Please no wobble (Solved!!!)

[Superluminaut]

I'm kind of stunned that an artifact of the original game's game engine limitations were carried over to the sequel.

Wobbly rockets were always a bug. The KSP1 devs spent years trying to minimize the problem, even bringing on new members to work on the issue.

Most people don't like them, and they introduce an un-intuitive stumbling block for new players or people that want to learn space. eg. Rocket veers off course on launch. Why? The control part wobbles away from the heading, resulting in SAS shenanigans and off COM thrust.

Furthermore, everyone gets rid of wobble to the best of their ability by adding struts, resulting in a higher part count. So you have a feature, that only produces greater part counts. Why?

To keep the destructive effect, just define stress tolerances for parts at which they explode, disconnect, crumble or shear.

Also, real rockets don't wobble.

Scott talking to KSP1 devs about the wobbly rocket bug.  They go off on tangents but almost the entire rest of the interview is on the topic.

KSP1 dev describes his next gen parts physics sim in which wobbliness is not a feature.

---

If you're new to KSP, wobble is that wet noodle, jello rocket thing.

 

Please get rid of wobble all together. Thank you.

Development on the issue:

 

 

Quote

From the June 16th Development Update:

A Word About Wobbly Rockets
Our team shares the community view that overly-wobbly rockets are a major issue in KSP2 (it is number 10 on our top-ten issues list). We have introduced a number of mitigations to address aspects of that issue (altering inertia tensor values to decrease joint issues that emerge when high-mass and low-mass parts are connected, introducing various bespoke multi-joint augmentations to areas of known over-flexibility), but we still see this as an area where major improvement is needed. For the record, this is our official view on what a successful implementation would look like, and against which we continue to measure the effectiveness of ongoing mitigation work:

• For inline parts that are connected serially, in most applications there should be little to no flexing. This is especially true when neighboring inline parts are the same core size.
• For radially-attached boosters or cantilevered subassemblies with single-point radial connections, some flexibility is expected. There are some applications for which manually-applied struts should be required.
• Wings should not require struts to stay rigid.
• Docking two vessels in orbit should result in a strong, non-wobbly connection that doesn’t fold on itself as soon as the player tries to move the resulting vehicle.
• Wobbly rockets are sometimes fun and funny. A big part of what originally got many of us hooked on the original KSP was the silliness and emergent problem solving that came from playing "World of Goo" with rocket parts. Broadly, we see this as part of the Kerbal DNA, and want to preserve it in some form. Whether that means limiting wobbliness to certain types or sizes of parts, or relegating certain behaviors to player settings, is the subject of ongoing internal discussion. We of course are following community conversations with keen interest, and this is an area where Early Access participants can have a significant impact on the 1.0 version of KSP2.
• Joint physics impact CPU performance, and as we progress through the Colony and Interstellar roadmap milestones the part counts will increase dramatically. Any solutions we arrive at for the above requirements must accommodate this reality.
• We would like to move away from autostrut, or any other band-aid solution that involves hidden settings that automatically apply additional joints to make a vehicle more rigid. Whatever solution we arrive at, we’d like it to be predictable and transparent to all users. If over the course of Early Access we find that some form of autostrut is still necessary to allow the creation of ambitious vehicles, we’ll revisit this requirement.

As a person who has dive-bombed more than one physics meeting with an exasperated "can’t we just make the joints stiffer" comment, let me assure you that in true KSP fashion, this is not a problem with a simple remedy. We’ve got very capable people on the case, and we will arrive at a good solution.

On 3/25/2023 at 3:52 PM, Superluminaut said:

Nate Simpson - creative director

from the 3/24/23 AMA

On wobbly rockets

Quote

This is a really good example of how having something in early access helps us prioritize and focus on the right things. This is obviously a very hot topic within the community, and it's also something that is frequently discussed within the team. I will describe my general goal for this.

When something is very skinny and made of many stacked parts, it should wobble. I think that if we were to move completely beyond rigid body physics, that we would have kind of subverted one of the things that's very fun and funny about this game. Do we want larger vehicles, do we want our interstellar vehicles to be wobbling around, do we want stuff that's larger than, lets say the 3.75 core size, to be wobbling all over the place? We do not. We are not happy with the current wobbliness of the vehicles. This is an area of current focus and heavy iteration and testing. And it will get better.

Edited February 10 by Superluminaut
added topic development news

[Strawberry]

The Wobble seems to be intentional as theres code settings to change it, but please remove it it sucks (at least for things that feel structurally secure). 

[silent_prtoagonist]

Agreed, wobbly rockets are amusing once, and just frustrating every time after that. 

[MarcAbaddon]

I think you need the physics underlying the wobble for certain aspects of the game and there are rocket builds that should wobble. But it needs to be tuned down, or controllable via autostruts or something similar.

[Periple]

I'd be perfectly happy if they just made most joints rigid and left wobble to some very specific ones. Noodle rockets aren't fun. Radial decouplers maybe? But even then I don't see the point as I'll just autostrut as a matter of course to get rid of the wobble. (Once autostrut is available that is)

[RayneCloud]

We've seen some stronger and more advanced strut parts in the back end, so this might be some sort of design decision to force players to build smaller and more compact vehicles at the start of their campaign, using a Tech 1 strut to move to larger vehicles and a Tech 2 strut to go even larger. I think the problem is that, the joint rigidity is way to low even if this was a design decision. Though, if you fly the stock rockets the K1 and K2, they have no issues with wobble what so ever. (The K2 needed fixing, so I threw a stack separator in it.)

[MarcAbaddon]

If you know which parts are prone to wobble it helps somewhat. Fairings like to wobble for example.

[Nuke]

while ksp1 kind of solved the problem, and added autostrut for situations where it was bad (which i almost always turned on). that's kind of a kludge though. id do a system where its automatically enabled when connecting two parts of equivalent diameter.

[BowlerHatGuy3]

Jiggly rockets are absolutely intentional as the jigglyness would be a very easy fix, as in literally changing a few numbers.

I do think that they should stay for the most part to make sure that you don’t make your rocket too tall and so that your craft can be pulled apart by wind forces. But as they are in the game right now rockets are way too flaccid. 

Just adding autostrut is kind of a lazy fix imo. I’d much rather have them change the actual values.

 

Edited March 2, 2023 by BowlerHatGuy3

[Superluminaut]

4 hours ago, BowlerHatGuy3 said:

I do think that they should stay for the most part to make sure that you don’t make your rocket too tall and so that your craft can be pulled apart by wind forces.

 

Spring joints are one of the worst solution for this effect. A limit at which parts crumple would be a much more interesting height restriction. Rockets also don't bend under high drag, the weakest segments simply break free or shatter. See above video of the proton.

Edited March 2, 2023 by Superluminaut

[EvelynThe Dragon]

It's a bad design decision as implemented. Real structures do flex, and sometimes catastrophically, but there is no reason that the rockets in KSP2 should be overcooked noodles (or even al dente to be honest). Yes, I do think we should have conditions where structural failure can occur, but this is way over exaggerated to the point of it being ridiculous.

 

[Superluminaut]

3 hours ago, EvelynThe Dragon said:

Real structures do flex, and sometimes catastrophically, but there is no reason that the rockets in KSP2 should be overcooked noodles (or even al dente to be honest). Yes, I do think we should have conditions where structural failure can occur, but this is way over exaggerated to the point of it being ridiculous.

 

Right, everything is a spring, and maybe ksp could go that route if they modeled tensile and compression strengths in the monocoque structure. Then after some degree of flex --not much because materials are selected for rigidity-- the segment with the weakest compression strength would snap.

We want those dried spaghetti rockets please.

I suspect much simpler models would produce results that look just as good and require no simulation. Simply monitor angle of attack and drag, mass and gs, etc, and when a limit is met trigger an appropriate failure mode in the weakest element.

Edited March 2, 2023 by Superluminaut

[Superluminaut]

Here is a poll on rocket flex.

 

[Bej Kerman]

18 hours ago, Superluminaut said:

an artifact of the original game's game engine limitations

What?

[Superluminaut]

4 minutes ago, Bej Kerman said:

What?

Check the dev interview linked in the op.

Unity had some unknown variables set somewhere in the code that limited the spring simulation, spitting out garbage that resulted in what the community has called wobble.

Notice that in the video they are playing a fairly developed version of ksp and have only found a fix at that point.

[Strawberry]

12 hours ago, RayneCloud said:

We've seen some stronger and more advanced strut parts in the back end, so this might be some sort of design decision to force players to build smaller and more compact vehicles at the start of their campaign, using a Tech 1 strut to move to larger vehicles and a Tech 2 strut to go even larger. I think the problem is that, the joint rigidity is way to low even if this was a design decision. Though, if you fly the stock rockets the K1 and K2, they have no issues with wobble what so ever. (The K2 needed fixing, so I threw a stack separator in it.)

Oh this sounds like a terrible design choice, seems much more enjoyable to just have you build smaller parts on the vab through height and mass limits.

[Bej Kerman]

1 minute ago, Strawberry said:

Oh this sounds like a terrible design choice, seems much more enjoyable to just have you build smaller parts on the vab through height and mass limits.

Honestly, dynamic limitations introduced by your designs and their physics sounds a million times better than arbitrary limitations imposed by fixed values that have no reason to exist besides Da Balance

[Strawberry]

Just now, Bej Kerman said:

Honestly, dynamic limitations introduced by your designs and their physics sounds a million times better than arbitrary limitations imposed by fixed values that have no reason to exist besides Da Balance

The dynamic limit is do you have to attach 50 struts to your rocket manually or 5 to your rocket manually.

[Bej Kerman]

2 minutes ago, Strawberry said:

The dynamic limit is do you have to attach 50 struts to your rocket manually or 5 to your rocket manually.

The dynamic limit is mess around and find out until you find something that works.

I would hate to be fighting arbitrary limits that have nothing to do with the rocket itself. Noodles aren't pretty but joint stress can at least be worked into something more realistic, and either way it gives you an engineering problem rather than a "make sure this number doesn't get bigger than x" problem.

[Strawberry]

Just now, Bej Kerman said:

The dynamic limit is mess around and find out until you find something that works.

I would hate to be fighting arbitrary limits that have nothing to do with the rocket itself. Noodles aren't pretty but joint stress can at least be worked into something more realistic, and either way it gives you an engineering problem rather than a "make sure this number doesn't get bigger than x" problem.

The finding around is attaching 6 struts or 60 struts, the way people will treat this is mostly just brute force as its easier then redesigning your rocket. The purpose of a game is to push people towards having fun, not pushing people towards doing tedious mind numbing stuff.

Mass and height do very much involve the rocket themselves, and having to deal with mass and height limits changes your rocket design in a much more interesting way then just attaching even more struts to your rocket. Weak struts can be dealt with easily albeit tediously with just adding more struts, height limits can only be dealt with by reducing the height.

[Bej Kerman]

2 minutes ago, Strawberry said:

The purpose of a game is to push people towards having fun, not pushing people towards doing tedious mind numbing stuff.

Such as making x number go down until it's below y? Believe me, a game about engineering of all things shouldn't resort to fixed values. If you are brute forcing stuff, that's on you and isn't an indication of how others might choose to make a rocket that stays together.

Edited March 2, 2023 by Bej Kerman

[Strawberry]

Just now, Bej Kerman said:

Such as making x number go down until it's below y? Believe me, a game about engineering of all things shouldn't resort to fixed values. If you are brute forcing stuff, that's on you and isn't an indication of how others might choose to make a rocket that stays together.

I mean making people consider how you can reduce weight of a spacecraft is like yeah a good thing. Honestly I dont see why you wouldnt brute force it with struts, its the faster and ideal solution, it saves you time and doesnt mean you have to rebuild everything youve just spent the past half hour putting together.

[Bej Kerman]

Just now, Strawberry said:

I mean making people consider how you can reduce weight of a spacecraft is like yeah a good thing. Honestly I dont see why you wouldnt brute force it with struts, its the faster and ideal solution, it saves you time and doesnt mean you have to rebuild everything youve just spent the past half hour putting together.

It's a better alternative to fixed limitations that have no reason or rhyme behind them.

[Strawberry]

1 minute ago, Bej Kerman said:

It's a better alternative to fixed limitations that have no reason or rhyme behind them.

The reason is "This is a tiny VAB and launchpad, it is not large enough to build these things yet", or alternatively "We do not want you to brute force game progression, work around these game limitations by limiting your scope of your missions before you can pass them."

[Bej Kerman]

3 minutes ago, Strawberry said:

The reason is "This is a tiny VAB and launchpad, it is not large enough to build these things yet"

That's not a reason, the VAB isn't tiny and and these limitations aren't realistic.