Briefing on SRB based CEV at NPS with Scott Horowitz

Scott Horowitz, former shuttle pilot and astronaut, came to the Naval
Postgraduate school today to teach a lecture to my class of Space
Systems Engineers & gave a very convincing argument for the SRB to
orbit option. (He works for ATK, but keep reading.) There were quite a
few details that I found interesting in his talk, and many of my intial
skepticisms about the idea were soundly refuted.

Here's a synopsis:
The summer of 2003, after OSP was announced and most of the spaceflight
office at Johnson was shaking their heads, several of the engineers
started looking at this option as an alternative after OSP was
inevitably cancelled.
What they determined from an overall perspective (much like the SpaceX
Futron study) is that maximizing reliability on larger systems means
minimizing component integration complexities--something the shuttle is
particularly ghastly at, and even EELVs like the Delta-IVH, with 3
cores, are bad about. (witness the first delta IV H failed!). That
means best reliability comes with 1 first stage engine, 1 separation
event, and 1 second stage engine.
The problem is that no liquid engines currently exist that are powerful
enough to (with one engine) thrust the first stage of a manned capsule
off the pad. Solution--make the first stage an SRB (3 million lbs of
thrust).

I've heard all this before, and I had some preconceived skepticisms,
among them:
1. Didn't an SRB failure cause the Challenger accident?
2. Don't SRB's cost alot?
3. Isn't that too much thrust to handle for a small vehicle?
4. thus Wouldn't extensive modifications to the burn profile be
required, costing even more?
5. Does the steerable nozzle have enough control authority?

Scott Horowitz was pretty convincing that these are not legitimate
concerns, for the following reasons:
1. Yes, but the explosion happened because hot gas burned through the
LOX/H2 tank. Solid rockets don't explode. If there wasn't a fuel tank
in the way, the Challenger probably still could have made orbit, as
most of the thrust was still going out the nozzle. Solid boosters DONT
explode like Lox/ H2, they just leak hot gas. A capsule on top of a
worst-case failed SRB (case burst, which has never happened in 240 some
SRB flights), even on the ground, where propogation is the worst, sees
only a 10 psi overpressure. Very survivable.
2. Yes, the shuttle contract costs the govt a fixed amount--but it
generally operates at ~15-20% of capability. In other words, if the
line is going to stay open through 2010 for shuttle anyway, why not use
that extra capacity to built boosters?
3. No, just put a bigger second stage on top of the SRB and
voila--thanks to the rocket equation, G loading is as variable as you
want to make the second stage. With an RP-1/Lox upperstage, Isp ~340,
and 250,000lbs propellant, a 28,000 lb payload (Astronauts) see a max
of 3Gs of acceleration, and a max Q about the same as the shuttle.
4. No, all proposed schemes use the exact same burn profile as the
shuttle SRB. In fact, it is a shuttle SRB with different GNC software
and an interstage instead of a nosecose.
5. Absolutely, in fact, it can launch under worse conditions than the
shuttle. Finite element analysis modeling shows that with the ker/lox
upperstage, CG, or rather cm is in about the same place as the shuttle,
and even under worse case wind conditions, etc, the steerable nozzle
has more than enough control authority to handle it, and lots of margin
to spare.

Some other points he made:
6. the reliability and safety of the SRB is extremely well known--thus
the system reliability depends nearly entirely on the reliability of
the second stage. For the 'best fit' notional model, he was using a
Saturn V J2S, although it works with a J2. That would require
re-developing the big ker/lox engine--big deal and much moolah, but it
would probably be a popular project. Also, rocketdyne has kept much of
the J2 knoweledge in working order, even using the turbopumps on the
X-33 linear aerospikes. It could probably be done pretty quickly.
7. One of the first people outside-of-NASA people briefed in on this as
a "reality check" was the then-head of APL, Mike Griffen. And he
apparently was very impressed by its possibilities for success. So
support inside NASA should not be a problem.
8. We should know if this is going forward within the next 2 months,
because that's the time limit to complete a people carrying version by
2010. If a decision is not made by then, it's not going to happen. So
standby...

Tom Cuddihy

Apparantly he left out the part where large segmented solids have gone
"boom" in the past. See Titan IV failure caused by a design flaw in the
propellant grain at the joints. The joints in large segmented solids are
definately a problem and can go "boom".

Jeff
--
Remove icky phrase from email address to get a valid address.

Apparantly he left out the part where large segmented solids have
gone
"boom" in the past.

That has never happened with the Shuttle motors in either flight or
groudn test.

wrote:
Apparantly he left out the part where large segmented solids have
gone
"boom" in the past.

That has never happened with the Shuttle motors in either flight or
groudn test.

Right.

The Challenger type burnthrough would be relatively harmless on
a single SRB first stage; there's nothing for it to cut into and
cause to fail, other than slowly eroding a hole in the SRB.
Which the Challenger burnthrough would not have done fast enough
to cause a flight accident, if I recall the analysis right.

You have to have enough control authority if there's sideways
thrust and loss of thrust due to such a burnthrough, but that
should be manageable.

Having the nozzle depart would likewise be spectacular but
fairly safe.

But... large SRBs *have* catastrophically dissassembled
themselves or outright exploded recently. Any large SRB
that got an internal flow disruption due to propellant
grain structural failure could CATO.

Can an overall launcher be safe enough to be escapable
in case of CATO? Probably. Big bad day though.


-george william herbert

wrote:
Apparantly he left out the part where large segmented solids have
gone
"boom" in the past.

That has never happened with the Shuttle motors in either flight or
groudn test.

True, but "The shuttle RSRMs have never gone BOOM" is a much weaker
statement than "solid rocket motors don't go BOOM." I seem to recall a
Delta solid going BOOM and raining bits of GPS satellite all over a
bunch of parked cars a few years back as well.

Without hearing Horowitz's defense of the SRBs word-for-word it's hard
to be specific, but claiming that "if there wasn't a fuel tank in the
way, Challenger would still have probably made orbit, as most of the
thrust was still going out the nozzle" strikes me as verging on
dishonest. If the contents of the ET were somehow magically rendered
inert, the flame from the SRB was still going to cause failure of the
aft attachment point and subsequent mechanical failure of the entire ET
and aerodynamic destruction of the stack.

-jake

Tom Cuddihy wrote:
Scott Horowitz, former shuttle pilot and astronaut, came to the Naval
Postgraduate school today to teach a lecture to my class of Space
Systems Engineers & gave a very convincing argument for the SRB to
orbit option. (He works for ATK, but keep reading.) There were quite a
few details that I found interesting in his talk, and many of my intial
skepticisms about the idea were soundly refuted.

Too bad nobody told me it was happening, I would have driven down
and offered some useful criticism...

Here's a synopsis:
The summer of 2003, after OSP was announced and most of the spaceflight
office at Johnson was shaking their heads, several of the engineers
started looking at this option as an alternative after OSP was
inevitably cancelled.
What they determined from an overall perspective (much like the SpaceX
Futron study) is that maximizing reliability on larger systems means
minimizing component integration complexities--something the shuttle is
particularly ghastly at, and even EELVs like the Delta-IVH, with 3
cores, are bad about. (witness the first delta IV H failed!). That
means best reliability comes with 1 first stage engine, 1 separation
event, and 1 second stage engine.
The problem is that no liquid engines currently exist that are powerful
enough to (with one engine) thrust the first stage of a manned capsule
off the pad.

This is not true. Both the Atlas V and Delta IV boosterless models
have enough weight to orbit a reasonably designed midsized capsule
with 4-6 seats. Both are single engine first stage launchers.

The other peak in reliability should come around 5 or more engines
per stage, as Saturn V's S-IC, S-II, and the upcoming Falcon V
first stage all do. But it's harder to engineer and will have
more soft failures than single engine stages.

Solution--make the first stage an SRB (3 million lbs of
thrust).


I've heard all this before, and I had some preconceived skepticisms,
among them:
1. Didn't an SRB failure cause the Challenger accident?
2. Don't SRB's cost alot?
3. Isn't that too much thrust to handle for a small vehicle?
4. thus Wouldn't extensive modifications to the burn profile be
required, costing even more?
5. Does the steerable nozzle have enough control authority?

Scott Horowitz was pretty convincing that these are not legitimate
concerns, for the following reasons:
1. Yes, but the explosion happened because hot gas burned through the
LOX/H2 tank. Solid rockets don't explode.

Large solid rockets can catstrophically fail, and have both
in ground test and flight recently. Titan IV had a really truly
spectacular ground test failure. A Delta II had a GEM let go
and blow chunks of the launch vehicle all over the pad, blockhouse,
and parking lot (and the Cape's air force museum).

If there wasn't a fuel tank
in the way, the Challenger probably still could have made orbit, as
most of the thrust was still going out the nozzle. Solid boosters DONT
explode like Lox/ H2, they just leak hot gas. A capsule on top of a
worst-case failed SRB (case burst, which has never happened in 240 some
SRB flights), even on the ground, where propogation is the worst, sees
only a 10 psi overpressure. Very survivable.

10 psi overpressure is a lot. Well built concrete
buildings get knocked down by 10 psi overpressure.

It won't kill the people, but it may damage a capsule
beyond safe escape or post-escape landing.

2. Yes, the shuttle contract costs the govt a fixed amount--but it
generally operates at ~15-20% of capability. In other words, if the
line is going to stay open through 2010 for shuttle anyway, why not use
that extra capacity to built boosters?
3. No, just put a bigger second stage on top of the SRB and
voila--thanks to the rocket equation, G loading is as variable as you
want to make the second stage. With an RP-1/Lox upperstage, Isp ~340,
and 250,000lbs propellant, a 28,000 lb payload (Astronauts) see a max
of 3Gs of acceleration, and a max Q about the same as the shuttle.
4. No, all proposed schemes use the exact same burn profile as the
shuttle SRB. In fact, it is a shuttle SRB with different GNC software
and an interstage instead of a nosecose.

What, no gentler thrust termination system?

5. Absolutely, in fact, it can launch under worse conditions than the
shuttle. Finite element analysis modeling shows that with the ker/lox
upperstage, CG, or rather cm is in about the same place as the shuttle,
and even under worse case wind conditions, etc, the steerable nozzle
has more than enough control authority to handle it, and lots of margin
to spare.

Some other points he made:
6. the reliability and safety of the SRB is extremely well known--thus
the system reliability depends nearly entirely on the reliability of
the second stage. For the 'best fit' notional model, he was using a
Saturn V J2S, although it works with a J2. That would require
re-developing the big ker/lox engine--big deal and much moolah, but it
would probably be a popular project. Also, rocketdyne has kept much of
the J2 knoweledge in working order, even using the turbopumps on the
X-33 linear aerospikes. It could probably be done pretty quickly.

There has to have been some error there; the J-2s were LOX/LH2
upper stage engines, not lox/kerosene as described elsewhere
and partly here.

J-2 isn't a bad choice, nor is LOX/LH2 for the upper stage,
but clarification would have been nice.

7. One of the first people outside-of-NASA people briefed in on this as
a "reality check" was the then-head of APL, Mike Griffen. And he
apparently was very impressed by its possibilities for success. So
support inside NASA should not be a problem.
8. We should know if this is going forward within the next 2 months,
because that's the time limit to complete a people carrying version by
2010. If a decision is not made by then, it's not going to happen. So
standby...

This launcher may well happen, but I remain unconvinced
that it's really a better choice than either of the two
EELVs in their boosterless configuration.

I do have this crazy idea to see what happens if you stack
a Delta IV core on top of a SRB, though...


-george william herbert

"Tom Cuddihy" wrote in message
oups.com...
Scott Horowitz, former shuttle pilot and astronaut, came to the Naval
Postgraduate school today to teach a lecture to my class of Space
Systems Engineers & gave a very convincing argument for the SRB to
orbit option. (He works for ATK, but keep reading.) There were quite a
few details that I found interesting in his talk, and many of my intial
skepticisms about the idea were soundly refuted.

Here's a synopsis:
The summer of 2003, after OSP was announced and most of the spaceflight
office at Johnson was shaking their heads, several of the engineers
started looking at this option as an alternative after OSP was
inevitably cancelled.
What they determined from an overall perspective (much like the SpaceX
Futron study) is that maximizing reliability on larger systems means
minimizing component integration complexities--something the shuttle is
particularly ghastly at, and even EELVs like the Delta-IVH, with 3
cores, are bad about. (witness the first delta IV H failed!). That
means best reliability comes with 1 first stage engine, 1 separation
event, and 1 second stage engine.
The problem is that no liquid engines currently exist that are powerful
enough to (with one engine) thrust the first stage of a manned capsule
off the pad. Solution--make the first stage an SRB (3 million lbs of
thrust).


The RD-171 has 740 tons of thrust, so I suppose this means no U.S. engine. A
good reason for bringing back the F1.

I've heard all this before, and I had some preconceived skepticisms,
among them:
1. Didn't an SRB failure cause the Challenger accident?
2. Don't SRB's cost alot?
3. Isn't that too much thrust to handle for a small vehicle?
4. thus Wouldn't extensive modifications to the burn profile be
required, costing even more?
5. Does the steerable nozzle have enough control authority?

Scott Horowitz was pretty convincing that these are not legitimate
concerns, for the following reasons:
1. Yes, but the explosion happened because hot gas burned through the
LOX/H2 tank. Solid rockets don't explode. If there wasn't a fuel tank
in the way, the Challenger probably still could have made orbit, as
most of the thrust was still going out the nozzle. Solid boosters DONT
explode like Lox/ H2, they just leak hot gas. A capsule on top of a
worst-case failed SRB (case burst, which has never happened in 240 some
SRB flights), even on the ground, where propogation is the worst, sees
only a 10 psi overpressure. Very survivable.
2. Yes, the shuttle contract costs the govt a fixed amount--but it
generally operates at ~15-20% of capability. In other words, if the
line is going to stay open through 2010 for shuttle anyway, why not use
that extra capacity to built boosters?
3. No, just put a bigger second stage on top of the SRB and
voila--thanks to the rocket equation, G loading is as variable as you
want to make the second stage. With an RP-1/Lox upperstage, Isp ~340,
and 250,000lbs propellant, a 28,000 lb payload (Astronauts) see a max
of 3Gs of acceleration, and a max Q about the same as the shuttle.
4. No, all proposed schemes use the exact same burn profile as the
shuttle SRB. In fact, it is a shuttle SRB with different GNC software
and an interstage instead of a nosecose.

Isp of 340 seconds is very high for a gas generator cycle lox-kero engine.
You would need an expansion ratio of about 200. A version of the Atlas
booster engine with extended nozzle would have the right thrust, but would
be about 18 feet wide at the nozzle exit, too large to fit on any
reasonably-sized upper stage.
It's more likely that the Isp would be around 330 seconds, and you'd be down
several thousand lbs of payload.

5. Absolutely, in fact, it can launch under worse conditions than the
shuttle. Finite element analysis modeling shows that with the ker/lox
upperstage, CG, or rather cm is in about the same place as the shuttle,
and even under worse case wind conditions, etc, the steerable nozzle
has more than enough control authority to handle it, and lots of margin
to spare.


This is very sensible. The SRB has plenty of thrust to lift the heavier
upper stage.

Some other points he made:
6. the reliability and safety of the SRB is extremely well known--thus
the system reliability depends nearly entirely on the reliability of
the second stage. For the 'best fit' notional model, he was using a
Saturn V J2S, although it works with a J2. That would require
re-developing the big ker/lox engine--big deal and much moolah, but it
would probably be a popular project. Also, rocketdyne has kept much of
the J2 knoweledge in working order, even using the turbopumps on the
X-33 linear aerospikes. It could probably be done pretty quickly.
7. One of the first people outside-of-NASA people briefed in on this as
a "reality check" was the then-head of APL, Mike Griffen. And he
apparently was very impressed by its possibilities for success. So
support inside NASA should not be a problem.
8. We should know if this is going forward within the next 2 months,
because that's the time limit to complete a people carrying version by
2010. If a decision is not made by then, it's not going to happen. So
standby...

Tom Cuddihy


Murray Anderson

George William Herbert wrote:

I do have this crazy idea to see what happens if you stack
a Delta IV core on top of a SRB, though...

Maybe 15 or so metric tons to LEO, assuming you use
the RS-68 somehow. At 26-27 tons empty, CBC is a bit
on the heavy side for an upper stage.

An S-IVB with a J-2 would be almost perfect. S-IVB
only weighed 10 tons (or a bit less) at burnout. I
get about 18 tons (about 40,000 lbs) to LEO with this
configuration. It's kind of strange how this works
out. You get a Saturn IB, basically, but without all
of the mad plumbing. It's as if the SRBs were
designed to provide the same total impulse as an S-IB
stage.

An RP/LOX second stage, fitted with a notional,
throttlable RS-27A would be able to orbit maybe
10 tons (22,000 lbs). Not too bad, and maybe
cheaper than an S-IVB type second stage, but
an Atlas V Medium could do the same for probably
less money.

To really cut costs, NASA should abandon the VAB and
stack its SRB-based single-stick launcher on a pad.
How about recently-vacated Pad 40?

- Ed Kyle

George William Herbert wrote:

There has to have been some error there; the J-2s were LOX/LH2
upper stage engines, not lox/kerosene as described elsewhere
and partly here.

J-2 isn't a bad choice, nor is LOX/LH2 for the upper stage,
but clarification would have been nice.

That's my error,not his. We were doing quite a bit of different
possibility discussions and I'm not sure how I got that mixed up. His
notional case used a J2--yes, LOX/H2, and there was another one he had
analyzed that was ker/Lox. I got mixed up somewhere in there & didn't
look up which was which.

Jake McGuire wrote:

wrote:


Apparantly he left out the part where large segmented solids have


gone
"boom" in the past.

That has never happened with the Shuttle motors in either flight or
groudn test.



True, but "The shuttle RSRMs have never gone BOOM" is a much weaker
statement than "solid rocket motors don't go BOOM." I seem to recall a
Delta solid going BOOM and raining bits of GPS satellite all over a
bunch of parked cars a few years back as well.


I recall lots of liquid boosters raining payload all over the landscape.
In fact, it's one of my current tasks... working on a historical summary
of launch vehicle reliability. A whole fo of rockets going BOOM for a
whole lot of reasons.

Without hearing Horowitz's defense of the SRBs word-for-word it's hard
to be specific, but claiming that "if there wasn't a fuel tank in the
way, Challenger would still have probably made orbit, as most of the
thrust was still going out the nozzle" strikes me as verging on
dishonest.

It's true. The analyses showed that Challenger should have made orbit,
had the burn-though been outboard. Performance would have been affected,
and thrust vectoring systems on both the SRBs and the SSMEs would ahve
been nearly maxed out, but Challenger would ahve made orbit.