Jhula principle - increase speed by applying brakes.

[xnonix]

Yep, Ubuntu. Humanity to Others. 2 Months without window$ here.

[xnonix]

Well, I found the bug.
The thing is the rigid joint position. In the model I post below there are 2 "hammers". the position of the rigid joint is different between them. The one on the left is the buggy one and the other on the right is the correct simulation of this system. So it STOPs again.

Cheers,
xnonix

[prajna]

Well done, xnonix.  What a crap bug.  It is nice that there is a workround but it won't always be possible to use it. I intend to keep thinking along these lines though, I can't help thinking that we might be able to unbalance something in this model.

[hartiberlin]

Well, I found the bug.
The thing is the rigid joint position. In the model I post below there are 2 "hammers". the position of the rigid joint is different between them. The one on the left is the buggy one and the other on the right is the correct simulation of this system. So it STOPs again.

Cheers,
xnonix

Hi xnonix,
can you please exactly describe, what you mean ?
Which rigid joint do you mean ?
Between the steel part and the plastic part ?

If I set a few more rigid joints there,the left unit still accelerates...

Why is the right one working ?
It has the same rigid joint as the left one, as I see...
but it is placed just a bit more into the center...
So what can we exactly do with rigid joints and what not
to avoid this program bug ?

2.So how do we now design a joint, that can only turn
in one direction without this error ?
Please post exact details.
Many thanks.

Regards, Stefan.

[hartiberlin]

I saw just, the model harti_swinger01.w2d  has the same problems with this joint.
So how do we put now 2 weights together without stumbling onto this bug ?