Joint Issues Since 3.8.0

[Karhgath]

Everything worked fine in 3.7.5, but since 3.8.0 it seems joints are more elastic or soft. However, things initially seemed worse in 3.8.1 and are still present in 3.8.2, but could have started in 3.8.0.

3.8.0
Craft A (Alpha TV) is mostly fine, you can see it bounce on the pad if you warp and stop, but it does not collapse and even during launch it looks mostly ok.

Craft B (Alpha MkII Luna) however bounces more after warp on the pad, and during flight it seems the mass and pressure compresses parts together. As soon as it stops accelerating, it visually bounces a few time until back into place.

3.8.1+

Craft A (Alpha TV) was the worst offender on the pad initially ... but is inconsistent. This morning it mostly behaves, so I cannot always repro. Sometimes joints will cause gaps between parts after load, and worse is when you warp and stop on the pad, the whole craft could compress and collapse in itself. As soon as you launch, it seems to "fix" itself, but exerts forces in doing so which can mess up MJ and ascent initially. Then during flight it will cause RUD under phys warp 3x, joint seems weaker as I could do this before to speed up the launch (at least until close to max q)

Craft B (Alpha MkII Luna) does not seems to be too bad on the pad for some reason, and weirdly does not seems to show the same compression behavior as above in 3.8.0... but if you do any physics warp under stress in flight, the joints seems weaker and could break immediately - not always. I saw it compress together similar to above once that way, other times is was just RUD. Normally you should be able to warp to 2x or 3x with this craft.

Then on phys warp 4x on the pad this happened... ragdoll effect - it stayed together like that and swayed/bounced, then when warp stopped it "bounced back" together as it wobbled.

However, I thought it was more predictable behavior, but seems it is not. I tested other crafts, this one under Phys Warp on the pad started sinking into the ground and gaps started to appear.

So from the above, since it is now clear that it is not consistent, it is hard to say if 3.8.1 is worse than 3.8.0 but it feels that way, could be just a coincidence tho. I tested normal warp on pad in 3.8.0 but not phys warp - I will try to test later if I can.

KJR Issue Crafts.zip

[siimav]

Btw, are those tests done in a career game or in sandbox. Stringer mass numbers can affect which extra joints are formed.

[siimav]

Now this is mighty interesting:

[KJR] Avionics [Procedural] connected to ground
[KJR] Pioneer Adjustment Motor connected to ground
[KJR] Communotron 16 connected to ground

It appears that in some cases part physicsMass isn't initialized yet by the time the joints are being created. I started using this field in 2a81c73.

[siimav]

I bet it's the damned B9PS firing the vessel modified event too soon when the vessel has parts like the Agena.

[siimav]

Please try v3.8.3

[Karhgath]

I can confirm that the ragdoll physics effect seems to be fixed, thanks!!

The compression issue still happens tho, but I saw it happened in 3.7.5 as well on these specific launchers, not a new thing most likely - seems like the .010m ring RCS tank in between the avionics and main tank is the culprit, effect goes away if I remove it and other than visual it does not seems to cause any side effects, it reacts normally on decoupling.

The main problems like ragdoll, bouncing, sinking on the pad, etc. all seems resolved for me.