Idea for modular booster | Booster | Forum

I know we're supposed to be working hard on organisational/legal stuff at the moment, but I can feel all the NPC complications sapping my enthusiasm, which I don't want to see happen, so here's some technical stuff to keep people motivated.

Elsewhere I mentioned that I felt a little uneasy about Selene 1 and 2 consisting of stages of a single modular booster, hence requiring us to develop mechanically complicated gimballed nozzles which would eventually be of no further user, since Selene 3 and 4 have clustered stages allowing us to use differential thrusting (which I feel should be much simpler mechanically).

My first proposed solution to this was to make our modular boosters smaller and have Selene 1 and 2 be clustered designs too.

Perhaps a better idea would be to take a sort of nested approach.  Each modular booster could in fact be a cluster of 4 narrower boosters mounted in a single cylindrical shell.  When that shell is used standalone, as per Selene 1 and 2, we can use differential throttling of the interior engines to achieve guidance – no gimballing required.  When a few of these modules are linked together, as per Selene 3 and 4, we set it up so that all 4 sub engines in each module can be throttled simultaneously via a single control signal, and then we can do differential throttling of the overal structure.  I hope this explanation is clear, let me know if not.

Potential advantages to this approach, other than removing the need for us to ever gimbal anything:

• Overall reduced scale of construction.  Our fuel grains will be smaller and hence easier to cast, high-pressure combustion chambers will be smaller and hence easier to manufacture, etc., etc.
• The individual engines will be long and thin, i.e. they will have a high length:radius ratio, which, if I recall (brmj seems to be most on the ball about this, hopefully he can clarify) is the ideal situation for a single port hybrid engine?

The biggest concern would be the possibility for an increased chance of failure.  I don't know enough about how failure modes for hybrid engines scale with size, and whether one big engine or four smaller ones would be more reliable.  It is worth noting that if the 4 subengines were arranged in a square pattern and one of them failed during a Selene 1 or 2 style flight (so there were no other clusters involved), we could switch off the one diagonally opposite it and still have our total thrust directed upward.  We wouldn't have enough thrust to reach our intended altitude, but we could probably manage a safe abort since our rocket wouldn't be flying at a weird angle.

What do people think?

The exact number of sub-engines and their arrangement is certainly something we should come to a principled decision about.  I chose 4 because it was the smallest number where you could compensate for an engine failure (with 3 in a triangle, if one fails there is no way to keep the rocket going straight).

I think it may actually be too early to start analysis of this.  An important factor in the decision, and one we can't forsee until we actually do some test burns, is how much thrust the engines produce.  It may be that we can produce appropriate sized sub-engines which put out enough thrust that 6 of them together is too much total thrust for our needs, but 4 of them is about right.

There is also the issue of which arrangement is easiest for differential thrusting.  4 engines in a square configuration makes for very easy control because there is a natural set of 2D axes you can easily pitch and yaw about, the lines connecting opposite engines.  6 engines in a hexagonal configuration may require active control of multiple engines to achieve timely rotation in certain directions.  This may not be too big a problem, though.

Basically, as you said, there are lots of factors to weigh up.  It may be that wasting a litle bit of space by using 4 engines is a small price to pay for other advantages, or it may not.