Am I the only one stuck with wet noodle rockets in 1.0.5?

[Jouni]

[quote name='SlabGizor117']Rockets are not skyscrapers. They're not structures, and they're never in storms earthquakes or large waves. If they had no way to make rockets as rigid as they are, they would bend into an L as soon as the engines fired. Skyscrapers and other tall structures are made to bend because it's easier and cheaper to take the path of least resistance and let it have a little bend, rather than just pour a foundation 3 times as big and add twice as much structural rigidity.[/QUOTE]

Rockets are made as lightweight as possible and barely strong enough. That's actually not too hard, because the forces involved are not that high by structural standards. We're talking about something like 0.5 MN/m2 at launch, which is similar to the weight of a highrise apartment building. The vertical loads of skyscrapers tend to be higher.

[Aser]

[quote name='A_name']ITT: A bunch of people saying OP is designing and/or flying his rocket wrong, but not bothering to explain WHY or what would be the "CORRECT" way to do it.

Just install KJR.[/QUOTE]

The "correct way"(s) was mentioned several times.

1. Lower Gimbal
2. Use struts
3. Avoid parts with weak joints


[quote name='Jouni']Rockets are made as lightweight as possible and barely strong enough. That's actually not too hard, because the forces involved are not that high by structural standards. We're talking about something like 0.5 MN/m2 at launch, which is similar to the weight of a highrise apartment building. The vertical loads of skyscrapers tend to be higher.[/QUOTE]

That still doesnt mean that rockets that bend are a good idea. :P

[I_Killed_Jeb]

[quote name='A_name']ITT: A bunch of people saying OP is designing and/or flying his rocket wrong, but not bothering to explain WHY or what would be the "CORRECT" way to do it.

[B]Just install KJR.[/B][/QUOTE]
this is the answer. so much grief (and partcount/performance) saved

[Snark]

Making something that's not strong enough (and is therefore loose and floppy) is bad.

On the other hand, making something [I][U]stronger[/U][/I] than it needs to be is bad, too, because structural reinforcement costs mass, and every kilogram counts when you're going to orbit.

KSP models this, somewhat, by having struts. You can make a ship stronger by adding struts, but that costs weight and drag, so it's a matter of finding the right balance. Over-strutting is as bad as under-strutting.

I think one of the places where it breaks down (i.e. where KSP diverges from real life in a way that's severe enough that it breaks immersiveness) is when you have a rocket that's a single tall stack. If this were real life, you could make the rocket stiffer by adding internal reinforcements (at the cost of mass, of course). KSP struts don't work well there, because there's no such thing as an "internal strut", and you can't easily attach them to the outside if there's just a central stack-- there needs to be some sort of "outrigger" to which you can attach an endpoint.

I don't like the KJR option, myself, because it goes too far the other direction and removes a major element of challenge from the game. What I'd like to see instead would be some sort of mechanism that allows stiffening parts of a rocket (at the cost of funds and mass, but without the external-component-bolted-on requirement of current struts). I can imagine various ways one might go about that.

[Superfluous J]

[quote name='A_name']ITT: A bunch of people saying OP is designing and/or flying his rocket wrong, but not bothering to explain WHY or what would be the "CORRECT" way to do it.

Just install KJR.[/QUOTE]

Weird. I saw multiple posts - including one of my own - that detailed exactly what was wrong and how to fix it without installing any mods.

I've never had KJR installed. Right when I was about to, they strengthened up joints enough that every rocket I few was controllable enough to fly.

[Alshain]

[quote name='A_name']ITT: A bunch of people saying OP is designing and/or flying his rocket wrong, but not bothering to explain WHY or what would be the "CORRECT" way to do it.

Just install KJR.[/QUOTE]

1. Don't stack incredibly large stuff on incredibly smaller stuff.
2. Struts.
3. Send up large payloads in smaller dockable pieces
4. Win.


KJR is completely unnecessary if you build rockets and not Whackjobian monstrosities (actually, not even sure HE uses KJR). In time new players will start to realize where key places that wobble are and strut them. If you use KJR as a crutch you will never learn to be better. It's your choice of course. Edited November 26, 2015 by Alshain

[Plusck]

I think it's perfectly realistic that large stacks of short flat parts start to wobble as soon as you start changing direction, since the rocket gimbal at one end and the reaction wheels on the other each have lots of joints to get passed through before they can affect the momentum of the parts in the middle.
Fixing this with a mod is kinda an easy way out.
Also, drag can have a massive influence on wobbliness. The general consensus is that drag needs to be sorted, but again it's possible to design away most problems or work around them.

[quote name='A_name']ITT: A bunch of people saying OP is designing and/or flying his rocket wrong, but not bothering to explain WHY or what would be the "CORRECT" way to do it.

Just install KJR.[/QUOTE]

And OP hasn't posted a pic of the rocket in question, so it's hard to be more specific than all the answers already given, which are:

1. Lower gimbal
2. Use longer single parts
3. Move the "control from here" part you are using down the stack if possible
4. Use smaller reaction wheels, perhaps radially attached
5. Use struts.

What else?

6. Check connections if using a tricoupler - parts can appear to be connected together but in fact not be.
7. SWITCH SAS OFF before making any movements. Move gently. Wait for stability. Switch on again.
8. Similar (or in conjunction with) the previous point, use custom action groups to toggle engine gimbal. Since your reaction wheels are generally close to your pod, SAS will think you have changed direction before the rest of the craft has straightened, but this will just cause a mangeable bend as long as the engine is not inducing a bend in the opposite direction at the same time.
9. Slow down below 15-18,000m. Drag effects will only exacerbate wobbliness. Try sticking to 140m/s under 8-10,000m and 180m/s under 13-15,000m. Yes it is inefficient but if your payload is the main reason you have so many stacked parts, that's just part of the cost of that payload.

[Jouni]

[quote name='Aser']That still doesnt mean that rockets that bend are a good idea. :P[/QUOTE]

KSP rockets don't bend significantly, unless you intentionally (or unwittingly) abuse the physics simulation. If you have a stack of appropriately sized parts, it behaves pretty much as it should. If you use too small parts in the middle of a stack, you're deliberately building a nunchaku. There are a few problems with decouplers in tall rockets, but otherwise the bendy joints don't restrict reasonable designs.

[COLOR="silver"][SIZE=1]- - - Updated - - -[/SIZE][/COLOR]

[quote name='Snark']I think one of the places where it breaks down (i.e. where KSP diverges from real life in a way that's severe enough that it breaks immersiveness) is when you have a rocket that's a single tall stack. If this were real life, you could make the rocket stiffer by adding internal reinforcements (at the cost of mass, of course). KSP struts don't work well there, because there's no such thing as an "internal strut", and you can't easily attach them to the outside if there's just a central stack-- there needs to be some sort of "outrigger" to which you can attach an endpoint.[/QUOTE]

Even single-stack rockets aren't that bad, as long as your decouplers and fairing bases hold. For example, this rocket has roughly the same dimensions as Falcon 9 v1.1:

[IMG]http://jltsiren.kapsi.fi/ksp/1.0.5/single_stack.jpeg[/IMG]

It won't bend, it won't wobble, and there are only four struts in the entire rocket. This picture of an earlier rocket shows the only nontrivial design choice:

[IMG]http://jltsiren.kapsi.fi/ksp/1.0/needle_4.jpeg[/IMG]

[Snark]

[quote name='Jouni']KSP rockets don't bend significantly, unless you intentionally (or unwittingly) abuse the physics simulation.

Even single-stack rockets aren't that bad, as long as your decouplers and fairing bases hold. For example, this rocket has roughly the same dimensions as Falcon 9 v1.1:
...
It won't bend, it won't wobble, and there are only four struts in the entire rocket.[/QUOTE]

Yah, 3.75m stacks tend to hold up pretty well-- they're very stiff.

The main pain point is with the smaller stacks. I can build a 1.25m rocket that doesn't look ridiculous, and based on "common sense" would seem to be pretty reasonable, but which will noodle quite a bit on launch.

I rarely have to strut anything on a 3.75m stack, but 1.25m often needs it.

[I_Killed_Jeb]

[quote name='5thHorseman']Weird. I saw multiple posts - including one of my own - that detailed exactly what was wrong and how to fix it without installing any mods.

I've never had KJR installed. Right when I was about to, they strengthened up joints enough that every rocket I few was controllable enough to fly.[/QUOTE]
I installed when the new aero hit for several reasons - first, it annoys the hell out of me that struts are so draggy, when they're really only there to preserve a "lol so kerbal" philosophy; second, stuff in fairing should be stiff.

[Jouni]

[quote name='Snark']I rarely have to strut anything on a 3.75m stack, but 1.25m often needs it.[/QUOTE]

I've had the opposite experience. The 1.25 m engines are so weak that the rockets are rarely large enough to bend and wobble. It might happen if you use SRBs as main engines, but I prefer using them as boosters.

In a 3.75 m rocket, all interstages in the main stack are potential weak links. The joints between the upper stage and the payload are the worst, but the service bay trick solves the problem rather nicely. If you add enough boosters, the interstage between the upper stage and the lower stage becomes the next weak link. Finally, if the rocket is really large, the link between a fairing base and the rocket below it may break without warning.

2.5 m stacks are the worst, however. The rockets can easily be as tall as any 3.75 m rocket, but the joints are much weaker.

[Red Iron Crown]

[quote name='Snark']I think one of the places where it breaks down (i.e. where KSP diverges from real life in a way that's severe enough that it breaks immersiveness) is when you have a rocket that's a single tall stack. If this were real life, you could make the rocket stiffer by adding internal reinforcements (at the cost of mass, of course). KSP struts don't work well there, because there's no such thing as an "internal strut", and you can't easily attach them to the outside if there's just a central stack-- there needs to be some sort of "outrigger" to which you can attach an endpoint.[/QUOTE]
This is the big point of contention I have with the current stock joint stiffness. Single stack rockets (by far the most common type of rocket IRL) are just about impossible to reinforce, if they approach the mass limit for the lifter engines of that diameter they get super wobbly. Though honestly this is as much the fault of the, uh, suboptimal SAS pointing algorithm and gimbals that go from one extreme of deflection to the other instantly.

[quote name='Jouni']KSP rockets don't bend significantly, unless you intentionally (or unwittingly) abuse the physics simulation. If you have a stack of appropriately sized parts, it behaves pretty much as it should. If you use too small parts in the middle of a stack, you're deliberately building a nunchaku. There are a few problems with decouplers in tall rockets, but otherwise the bendy joints don't restrict reasonable designs.[/QUOTE]
I consider not being able to use the small parts of the appropriate diameter anywhere in the stack to be a problem. If I want to make a recoverable 2.5m core stage, it would seem to be obvious that I should use the 2.5m probe core, battery, and perhaps a reaction wheel somewhere between the payload and the stage itself but doing so makes a bendy mess. I feel this behavior of smaller parts is the reason why we don't have the full complement of service parts in the 3.75m size, and why the 3.75m decoupler is physicsless.

There's also the issue of assembling things in orbit, the 2.5m docking ports are among the wobbliest parts of their diameter (and you must pair them!) and strutting for reinforcement is not an option in stock.

There must be some better middle ground between what we have now and the overstiffness that KJR delivers at default settings. I know the devs want to keep some wobbliness (I suspect because it makes comical failure videos that help promote the game); however, once a player makes it out of the "explosions and wobble" phase and begins to get a bit serious about making spacecraft the wobble (and requirement for part count inflating struts) becomes an annoyance and hindrance.

[Jouni]

[quote name='Red Iron Crown']There must be some better middle ground between what we have now and the overstiffness that KJR delivers at default settings. I know the devs want to keep some wobbliness (I suspect because it makes comical failure videos that help promote the game); however, once a player makes it out of the "explosions and wobble" phase and begins to get a bit serious about making spacecraft the wobble (and requirement for part count inflating struts) becomes an annoyance and hindrance.[/QUOTE]

Maybe joint rigidity should depend both on node size and on the heights of the connected parts. A joint between two X200-8 fuel tanks should be more rigid than a joint between two orange tanks, because there are more joints in a stack of X200-8s than in an equally tall stack of orange tanks.