Pendulum pseudostaging

[Nathair]

I've been playing with a ring of radially mounted engines fairly high up on a pencil rocket. Instead of actual staging I've been suspending fuel tanks below this ring like beads on a necklace connected with docking ports stabilized with a few struts(the dangling weight is not very stressy). Tanks are fuel piped from bottom to top and as each tank empties I just decouple the port and they fall away. All engines are firing all the time (and I can drop them in stages if I want as the twr gets higher and I don't need them.) This seems quite efficient to me but is it really?

This kind of thing;

[Vanamonde]

Everybody tries that idea sooner or later. Efficiency isn't the problem. Steering tends to be a nightmare and structural failures are common. Also, thrust can get out of hand because it has to start with what it needs to lift the whole mass, but that becomes proportionally stronger as you discard tanks, and can lead to the ship damaging itself with excess G loads. If it could be made to work, though, it would be more efficient than discarding engines with expended stages as in more conventional staging methods.

[Kasuha]

Fun fact is, with engines mounted on girders you probably don't need all those fuel lines from the central tank to your engines. They'd draw the fuel from it through those girders anyway.

Instead of docking ports, it is better to use just standard 1.25 m docking ports in between stages. They also allow you to put struts and fuel lines around and are easier to stage. If you have trouble installing struts and fuel lines in the gap, just add some spacer part, install them and remove the spacer.

In general though, it is better to keep the thrust at about twice the local gravity and that means you can drop unnecessary engine mass together with unnecessary empty fuel tanks.

[celem]

I always rather liked this vertical staging idea. Its actually quite effective provided you do as you suggested and drop some of the engines themselves as the TWR climbs.

I regular used this in 0.22 with NERVAs on interplanetary designs.

If the docking ports get too swingy and the tail starts to whiplash then consider moving to full-on 2.5m decouplers rather than the ports. You will no doubt have tried this and realised you cant stitch the fuel lines across the decouplers due to their high profile, however there are a couple of tricks to get around this;

Firstly, if you have the KW Rocketry modpack then the slim profile decouplers included do allow for a vertical stitch like this.

If you dont have this mod then you can put a tiny fuel can radially-mounted just above each joint and pump the fuel across the decoupler via this. You can even tweak the radial can to be empty at launch to save weight, the lines will work anyway since they just define fuel-draw paths and dont actually pump anything. Note that while the oscar B or toroidal tanks are tempting for this, they tend not to work very well since they dont attach well to radial decouplers and/or dont like fuel lines. Use the tiniest 1.25m tank. You can jettison each of these helper tanks along with the tank below since you emptied them in the VAB anyway.

Thirdly, and this one makes for some monstrosities. Rather than your single 'beaded' stack, have 2 or more such stacks up the middle with your ring of engines up top. This then lets you run your fuel-lines up the stack without the need for mod decouplers or radial pumping points. You run each line to the tank above it on the next stack across, giving a 'helter-skelter' pattern of fuel lines. I've done this one a couple of times using 6/8 of those vertical staging stacks and it works like a charm, though is pretty hvy on part-count.

Edited March 6, 2014 by celem

[Claw]

I've done similar with feeding fuel upward, but I also found it to be a bit more efficient to dump engines along the way. Here is one I did for a challenge using radial engines. I had a taller design that was similar, but I don't have pictures. Basicall what I did was put late stage engines up high and early stage engines down low. They all fire together, but they are misaligned from each other to not interfere with the thrust exhaust.

I do like this idea, I should try it more.

The one below is only two stage. All the fuel feeds to the second stage and all engines burn at launch. Nine engines to start, with the first stage dropping five of them.

Javascript is disabled. View full album

[cantab]

Watch that the centre of thrust isn't above the centre of mass. That's not a deal breaker but I believe it does cause control screws.

[MarvinKitFox]

That is simply Asparagus staging, with the engines separated from the fuel tanks so you are not forced to drop them when you drop the tank, but rather when you want.

Disadvantage: More structure, more decouplers/docking ports needed.

Big disadvantage: Much greater stresses that need to be propagated around your vehicle. In normal staging, each tank is lifted by a rocket engine directly beneath its center of mass. The *only* stresses you need to accommodate are from uneven tank emptying(due to fuel lines), uneven engine specs and uneven mass distribution of payload. With your scheme, *all* the energy produced by the engines has to skitter over multiple tanks, girders, docking ports and fuel lines, before pulling at the bottommost tank. In real life, your mass penalty due to structural and control linkages would override the benefits of this layout. In KSP, it is more forgiving, but you will find your design tops out at a much lower tonnage before becoming Kraken bait, and even for a smaller vehicle the maxx survivable G-forges (due to TWR) will be less. Much less.

Big Advantage: As your asparagus is hanging vertically, you can drop a single tank at a time and still maintain radial symmetry. Normal asparagus requires you drop 2 (or 3,4,etc) tanks from opposite sides at the same time.

Advantage: You can pick and choose *which* engines to drop, and when.

Build your lifter segment with a couple different engines for different altitude + thrust profiles. Drop the Mainsails first, keep your nukes on forever. You can tailor your ship to have the very best ISP practical at any given time, yet still allow it good enough TWR.

Big advantage (for me, at least): You can drop your stages without the least bit of worry about it running into the rest of your vehicle. When my designs get complex, that is the #1 killer of heavy launchers. Apparently dropping a spend SRB casing 0.5m away from the top of a 400m-long behemoth is... not very smart.

Edited March 6, 2014 by MarvinKitFox

[MBobrik]

It is not very efficient because if you drop only tanks without the engines, your TWR goes up unnecessarily. You would be better off dropping both an empty tank and the corresponding amount of engines so that your TWR remains more or less the same ( which is what the asparagus staging does ) and you will not have to carry unnecessary dead weight engine mass with you.

[Kryxal]

If you dont have this mod then you can put a tiny fuel can radially-mounted just above each joint and pump the fuel across the decoupler via this. You can even tweak the radial can to be empty at launch to save weight, the lines will work anyway since they just define fuel-draw paths and dont actually pump anything. Note that while the oscar B or toroidal tanks are tempting for this, they tend not to work very well since they dont attach well to radial decouplers and/or dont like fuel lines. Use the tiniest 1.25m tank. You can jettison each of these helper tanks along with the tank below since you emptied them in the VAB anyway.

Might it be a better idea to have the tiny cans BELOW the joint, so you automatically drop them with the big tank they're giving a route to?

[Red Iron Crown]

I've tried making a similar lifter, with the difference being I had some of the engines mounted to lower tanks to stage them off as mass falls away.

In theory, it should be more efficient than asparagus because you're staging away one empty tank at a time rather than two, meaning you're carrying less dead weight.

In testing, I found that having to have two engines in the final stage reduced its delta-V enough to negate the advantage of more granular staging, so now I stick with asparagus staging.

Fun to experiment with, though.

[celem]

Might it be a better idea to have the tiny cans BELOW the joint, so you automatically drop them with the big tank they're giving a route to?

Heh yeah thats a good point. I originally had the radials above the joins as it was a design I came up with for 0.22 when I couldnt empty the radials, they dropped a second or two after the tank below them. If emptying the radials prior to launch in 0.23+ then below the join is viable (and saves a part in the decoupler if you weld it straight to the other tank)

[angeldust]

I think "pendulum staging" is really effective with landers. I sometimes build heavy landers with three stages. A deorbiting stage, a landing stage and a takeoff stage. The trick is to just give the takeoff stage the engines. So you won't carry extra weight in engines on the other stages. the deorbiting stage is essentially just a tank, which gets dropped before the final stage of descent, then the landing stage with landing gear and rover or some other fancy stuff + fuel, which gets left behind when taking off, so the takeoff stage is just capsule, tank and engines. I think this is rather effective and lots of fun to design

[Kasuha]

I think "pendulum staging" is really effective with landers.

That's very good point. On landers you usually need quite the opposite to lifters - you need your TWR to rise as you are slowing down and getting near the ground. Especially if your initial TWR is low.

Once upon the time I had to learn the very hard way that asparagus staging is a very bad idea for a lander, particularly for its descent part.

[Red Iron Crown]

It's not always the case that asparagus staging has a falling TWR, though, is it? If your core stage weighs less than your outer boosters (via having enough less fuel to offset payload mass) wouldn't TWR increase as you stage?

[Kasuha]

What I meant was the "classic simple" asparagus, i.e. about equally sized fuel cans and engines on each of them. In fact, what is presented in the first post of this thread is also kind of generic asparagus staging, actually exactly the case where your TWR grows as you drop stages.

[angeldust]

well the concept of "why do I have to stick my engines at the bottom" in general seems like a good idea, until your craft starts lashing out like a psychopathic python. But maybe a mixture? Does anyone have experience with "normal" rockets, where a higher stage with radial engines fires along the big ones on the bottom? Does it have any advantages?

[Kasuha]

Gimbal control in game assumes engines are below CoM. When they're placed above, SAS functionality turns gimbal the wrong way, increasing the error instead of compensating it.

If you disable gimbal on all engines above CoM and steer the rocket only using torque, things are fine. And part joints are often more forgiving on pull than on push, especially for short parts such as decouplers ort docking ports.

[LordQ]

Personally, I don't really see myself trying this too much as it would probably introduce a tonne of drag with FAR installed.

[Red Iron Crown]

Has better gimbal control already been suggested to the devs? Seems like it wouldn't be too hard given that CoM is already calculated, and pull configurations are common enough to justify it, I think.