Figured out how to chain quad/tri/bi-couplers.

[arise257]

I've been toying around with a new engine design for my space station. In doing so, I figured out how to link (or chain, or stack) quad/tri/bi-couplers together. With the right symmetries and designs, you can create very powerful thrusts beneath a single stack of fuel.

This is especially helpful if you don't like and/or lack space or thrust for radial expansion. Also if you like to keep things stock, here's a way to do things without kludging. This particular design uses 12 engines, 8 being mini-poodles (for boost), and 4 being nukes (for sailing) -- isolated by different action groups.

What do you all think? Also, am I the first to do this?

[kurja]

wow, nice. I've been wanting to make something like this but could never make it work properly. Are you using docking ports in there to make all those adapters connect?

[boomerdog2000]

I used a similar (albeit simpler) technique to create a quad coupler out of 3 bi-couplers before I unlocked the quad in my career save. Symmetry was a little weird but she flew.

[aleis]

I've done a similar thing trying to launch 2 solar panel booms up to my station in a single launch but it was to put one on either end of the booms to stabilize them. That required docking ports particularly since i wasn't keeping the Bi-coupler. Takes a Lot of fiddling to trick KSP into accepting a double ended connection.

[arise257]

wow, nice. I've been wanting to make something like this but could never make it work properly. Are you using docking ports in there to make all those adapters connect?

No docking ports necessary. What you do is chose a starting coupler and mount it right-side-up to your stack. Then flip your choice of couplers up-side-down and mount it on one of the slots with symmetry on. Then connect another coupler of your choice to the up-side-down free slot. You'll have to conduct plenty of tests to figure out where to place your fuel lines, I've never had every engine light up on the first go.

[arise257]

I used a similar (albeit simpler) technique to create a quad coupler out of 3 bi-couplers before I unlocked the quad in my career save. Symmetry was a little weird but she flew.

I designed the same shenanigans when I was experimenting with this technique. You dodge the fuel flow issues I have with this design (note the fuel lines leading directly to engines), but you don't get moar powah.

[Pecan]

It's harder, heavier and physically larger/wider doing things this way than with struts, etc. but I like how a little of it looks. Fuel will only automatically flow in the narrow to wide direction. Use fuel-lines for engines connected the other way (eg; the nukes in your picture).

[Dispatcher]

You are more patient than I am.

[arise257]

It's harder, heavier and physically larger/wider doing things this way than with struts, etc. but I like how a little of it looks...

I generally agree that it's more difficult for a novice builder to wrap their heads around. I can't agree with this being heavier and necessarily larger or wider. Here is why:

1. This is significantly lighter than radial attachments. If you attach engines radially and buttress them with struts, you will also have to add a small fuel tank above the engine and run a fuel line to it. With just a fuel line (.05) and two struts (.05 x 2 = .1), a single radial expansion weighs .150 tons. That doesn't even include a decoupler for asparagusing/staging, nor the fuel tank (minimum .025+.07=.095 tons), which brings your total for one radial expansion (sans engine) to .245 tons. In my design above, I attached 12 engines, so let's say you put 12 radial attachments on your rocket. 12 x .245 = 2.94 tons. My engine assembly has 1 quad-coupler (.175), 4 bi-couplers (.1x4=.4), 4 tri-couplers (.125x4=.5), and 4 fuel lines (.05x4=.2). That equals 1.275 tons. My design is less than half the weight, and I used the smallest parts possible to calculate radial expansion like you suggest.
   
2. To the point of whether or not this is wider or larger than radial expansion, I would argue it can be, if you design it that way. I could see some clever ways to keep the cluster condensed. What you see here isn't outrageously large. It's also at the bottom of the mass in general. I would argue this is far more true-to-life than the asparagused flying pyramids most people wind up designing.
   

I don't mean to come off as holier-than-thou, just defending my design.

[thomasjkenney]

If you attach engines radially and buttress them with struts, you will also have to add a small fuel tank above the engine and run a fuel line to it. With just a fuel line (.05) and two struts (.05 x 2 = .1), a single radial expansion weighs .150 tons. That doesn't even include a decoupler for asparagusing/staging, nor the fuel tank (minimum .025+.07=.095 tons), which brings your total for one radial expansion (sans engine) to .245 tons.

You can do radial engines without extra fuel tanks. Use a Modular Girder outward, then a Modular Girder Adapter downward, attach an upside-down Rockomax Adapter #2, then a Skipper or Mainsail. Run the fuel line from the central tank out to the Rockomax Adapter #2. I'm at work right now, so can't post a screenshot, but I'll do so when I get home.

[arise257]

You can do radial engines without extra fuel tanks. Use a Modular Girder outward, then a Modular Girder Adapter downward, attach an upside-down Rockomax Adapter #2, then a Skipper or Mainsail. Run the fuel line from the central tank out to the Rockomax Adapter #2. I'm at work right now, so can't post a screenshot, but I'll do so when I get home.

If I'm imagining this correctly, then your weight per engine mount is .555 tons. Multiplied 12 times = 6.660t. More than 5 times as heavy as what I've done, and more than double the weight of my lightest possible alternative.

Edited March 5, 2014 by arise257
forgot quote

[thomasjkenney]

If I'm imagining this correctly, then your weight per engine mount is .555 tons. Multiplied 12 times = 6.660t. More than 5 times as heavy as what I've done, and more than double the weight of my lightest possible alternative.

Yes, heavier (your design in OP is quite elegant, btw!), but it's flexible. The ship I used this design on was 7 x Rockomax tank stacks (1 center plus 6 surround), each with a Skipper engine. Then I radial'ed (as per my description) 12 more Skippers around the outside. I think I actually used the longer girder segment for the initial attachment. I'll post a screenshot later. Don't want to hijack this thread, though, as it's about bi-couplers and such. Maybe I'll start a Radially Attached Engines thread.

[arise257]

I'll post a screenshot later. Don't want to hijack this thread, though, as it's about bi-couplers and such.

Your post is welcome. I shared my design in hopes that people would poke at it or think of alternatives. If there's a more lightweight, stock, non-clip method to achieve something similar, everyone wins.

[thomasjkenney]

OK, I'll post here, then. Thanks!

Here's the Skipper design, which I used in Career with only about half the tech tree filled in:

Not quite the same as the behemoth I did later, but I can't find the .craft file (got 2 machines running KSP, so stuff spread across drives). I use a similar design with my interplanetary tug craft:

In the tug, I have a Rockmax #2 Adapter and a big reaction wheel, and 'inside' the reaction wheel I place 4 x large girders as the crossbeams. In this design, they do 'clip' through the reaction wheel and the topmost fuel tank, but it's exactly the right length for the job.

[irongut]

Nice design arise. I use those couplers a lot and have wondered if you could chain them before but didn't put the work in to figure it out. I may use your technique at some point.

[Accelerando]

That's beautiful.

[Whackjob]

Most I've ever put under one stack was thirteen mainsails. It's tough to build a stack that can handle that level of force.

#EDIT: I built it again for teh lulz.

Not a bad speed. This is with RSS and FAR. Thirteen wonderful mainsail engines.

Edited March 5, 2014 by Whackjob

[celem]

Nice work with the coupler-chaining there. For some reason it never occured to me to do this with engines.

I did however use the same principles on the nose-end of several rockets to dramatically increase the number of 'payload slots'.

Dont have a pic right here, but imagine your setup inverted on the nose of a rocket with each engine replaced by an independant satellite on a port/decouple. Was loads of fun, loved it, but never found any way to make it work with FAR, just too much drag.

[Red Iron Crown]

This technique certainly is aesthetically pleasing, but I'm pretty sure it's inefficient.

Imagine I built the same rocket but attached the nukes directly to the quadcoupler beneath the core stage and put the LV909s on bicouplers under the radial stages. It would have the same thrust but less mass, thus better TWR and deltaV.

Or is there some other benefit I'm missing?

[Pecan]

...I would argue this is far more true-to-life than the asparagused flying pyramids most people wind up designing...

Are you calling my ship a dog? (Non-stock Procedural Fairings on the payload, but you get the point).

Bet my way of putting 12 engines on your ship is lighter than yours ;-0 All stock, no clipping - not even that allowed by the editor.

Cubic octhangol struts and fuel-lines are also massless in flight (the VAB lies). I don't see any decouplers on your original post so I haven't put them here and, since you have tanks on the radials there's already somewhere there to attach the engines*. I used the quad coupler for the nukes but that isn't necessary either; octhagonal struts are again the usual and massless way to do it - really, they're the king of structural parts.

[*Oops, I was trying to knock this together quickly and not paying attention. I should have only put the FL-T100 tanks on every other (symmetry 4) engine-mount. Nevertheless the RCS tanks would have served the same purpose.]

Edited March 5, 2014 by Pecan

[kurja]

No docking ports necessary. What you do is chose a starting coupler and mount it right-side-up to your stack. Then flip your choice of couplers up-side-down and mount it on one of the slots with symmetry on. Then connect another coupler of your choice to the up-side-down free slot. You'll have to conduct plenty of tests to figure out where to place your fuel lines, I've never had every engine light up on the first go.

hm now that I look at it again this should really work... usually when one puts adapters together so the stack branches out and then merges again there are all sorts of problems but this seems to avoid the issue entirely, must try =)

[arise257]

This technique certainly is aesthetically pleasing, but I'm pretty sure it's inefficient.

Imagine I built the same rocket but attached the nukes directly to the quadcoupler beneath the core stage and put the LV909s on bicouplers under the radial stages. It would have the same thrust but less mass, thus better TWR and deltaV.

Or is there some other benefit I'm missing?

It's probably just as efficient, but I can't do the numbers right now. Something tells me your weight would be nearly identical. Read my previous posts, if it beats 1.2t with no engines (in VAB), then yes, it's more efficient from a TWR perspective. It's tough to beat the efficiency of multiple engines beneath a stack because there are no off-prograde forces acting on your rocket/engines. Every wobble/twist to the side is energy not spent going up/forward, and that's affecting your deltaV. I understand those forces are small, but over the course of a mission, that's thousands of meters peeled off your potential. I think that's really the point of this design - getting everything firing beneath the stack, and having massive thrust carried up to orbit - with minimal weight spent.

From a fuel consumption perspective, I'm not sure it gets significantly better than this. I'll take that back if someone can show me why though.

Edited March 5, 2014 by arise257

[arise257]

... usually when one puts adapters together so the stack branches out and then merges again there are all sorts of problems but this seems to avoid the issue entirely, must try =)

Exactly the reason why I started experimenting with this. You can't stack stacks that terminate in multi-couplers, so you can't get hella thrust in orbit without a weight or stability penalty. Nothing saddens me more than carrying a full tank/engine up there without it doing any work - that's a waste of fuel.

[arise257]

...I did however use the same principles on the nose-end of several rockets to dramatically increase the number of 'payload slots'. Dont have a pic right here, but imagine your setup inverted on the nose of a rocket with each engine replaced by an independant satellite on a port/decouple...

Great idea, I'll implement it on my popcorn-style probe launcher. That should take it from 8 probes to 24 very easily.

[Red Iron Crown]

It's probably just as efficient, but I can't do the numbers right now. Something tells me your weight would be nearly identical. Read my previous posts, if it beats 1.2t with no engines (in VAB), then yes, it's more efficient from a TWR perspective. It's tough to beat the efficiency of multiple engines beneath a stack because there are no off-prograde forces acting on your rocket/engines. Every wobble/twist to the side is energy not spent going up/forward, and that's affecting your deltaV. I understand those forces are small, but over the course of a mission, that's thousands of meters peeled off your potential. I think that's really the point of this design - getting everything firing beneath the stack, and having massive thrust carried up to orbit - with minimal weight spent.

From a fuel consumption perspective, I'm not sure it gets significantly better than this. I'll take that back if someone can show me why though.

I'm talking about the exact same rocket as you show in the OP, just without the inverted bicouplers for mounting the nukes (they go straight on the quad coupler in the core) and replacing the tricouplers the LV909s are mounted to with bicouplers. Removed four parts, replaced four others with lighter parts. Same engine combination. Same thrust. Less dead mass. More TWR. More dV.

I'm at work now so I don't have access to KSP, I'll see if I can whip up an example tonight. Can you link to the .craft file for the ship in the OP?

I'm not convinced that those inverted bicouplers are actually making two connections to the main ship, either. As I understand it, that's not possible without docking ports or struts in the current game, due to the tree structure used for craft.

Edited March 5, 2014 by Red Iron Crown