Are larger manned launch systems more dangerous?

In the past, as I've pondered the various decisions made in the design
of the shuttle and various proposed replacement systems, I realized
the other day that I've been using various assumptions that are, at
best, questionable.

A major one for me has always been that the shuttle is more dangerous
because of its heavy-lift capability, and I don't know where I got
that idea. I guess my assumption was that bigger boosters mean higher
loads, more stresses, more fuel energy that can be released in case of
a booster accident, etc. This lead me to the conclusion that the
shuttle program (or any replacement) might be inherently safer if you
have a manned system just big enough to get your people to and from
orbit, and then pair this with a separate unmanned heavy-lift cargo
system (which could potentially also have on-orbit manned capability,
like a separate version of the Shuttle's mid-deck, or Spacelab).

But having realized that I've been making that assumption, it now
seems very dubious, and even if there is ANY truth to it, it could
lead to much different conclusions.

Looking at real-world systems the Shuttle isn't necessarily any more
dangerous than Soyuz. It just seems so because there have been more
flights, and the larger crew capacity gives a higher body-count when
there isn an accident. The Saturn V Apollo (based on a VERY limited
set of data points, of course) seemed like a very reliable system in
spite of its size. So it's also possible that a case can be made for
"bigger is better."

It seems that even if system size is a factor, it could easily drop
below the noise level when compared to other design decisions (to use
solid-rocket boosters, or to not provide crew pressure suits for
launch and entry, for instance).

If system size has any importance, it could have more to do with
flight rate rather than the factors I mentioned above. Build only a
few vehicles, and have a low launch rate, and you don't have a chance
to work the bugs out of the system. Less obvious problems like
O-rings or cracked panels bite you in the ass, and since you have so
many eggs in so few baskets, you get bit hard.

Smaller craft, lots of vehicles, higher flight rate, and you have a
chance to work out even low-probability failures. Some failures are
almost inevitable, but you lose less with individual failures. On the
down side though, high production and flight rates give people every
opportunity to become complacent and sloppy.

But again, the flight-rate assumptions may be badly flawed as well.
So I ask the group-mind, do you see any possible correlation between
vehicle size and safety?

For that matter, how about the mix of astronauts with heavy-lift? If
you had a shuttle-sized vehicle that took 50 people into orbit rather
than a dozen (not that there's any requirement for that now, just a
hypothetical), is there any reason it would be more or less safe than
the current mixed system?


-------------------------------------------------
J. Steven York's Multiplex of the Mind
http://member.newsguy.com/~jsteven/

J. Steven York wrote:

But again, the flight-rate assumptions may be badly flawed as well.
So I ask the group-mind, do you see any possible correlation between
vehicle size and safety?

The rate of some failures, such as a TPS failure such as destroyed
Columbia, could be proportional to the size (area?) of the relevant hardware,
all else being equal.

Paul

The rate of some failures, such as a TPS failure such as destroyed
Columbia, could be proportional to the size (area?) of the relevant hardware,
all else being equal.


With a optimum designed manned launcher size shouldnt matter much, thats my
hunch anyway its a interesting question
HAVE A GREAT DAY!

"J. Steven York"
So I ask the group-mind, do you see any possible correlation between
vehicle size and safety?

The group mind is what is called for to come up with an answer that would
do justice to this interesting question.

My meager individual mind, upon consultation, opines that the Shuttle
suffers from a full-mode escape system (there is a lot of the flight that
the passengers are along for the ride no matter what).

The Russian proposed Kliper has room for six, small payloads, and a
capsule shape that nonetheless enters the atmosphere nose first.
Launching on something similar to an Ariane 5, it would presumably have
an escape rocket in case one was needed.

The Russian Space Agency/Engergia will be able to launch this thing all
day and give you change back for your $100 million.

(bob haller) writes:

The rate of some failures, such as a TPS failure such as destroyed
Columbia, could be proportional to the size (area?) of the relevant hardware,
all else being equal.

With a optimum designed manned launcher size shouldnt matter much, thats my
hunch anyway its a interesting question

This sounds like the sort of statement a non-engineer manager would
make. Completely devoid of any actual engineering content.

Paul's original comment was actually had engineering content. Given a
random distribution of shedding foam, rain, or orbital debris, damage
to a fragile, external system like the shuttle tiles will be roughly
proportional to the area of the tiles. That statement can easily be
backed up with an analysis and test data, where your statement can't.


Furthermore, as Henry has pointed out in the past, if you have a
semi-ballistic capsule design, increasing the size of your capsule,
while holding the mass of the payload inside constant, makes things
easier. You end up with a less dense vehicle that you can more easily
control the center of gravity. Offsetting the center of gravity
results in more hypersonic lift, which results in a re-entry
trajectory that exposes the occupants to less G force. This is a fine
example of "size matters" when designing a re-entry vehicle.

Jeff
--
Remove "no" and "spam" from email address to reply.
If it says "This is not spam!", it's surely a lie.

J. Steven York wrote:
This lead me to the conclusion that the
shuttle program (or any replacement) might be inherently safer if you
have a manned system just big enough to get your people to and from
orbit, and then pair this with a separate unmanned heavy-lift cargo
system (which could potentially also have on-orbit manned capability,
like a separate version of the Shuttle's mid-deck, or Spacelab).

Henry is fond of saying that if a booster isn't safe enough to carry
people, then it isn't safe enough to carry multi-million or billion
dollar payloads. There is a certain amount of truth to this, even if
it does seem to 'place a price on human life'. (Something done on
daily basis by all manner of people and organizations, just not as
nakedly or as publicly as in space travel.)

The Saturn V Apollo (based on a VERY limited set of data points,
of course) seemed like a very reliable system in spite of its size.
So it's also possible that a case can be made for "bigger is better."

The conclusion that the Saturn was reliable is an extremely shaky one,
not just because of the paucity of data, but because over half the
flights had significant problems. Most people regard success on a
binary scale, flights are a success or not, missions are completed, or
not, but that's greatly oversimplifying matters. In reality is a
sliding scale with many intermediate positions and shades of grey.

Smaller craft, lots of vehicles, higher flight rate, and you have a
chance to work out even low-probability failures.

Smaller craft are no guarantee of having a higher flight rate, this
can be clearly with Soyuz, which despite being far smaller than
Shuttle, and having a much longer service life, has actually flown
fewer flights.

But again, the flight-rate assumptions may be badly flawed as well.
So I ask the group-mind, do you see any possible correlation between
vehicle size and safety?

I don't see that assumption as being flawed. You can't work out the
bugs unless you fly the craft.

D.
--
Touch-twice life. Eat. Drink. Laugh.

jeff findley wrote:
Furthermore, as Henry has pointed out in the past, if you have a
semi-ballistic capsule design, increasing the size of your capsule,
while holding the mass of the payload inside constant, makes things
easier.

However, that requires handwaving away the issue of booster diameter.
There's going to be hard limits on how far you can 'hammerhead' a
given booster.

You end up with a less dense vehicle that you can more easily
control the center of gravity.

That's a qualified 'maybe' at best, and more handwaving on Henry's
part. As you get larger, the weight of structure and TPS starts to
dominate over that of systems and payload. It also matters what kind
of payload(s) you are carrying and whether or not you are returning
them as well. If you aren't, you have to deal with the issue of
ensuring that both launch and recovery CG's are within limits.

Offsetting the center of gravity results in more hypersonic lift, which
results in a re-entry trajectory that exposes the occupants to less G
force. This is a fine example of "size matters" when designing a
re-entry vehicle.

Again, not completely true. Henry bases his comments on the problems
they had with the Apollo capsule. A 2004 model CM would likely have
far fewer problems because the systems would be lighter and smaller,
giving you more options to move them about to optimize your CG. In
this instance it not size that matters, but your weight and volume
constraints.

D.
--
Touch-twice life. Eat. Drink. Laugh.

(Derek Lyons) writes:

J. Steven York wrote:
The Saturn V Apollo (based on a VERY limited set of data points,
of course) seemed like a very reliable system in spite of its size.
So it's also possible that a case can be made for "bigger is better."

The conclusion that the Saturn was reliable is an extremely shaky one,
not just because of the paucity of data, but because over half the
flights had significant problems. Most people regard success on a
binary scale, flights are a success or not, missions are completed, or
not, but that's greatly oversimplifying matters. In reality is a
sliding scale with many intermediate positions and shades of grey.

This is true. Partly because the thing was so big and expensive and
partly because of the goal to land on the moon "before the decade is
out", Saturn V was used for manned missions more than a bit early.
There were only two unmanned test flights (Apollo 4 and Apollo 6)
before Apollo 8. Apollo 6 had severe pogo. It would have been nice
to have at least one more test flight to verify that the severe pogo
problem was completely solved.

Based on the total number of Saturn V flights (only thirteen total),
one could easily argue that the program ended before NASA could be
certain that it was completely debugged.

Jeff
--
Remove "no" and "spam" from email address to reply.
If it says "This is not spam!", it's surely a lie.

(Derek Lyons) writes:

jeff findley wrote:
Furthermore, as Henry has pointed out in the past, if you have a
semi-ballistic capsule design, increasing the size of your capsule,
while holding the mass of the payload inside constant, makes things
easier.

However, that requires handwaving away the issue of booster diameter.
There's going to be hard limits on how far you can 'hammerhead' a
given booster.

Perhaps, but that depends on the booster.

A quick search turns up a Delta V Heavy payload shroud with a diameter
of five meters (16.5 feet). That will allow for quite a bit more
internal volume than the Apollo CM's 3.9 meter diameter allowed.

You end up with a less dense vehicle that you can more easily
control the center of gravity.

That's a qualified 'maybe' at best, and more handwaving on Henry's
part. As you get larger, the weight of structure and TPS starts to
dominate over that of systems and payload. It also matters what kind
of payload(s) you are carrying and whether or not you are returning
them as well. If you aren't, you have to deal with the issue of
ensuring that both launch and recovery CG's are within limits.

The weight of the structure and TPS will scale more closely with the
area of the vehicle than the volume. The volume of the vehicle will
increase faster than the area, thanks to the square-cube law. It is
volume that will give you flexibility in moving internal systems and
payload.

Offsetting the center of gravity results in more hypersonic lift, which
results in a re-entry trajectory that exposes the occupants to less G
force. This is a fine example of "size matters" when designing a
re-entry vehicle.

Again, not completely true. Henry bases his comments on the problems
they had with the Apollo capsule. A 2004 model CM would likely have
far fewer problems because the systems would be lighter and smaller,
giving you more options to move them about to optimize your CG. In
this instance it not size that matters, but your weight and volume
constraints.

Note that the same applies to the structure and to the TPS. This is
especially true when you consider the Apollo CM's TPS was over
designed and therefore overweight. The engineers at the time chose to
be conservative with the design because they didn't have the heating
hard data (for such a fast re-entry) they needed to shave the weight
of the TPS down to a bare minimum.

A modern capsule design capable of re-entry from lunar missions will
have lighter structure and TPS than that of the Apollo CSM.


I'll agree that the devil is in the details. Worst case, if you have
empty volume, you can add ballast to shift the CG. According to the
old STS Newsref web site:

Lead ballast in the nose wheel well and on the X o 378 bulkhead
provides weight and center-of-gravity control. The nose wheel well
will accommodate 1,350 pounds of ballast, and the X o 378 bulkhead
will accommodate a maximum of 2,660 pounds.

Ballast isn't an elegant solution to keep the your CG within limits,
but it is a solution you can implement if you've got empty volume in
your design. The more empty volume you've got, the more opportunity
you've got to fill that space with less dense ballast that could be
potentially useful as opposed to something far more dense, but far
less useful, like lead ballast.

Useful ballast would include things like survival gear for "off
nominal" landings. This could mean things like bigger and better life
rafts or a larger food and drinking water supply.

Jeff
--
Remove "no" and "spam" from email address to reply.
If it says "This is not spam!", it's surely a lie.

jeff findley wrote:
Based on the total number of Saturn V flights (only thirteen total),
one could easily argue that the program ended before NASA could be
certain that it was completely debugged.

I think it's safe to say the biggies (Pogo!) were mostly sorted out,
as were operational issues. Whether any other low hanging fruit
remained, I don't know. This of course doesn't mean that the SV was
free of systemic (as in SRB O-ring) problems that simply never came up
due to the limited number of flights, despite having a non-trivial
chance of occuring.

Most peoples impression and arguments are based on the fact that on
the two flights with the most severe problems (6 and 13), there were
direct and primary errors in manufacture and handling that lead to the
problems. This is emotion, not engineering.

D.
--
Touch-twice life. Eat. Drink. Laugh.