Have radial connections got weaker in 0.23.5?

[katateochi]

Has anyone had any experiences that suggest radial connections might now be weaker? I've been loving the stronger stacked connections for rocket builds but I've just come back to a rover that I was working on in 0.23 and I'm now wondering if radial connections are now weaker.

I'd been working on a rover in 0.23.0 that was designed to be very rugged, in tests it was proving to be pretty indestructible in crashes/rolls at around 40-50ms. It has I-beams that protect it as it tumbles along the ground and even when flipping over many times it usually only lost a light or two. I could basically roll it down most hills and after repairing the wheels it could carry on again.

However now in 0.23.5 its roll bars are extremely flimsy and just shear off, even at low speeds. The wheels are also much more prone to falling off completely too. I've heard some people saying that craft from previous versions had issues so I rebuilt it completely from scratch but the roll bars still fall off in low speed crashes. The main body and roll bars are stock parts, but it also had a number of mod parts so I did another rebuild in my completely stock install. But the roll bars still broke off just as easily.

Have radial connection gotten weaker? Has anyone else had similar structural problems in 0.23.5 with designs that had worked before?

Here's a few pics of the rover;

Javascript is disabled. View full album

[Geschosskopf]

I have found that radial connections are MUCH stronger in 23.5 than before.

Such as...

New Joints 1 by TheBullethead

It's also possible to make a 15-wide thing like this with orange tanks and Mainsails. Each radial tank is attached to the next inboard by 1 radial decoupler and 2 struts. Period. Yet the innermost radial joints can still hold up all the mass outboard of them, and the whole thing can hang on 2 launch clamps in the middle without anything even sagging, let alone breaking. This wasn't at all possible before 23.5

[Dirty Bob]

It could be the effect of impacts has changed. I've had very little time with the new update, but I have noticed rolls seem to cause more damage than before. At the same time, straight impacts seem to cause less damage, such as tipping over a lander.

By the way, I love your design. That's a very good looking vehicle.

[smysha]

Weaker, stronger... I have no idea. I still strut-connecting s...t out of stages:)

[Kerbonautical]

Mine's behaving the same way actually. Sometimes I build a light shuttle out of 5 small fuel tanks and 4 engines which would not typically require any struts, but suddenly they do. The struts themselves however have a much larger effect than they used to.

[Spartwo]

My rover can withstand 10Gs of pressure now,so I'd say they got stronger.

[helaeon]

Yes it seems like radially connected things that were just kinda glued on are weaker and need strutting, but not very much. It's like these parts are connected at a single point or on a little line so there's not much surface area of connection.

Things that are designed to be connected radially are quite a bit stronger. These are parts that seem to have a pretty good connection surface, like radial decouplers, standard docking points, etc.

Things connected via node are a ton stronger.

[katateochi]

Thanks for the replies guys, sorry I didn't reply sooner I've been away for a few days.

It sounds like quite mixed feeling re radial strength. Perhaps there are more noticeable differences between parts now, some like decouplers seem really strong, possibly other connections are not as strong. I'm going to try and make the roll bars out of radial decouplers or perhaps use radial decouplers as the connection points for the I-beams. In my original design the I-beams are connected to the rover via the small cube struts and in the crash reports it is the connection between the strut and I-beam that breaks. The cube struts are also radially attached but they've only broken off a couple times. I'll report my findings as soon as I can (if someone could launch a rocket at my boss that would really help!)

[NecroBones]

I think it's a mix of effects. The breaking strength of joints is considerably stronger now, but the joints are still bendy. You can see a 90-degree bend without a joint letting go now. What I'm not sure of, is whether the bendy/twisty aspect has changed.

[katateochi]

After trying a number of things I've found the problem with my design was due to the small cube struts that where used as the mounting point for the I-beam roll bars. By replacing them with radial decouplers the roll bars returned to being strong enough to withstand crashes. I didn't like the look of the decouplers thou so I tried a couple other parts. It seems that basically any other radial part worked fine and in the end I've used M-beams as the mounting points. Even attaching two M-beams together is stronger than when using the small cube struts.

Here's a couple close ups of the new connection for the roll bars;

So what's interesting is the cube strut is one of the parts that doesn't have a physics significance. In the past I've had issues with the octagonal strut and more recently the new decoupler from the NSAS parts which also don't have physics significance. So it could be something to do with that and the way those parts are simulated.

[boxman]

I heard someone claim that their older ships does not seem to have stronger joints in 23.5 but yet when they rebuilt it the exact same way it suddenly benefited from the stronger connections. I dont know how true this is, but it is worth a try.

[katateochi]

I heard someone claim that their older ships does not seem to have stronger joints in 23.5 but yet when they rebuilt it the exact same way it suddenly benefited from the stronger connections. I dont know how true this is, but it is worth a try.

Yeah I'd heard that too, so I re made it from scratch, I also re made the main section in my pure stock install, but the weak links between the small cube struts and the I-beams remained.

One thing I forgot to say in my last post. Once I'd got the roll bars working again I started added more gear to the rover. At one point I attached a comms DTS-M1 dish to one of the I-beam roll bars. The result was that roll bar would get knocked off quite easily while the others stayed on. Interesting?