.
2. Elimination of Shuttle-related load structures (probably offset
by other changes for a pure in-line load)
3. Stretch tank somewhat for additional propellant load (probably
more advantageous for denser propellants?)
4. As a last resort, nested bulkheads to eliminate intertank
structure.
I wouldn't want to get too carried away with tweaks, just make
changes that would be compatible with existing ET manufacturing
tools and infrastructure.
COBRA or the RS-83 may have weight/ISP advantages over the RS-68
at a lower cost than a SSME, but let's not go there just yet.
--Damon
Josh Hopkins
>
> (* Contrary to popular belief, it is harder to get high delta-V with
> LH2 than with kerosene. When you look at *vehicle* performance rather
> than *engine* performance, LH2's high Isp is not enough to make up for
> its large, heavy insulated tanks and its poor engine T/W. LH2 stages
> have low *mass* but not particularly high delta-V.)
>
Then Ed Kyle said
> To me, it seems that while both engines offer expendable SSTO
> possibilities, the hydrogen engine provides a bit more
> "wiggle-room" in the dry mass department.
.followed by lots of equations in several posts based on the mass fraction
of existing stages minus engines.
Then Jim Davis said
>I draw the opposite conclusion from your numbers, Ed.
Followed by more equations in more posts, based on the assumption.
>In the case of rocket propellant tanks the internal
>loads the tanks are designed to resist are not driven by the
>hydrostatic loads. There probably is a size where hydrostatic
>forces are the design driver but we've yet to see it.
If I may interject, I think the reason that Ed and Jim are coming to
different conclusions from the same set of numbers is that you are each
making some oversimplifying assumptions that are incorrect. This is
exacerbated by the fact that using mass fractions in the rocket equation is
of limited accuracy, particularly when at the sensitive fringes of a
high-delta-V case.
First, Jim's:
The mass of tanks is *not* driven solely by the ullage pressure load.
Consider the following from "Techniques for the Determination of Mass
Properties of Earth-to-Orbit Transporation Systems, MacConochie and Klich,
1976, NASA TM-78661. The text derives the structural equations explaining
why tanks *would be* driven only by volume and ullage pressure, and then
adds:
"The above statement is valid assuming uniform pressure throughout the tank
with no insulation, no changes in non-optimums or tank loads with size. In a
gravity field, or under the influence of engine thrust, however, tank
hydraulic pressures increase as a function of tank dimension along the line
of action of thrust or gravity axis. At the bottom of a tank containing a
dense propellant, such as LOX, the hydraulic pressure alone at the bottom of
a 60 foot tank under 1.3 g's acceleration is 297 kPa (43 psi) compared to
the shuttle external LOX tank design ullage pressure of 262 kPa (38 psi). It
can be seen that the tank mass is, therefore, sensitive not only to tank
wetted area and wall thickness due to ullage pressure, but is also affected
by hydraulic head linearly increasing from the free surface of the fluid."
It is for this reason that the wall thickness of many launch vehicle tanks
is greatest at the bottom and decreases with height, and why the aft
bulkheads of these tanks are much heavier than the forward bulkheads. For
example, the S-IC LOX aft bulkhead was 65% heavier than the forward
bulkhead. Tanks can also be sized in part by loads imparted from upper
stages.
In my experience, it is not a bad first-order assumption to assume that *for
a given propellant and configuration* the tank mass scales with volume,
(assuming your starting scaling point accounted for the head pressure)
because the head effect described above is only a 1/3 power with volume
(height is linear vs volume is a cube), and to some extent it is mitigated
by various "non-opt" tank parts that become more efficient with size. Tank
insulation (if present) scales only with surface area, and various
"non-optimum" tank parts like access ports, mounting brackets, stiffners,
can become a smaller percentage of the mass as the scale increases.
However, you *can not* assume that a hydrogen tank and a LOX or kerosene
tank will have similar mass per cubic foot. On the plus side, hydrogen
imparts very little head pressure, and generally doesn't need baffles, on
the minus side it does need insulation. The mass per unit volume factor is
generally a little bit lower for hydrogen tanks. The very un-dense nature
of hydrogen still overwhelms this modest savings and does make hydrogen
tanks much heavier on a mass of tank / mass of propellant basis. (As an
aside, I really wish there were a good antonym for "dense").
The broader error, though, is the assumption that nonpropulsive dry mass =
tank. There are a number of other components to any launch vehicle stage,
most of which clearly don't scale with directly volume, though many of them
do have propellant-related scaling effects.
Consider the following mass breakdowns of the S-IC and S-II stages, from
"Mass Breakdown of the Saturn V" John Whitehead, AIAA 2000-3141.
(my apologies if the tables come out formatted poorly)
S-IC
Metric Tons % of burnout % of dry
Engines 42.5 26% 33%
TVC 3.9 2% 3%
Fairings/Fins/
Firewall/Paint 7.7 5% 6%
Thrust Structure 21.5 13% 16%
Electrical 3.3 2% 3%
Fuel System 6.1 4% 5%
Kerosene Tank 10.2 6% 8%
Residual Kerosene 12.5 8%
Intertank 6 4% 5%
Oxidizer System 10.7 7% 8%
Oxygen Tank 16.1 10% 12%
Residual Oxygen 17.3 11%
Forward Structure 2.4 1% 2%
Misc Residuals 1.1 1% 1%
Total at Burnout 161.3
Total Dry Only 130.4
S-II
Metric Tons % of burnout % of dry
Engines 8.15 20% 22%
TVC 0.51 1% 1%
Fairings/Insulation/Paint 1.02 2% 3%
Thrust Structure 3.31 8% 9%
Purge and propellant systems 3.34 8% 9%
Aft Structure 1.9 5% 5%
Electrical 2.7 6% 7%
Tanks & Insulation 13.48 32% 37%
Residual Hydrogen 2.14 5%
Residual Oxygen 3.17 8%
Forward structure 1.84 4% 5%
Total at Burnout 41.56
Total Dry Only 36.25
Observations:
Hydrogen engines are generally have much lower thrust/weight than kerosene
engines.
TVC and thrust structure are functions of thrust and some other things, but
not particularly related to propellant density.
Pressurization (I think it's covered here under "propellant systems" and e.g
"LOX system") is obviously a function of volume, but there are other
complicating factors as well. Hydrogen tanks are often pressurized just by
heating extra hydrogen, which is very light both in terms of the gas
required and the hardware compared to pressure bottles carrying a separate
pressurant. However, those tanks are still very big, which hurts.
Feed systems (also covered under e.g. "LOX system" I think) are indirectly
correlated to propellant density, but not necessarily linearly or
exclusively. (For example, line lengths obvious depend on tank geometry -
big aft hydrogen tanks can cause very long LOX lines). In general hydrogen
feedlines are heavy because of their insulation requirements.
Intertanks when present are typically non-pressurized structure, so their
mass is a function of line loads (i.e. heavy compact propellants are bad)
but also geometry (small propellant tanks with short domes mean short
intertanks which is good, but also small diameters, which is bad).
Electrical systems and many misc structures are only weakly related to
propellant choice.
Residuals can be a significant portion of burnout mass, but their mass is
not easily correlated to propellant type. It's not uncommon to assume that
residuals are some fixed percentage of gross propellant weight, which would
suggest that LOX/kerosene is bad because of its higher propellant weight
relative to hydrogen. However, residuals depend on any number of factors,
including the engine cycle, mixture ratio, feed line geometry, and how good
the propellant utilization system is.
So, summarizing the above: a significant portion of the inert mass of a
stage does not scale linearly with propellant volume. However, propellant
density and other characteristics do have complex effects on many of these
parameters. In general, hydrogen causes heavier systems, but not in a way
that's as straightforward as Jim is suggesting.
Now, I think Ed also is missing some of the details, that drive the
conclusion in the opposite direction.
First, I think the average Isp number for the RS-68 is incorrect. According
to "Propulsion for the 21st Century-RS-68" B. K. Wood, AIAA 2002-4324, the
sea level Isp is 357 an the Vac Isp is 409. So the average is 383 s, not 393
s (assuming by "average" you mean the arithmetic average of the two values,
not some sort of mission averaged value.)
Second, different engines many or many not include various engine- related
elements. For example, gimbal actuators, the power source to run those
actuators, pneumatics to run valves, engine control avionics, may or may not
be part of a given engine. This makes it harder to compare engines on an
apples-to-apples basis. In this particular case, gimbal actuators and the
hydraulic power to run them are built into the RD-180, but AFAIK they are
not included on the RS-84.
Third, there are many elements of the propulsion system that are not part of
the engine, but nevertheless do need to be scaled with thrust. Ed's list of
kerosene stages includes only one with a stage T/W < 1 and some are over 2,
but the list of hydrogen stages includes only one with a T/W > 1. Much of
the weight to adjust for this is captured by calculating the engines
separately as you are doing, but the mass of the thrust structure, TVC, and
feed systems would also have to be adjusted accordingly. If you imagine an
S-II stage with a 50% higher thrust/weight, and scale up the thrust
structure, TVC accordingly, and scale up the purge and propellant systems by
25%, by my calculations that shifts your Ms term (burnout wt -)/ gross wt
from 0.0579 to 0.0736 if residuals are excluded, or 0.0736 to 0.0834 if
residuals are included. So, in other words, it makes a significant
difference.
Fourth, as you've all said, the details matter. All but one of the hydrogen
stages on Ed's list are second stages that go to orbit, but none of the
kerosene stages do. One would expect that mass reduction has been a higher
priority on the hydrogen stages - certainly that's known for the S-II vs
S-IC. Several of the hydrogen stages have unusual structural arrangements.
Both the Ariane EPC and the ET have most of their thrust input from the
strap-on boosters toward the top of the stage, immediately below the LOX
tank, rather than at the bottom of the stage. This makes the hydrogen tank
much more efficient because it doesn't carry the heaviest loads from the LOX
tank. In fact, my understanding is that the Ariane 5 EPC hydrogen tank
actually tapers backwards - the tank wall is thickest at the top and thinner
at the bottom.
So far, I haven't offered my own answer, other than to say "it's more
complicated than that."
One way to capture at least some of those complications is simply to plot a
large number of actual data points from existing stages that must account
for these issues, rather than trying to derive an answer yourself. A few
years ago, I put the following data together. I haven't double checked it
recently, and there's bound to be some inconsistencies in the data (e.g.
some stages probably include interages and some don't, some account for
propellant residuals and some don't, some stages weren't yet built so data
may be wrong).
If you plug it into your plotting program of choice, and plot ideal velocity
vs T/W to address the thrust/weight issue above, you'll see some things.
First, that there is plenty of scatter, so there's still room for
disagreement about what to infer. Second, that the dense fueled stages are
certainly no worse than hydrogen (which would come as a surprise to some),
and when staged combustion engines are used, kerosene comes out a little bit
better at any given T/W level. This is particularly true considering that a
kerosene ends up going a little faster for a given ideal delta V than a
hydrogen stage.
The difference does not appear to be enormous. If I were trying to build an
expendable SSTO I would consider both propellant options. But I think
Henry's point is correct.
Stage Mass Fraction Vacuum T/W Isp Ideal DV
(ft/sec)
LOX/Kerosene
Angara 0.929 1.5 338 28680
Atlas IIA 0.938 1.4 296 26516
Atlas III 0.930 2.2 338 28925
Atlas V 0.932 1.4 338 29171
Delta II 0.940 0.9 302 27332
Zenit 0.921 2.3 338 27543
Saturn V S-IC 0.941 1.6 304 27686
UDMH/N2O4
Proton 0.932 2.3 316 27353
Ariane 4 0.930 1.1 278 23805
Long March 2C 0.935 2.0 261 22899
Long March 2E 0.952 1.5 261 25404
Titan II 0.967 1.8 296 32575
Titan IV 0.947 1.5 302 28579
Kosmos 0.939 2.0 291 26241
Cyclone 0.936 2.4 301 26599
LOX/LH2
Delta 4 0.882 1.5 409 28118
Ariane 5 0.927 0.66 430 36315
H-II 0.880 1.1 445 30381
H-IIA Initial 0.880 1.0 429 29289
H-IIA Upgraded 0.880 1.0 440 30040
H-IIA LRB 0.850 1.9 440 26878
(Delta V is ideal delta v capability of the stage without a payload, usually
ignoring residuals and other details. Note that these are only abstract
measures of capability - I am not suggesting these stages could actually
perform this well in flight. They are stages, not vehicles, and many
performance issues have been glossed over.)
Josh Hopkins
Jim Davis
> First Henry Spencer wrote
> Then Ed Kyle said
> Then Jim Davis said
Josh, that was a great post. I'm sure Ed and Henry agree.
Thanks,
Jim Davis
ed kyle
>
> First, I think the average Isp number for the RS-68 is incorrect. According
> to "Propulsion for the 21st Century-RS-68" B. K. Wood, AIAA 2002-4324, the
> sea level Isp is 357 an the Vac Isp is 409. So the average is 383 s, not 393
> s (assuming by "average" you mean the arithmetic average of the two values,
> not some sort of mission averaged value.)
>
You're correct. It seems I was using outdata data. It turns
out that Boeing-Rocketdyne did publish a series of papers in
2002 that gave the 357s/409s ISP values for RS-68. See also,
for example, the 2.6 MB file:
These data, presented by the people who should know, are surely
more accurate than the 2001-ish 365s/410s ISP values that
Boeing stubbornly continues to present on its corporate web site.
The 2002 ISP values were given for 299 and 341 metric tons of
thrust, respectively, while the 2001 data was provided for
294.8 and 337.9 metric tons of thrust. I wonder if Boeing had
to add thrust at some point, sacrificing ISP in the process?
I also wonder if ISP varies with engine throttling.
Thanks for your message!
- Ed Kyle
Heinrich Zinndorf-Linker
>> First, I think the average Isp number for the RS-68 is incorrect. According
>> [...]
>> sea level Isp is 357 an the Vac Isp is 409. So the average is 383 s, not 393
>> s (assuming by "average" you mean the arithmetic average of the two values,
>> not some sort of mission averaged value.)
>You're correct. It seems I was using outdata data. It turns
>out that Boeing-Rocketdyne did publish a series of papers in
>[...]
I think, he is NUMERICALLY correct. But the simple average of sea isp
and vac isp is a unusable number, because that engine is designed to
be used in a specific mission profile. So the only number that counts
is the effective isp that is achieved during its use. And it is an
undeniable fact, that 393s is the real answer, if you have to
calculate with it - in a single core launcher configuration.
It's another game, when you want to use the numbers for calculation of
the Heavy configuration. The outer (booster) stages will have a lower
effective isp, simply because they have an other flight profile than
the core stage (which itself has a slightly higher one compared to a
single core configuration).
cu, ZiLi aka HKZL (Heinrich Zinndorf-Linker)
--
/"\ ASCII Ribbon Campaign
mailto:hein...@zili.de \ /
http://zili.de X No HTML in
/ \ email & news