[1.3] Kerbal Joint Reinforcement v3.3.3 7/24/17

[crazyewok]

This mod is great but it still is having problems with large rockets. I have a 100 meter tall 10 meter wide rocket and as soon as it appears on the pad it starts swaying and bending.

Same. Mine nothing complicated. But it still wobbles and fall apart like made of glue

[Surefoot]

I'm gonna guess that's not KJR, since it sounds like a center of mass issue, and KJR doesn't mess with that...

I was suspecting a kind of induced shift due to propagating stresses, for example, but the design is totally symmetric, i fail to see any reason why the stress would be unbalanced to start with.

What do all of these vehicles have in common? Do all these vehicles use fuel lines? Does the issue continue if you remove the fuel lines and manually pump fuel (if necessary)?

Ohh yes i didnt try that yet (yes they have fuel lines). Otherwise nothing in common apart from being a symmetrical radial tank design. I even changed the radial attachment, trying different solutions.. Good idea, i'll have a shot at that one (and report on MFT thread, sorry for the hijack).

Does removing MFT fix the issue?

I know that I had some weird MFT-caused CoM changing with v4, but v4.1 fixed that for me.

Good point, i'll see without MFT (i have to remove stretchytanks and toroidal tanks too then do i)

[chrisb2e9]

Hello!

I grabbed this mod a few days ago hoping that it would solve my problem with noodle rockets...

In some ways it did. In some ways it did not.

Take a look at this one for example, I have three SR docking ports. Not an easy feat to get them all to dock.... but I managed.

Problem is that it was still just a loose noodle even while simply turning the ship. The problem from what I could see, was connections between things like the SAS modules, the battery, the... computer... control.... thing... attached to the battery. Basically the smaller parts on the rocket.

The docking port section was rock solid. As were the larger parts.

What do I need to edit in the config file to make the smaller connection not wobble,(or wobble as little as possible)

Thanks!!!

And thanks for making great mods like this one!

[Ringkeeper]

Problem is that it was still just a loose noodle even while simply turning the ship. The problem from what I could see, was connections between things like the SAS modules, the battery, the... computer... control.... thing... attached to the battery.

]

The 2 round batteries , the big SAS sometimes and definitive the Sr. Docking Port (worst off all) have some huge problems. It seems the connection between them and other modules doesnt even exist.

But i dont think its a problem of the mod.

[TMS]

The mass at either end of that craft, with only a weedy corridor connecting it, will pretty much ensure wobble - mod or not.

[chrisb2e9]

]

The 2 round batteries , the big SAS sometimes and definitive the Sr. Docking Port (worst off all) have some huge problems. It seems the connection between them and other modules doesnt even exist.

But i dont think its a problem of the mod.

The mass at either end of that craft, with only a weedy corridor connecting it, will pretty much ensure wobble - mod or not.

So there really isn't much I can do then is there? I'm doing a mission to Jool, maybe I should just fly each ship there separately? And refuel the lander as needed once both are there...

I would rather fly them as one ship. less work for me. But even at 1m/s acceleration. It broke apart. not from wobbling, but the battery connecting to the remote guidance unit crushed into each other and broke. The thing weighs in at 1700T. I guess I just have to fly em separate?

i used to use docking struts. But they don't seem to work like they used to.

[Neutrinovore]

So there really isn't much I can do then is there? I'm doing a mission to Jool, maybe I should just fly each ship there separately? And refuel the lander as needed once both are there...

I would rather fly them as one ship. less work for me. But even at 1m/s acceleration. It broke apart. not from wobbling, but the battery connecting to the remote guidance unit crushed into each other and broke. The thing weighs in at 1700T. I guess I just have to fly em separate?

i used to use docking struts. But they don't seem to work like they used to.

I find Quantum Struts to be quite useful in situations like this. Any particular reason why you wouldn't want to use those?

[chrisb2e9]

I find Quantum Struts to be quite useful in situations like this. Any particular reason why you wouldn't want to use those?

I had been using the docking strut mod. which is essentially the same thing but I always thought it looked better. I was under the impression that quantum struts doesn't work with .23

Am I wrong on that?

[Neutrinovore]

I had been using the docking strut mod. which is essentially the same thing but I always thought it looked better. I was under the impression that quantum struts doesn't work with .23

Am I wrong on that?

They work fine for me. Oh, well, I don't know if they're integrated into career mode, though, I only ever play in sandbox. But yeah, functionally they work perfectly. And as far as how they look, if you don't like the 'blue laser beam' look of the 'Quantum Strut', just use the 'Strut Gun'. It works exactly the same, it's just that the 'beam' is grey instead of a blue laser beam. This way it looks exactly like a regular strut.

[chrisb2e9]

They work fine for me. Oh, well, I don't know if they're integrated into career mode, though, I only ever play in sandbox. But yeah, functionally they work perfectly. And as far as how they look, if you don't like the 'blue laser beam' look of the 'Quantum Strut', just use the 'Strut Gun'. It works exactly the same, it's just that the 'beam' is grey instead of a blue laser beam. This way it looks exactly like a regular strut.

I didn't know about the strut gun. I'll give it a try. Currently I decided to send both ships on their own and connect at Jool to refuel the lander as needed.

Also decided to make a fueling station just outside Minmus orbit so that they can fuel there before making the trip.

[Castun]

Strut guns are awesome. I've built some pretty long ships using them together with KJR.

[Galane]

One thing about the quantum struts and strut guns is they're very rigid, don't seem to have any flex at all like the stock struts. They must be employed carefully because on the launch pad, if your rocket's parts sag or slump a bit, these struts can push connections apart.

That rigidity makes them very useful for docking, especially when you need to move things that are docked together. Action grouping the struts may not be required, unless they fail to automatically engage upon docking or you want to turn them off to save a bit of power.

There's a "feature" with the physics which can be exploited with quantum struts, possibly with strut guns. Put one of the little truss "struts" (such as the cubic octagonal strut) on your rocket then a quantum strut somewhere so that when it's on, the beam hits the little truss. Your ship's rotation is locked solid without needing to use any reaction wheels or RCS thrust.

Action group a toggle on the strut and you don't have to bother with what that ship will be doing while docking another ship to it.

[Starwaster]

So there really isn't much I can do then is there? I'm doing a mission to Jool, maybe I should just fly each ship there separately? And refuel the lander as needed once both are there...

I would rather fly them as one ship. less work for me. But even at 1m/s acceleration. It broke apart. not from wobbling, but the battery connecting to the remote guidance unit crushed into each other and broke. The thing weighs in at 1700T. I guess I just have to fly em separate?

i used to use docking struts. But they don't seem to work like they used to.

Seems like periodically I have to keep reposting this, so here we go again.

There are two properties in each part's configuration: breakingForce and breakingTorque. If those properties are not set in the config file they default to a value that is too low even for normal usage (based on my emperical observations and having parts like engines spontaneously disconnect when forces operating on the ship should have been non-existent)

This fix requires ModuleManager 1.5.x and above. Make a text file named breakingforce_fix.cfg and put it somewhere in your Gamedata folder. (I have a tweaks folder just for this purpose)

breakingforce_fix.cfg

Cut and paste the following code into the text file and save it. Must have an extension of .cfg!

@PART[*]:HAS[~breakingForce[]]
{
	breakingForce = 200
}
@PART[*]:HAS[~breakingTorque[]]
{
	breakingTorque = 200
}

[Zander]

Starwaster will that also fix my problem?

[ferram4]

@Zander: Your issue, the 100m length + 10m diameter rocket swaying is going to be difficult to deal with no matter what; part mass goes with length^3 but connection strength only goes with length^2, so there's really nothing you can do, beyond me looking at more and more convoluted ways to try and make the physics initialization gentler. At a certain point I'm going to be messing with forces that I shouldn't be messing with.

Also, breakForce controls what force is required to break the joint, but it doesn't control the stiffness at all.

Besides that, Starwaster's fix is overriden in KJR v1.6, where KJR sets the breakForces itself, unless the breakForces would be smaller than the originally set forces. If anyone's using KJR v1.6 and they're still dealing with parts breaking the issue is that their designs are garbage. Beyond that point all that's really left in their arsenal is cheating joints into indestructibility.

[Zander]

@Zander: Your issue, the 100m length + 10m diameter rocket swaying is going to be difficult to deal with no matter what; part mass goes with length^3 but connection strength only goes with length^2, so there's really nothing you can do, beyond me looking at more and more convoluted ways to try and make the physics initialization gentler. At a certain point I'm going to be messing with forces that I shouldn't be messing with.

Also, breakForce controls what force is required to break the joint, but it doesn't control the stiffness at all.

Besides that, Starwaster's fix is overriden in KJR v1.6, where KJR sets the breakForces itself, unless the breakForces would be smaller than the originally set forces. If anyone's using KJR v1.6 and they're still dealing with parts breaking the issue is that their designs are garbage. Beyond that point all that's really left in their arsenal is cheating joints into indestructibility.

Well we have an unrealistic situation when there's any wobble at all in joints. In real life if the joint separates at any point even by an inch the entire rocket will rip apart. Minor flexing of course does occur but not at the joints. Anyways the problems with my rocket only occurs after something is decoupled.

Its really weird. The thing is perfect until i set off a decoupler of some kind then the whole things goes like spaghetti. So if i don't use the launch clamps, and just launch like that its fine. But if i use launch clamps separating them will ruin the rocket. Can this be fixed?

[AbeS]

I think someone mentioned that the invisible struts disappear when you decouple, maybe that's it

[ferram4]

@Zander: I think I see the issue... the joints are breaking everywhere regardless of whether they should. I'll look into fixing it.

But other than that, there's only so much we can do in the game. Unless you would like to throw the lag through the roof with many more physics calculations. The problem is that rigid bodies connected by joints is not a good way of modeling soft body dynamics in a rocket.

[Galane]

The problem is that rigid bodies connected by joints is not a good way of modeling soft body dynamics in a rocket.

Have a look at http://www.rigsofrods.com

Rigs of Rods is an open source vehicle simulator licensed under the GNU General Public License version 3. What makes Rigs of Rods different to most simulators is its unique soft-body physics: vehicles, machines, objects, etc. are simulated in real-time as flexible soft-body objects, giving the simulation an extremely accurate behavior which entirely depends on the physical construction of the vehicles or objects you create.

Features

Soft-body physics. Objects according to their weight distribution, construction, and/or suspension (in the case of vehicles).

Advanced flight model based on blade element theory. It allows the accurate simulation of any airplane, based entirely on its physical dimensions and wing airfoils, similar to X-Plane.

Accurate buoyancy model based on elemental pressure gradients, enabling boats with complex hulls to move realistically in the swell.

Basic support for dual-core processing. More multithreading and CUDA support is planned.

Basic support for scripting using AngelScript.

Based on the OGRE Graphics Engine.

[Starwaster]

@Zander: Your issue, the 100m length + 10m diameter rocket swaying is going to be difficult to deal with no matter what; part mass goes with length^3 but connection strength only goes with length^2, so there's really nothing you can do, beyond me looking at more and more convoluted ways to try and make the physics initialization gentler. At a certain point I'm going to be messing with forces that I shouldn't be messing with.

Also, breakForce controls what force is required to break the joint, but it doesn't control the stiffness at all.

Besides that, Starwaster's fix is overriden in KJR v1.6, where KJR sets the breakForces itself, unless the breakForces would be smaller than the originally set forces. If anyone's using KJR v1.6 and they're still dealing with parts breaking the issue is that their designs are garbage. Beyond that point all that's really left in their arsenal is cheating joints into indestructibility.

I didn't realize that you were setting breakingForce / breakingTorque directly now now. Even so: unless the breakForces would be smaller than the originally set forces.

The config I posted is adding those values into the part's config node so unless I'm badly misinterpreting what you're saying, those values would become the bare minimum since you're not replacing those values with smaller values. Basically I'm not clear how that would be overriding them nor how it obviates the need for the fix I posted. Bottom line is, if those properties aren't present then default minimums are used that aren't even enough to stop unstrutted parts from spontaneously falling off in solar space.

[chrisb2e9]

If anyone's using KJR v1.6 and they're still dealing with parts breaking the issue is that their designs are garbage.

Looking at the picture that I posted specifically, where I applied 100% thrust, 4 engines with a thrust of 200 each for a total of 800. in less than a second the ship broke in two, where the remote controller(cant remember what it's called again) connects to the large battery. not from bending, but from being pushed into it.

The max acceleration at that time, according to mech jeb was only 1m/s.

If the design was garbage, than it was garbage. I'm fine with that, I can redesign. It just seems to me that I was dealing with a small amount of thrust, and with the exception of being bent too far, the smaller connections should have been able to handle the strain(a thrust of 800 is pretty small after all).

I had been changing things in the .cfg file. I'm just wondering if I changed something in the wrong direction to cause this to happen. Or if it's just, as you put it, a garbage design?

[ferram4]

800 kN is quite a lot of thrust to put through a 2.5 m connection in space, especially if there are very large masses on either end of the connection that can swing around, especially if the joint is at the CoM of the mess.

If you decreased the breakStrengthPerUnitArea value then things probably broke there. If not, changing some of the numbers can cause the joints to glitch out, resulting in the parts teleporting slightly between frames, causing very large forces to be applied to the part.

[Zander]

Looking at the picture that I posted specifically, where I applied 100% thrust, 4 engines with a thrust of 200 each for a total of 800. in less than a second the ship broke in two, where the remote controller(cant remember what it's called again) connects to the large battery. not from bending, but from being pushed into it.

The max acceleration at that time, according to mech jeb was only 1m/s.

If the design was garbage, than it was garbage. I'm fine with that, I can redesign. It just seems to me that I was dealing with a small amount of thrust, and with the exception of being bent too far, the smaller connections should have been able to handle the strain(a thrust of 800 is pretty small after all).

I had been changing things in the .cfg file. I'm just wondering if I changed something in the wrong direction to cause this to happen. Or if it's just, as you put it, a garbage design?

No offense but looking at that huge unwieldy thing, if it DIDN'T break when you applied any force to it then ferram is doing something in his mod that he shouldn't be.

I wanna see how you got that thing into space anyways. nose cones and interstage sections that don't properly cover whats below them.. so then hypersonic air is entering the interior of your rocket. yeah that thing should have ripped apart seconds after launch.

[DerekL1963]

If the design was garbage, than it was garbage. I'm fine with that, I can redesign. It just seems to me that I was dealing with a small amount of thrust, and with the exception of being bent too far, the smaller connections should have been able to handle the strain(a thrust of 800 is pretty small after all).

It's the bent too far that's causing the problem... once it starts to bend, even a small amount of thrust tends to increase the bending (and thus the stress on the joint) in a vicious positive feedback loop until the joint breaks. You're ship is simply too heavy.

[Garek]

You could try adding truss structures alongside the middle section to relief it of some of the stress.