Stability problems with excessively huge rockets

[Gingham]

Figured I\'d give KSP a go again now that I\'ve got a new graphics card installed just to see how many fuel tanks, engines and boosters I could slap onto a rocket before the game started lagging out on me.

What\'s happened instead is KSP has frustrated me to no end because none of my stupidly huge designs fly in a straight line, with everything perfectly balanced and symmetrical I still see with every launch my rockets deviating off course and going into a nosedive. Doesn\'t matter how many SAS modules I spam all over, doesn\'t matter how many RCS modules I use, doesn\'t matter if I put [gmod]Please do us a favor, and stop with the profanity yeah? -DR[/gmod] of boosters at the TOP of my rocket, I still get these same goddamn inertia coupling effects.

What the hell is going on here and how do I avoid it?

[togfox]

AV8 winglets fixes mine everytime. I don\'t even use SAS and RCS.

[Gingham]

Placed over a hundred AV-8 winglets on my 'Heavy Lift Vehicle'

It did absolutely nothing

[togfox]

pics?

[leopardenthusiast]

Tell me, are they bending?

I find that happens with particularly huge ships. Just about the only thing you can do is add more struts and hope that it\'s stable before your ship is unable to lift the weight of the struts.

Of course, it\'s also possible that you just need a lucky shot with design.

Finally, if you\'re not using ASAS, you might as well just give up right now. I seriously hope that you are though.

[Gingham]

Adding struts does not help either

Here is a pic:

[leopardenthusiast]

Adding struts does not help either

Here is a pic:

All I can say is that the ONLY way you\'re getting that anywhere but into a large crater of its own creation is to use the NovaPunch heavy struts or some other insanely strong strut.

[Gingham]

It lifts off just fine if I strip it of all boosters. Even goes straight, just not very fast at all.

[Khrissetti]

AV8 winglets fixes mine everytime. I don\'t even use SAS and RCS.

I tend to find that any more than three actually destabilises the rocket.

[invalid8]

Do you really think something like that would be stable IRL?

[UmbralRaptor]

Vectored thrust + ASAS?

[Markus Reese]

First, this is a civil forum, watch the language.

Definitely all about not enough structural connections. Even a slight deflection of your engines will cause it to go out of control. You also need the 45 engine nozzles since they can give control with their vectoring. Second, I don\'t seen any vertically connecting structural links. By connecting the tops of lower ones past the couplers, it will help stabilize by tying everything into the core more directly. Lastly, large rockets must have the ASAS since control is so slow manual, any deviation the ASAS will need to take care of right away.

[Gingham]

By connecting the tops of lower ones past the couplers, it will help stabilize by tying everything into the core more directly

Problem solved, many thanks to you good sir.

[EndOfTheEarth]

Looking at that rocket, I think you\'re losing stability in the connectors (or lack thereof) between outer and inner tanks. As a result, the tanks will swing back and forth during takeoff, and this swiniging action will cause the rocket to tip over from uneven thrust.

The way to stop this from happening is to attach connectors to the layers of boosters in a ring-like fashion, where all the outer boosters are linked by connectors.

That, and in weak places, like near stage decouplers, having vertical connectors will help you too. The big idea is to prevent as much from swinging as possible.

After that, I see a lot of SRBs on that thing. For whatever reason, SRBs don\'t like to go straight. You compensate by adding a few SAS modules on that part of your rocket.

[Markus Reese]

Problem solved, many thanks to you good sir.

Glad I could help. Just remember that since KSP is a physics game, it can do odd things. Often a problem requires experimenting or an unusual solution to a buggy problem. If you have a specific design not working right, sometimes helps to take a break. Post problem, and ideally the screenshot of any failure that happens.

[Gingham]

Now the problem I\'m having is attempting to make a modified version of the above (with a munlander payload) into something a bit more maneuverable.

The basic staging is as such:

Stage 1: A whole mess of liquid fuel engines, just enough to get me into orbit around kerbin

Stage 2: Transfer stage, used to get me into a rendesvouz trajectory with the mun, and for the powered descent of the same.

Stage 3/4: The Munlander, pimped out with extra fuel tanks

This thing is an absolute pig to fly, unwieldy as all hell in trying to achieve orbit and even with just stage 2 remaining it takes me almost quarter of an hour to make course corrections for my mun transfer. Slapping 24 RCS modules on the upper stages did very little to solve this.

[rkman]

I would add another stage and decrease the size of 1st stage.

1st stage would then get you nearly out of the atmosphere but not yet in orbit, so that much of the mass is shedded before you need orbital maneuvering. Maybe add a few more SAS units for faster response in attitude control.

[Trekkin]

How do you mean powered descent into the Mun, Gingham? The deorbit burn around the Mun, if your transfer orbit was small enough, is trivially small.

Incidentally, slapping on RCS modules wherever they fit rarely helps. They need careful positioning to maximize torque, or they just spew forth fuel to no appreciable result.

As far as making your rocket more maneuverable: you\'ve overbuilt by about an order of magnitude, but that\'s apparently what appeals to you. That said, the poor maneuverability is kind of inevitable. If you\'re up for it, I\'d combine stages one and two, so that you don\'t stop burning in Kerbin orbit until your apogee is around 11 million and then circularize; it\'s easier to maneuver under power, I\'ve found, and you\'ve certainly got the capacity to make it work.

[rkman]

Do you use Mechjeb?

It does not utilize SAS units except for the small SAS that\'s in the command pod.

Also there seems to be a bug in stock KSP where the command pod reduces the effectiveness of SAS units.

see

http://kerbalspaceprogram.com/forum/index.php?topic=14066.msg241661#msg241661

and

http://kerbalspaceprogram.com/forum/index.php?topic=16127.msg241417#msg241417

[Gingham]

How do you mean powered descent into the Mun, Gingham?

Stage 1 typically burns out as I reach 2300m/s going horizontally around kerbin, stage 2 typically burns out as I near 10,000m above the munar surface.

If you\'re up for it, I\'d combine stages one and two, so that you don\'t stop burning in Kerbin orbit until your apogee is around 11 million and then circularize; it\'s easier to maneuver under power, I\'ve found, and you\'ve certainly got the capacity to make it work.

In that case I\'d have to start over from scratch I think, but I\'m not keen on that thought. I strongly suspect that struts linking structural elements horizontally induce instability because there\'s no way to align them properly. (or is there??)