Space Question

[cheshirenoir]

This one goes out to the Space Shot lovers out there (Yes, krjalk, I am looking at you)

So I have been looking at rockets again, specifically multithruster rockets similar to what Armadillo are eventually aiming to build. Now the Soviets played around with these and had no end of problems, most commonly that variences between thrusters couldn't be countered fast enough and the stack tended to fall over. (There was also some pogoing, but I wonder if it was related)

Now the way I have been thinking about it, we're basically balancing a long thin construction on a relatively small vector. It's like balancing a pencil on the tip of your finger. It is inherently unstable so you are having to continuously "tweak" your engine output.

Now what stops you mounting some Peroxide thrusters higher up your stack and rather than trying to balance everything from below, you do your "quick" adjusts from the top of the stack, which means your primary rockets don't need to be anywhere as sensitive in terms of response time?

I am sure there is something I'm missing (The obvious one is that you are putting stresses across the entire stack frame that may make your stack snap in two) but what?

Comments

[krjalk.livejournal.com]

As I understand it he N-1 failures were mostly due to combustion instabilities due to the fiendishly complicated plumbing, with the attitude control issue being a consequence of that rather than a root cause in and of itself. Multi-engine vehicles are tricky, but it's a balancing act against the trickiness of handling fewer very large engines which have their own issues. The Soyuz has certainly been reliable for nearly half a century, and SpaceX just successfully test fired a five engine cluster on their Falcon-9 test stand with a full 9 engine firing coming up in the northern autumn.

The top-of-my-head issues I can see with your thruster proposals -

Your first point about cross-stresses certainly has validity. The walls on these puppies are wafer-thin, to the point that some of them can't stand unsupported on the launch pad unless fully fueled. The weight budget for strengthening the stack may make it uneconomical in terms of reduced payload.

Another point is the responsiveness of the thrusters. If there is too much of a lag in getting full thrust out of a pulse from the stabilising rocket then gimballing the main engine may react quicker.

A third possibility is that they might not be powerful enough. Gimballing the main engines by a small amount provides a fairly high degree of angular momentum control, particularly given that it is occurring at the fulcrum. Think of it like this, in terms of your pencil analogy - which would give a greater control authority, moving your finger at the fulcrum point, or having a ring of people blowing on it at the tip? It could be that in order to get enough control the thrusters would have to be powerful enough that the required fuel may eat into the payload margin again.

That's all I can think of right now, and that's obviously without referring to any texts or doing any sums, but I can't think of any main launch vehicle that has ever been proposed using this method. The Apollo LM used a combination of attitude thrusters and main engine gimballing during the lunar landings, but that's a way different environment than an Earthside launch.

[cheshirenoir.livejournal.com]

Thanks for that.

You have answered my question, and then some.

Mind you, it was discussions with you about VentureStar and the Clipper that re-energised my interest in space tech.

[discordia13.livejournal.com]

Acoustic vibration would be a major problem. I could see tall stacks having interesting waveforms in addition to the issue of protecting the lower stack from exhaust heating.

[rdmasters.livejournal.com]

K has it right about the N1 failures - they was essentially plumbing issues (in one case a single loose nut came adrift and screwed a turbopump - and when it came apart, so did everything else).

Now the thing is, there are birds that do the thruster trick - but in conjunction with gimballing or throttling. DC-X, for example.

As you suspected, the primary problem is that you have (with your typical tall stack), a relatively flexible, but compressively strong structure. Provided the forces are mostly compressive, the whole thing is OK, but the moment you apply a side vector (especially near one end, away from the primary force line), you will start getting structural issues.

In a short-stack or rigid system, though, it is viable, as the DC-X, or any number of air-air missiles with front guidance fins have demonstrated.

As far as your analogy goes, try this little experiment for size:
Take a 1m mailing tube, with its endcap in. Balance it on four fingers (two from each hand - index and middle). Now lift it. You get quite a bit of control, don't you?

[rdmasters.livejournal.com]

Oh, and of course, once you are in free-fall and outside of an atmosphere, all bets are off!

[decrypt-era.livejournal.com]

All the engineering concerns raised by the others are valid,
but i thought it might clarify things somewhat
if we take this back to physics fundamentals.

Firstly, you're right, a rocket is an inherently unstable design.
Ideally, we'd like to apply the accelerating force above the centre of mass.
But in practical terms,
this means pointing the exhaust back at the vehicle.
We could tilt the thrusters outward slightly,
but this translates to wasted thrust, ie: wasted reaction mass.
We could mount the thrusters on arms,
but they'd hafta be strong enough to transmit the force,
and include fuel pipelines - again, all extra mass.

Secondly, you're right, fuselage strength aside,
even small thrusters near the top of a rocket
would add greatly to th leverage available to turn the vehicle.
You'll notice many military missiles
(esp. ones aimed at moving targets)
have puta-controlld canards etc for this purpose
(though, as the rocket and its thrust get larger,
fins have less and less relative effect, of course).
But, are radical manoeuvres and the extra mass they entail
really necessary for a vehicle designed to arc gracefully into orbit?

As Tsiolkovsky pointed out so elegantly,
each extra kilo carried translates exponentially
into reaction mass on the launch pad,
so th engineers ask of each and every gram:
are you necessary?

At core, th answer to yr question is to note
that some things humans find hard, computers find really easy (& vice versa).
I'v seen footage of robot arm/sensor combos
which could do the pencil balancing trick so responsively
you'd swear th thing was glued there.
The calculations aren't that complex,
all it takes is really fast reflexes.
It turns out that a more stable application of force
is simply not necessary in spacebound rockets.
The primary design problem is how to make a machine
which can reliably channel high temperatures and pressures,
and yet be as light as possible.
Cutting this tradeoff too close is what makes rockets go boom.