Whacky KSP Physics - Tension, Torsion and Shear forces

[Kobymaru]

Hi!

I recently put my ship under rather large forces accidentally, and I found that it was quite stable. More stable than I'd like it to be, actually. So I set out to test some types of stresses. Since I'm neither a physicist, nor a materials-study-guy, I probably misnamed them. In that case, please ignore the names and just look at the pictures.

Here is the first, I call it "tension".

It's simply a large pull along the stack axis. Here are the pictures:

The setup: http://i.imgur.com/xNwq0cT.png

The hinge: http://i.imgur.com/f7DBrHy.png

The result: http://i.imgur.com/LFjP2W5.png

I would expect the connection to simply disconnect. But it seems that there is no such thing as "tearing" due to tension forces. All joints behave somewhat like a spring.

Here is something that I call "shearing".

It's a pull perpendicular to the stack axis.

I would expect the connection to fail.

The setup: http://i.imgur.com/7agjKg8.png

The hinge: http://i.imgur.com/CA2A7jG.png

The result: http://i.imgur.com/BpYzOy2.png

But no! It seems infinitely strong. Instead, we got us a nice little pendulum. Well, at least the connection exerts some torque to get back into the original position.

Here is the only one that actually worked, I call it "torsion".

The setup: http://i.imgur.com/uB4VJnb.png

The hinge: http://i.imgur.com/VKuw4RS.png

The result: http://i.imgur.com/PY4dB0k.png

This works with most parts, but not with all. There was no tearing with the "Modular Girder Segment", this simply behaved like a "wire" that could be wound up indefinitely.

Do you have other types of stresses that some parts seem to withstand infinitely? If so, please post your results here.

So, what do you guys think? Will this ever be fixed? Does Squad care about stuff like this? If they do fix it, will it open a huge can of bugs?

[Xaiier]

I think to an extent, these cases are things that don't really happen that often in designing rockets and such. It's generally assumed that when two parts are connected they become one and are appropriately supported, and so snapping along part borders would ruin that element of the visualization. That being said I can definitely see the argument to include the snapping, but it would require precise balance to prevent it from being too annoying. FAR's aerodynamic disassembly is probably the closest thing we have right now to structural failures like this, and it can definitely be cool, in moderation.

Also, I totally kept mashing "Close" in the last screenshot and wondered why KSP wasn't responding...

[Laie]

Good work thinking up and doing these tests, but they can only be a beginning, not the end.

I would expect the connection to simply disconnect. But it seems that there is no such thing as "tearing" due to tension forces. All joints behave somewhat like a spring.

Oh yes, ships can be torn. "Ripped into two by parachutes" is a common failure mode for landers. I think 2000kn of pull / 200t of weight are a common breaking point; adding another half kerbodyne tank to your first setup might be enough to tear the connection.

Also, parts matter. Most structural parts are quite flexible; the engines on this setup would start swinging wildly at the slightest pretext. Additional thrust attached to the outriggers would make the I-Beams bend visibly. Using FLT-800 tanks for the same purpose, I got a much more rigid connection. It would break a little sooner, but hardly bend until it's breaking point.

[rkman]

Here is the first, I call it "tension".

The result: http://i.imgur.com/LFjP2W5.png

My theory (based on experience but not on extensive testing) is that the elasticity of connections between parts depends in part on the mass of (one of) the parts. If true then a substantial increase of the dry mass of the Oscar tank should reduce the result of the tension test.

[Kobymaru]

Good work thinking up and doing these tests, but they can only be a beginning, not the end.

I Agree

Oh yes, ships can be torn. "Ripped into two by parachutes" is a common failure mode for landers.

TBH, I have seen this only once when having FAR installed. When it wasn't, I always got the stretchy-connections thing even when going far above the 15g mark.

I think 2000kn of pull / 200t of weight are a common breaking point; adding another half kerbodyne tank to your first setup might be enough to tear the connection.

Nope! Turns out, parts are held together either with an invisible, super strong glue or simply MAGIC.

http://i.imgur.com/lA3k5C1.png

http://i.imgur.com/mNg2hlJ.png

My theory (based on experience but not on extensive testing) is that the elasticity of connections between parts depends in part on the mass of (one of) the parts. If true then a substantial increase of the dry mass of the Oscar tank should reduce the result of the tension test.

How could one go to test that?

Also, does anyone have a nice Idea for a proper shearing test? From the torsion testing, I found out that there is a shear-breaking point with the launch stability enhancers attachment, so how could I properly test shearing?

[Starwaster]

Nope! Turns out, parts are held together either with an invisible, super strong glue or simply MAGIC.

A Kraken did it.