What is the point of keeping "wobbly rocket syndrome" stock?

[Razhan]

People complaining of wobble never played pre-0.23.5. Rocket strut-exoskeletons are where it's at.

Stitch Time!!

[brdavis]

I'm pretty sure most do have one single sturdy connection point. The Shuttle SRB's connected to the ET in one point.

Actually each SRB attached to the ET at two points, fore and aft, and those were not simple single-point connections either… and the upper one connected to a very heavy beam through the inter tank section of the ET, a beam that flexed during thrust. and the SRB's themselves flexed during use (remember the "puffs of black smoke" from the Challenger SRBs?). One of the Apollo Saturn V's had POGO vibrations strong enough to break the interstage fairing around the LEM and have it fall out of the launch vehicle.

Real rockets flex. A lot. In dangerous and difficult ways. I think assuming that the developers have "left this in to mimic real life" might be stretching it, but it would probably be a mistake to say that it's "not fair" or "not realistic".

Since complete realism is clearly not the focus of the game, why not just add rigid body physics for performance increase and part count decrease.

Why not just turn off aerodynamic heating? Or properly modeled something else? There's a balance. Where that balance has been chosen… is not a simple process.

- - - Updated - - -

I also can't spend the time reading it. Sorry.

Too bad - it's actually a well-written and very informative post, with a lot of good details… about KSP, and the real world. Thank you PB666

IMHO if there is sooo much specificity required in stock play to address "wobble" then the entire concept of game play and fun has been completely lost.

Well, I guess to offer another IMHO… I don't have a problem with it. Yes, there's some wobble… but less (far less) than I've dealt with in the past. And it's something that I can deal with using strategies that mimic (and teach) real-world rocketry. Resonance is something that happens in KSP… and in the real world. Detuning a stack is something that, just like delta-v, Isp, or proper staging, I can have fun with, and learn from, KSP.

[Zucal]

Well, I have seen people defend the game's bugs and glitches as being "part of the charm". I think they need immediate medical help.

Really? I love it when my game crashes literally every four minutes! It's so Kerbal!

[Mister Kerman]

I'm not good at making efficient rockets so I end up making them too large with lots of wonky BS hanging off of them, but even still, I haven't come across a wiggly rocket problem that couldn't be tied down rock solid with just 2 struts. The SAS/gimbal wiggle problem is easily dealt with as well.

[Jarin]

Really? I love it when my game crashes literally every four minutes! It's so Kerbal!

Bad experience with the heating memory leak? Or just loads of mods?

[PB666]

1. Not everyone uses MechJeb

2. Not everyone can spend 10 hours on every craft and lifter.

3. Have you ever seen a booster without struts in stock KSP work? And it's not a design flaw because real boosters don't use struts. If boosters are a "design flaw" then the Delta 4 Heavy would have never existed.

4. Joints in KSP are all the same so 3.75 m parts are the same as 1.25 m parts. So building larger rockets is a "design flaw" too.

1. Eventually you need to take your space agency out of the stone age. When the Soviets failed several times to dock their rockets the got unmanned vessels to do it and they insisted that there has to be an experienced astronaut on every manned mission. You never know they might have trained these cosmonauts by letting them see the machine do it.

2. Are you kidding? 10 hours? lol. BTW it I have parts for a given form factor I use the more than the stock parts, generally. So once you make a part you can use it on future builds.

3. I have many lifters without struts. My basic logic for struts on boosters is this

1. Struts go from booster to booster.

2. Strut only if the booster never exceeds 100 m/s at 1000 feet or the equivilant IAS at higher altitude.

4. Yes but not the same, the number of intersegment joints per meter of stack is less for 3.75 than for short rockets. And as I said its all about segmental flexibility.

Hold a fishing rod and put a small weight on the top, holding the rod strait up move the base in a circle. If the weight on the top is high enough and if it is flexible enough the top of the rod will have a radius of rotation greater than the base causing the rotation, these is a recipe for resonance.

As your rocket goes up, the mass at the bottom decreases, the gimble is further from the center of mass, the amount of mass it has to move is less, therefore its torque is greater by distance and its acceleration is greater by the F/M mass has decreased. So tuning the gimble during flight can fix this problem.

I just completed a 6 stage rocket, with 13000 DV with 4 "Kickback" attached with TT38K and with the equivilant of two FL-T800 (one an areo) with no strut connectors and only 4 fuel lines. So... don't give me this gripe that boosters must have struts, they do not. Should they be struted, if you don't mind the added weight.

[clivman]

@brdavis I'm a RO person and can't wait for it to come out, that was sarcasm(sorry) complaining about the non-realism crowd. Yes, rockets wobble in RL, but as you said there should be a balance, and KJR has that balance, I just hate the wobble in stock right now and don't know why the devs would leave it this way.

@PB66 So your saying mods should be necessary to play the stock game? I get that the devs want to keep wobble in, but why don't they bump joints up to realistic levels?

And they are not realistic right now, otherwise KJR wouldn't be required for RO.

Edited June 17, 2015 by clivman

[PB666]

PB666... I can't requote such a big post.

I also can't spend the time reading it. Sorry.

IMHO if there is sooo much specificity required in stock play to address "wobble" then the entire concept of game play and fun has been completely lost.

Something is clearly broken... And it's likely game physics.

Life is long also, you'll figure it out.

The problem is not wobble, the problem is that many folks want to do what NASA hasn't even done, and they want it to be easy. Yeah, sure 23.5 had problems, mainly not enough large diameter stuff.

And if you think wobble is broken game. in V 22 I had an oversize tank that kept colliding on the launch pad. I removed the engine and launched the tank. The tank jumped up and down until it jumped about 150 meters in the air and then crashed just shy of the water. Since 23.5 the physics has been decidedly tame.

You can deal with the wobble, but the reason it easier for the folks who got the game version < 23.5 is because before 23.5 the problem was a magnitude worse, and the hard lessons taught them how to solve problems. Which means you need to learn those lessons.

And finally, if something is clearly a bug (e.g. fired players not leaving the crew lineup, terrain fall throughs, etc) I reported my fair share of bugs. I reported recently on my ship that blew up the Mun. But when someone is basically having their stuff malfunction when they are do a V2 manned rocket launch and have not gone about reducing the drag of their set-up, im wonder why their rocket didn't blow up earlier in the flight. Watch the space shuttle video, see how they control the speed in launch. You see a rocket flying up from the ground, but the SSME are constantly changing parameters during flight.

Oh and learn to use the QUOTE tags.

[Felger]

This should probably still work, haven't tested it recently.

https://youtu.be/cWXghLnilrY

[n0xiety]

Wobbly rockets cause kraken eventually. This is a fact... You will experience it a few times when you launch 1000+ rockets and it may even break your save(Done that, been there...). So yes i don't think something that causes bugs should be a part of the game.

[Scarecrow88]

Whilst a small amount of wobble may be charming in a Kerbal sort of way, enough wobble to destroy a ship is too much. There are a lot of seasoned players on this forum who have been playing KSP for quite a while, and know of all the workarounds and tricks to circumvent such problems. Personally I don't think you should be required to do that. There are a lot of new players coming to this game all the time and not all of them will be on a forum like this to get answers when they encounter problems such as the wet noodle rocket syndrome.

I have no idea what causes the problem - weak joints, excessive gimbal, whatever - but it largely goes away if you use the KJR mod, so it's obviously fixable, and I just think Squad would have a much better game if something like KJR was included as stock.

[brdavis]

...that was sarcasm (sorry)…

No worries, I just missed it.

Yes, rockets wobble in RL, but as you said there should be a balance, and KJR has that balance…

It has that balance for you (and, to be honest, many others). For some of us, it's not been a problem in stock, and we deal with it OK. My balance may be different than your balance… which is why the conversation is good to have.

A similar thing seemed to happen when people first ran into the new Stock Aero model, and all their rockets started flipping out… especially when they turned to 45 degrees at 10 km or so. Flipping out made things harder… and I loved it. Because, honestly, that's how the real world works, and I get a bit of a thrill trying to solve simplified versions of real world problems. Like a shifting CoM relative to the CoD, or POGO and bending moment resonance.

I've had rockets on SAS ascent that start wobbling… and I have to take manual control to reduce it before tripping SAS back on. To me… that's a game feature, not a bug. I'd honestly miss it if it just "went away".

I just hate the wobble in stock right now and don't know why the devs would leave it this way.

I get that, and understand. But the reason they devs "left it that way" may very well be for the reasons I mentioned above (their preference point may be different from yours), or even because they aren't done but have more important tasks at the moment.

My priority list seems to be different than yours in this regard… are we so sure that the devs priority list isn't likely different (perhaps for very good reasons) than yours or mine?

...why don't they bump joints up to realistic levels? And they are not realistic right now, otherwise KJR wouldn't be required for RO.

They may be quite realistic for the stock game. RO is not the stock game… and it may indeed require KJR. But Squad isn't developing RO.

[Jouni]

@brdavis I'm a RO person and can't wait for it to come out, that was sarcasm(sorry) complaining about the non-realism crowd. Yes, rockets wobble in RL, but as you said there should be a balance, and KJR has that balance, I just hate the wobble in stock right now and don't know why the devs would leave it this way.

Stock rockets don't wobble significantly. Wobble can become a problem, if the rocket is designed badly or if you're using 5 m (or larger) mod parts.

[Renegrade]

I'm going to have to chime in with the people who played before 0.23.5: it really is much, much better now.

Of course I would have preferred the joints to be stiffer still, but not as physically strong. You can get some rather silly floppy dongers going on by stacking a lot of anythings (inline batteries, round8s, etc) which can survive insane forces but just won't stop flopping... It's still much, much better though.

(My old pre-ARM rockets tended to be 20% to 40% struts by part count)

[Bosun]

One thing I've thought several times in my newness here:

Is there a mod or upcoming patch idea for adding a double radial decoupler option? IE - having decouplers that come as a set of 2, or even three for larger Kerbodyn tanks, that when you place them, they put two decouplers for each one you place, being set on the same radial as each other, a bit apart?

In all my watching of NASA bits, there are several times where I swear I see them use multiple de-couplers on one booster or engine, because of exactly the kind of structural integrity you'd want.

I actually tried to do this in KSP when I first started building rockets, only to be disappointed that you cannot attach a decoupler to an engine that's already attached to a de-coupler on the engine you're trying to couple too. (No no no, YOU stay here. Make sure. He doesn't leave.)

[Wallygator]

Life is long also, you'll figure it out.

The problem is not wobble, the problem is that many folks want to do what NASA hasn't even done, and they want it to be easy. Yeah, sure 23.5 had problems, mainly not enough large diameter stuff.

And if you think wobble is broken game. in V 22 I had an oversize tank that kept colliding on the launch pad. I removed the engine and launched the tank. The tank jumped up and down until it jumped about 150 meters in the air and then crashed just shy of the water. Since 23.5 the physics has been decidedly tame.

You can deal with the wobble, but the reason it easier for the folks who got the game version < 23.5 is because before 23.5 the problem was a magnitude worse, and the hard lessons taught them how to solve problems. Which means you need to learn those lessons.

And finally, if something is clearly a bug (e.g. fired players not leaving the crew lineup, terrain fall throughs, etc) I reported my fair share of bugs. I reported recently on my ship that blew up the Mun. But when someone is basically having their stuff malfunction when they are do a V2 manned rocket launch and have not gone about reducing the drag of their set-up, im wonder why their rocket didn't blow up earlier in the flight. Watch the space shuttle video, see how they control the speed in launch. You see a rocket flying up from the ground, but the SSME are constantly changing parameters during flight.

Oh and learn to use the QUOTE tags.

Thanks I already know how to use quote tags - I just chose to not use them on your post. Nothing personal.

From my recollection, Rockets typically do not fail due to "wobbling". It is usually due to faulty guidance, propulsion or human error. However, If a rocket is not structurally sound it will likely disassemble itself as it leaves the pad - they did that in the 50's a few times.

Whatever, keep your wobble, or not. its ok.

[Juicy]

You can reduce the wobble somewhat by removing some of the moving fins. There's a fair chance your rocket's fins are "overmaneuvering". So if you have lots of fins, replace them with rigid fins, and keep maybe 4 moving ones for maneuvering.

I didn't test any rigid-fins-only/reaction-wheels-maneuvering-only setup yet. I've also heard of people having success by NOT using gimballed rocket engines.

[Jouni]

Of course I would have preferred the joints to be stiffer still, but not as physically strong. You can get some rather silly floppy dongers going on by stacking a lot of anythings (inline batteries, round8s, etc) which can survive insane forces but just won't stop flopping... It's still much, much better though.

I think the problem here is that the parts themselves are rigid.

Large structures generally bend under stress (think about planes, ships, and buildings, for example). In KSP, this bending has to happen at joints, because the parts themselves don't bend. But because ships built from small parts have more joints than ships built from large parts, it's impossible to set the ridigity of the joints right. No matter what you do, either some ships become too rigid, or others become too floppy.

[tutike2000]

I can make any design rock solid with just a few of well placed struts...

Well yeah, but won't that increase part count and therefore decrease FPS ?

My biggest pet peeve is that docked vessels tend to have a positive feedback wobble - the amplitude of the oscillations get bigger and bigger and bigger, the ship making wild 90 degree bends and eventually snaps.

I've found that this only happens when controlled from certain points. If I switch to 'control from here' eventually I find a command pod or docking port that will get SAS to dampen the oscillations instead of amplifying them.

Edited June 18, 2015 by tutike2000

[Jouni]

My biggest pet peeve is that docked vessels tend to have a positive feedback wobble - the amplitude of the oscillations get bigger and bigger and bigger, the ship making wild 90 degree bends and eventually snaps.

I've found that this only happens when controlled from certain points. If I switch to 'control from here' eventually I find a command pod or docking port that will get SAS to dampen the oscillations instead of amplifying them.

That's a consequence of using a single PID controller.

You have two objects with different periods of oscillation connected by a and bendy joint. If the joint bends, the docked objects quickly end up in a situation, where the objects are in opposite phases of oscillation. The PID controller observes the phase of the control part, and issues commands to all parts generating torque according to that phase. But because the parts in the other object are in the opposite phase of oscillation, the PID controller gives them exactly the wrong commands. In many cases, the end result is that the more force the PID controller uses, the stronger the wobble becomes, and the more force the controller will subsequently use.

There are many partial workarounds, such as using MechJeb, changing the control part, disabling some reaction wheels / RCS thrusters / engine gimbals, and using more and/or bigger docking ports. The real solution would be using a more complex stabilization system that doesn't assume that the ship is rigid.

[Warzouz]

I've tested some launch stages with dead weight payloads (mostly streamlined deactivated tanks). I didn't encounter wobbling when testing those rockets first stage (I remembered testing a launcher with a payload of 5 orange tanks stacked with a nose cone). But when I started to actually using them with real payload wobbleing came into play...

It mostly came from docking rings and overcompensating SAS. In beta 0.9, struts where free and dragless, it wasn't an issue. In 1.0, it's not.

Docking ring bending wouldn't be an issu if there were some "corset" parts that rigidify it during launch. Some kind of specific "separator/decoupler"

Fairing helps with drag, but the payload can wobble outside of it (which looks totally silly). As Fairing weight is really high, it could at least remove/vastly reduce wobble for parts inside it.

As I said on heat. This feature is interesting if there is an understandable way of managing it. In KSP, managing wobble is not easily understandable. Either there are some mission parts, either the mecanisme to handle it have to be more integrated in the gameplay, either it has to be removed. At least that could be some slider parameter when starting a carreer.

For now, it looks like a bug or, at least, a badly designed feature.

[Jouni]

As I said on heat. This feature is interesting if there is an understandable way of managing it. In KSP, managing wobble is not easily understandable. Either there are some mission parts, either the mecanisme to handle it have to be more integrated in the gameplay, either it has to be removed. At least that could be some slider parameter when starting a carreer.

Wobble is not always easy to understand, but it's intuitive, because it works more or less in the same way as for real structures. Try building something from Lego bricks, and see what happens, when you quickly lift the structure from ground. Repeat the same with different structures, and see which designs work and which don't.

[Juicy]

wtb Sepatronic Struts! Basically, struts that can detach themselves when you no longer need them.

Why? A lot of wobbling happens within a single stage. But, strutting up may prevent or block the use of components within that stage (dockingports and such).

Strutting to a different stage can be problematic when the struts prevent that stage from blasting away far enough for clearance.

Both issues could be solved with struts that detach themselves.

[clivman]

wtb Sepatronic Struts! Basically, struts that can detach themselves when you no longer need them.

Why? A lot of wobbling happens within a single stage. But, strutting up may prevent or block the use of components within that stage (dockingports and such).

Strutting to a different stage can be problematic when the struts prevent that stage from blasting away far enough for clearance.

Both issues could be solved with struts that detach themselves.

Or just make the joints made out of steel instead of green beans?

[anarkhon]

crafts wobbling in KSP,its not a feature just unfinished mess like "water".