Do all modules need to be steerable?

I don't remember seeing anything about his, but do all the module in a AA modular rocket need to be steerable?

I was thinking that if you only had steerable engines (gimbals or thottleable) around the outside, you could get away with (a lot) cheaper/lighter modules in the middle. Yes, you do need two sorts of module, but they are still very similar.

Perhaps you only need steerable in the corners of the module rack? Even lighter and cheaper.

I'm thinking along these lines... 0=engine X= steerable engine

XOX
OOO
XOX

or

XOOOX
OOOOO
OOOOO
OOOOO
XOOOX

or

XXXXX
XOOOX
XOOOX
XOOOX
XXXXX


James

We are hoping that simply differential throttling of the modules gives sufficient control authority, in which case steering is "free". It worked fine on our original little peroxide vehicles, but we won't know for sure on the big biprops until we try and fly the four module system that way.

John Carmack

How do you plan roll control with differential throttle though? I heard before that you were able to do this by canting your engines, but I can't see how that would apply a roll torque unless you canted them off to the sides of the CG in which case you'd get some interesting forces in flight I think.

If I remember correctly from the peroxide days, they had two pairs of engines, each slightly tilted a certain direction. Each pair (on opposite sides of the vehicle) would generate either the clockwise or counter-clockwise roll. So, you would have:


Increasing the throttle of both A-engines would e.g. generate a C-wise roll, whereas increasing it for the B-engines would give a CC-wise roll. For translation to the right, the throttle of the 1-engines (which are both pointing "outwards") would be increased (2 for the left, etc). I hope I made myself clear (and remember it correctly).

Cheers

That sounds fairly messy in terms of interaction between the motors, I guess that could be handled with software but it would still put fairly specific requirements on when you could roll and at what rate.

Probably a dumb question, but is it really for free? Would a module that was simply on/off be cheaper than one that had differential throttling? Or do you need the same plumbing and control systems for both?

You would still only need two module types, throttleable/steerable, and thrusty (Can I copyright that ?! Thrusty Module sounds like a pr0n star name though)

James

I assume that the two alternatives are:

- Differential throttle on some/all modules
- Gimbal on some/all modules

I'd think that when compared to actuator controlled gimbal, differential throttle would be practically free. Especially if you consider (or I assume) a few things.

One is that your vehicle already has multiple engines, so differential throttling is even a possibility and there is no need to radically adapt the design.

Two is that the only major requirement for a differential throttle system is a variable flow valve and servo system on your main propellant lines. I'm not exactly sure about their other vehicles but it looks like armadillo almost exclusively uses variable valves over open/close types, so it's likely that the option of throttlability would be built into even a gimbal module, making the throttling system pretty much exactly 'free'. Also I'd think at these valve sizes and mass flows it would be harder to get all on or off valves compared to motor actuated ball valves.

Wandering somewhat off-topic again. Another useful feature of differential throttling is that it's more compact, so it's aerodynamically more tidy. It also leaves more room for bigger engines, useful if it seems attractive to use stretched modules for the first stage, reducing base drag, increasing the the ballistic coefficient and making it easier to improve the drag coefficient without as big an impact on the module mass ratio. It's possible that might make enough of a difference to be able to shed a whole stage.

Ok, refering to my original diagrams, replace X with gimballed, and O with differential. Best of both worlds? (If differential alone doesn't give enough control)?

Should keep the costs to a minimum. Modules still almost the same, just a bit more hardware/software on the gimballed engines.

James

Landing a VTVL with half its engines unthrottleable would be all kinds of exciting.

Just turn them off?

You don't need half as many engines to land as take off.

Well, half may be an exaggeration but you know what I mean.

Would also mean that unthrottlable engines need smaller fuel tank (You dont need them on the way down), making them even cheaper to make.

I guess what I am saying is that as long as the modules all bolt together - ie fixing points are standard - they don't need to be the same - some for steering, some just for thrust on way up etc. Keep the number of different types small - two or three different ones - throttleable, non-throttleable big tank, non-throttleable small tank (boosters).

James

James,

The issue with something like that is that, while yes, you are improving efficency and the like, you are increasing development costs, and quite possibly opperational costs. The increase in efficency might be cost effective compared to the other costs, but they might not be. Historically, the push has been for better performance and efficency, at the expensive of cost. We see that clearly with the shuttle.

That was part of the idea of the original Otrag, as well as AA's plans - costs must be considered most important

There is exactly zero overhead to having a throttling engine versus a fixed thrust engine in our designs. Most engines are considered "not throttleable" for the simple reason that the designers are afraid that bad things will happen when they are throttled and don't want to test those conditions, rather than for any mechanical reason.

The idea of shutting down some of the engines to land is necessary if the mass ratios get much higher and / or the initial chamber pressures go lower. A vehicle with a mass ratio of 8 and an initial T/W of 1.5 and a 150 psi initial chamber pressure would not be able to land, because the chamber pressure would be below the critical sonic pressure, and you would get a blowtorch instead of a rocket engine, and negligable thrust. Engine efficiency at landing has dropped of significantly by then as well. If you had eight modules and only landed on four, you could double the ratio of initial thrust to minimum landing thrust.

John Carmack

Hello, John,

you say



This issue is something I am supposing since long. To some degree it seems to have to do with assuring against responsibility if throttling causes damagesm accidents or even death, to some other degree it seems to have to do with psychology which is frustrating.

So I find it very positive hwo you handle it. You are doing the tests that enhance and increase experience regarding application, using and handling those engines, the knowlege and the Know How. I feel tempted to call it a kind of stress- and crash-testing others should have done (in Germany particular companies, foundations, federations and customer-oriented agencies are doing it).

Very fine.

Can you imagine to write a documentation about it for others - without unveiling your technical and business secrets? It might further the whole business of private space vehicle development and design and earn you a bit more funds.

What about it?



Dipl.-Volkswirt (bdvb) Augustin (Political Economist)

Thanks John, an perfect answer to the question!!

James