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

[arise257]

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

That is a perfectly beautiful rocket, sir. I wasn't criticizing the technique itself, more some peoples' execution of it. With most (myself included), this sort of thing gets out-of-hand fast.

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.

Indeed, it looks like you've come up with an even more minimal solution than me. My question is, how does that fly? Any sway/wobble on those radial assemblies? It seems like you'd have some droopage/torsion acting on them everytime you made a piloting adjustment.

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.

I didn't know the cubic-octo-struts and fuel lines were massless in flight - thanks for the knowledge. I still think this winds up less efficient than my design because you wind up carrying empty tanks with no way to dump them (unless this is the final stage). Also, every part added adds drag, so that's something to consider, even if it is small.

With that said though, this construction is clever - a blend of two approaches - a compromise of sorts. If it's just as stable and weighs less, then I'd use it.

[arise257]

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 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.

I think you're misinterpreting the construction of my rocket. The quad-coupler (Q-C) is hosting 4 B-Cs, attached by one wide-end slot on the B-Cs. Those B-Cs are hosting a nuke on the narrow end. The other wide-end slot of the B-Cs hosts a T-C. The T-Cs hold 2 LV909s on their open slots. I will absolutely assure you that the inverted B-Cs are obeying the tree structure and are securely attached to the Q-C by one (valid) linkage. The only thing that could be vaguely defined as "dead weight" is the Q-C, since it's not directly hooked to an engine, just facilitating the connections.

If you put nukes on the end of the quad-coupler, you just ended the stack, and there are no points left to attach anything beneath. That basically kills it right there -- unless I'm misreading what you're trying to explain. I've made sure to use the absolute minimum of parts for this design to work, so removing anything should guarantee failure.

Where do I host .craft files to share? I'd be willing to pass this around for scrutiny.

Edited March 5, 2014 by arise257
further elaboration

[Red Iron Crown]

I can host the craft file in my dropbox for you if you'd like, pm me and I'll give you my email address to send it to me.

I was misinterpreting your construction. I thought the fuel tanks above the tricouplers were attached to the core stage radially, but if I'm reading you correctly, the tricouplers are connected to the bicouplers, which are in turn connected to the quadcoupler. I still think radial attachment with conventional coupler orientation will be lighter and more efficient, but it's not as clear cut as I thought. Looking forward to testing it.

[Vanamonde]

For file sharing, I like Mediafire.com. The interface is quite easy to use, though they they've allowed ads on their pages to become a real nuisance.

Also, it occurs to me that this should be in the gameplay and how-to section. And so, moved.

Edited March 5, 2014 by Vanamonde

[arise257]

I will post a .craft file tonight for any who are curious. I've done my own tests against my typical radial attachments and came out wayyy ahead of them. I will any alternative design mentioned and report back some results. Basic math should be able to rule out all assemblies which go over 1.2 tons, no engines. I'm predicting the main differences in what's left to be spin, stability, and durability.

[Claw]

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.

Thanks for this. I thought I was going crazy for a minute. I kept reading the thread and not understanding why you wouldn't just radially attach FL-T100s and go from there. Lighter and much less complex than an array of couplers.

However, I have used coupler ideas like this before and I admire the ingenuity. I have mostly used them to put engines in the position I wanted while providing more connection points for other fuel/payloads/intakes in front. It does require some fuel trickery though.

By the way, you can connect fuel lines to cubic struts and get the same feed effect to engines mounted to that strut. Also works if the strut is connected to a fuel tank, no fuel line needed. Probably not as big of a factor for a design like this, but just wanted to throw that out there. (I personally don't radially mount engines very often with just cubic struts, but that's just me.)

[Claw]

Indeed, it looks like you've come up with an even more minimal solution than me. My question is, how does that fly? Any sway/wobble on those radial assemblies? It seems like you'd have some droopage/torsion acting on them everytime you made a piloting adjustment.

There is very little sway or wobble because the connections aren't very long and the engines aren't all that powerful. If you have long side boosters, you'll get more wobble. Similar wobble problems would pop up with an elongated stacked coupler design.

[arise257]

Thanks for this. I thought I was going crazy for a minute. I kept reading the thread and not understanding why you wouldn't just radially attach FL-T100s and go from there. Lighter and much less complex than an array of couplers

If you look earlier in the thread, I did the math that justifies why I'm doing it my way. What I didn't know is that you could remove so much from a radial assembly and still make it work. So while the old way may be a little lighter...

However, I have used coupler ideas like this before and I admire the ingenuity. I have mostly used them to put engines in the position I wanted while providing more connection points for other fuel/payloads/intakes in front...

Thank you...Yes, there are real advantages to stacking engines down low where your other parts aren't.

By the way, you can connect fuel lines to cubic struts and get the same feed effect to engines mounted to that strut. Also works if the strut is connected to a fuel tank, no fuel line needed. Probably not as big of a factor for a design like this, but just wanted to throw that out there. (I personally don't radially mount engines very often with just cubic struts, but that's just me.)

Thanks for that additional info. I don't like to use those little cubic struts for anything but extending a cable strut down the length of fuel stacks. I'm a stickler for keeping things as close to realistic as possible.

[arise257]

There is very little sway or wobble because the connections aren't very long and the engines aren't all that powerful. If you have long side boosters, you'll get more wobble. Similar wobble problems would pop up with an elongated stacked coupler design.

Imagine replacing every one of the engines in my post with LV-T45's. This doesn't wobble or torque at all. Not in the slightest. Imagine doing that to a radial assembly like the one above. I smell a bad time.

I do agree that when you start pushing this out past what I've done, you end up with a major problem. That problem is: the rest of your rocket takes off, and your engines+assembly are left sitting on the launch pad. It happened to me a few times before I figured out what was going on.

[Claw]

If you look earlier in the thread, I did the math that justifies why I'm doing it my way. What I didn't know is that you could remove so much from a radial assembly and still make it work. So while the old way may be a little lighter...

I did, several times. While I understood the math you were doing, I wasn't catching onto what you were really trying to do. I was thrown off because you can attach FL-T100s radially and I wasn't understanding why all the decouplers were needed if you were just trying to create mount points.

And I'm not saying don't do it your way, I was just trying to understand what efficiency you were aiming for.

It's also very atypical for me to attach engines using cubic struts. In fact, I can only think of one real design where I used them to mount an engine for something other than testing purposes. (It was to get a radial engine away from the core so it wouldn't burn a kerbal in a chair...)

EDIT: For example, one of the things throwing me off is why are you using mono tanks on the radial portions? Does that just happen to be where you're storing the monopro?

Edited March 5, 2014 by Claw

[Pecan]

Like Claw I admire what you've done and might use the same sort of thing in some designs. Apart from anything else the parts just look more streamlined than girders & struts. In my first post I'd only intended to point out that this sort of design is neither lighter nor narrower than the 'conventional' way of doing things. There is far less torque and movement on the struts too - fewer joints to flex, nearer to CoM, etc. - so the thing is very stable. (Also the "calling my ship a dog" line was tongue-in-cheek, I know it doesn't look too bad but you said 'asparagus' like it was a bad thing *grin*.)

'Stack' decouplers placed sideways on the struts let you stage radials without much mass if you want to. It's a simple - although, I agree, slightly cheaty - way of doing things. All that aside you say you want things realistic - how long do you thing it would take a welder to stick one tank on the side of another, compared to all the couplers? Anyway - congratulations on what you have done. If there's anything you think I've missed from the alternative I've shown just let me know.

Have fun :-)

[SSSPutnik]

0.132 tons per mount. (Inclusive of fuel line).

[Red Iron Crown]

I assembled something similar to the OP (.craft):

Props to the OP for taking the time to assemble something like this, symmetry gets confused by the upside down couplers and it has to be assembled piece by piece. Upside down multicouplers don't flow fuel properly.

It seems to be a bit heavier and more flexible than traditional radial mounting plus a central quadcoupler, but it does spread out the nukes a bit, making them more resistant to overheating. The nodes above the outer tricouplers can accommodate things larger than 1.25m parts, if that is useful in your design.

It does have a certain aesthetic appeal. I think if I built it again I'd use the 2.5m tricouplers and put big nosecones on top of them.

[SSSPutnik]

Looks like a manga craft from below. Kind of cool. I've done something similar with a large carrier craft.

[arise257]

I assembled something similar to the OP (.craft):

Props to the OP for taking the time to assemble something like this, symmetry gets confused by the upside down couplers and it has to be assembled piece by piece. Upside down multicouplers don't flow fuel properly.

It seems to be a bit heavier and more flexible than traditional radial mounting plus a central quadcoupler, but it does spread out the nukes a bit, making them more resistant to overheating. The nodes above the outer tricouplers can accommodate things larger than 1.25m parts, if that is useful in your design.

It does have a certain aesthetic appeal. I think if I built it again I'd use the 2.5m tricouplers and put big nosecones on top of them.

Thanks! Actually, It's very finnicky, but I was able to get symmetry to work and built it pretty fast. You don't need the extra fuel lines running from the tank to the tri-couplers, it should feed fine without it.

Here's what I've figured out about storing things on top of the tri-couplers: don't use liquid fuel tanks (unless you run a line to feed your core). If you don't specify a feed direction, the LV909's suck it down after the core, so you'll have mistimed engine shutdowns. Also, it can't be super-heavy or else you'll have to strut it back to the core or other parts. I'm mounting tiny amounts of monoprop there, only because it's a super-convenient spot for this core-mounted modular space-station engine.

Also: OP WILL DELIVER. I fell asleep mad early last night, after Skyrimming til 5AM the previous day. The .craft will be up tonight.

[Red Iron Crown]

Thanks! Actually, It's very finnicky, but I was able to get symmetry to work and built it pretty fast. You don't need the extra fuel lines running from the tank to the tri-couplers, it should feed fine without it.

I couldn't get the bicouplers to go on with 4-way symmetry, they would show in the right places but some would be tinted red and wouldn't place. I've found that symmetry messes up this way sometimes when building something unusual, and sometimes camera position is a factor (it shouldn't be, but is).

If I don't connect the tricouplers to the central tank with fuel lines, the LV909s won't fire during testing. I suspect in your design they're feeding from the tanks mounted on top of the tricouplers. Upside down multicouplers mess with fuel flow, see Kasuha's excellent post about fuel flow, with testing, here.

[Red Iron Crown]

Now with moar boosters (4 Mainsails, 16 LVT-30s):

The two LVT-30s on each outer petal are powerful enough to cause significant flex, I think some mass on top of them and strutting would eliminate it.

A 28 LV-N cluster using the technique (.craft):

[arise257]

Here is one piece... I think this is the science module, with the launch platform attached. Sometimes I can't tell till I load them because I have too many ships.

Science

Here's the launcher by itself

Launcher