How should rockets flex?

[MechBFP]

8 hours ago, Pthigrivi said:

there's way to much flex in KSP2 right now, but there also shouldn't be zero. 

Yup, exactly. 

[TwoCalories]

I'd like it realistic: flex a little, break under major joint stress, and rip apart under extreme aerodynamic forces. In KSP1 I use Kerbal Joint Reinforcement, and it gets rid of all the wobble, but it's better than the alternative. With KSP2 being built from the ground up (and that's why it has glitches that weren't in KSP1 and took extra development time) the devs can rework some of the game design.

Edited July 5, 2023 by TwoCalories

[Pthigrivi]

This is interesting; the positioning optics for the Saturn V accounted for 1/3 of a meter of flex just from wind on the pad: 

https://www.tiktok.com/t/ZPRWmdXqd/

[cfds]

59 minutes ago, Pthigrivi said:

This is interesting; the positioning optics for the Saturn V accounted for 1/3 of a meter of flex just from wind on the pad: 

So a relative deflection of 0.3% to 0.4% (depending on the height of the sensor on the rocket and assuming no flex in the launch platform or the clamps)? Also known as "not visible to the naked eye"? Zero flex may not be entirely correct, but is far more correct than visible flex. Which, again, is not an intentional feature but a limitation of the engine.

Edited July 5, 2023 by cfds

[Alexoff]

5 hours ago, Pthigrivi said:

This is interesting; the positioning optics for the Saturn V accounted for 1/3 of a meter of flex just from wind on the pad: 

https://www.tiktok.com/t/ZPRWmdXqd/

Skyscrapers flex even more, it seems several meters at a height of 500 m. Because of this, it is difficult to set an even vertical and horizontal inside, the building constantly fluctuates in the wind. But without instruments from the outside and especially from the inside, it is difficult to notice. If skyscrapers were flexed like in KSP2, then they would not be built higher than 100 meters.

[PDCWolf]

8 hours ago, Pthigrivi said:

This is interesting; the positioning optics for the Saturn V accounted for 1/3 of a meter of flex just from wind on the pad: 

https://www.tiktok.com/t/ZPRWmdXqd/

Pad sway is not so much the structure flexing, as it is the entire rocket... rocking, as they're attached only a ground level, so exactly like a building. Funnily enough, Falcon 9 is considered to be the (or close to the) limit for tall vs thin.

8 hours ago, Pthigrivi said:

Im talking about things like the below (2:00 minutes into the Matt Lown video).

 

How does real life cope with an entire rocket stage + its payload inside a fairing? a single ring.

[Pthigrivi]

Yeah again I think we're in agreement here in principle. You can see that "forward load reactor", a structural ring thats in contact with the fairing. It's not just wobbling around on one point. And as you say Falcon is within but pushing the "tall and thin" limits, and it wiggles some, but not enough to ruin a flight. If a player made something that was say 3 times that long and skinny you'd expect that wobble to be much harder to manage. I think thats pretty close to how the game should perform. 

[PDCWolf]

2 hours ago, Pthigrivi said:

Yeah again I think we're in agreement here in principle. You can see that "forward load reactor", a structural ring thats in contact with the fairing. It's not just wobbling around on one point. And as you say Falcon is within but pushing the "tall and thin" limits, and it wiggles some, but not enough to ruin a flight. If a player made something that was say 3 times that long and skinny you'd expect that wobble to be much harder to manage. I think thats pretty close to how the game should perform. 

And my difference in thinking (that wobble should be completely gone) is that same "too tall and thin" structure could be applicable to other stuff, like a long truss to hold solar panels on a station/base, where it shouldn't wobble since it's not supposed to be the same stuff. However, in KSP1 wobble doesn't differentiate, and I seriously doubt devs are up to implementing that in 2, thus the best solution is that wobble be removed.

[TwoCalories]

6 hours ago, PDCWolf said:

Funnily enough, Falcon 9 is considered to be the (or close to the) limit for tall vs thin.

 

Tall needle-thin rocket go brrrrr.

Jokes aside, I actually never thought that rockets would sway on the launchpad from the wind. I always thought they were clamped down by the umbilical arms and stuff.

[mcwaffles2003]

looks possibly wobbly to me...

KSP 2 is far more exaggerated, but the 30cm sway of the Saturn V is doubtfully the limit of the wobble it can make without breaking as that's with a mild crosswind. 

Also, if we eliminate all flexability of joints that will limit our ability to make stuff like this

 

which would be sad..

[Alexoff]

48 minutes ago, mcwaffles2003 said:

Also, if we eliminate all flexability of joints that will limit our ability to make stuff like this

 

which would be sad.

Oh, those good old times when bugs in KSP were funny and allowed you to do incredible things! How few of these videos are now ...

Edited July 5, 2023 by Alexoff

[PDCWolf]

1 hour ago, mcwaffles2003 said:

Not only is that a nitpicked frame, it is also after the failure (pretty much 90º into the airstream as well). I could also nitpick 200 milliseconds before your frame to say it didn't wobble, it just snapped:

Here's the video for people to draw their own conclusions from (you can operate frame by frame with comma and period):

 

[mcwaffles2003]

20 minutes ago, PDCWolf said:

Not only is that a nitpicked frame, it is also after the failure (pretty much 90º into the airstream as well). I could also nitpick 200 milliseconds before your frame to say it didn't wobble, it just snapped

I wasn't trying to imply the rocket survived, only that they can bend and don't just instantly sheer. you'd have to be some level of insane to think that image is an attempt to show proof of flexibility without RUD in imminent future. Theres only 2 videos of rockets flopping all kerbal-esque through the atmosphere to my knowledge and the maiden flight of starship both didn't tear apart through aerodynamic forces nor was in high enough resolution to determine if it was significantly bending.

[Alexoff]

4 minutes ago, mcwaffles2003 said:

I wasn't trying to imply the rocket survived, only that they can bend and don't just instantly sheer.

The rocket practically did not bend, its fairing broke from overloads.

[RocketRockington]

1 hour ago, mcwaffles2003 said:

I wasn't trying to imply the rocket survived, only that they can bend and don't just instantly sheer. you'd have to be some level of insane to think that image is an attempt to show proof of flexibility without RUD in imminent future. Theres only 2 videos of rockets flopping all kerbal-esque through the atmosphere to my knowledge and the maiden flight of starship both didn't tear apart through aerodynamic forces nor was in high enough resolution to determine if it was significantly bending.

Picking the frame right before the fairing tore off isn't an example of 'bending'.  It's an example of something breaking off and it still looking like it was attached because you picked the exact right frame to maximize the visual.  It wasn't going to snap back into place elastically like a kerbal rocket.   The fact that people are going to such crazy lengths to 'prove' wobble is a thing is just a great demonstration that it isn't.

[Pthigrivi]

But yall, even this kind of misses the point. If you are asking to remove ALL flex from the game you are essentially removing internal physics from the game. You’ve removed structure from the game. Thats actually much worse than bendy rockets. This is baby-out-with-the-bathwater rock-bottom expectations. If instead you’re saying wildly unrealistic rockets should not flex but simply spontaneously self-destruct when too much force is put on them you’ve created another problem: you’ve removed a key visual cue to players about the nature of the failure. Was my staging off? Is it a bug? Was it aerodynamic stress overload? Who knows! Because there was no visible cue to the player to understand “this rocket is too spindly or poorly designed to be stable.” 
 

 

Btw Im an architect by training. In real life Everything moves. Buildings sway, and everything bends. Even the earth slumps with time and pressure. And especially in extreme situations you need to understand that. Hurricane uplift, worst-case lateral earthquake loads. Even in completely common conditions you need to understand snow-loads and resulting deflection and live loads from occupancy and storage.  Obviously KSP is simplified and not everything can be in scope but just the way KSP teaches orbital mechanics it also, in a loosey-goosey way, teaches statics and basic load-transferal principles. Nobody talks about this but its a critical aspect of KSPs appeal. 

Edited July 6, 2023 by Pthigrivi

[cfds]

The problem with visible bending is that it is not just a visual effect that gives the user feedback: It has knock-on effects on the physics computation, leads to offset thrust and incorrect heading/bearing values, especially if the satellite inside the fairing is the "control point" for SAS. The current single joint simulation has nothing to do with how materials behave in real life, and the "feedback" the user gets is not helpful: The problem is never that the rocket is "unrealistically skinny", the problem is that the rocket has too many joints.

The only way bendy rockets would be sensible is if the underlying physics simulation were a lot more complex: At least joints between parts would have to be some form of triangle lattice, ideally the parts themselves as well. KSP 1 did not do this because they simply did not have the capacities and we (begrudgingly) accepted it because of "small team of amateurs". That excuse does not fly anymore when a professional studio is doing the development.

[LoSBoL]

7 hours ago, RocketRockington said:

The fact that people are going to such crazy lengths to 'prove' wobble is a thing is just a great demonstration that it isn't.

Funny thing is that all arguments for full rigidity all come down to fear of missing out. Fear of not being able to launch or construct what you are able to in KSP1, based on no confidence in the development team being able to sort it out.

I can understand that fear, and wanting the bandage called full rigidity. I am however certain the devs do have an understanding that facilitating players in wanting to play in KSP2 what can be done in KSP1 and they'll come up with a solution.

[RocketRockington]

1 hour ago, LoSBoL said:

Funny thing is that all arguments for full rigidity all come down to fear of missing out. Fear of not being able to launch or construct what you are able to in KSP1, based on no confidence in the development team being able to sort it out.

I can understand that fear, and wanting the bandage called full rigidity. I am however certain the devs do have an understanding that facilitating players in wanting to play in KSP2 what can be done in KSP1 and they'll come up with a solution.

[Snip]  That's not most people's position at all.  Most people want to be able to do more than KSP1, have good frame rates with higher part counts, build bigger rockets or orbital structures, etc.  That's not fomo, that's having expectations based on it being a higher priced sequel.  The only 'fear' is that the ksp2 devs will keep shackling the game to some outdated idea of what 'keebal' means and what makes it fun.

Edited July 6, 2023 by James Kerman
Redacted by a moderator

[LoSBoL]

32 minutes ago, RocketRockington said:

[Snip]  That's not most people's position at all.  Most people want to be able to do more than KSP1, have good frame rates with higher part counts, build bigger rockets or orbital structures, etc.  That's not fomo, that's having expectations based on it being a higher priced sequel.  The only 'fear' is that the ksp2 devs will keep shackling the game to some outdated idea of what 'keebal' means and what makes it fun.

I didn't talk about most people's position, what most people want can be found in the outcome of the poll when concerning wobbly rockets.

[Alexoff]

7 hours ago, Pthigrivi said:

you’ve created another problem: you’ve removed a key visual cue to players about the nature of the failure. Was my staging off? Is it a bug? Was it aerodynamic stress overload? Who knows! Because there was no visible cue to the player to understand “this rocket is too spindly or poorly designed to be stable.” 

This can be solved in a bunch of ways. For example, to give the opportunity to see the destruction of the rocket in the slowmo after the fail. How it happens in real life. Or make normal damage reports. This is certainly more difficult than just adding a few zeros to the game parameters, much less solving the anti-aliasing problem, but it seems to me that it is still better than watching steel cylinders wobble with flammable fuel over and over again.

1 hour ago, LoSBoL said:

however certain the devs do have an understanding that facilitating players in wanting to play in KSP2 what can be done in KSP1 and they'll come up with a solution.

I'm just wondering - can the developers do something that you wouldn't approve of?

[LoSBoL]

14 minutes ago, Alexoff said:

I'm just wondering - can the developers do something that you wouldn't approve of?

Absolutely,  KSP¹ for me gave me the great experience of figuring out everything myself, at the time I started playing there was no TWR reading, no burn time indicator, no transfer window planner, no guidance on how to get to other planets, nothing. That was awesome, for me, because it matches my contemplation that 'space is hard'. So I loved the learing experience of needing to investigate to master.

So now KSP² is being developed, and with that a vastly better onboarding for new players, it saddens me somewhat that what I experienced, new players experience in a lesser form.

But that's me and my experience, and not everyone is waiting for that particulair experience, so I can understand as to why KSP2 is being developed with a mindset to help players better than KSP1.

So I have a choice, argument that I want that experience back for KSP2, or just not at all.

[PDCWolf]

10 hours ago, Pthigrivi said:

But yall, even this kind of misses the point. If you are asking to remove ALL flex from the game you are essentially removing internal physics from the game.

No, the physics are still there, as the construction is still a node tree with limited joint capacity. What isn't there is something that does not happen in real life nor is analogous to anything that happens in real life.

 

11 hours ago, Pthigrivi said:

This is baby-out-with-the-bathwater rock-bottom expectations.

Well, there's 1 for the "game too easy" complaints you never heard. Funny enough that you think wobble adds any sort of difficulty when all you have to do is right click -> autostrut.

11 hours ago, Pthigrivi said:

You’ve removed a key visual cue to players about the nature of the failure.

A "key" visual that relates to nothing in real life. Further on, if your rocket explodes with no reason, that's still a reason: You didn't stage, you didn't see heat indicators going off, and more importantly, in both games you get a huge screen in your face stopping action and telling you exactly why a part was destroyed. Don't forget that.

5 hours ago, LoSBoL said:

 fear of missing out.

Well, in KSP1 you have to auto-strut anything that isn't a short rocket, or it becomes spaghetti. It's not "fear of", it's already happened. Yes, KSP2 has a different dev team, but they haven't exactly proved themselves competent yet to trust them to not do another bandage mess when they want to call 1.0 and move on like the previous ones did.

 

 

[linuxgurugamer]

Vote this up if you think it make sense, after watching it

 

 

[The Aziz]

Most of the community thinks exactly that, and I see this video posted here for like 20th time.

Give it a break.