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

[Galane]

Quantum struts have no flex so placing them requires some care.

For example, if you have a rocket with an orange tank and four small diameter tanks around it on TT-70 decouplers, with a Rockomax adapter on top of the orange tank, then use quantum struts to brace from the tops of the small tanks up to whatever is on top of the adapter - the quantum struts can pop the upper part loose from the adapter when the rocket is loaded on the pad.

KJR alleviates that some by slowing the sudden jerk as the physics engine is enabled. It's still good practice to build rockets to take thrust both directions axially. Mostly what I've used the quantum struts for on launch vehicles is to stop sway of upper stages, and if I do mount them on side boosters I also use standard struts from them to the core to hold the side boosters against putting upward pressure on the quantum struts.

So far I've not had a problem using the quantum struts or strut guns to reinforce docking connections. I'm still using the original version, not the Quantum Struts Continued update.

The one thing you must not do with quantum struts or strut guns is mount them on or aim them at little massless parts like the small cubic and octagonal girders that are mis-named as "struts". The ship's rotation in all axes will lock and various other odd things may happen. On a test vehicle I built, I mounted quantum struts onto the little girder cubes. When I turned on the struts, the cubes flew away but the strut projectors stayed put and I could turn them on and off.

[Zyglrox]

I can't seem to understand how the recent versions of KJR work. So, I'm using full RSS/RO modpack, all latest versions. Some (not all) of my rockets are very fragile, what's strange is that they are not the largest and have relatively low parts count. For example (imagine Atlas V 501):

1. payload

2. decoupler

3. procedural structural element

4. procedural fuel tank

5. engine (upper stage)

6. decoupler

7. procedural fairing base

8. procedural structural element

9. procedural structural element

10. procedural fuel tank

11. procedural interstage adapter (for shrouding the engine)

12. engine (first stage)

For some reason, this rocket can't handle any control input at all, it immediately starts flexing and spinning around like it's made of rubber. This is not a "bad aerodynamic flip" thing, the rocket in this regard should be fine. There's no ASAS module which interferes with gimbaling. I tried to mess with different config options (particularly, use volume not area, mass ratio for stiffening, uncomment "a's") and it didn't help. Now I have been away from KSP for several months, but I remember older versions of KJR and similar rockets in the 0.22/0.23 working flawlessly, so I even put config values from an older version of KJR (angular section) and it still didn't help.

But if I remove the "9" part which is a simple 3.8x3.8m cylinder, the rocket suddenly becomes very rigid, as it should, and responds and behaves perfectly. I just can't understand why one simple part breaks the connections in such way, is it KJR, Procedural Parts or other issue? And how can I tune the KJR config to glue these things better?

[Galane]

Zyglrox, at least post a picture of the rocket, on the pad and while it's failing.

I've been working (slowly) on updating PolecatEZ's old ReStock config file parts pack to have sizes that work with the Kerbodyne parts and in testing ran into a similar situation where if I flipped the Kerbodyne to Rockomax adapter upside down, put a Skipper under that, then a Rockomax decoupler and another adapter, with the upsized Bigger Beastlier Quad Coupler with four short Kerbodyne tanks and four of the quad chamber engines under it...

The result is a violent sort of circular sideways shaking until parts start bending every which way while the rocket goes out of control and usually eventually overcomes KJRs strengthening and breaks apart. I tried bridging across the Skipper engine with some quantum struts but that had no effect.

If I simply put one of the large Kerbodyne engines in the middle, no problem!

Edited May 19, 2014 by Galane

[Zyglrox]

Here it is, no control input at all, just turned SAS on and going straight up:

The small reaction wheel in the payload section can be turned on or off - it makes no difference.

If I change the part with "KSP" letters from structural element into a fuel tank, no difference too.

If I remove that part, the rocket controls just fine with no flexing, as it should.

The result is a violent sort of circular sideways shaking until parts start bending every which way while the rocket goes out of control and usually eventually overcomes KJRs strengthening and breaks apart. I tried bridging across the Skipper engine with some quantum struts but that had no effect.

Yup, similar symptoms.

And yeah, struts between upper stage fuel tank and fairing base don't help too.

Edited May 19, 2014 by Zyglrox
quoting

[NathanKell]

Are you using the latest (v3.01) proc fairings? Or an old version? If old, then node sizes of the fairing bases will be incorrect. Try opening their part.cfgs and change the stack top and bottom nodes' sizes to be correct; in the 2.5m base, for example, change the seventh number (or add if only 6 numbers) for both stack_top and stack_bottom to 2. (Or, for RO's purposes, 3.)

[Zyglrox]

Yes, using 3.01. I tried to change node size to 3 anyway both in procedural fairings and procedural parts configs, to no effect.

Will try 2.xx version later to see if it's procedural fairings causes troubles or not.

Nope, with PF 2.4.4 and node size 3 it's the same.

Even without procedural fairing parts at all it's still the same.

Edited May 19, 2014 by Zyglrox

[ferram4]

@Zyglrox: Congrats, you have managed to build a rocket that is perfectly designed to break the physics engine in ways that KJR cannot hope to fix. The source of the issue is the same thing that caused the massive amount of wobbling in stock KSP pre-0.23.5; the maxAngularVelocity was clamping the torque that should have been applied through joints to counter angular deflections. However, raising that value doesn't help if the inertia tensor of your object is low as all hell. How heavy are those structural elements? Do they have any mass at all? How heavy are the decouplers? Do they also have absurdly low mass compared to the other parts?

This is what happens when you add more parts than you need to and place lots of lightweight parts together with heavy parts on the either end of the lightweight assembly. No amount of magic on my end will fix that without also making everything laggy as all hell (from all the joints) and without causing lots of problems with unorthodox designs.

To fix it: Get rid of the 2 structural elements at the top of the first stage; stretch the fuel tank to take up the cylindrical section, achieve the flare out with the procedural fairing, not with a conical structural element.

You may ask, why doesn't this happen with stock parts? The reason is that the part masses were chosen to work within the physics engine. Realistic tank and structural masses combined with lots of procedural parts does not tend to play well in physics engines.

[Zyglrox]

Achievement unlocked: Build a rocket that is perfectly designed to break the physics engine in ways that KJR cannot hope to fix.

@ferram4

Thank you, I didn't know mass is such a huge factor. I thought it only affects CoM and aerodynamics. The decoupler is 0.6t, structural elements are 1-2t each, so yes, they have some mass but of course it can't compare with the 284t in the fuel tank below. As I said earlier, removing any of the two structural elements makes the rocket behave as it should (and I also forgot that you can add extra radius to fairings), at least now I know what was broken, so thanks for the explanation.

[Galane]

Try making that structural part with KSP on the side heavier and see what happens. Make a copy of its cfg and change the part names in it so it doesn't interfere with the original.

I did a test last night with a Skipper sandwiched between two Kerbodyne tanks, using two of the Kerbodyne to Rockomax adapters and one Rockomax decoupler. Worked perfectly without any bridging struts. Need to do the same test with Tri and Duo adapters then edit the descriptions to give hints about what configurations to not try.

*I'm making a note, here. Use the Beastlier ReStock parts with extreme caution. Now on to expandifying batteries, RGU, reaction wheels and monoprop tanks to Kerbodyne diameter. Will do math to them to come up with reasonable numbers for increased mass/torque/charge. Should work just scaling up according to volume increase...

Edited May 19, 2014 by Galane

[AndersW]

In Stock, if you add a strut to stabilize something mounted with a radial decoupler it nulls out the force of the decoupler when it is activated. KJR fixes this, but adds a 2-3 second delay while physics is being loaded. You have said that this is part of the core functionality to prevent strengthened rockets from blowing themselves up on physics load. I was wondering if I could get the functionality of not nulling the decoupler force, without all the other stuff from KJR like the slowed physics load.

[raxo2222]

Wow this mod made my launch much more stable

[akman]

Is the mod maintained anymore?Will there be new versions?

[softweir]

Ferram is very active and updating other mods (especially FAR), but this mod is pretty well "finished" and is unlikely to be updated except to maintain compatibility with KSP updates.

[FiresThatBurn]

Could we have a download mirror? Mega? Mediafire won't work

[jonnyt21]

My KJR suddenly seems to have stopped working and my rockets are starting to wobble again, any suggestions please?

[tigerxchaos]

My KJR suddenly seems to have stopped working and my rockets are starting to wobble again, any suggestions please?

Same thing has been happening to me. I tried reinstalling KSP from scratch and attaching the latest version of KJR, with no luck - struts are once again very necessary on all my builds.

[jonnyt21]

Same here, reinstall seems to do nothing. Could other mods be interfering I wonder?

[Agathorn]

I am using KJR without RSS right now, so I don't know if that might be my problem, but.. it doesn't seem to be working. In fact quite the opposite. My rockets are far more fragile than they were in the past last time I played KSP (about 3 months ago?).

Even the simplest rocet - Pod->Heatshield->Tank->Engine... The parts just slide apart when under acceleration. If during launch I bank the rocket even 20-30 degrees, suddenly the whole thing just falls apart, as the fuel tank slides out from under heatshield, and then just rams up against it - but no longer connected and no longer under control.

[AdmiralTigerclaw]

I am using KJR without RSS right now, so I don't know if that might be my problem, but.. it doesn't seem to be working. In fact quite the opposite. My rockets are far more fragile than they were in the past last time I played KSP (about 3 months ago?).

Even the simplest rocet - Pod->Heatshield->Tank->Engine... The parts just slide apart when under acceleration. If during launch I bank the rocket even 20-30 degrees, suddenly the whole thing just falls apart, as the fuel tank slides out from under heatshield, and then just rams up against it - but no longer connected and no longer under control.

Are you running with the newest F.A.R.?

It now includes aerodynamic stress that can tear your rockets apart.

[Agathorn]

Are you running with the newest F.A.R.?

It now includes aerodynamic stress that can tear your rockets apart.

I am yes, but if that is what it is, then it seems very excessive. I can't even get a rocket out of the atmo because they all fall apart!

[AdmiralTigerclaw]

I am yes, but if that is what it is, then it seems very excessive. I can't even get a rocket out of the atmo because they all fall apart!

Atmospheric stresses are phenomenal at sea level. And your comment about a 20 degree 'bank' sounds excessive for a rocket. Bring up FAR and open flight data during a launch. Look at the value known as Q. This is the dynamic pressure on the hull. You don't really want to do anything 'aggressive' while it's over 20 Pa. Heck, my tests for an SR-71 design tended to come apart spectacularly around 11 Pa if I attempted a pitch greater than 10 degrees AoA.

As a test, launch a rocket dead vertical and keep it going until you get above 70 km, then just turn it, see if it falls apart there. We need to narrow down what causes you to break up and determine if it's the mod that's at fault, your engineering, or your piloting. As I've never encountered the problem you're having.

In a second test, open the FAR in-flight data window and watch the 'nominal' condition. If you have an aerodynamic failure, it flashes 'AERODYNAMIC FAILURE' in yellow at you.

Rockets are long tubes filled with canned kaboom after all, they don't like being broken over the proverbial knee with atmospheric pressure at the side.

[tigerxchaos]

Are you running with the newest F.A.R.?

It now includes aerodynamic stress that can tear your rockets apart.

I installed FAR a few days ago. The loss in KJR functionality has been going on for at least a month or so, despite a recent clean install. It's been one of my favorite mods for months, too.

[VaporTrail]

Is that measurement in Pa, kPa, or MPa?

I can't see a measurement of 20 Pa being enough to break a rocket in half... (20 newtons per m^2, with a standard atmospheric pressure being 101,325 Pa).

Just doesn't seem to jive to me... sounds like someone could sneeze on a rocket and bend it in half.

Edited June 28, 2014 by VaporTrail

[Starwaster]

I am yes, but if that is what it is, then it seems very excessive. I can't even get a rocket out of the atmo because they all fall apart!

That seems a bit bizarre to not be able to get anything to orbit. What exactly are you trying to launch? Unless it's something excessively massive or over engineered. And there's also such a thing as too many struts. (yes, I know that sounds sacrilegious but it's true)

Try simplifying your current designs and do away with the struts. Use small amounts of them strategically.

[ferram4]

Everyone complaining about KJR not working will need to post example crafts with the issue, as well as a full list of all mods they are running, and full reproduction steps to cause the issue (which means, saying, "just launch and watch it come apart," better result in it coming apart with no other inputs from the player after launch. If you need to turn the craft to the side, make a sudden gravity turn, fly a loop-de-loop, or whatever to cause the issue, you better tell me; anything left out I won't do at all.) Further, I will need copies of the output_log.txt from the KSP_Data folder. Without all of this I will not be able to reproduce and fix any issues.

@VaporTrail: It is measured in Pa in that window, though you'd be hard pressed to see 20 Pa for anything other than the first fraction of a second of launch. AdmiralTigerclaw is listing values in kPa, but is missing the prefix for his units. And you'd be surprised how little dynamic pressure it takes to tear something apart; you need to remember that static pressure is quite often felt on both the inside and outside of the vehicle, resulting in a net force on the structure of 0, but dynamic pressure results in different forces across the vehicle (measured using lift and drag coefficients), resulting in bad times. And 20 kPa with a rocket sideways at supersonic velocities are more than enough to cause it to break up. For an SR-71, it's even worse, because that has lifting surfaces to toss it around.