MotorTorque / BendConstant relationship?

[Xray]

When I set up an axel (hinge) so that its rotation is controlled with the BEND function, the torque seems to be oddly dependent on the values of motorTorque AND bendConstant. I found that while using Bend, not only does the Bend Constant have an influence on the torque, but the Motor Torque also has a major influence on the torque (even though the Motor function has been disabled). But what's so surprising and confusing is the relationship between the two values. It doesn't make sense! Take a look at my attached sample scene. I used the second hand of my Thyme Clock scene as a test sample. With gravity turned on, I increased the weight of the second hand so that there is a significant load on the axel during movement.

I show five different hands with five different values for Motor Torque and Bend Constant. The strangest values are for hand numbers 3 and 4 (reading left to right). Hand #3 has a Motor Torque value of 1000, and a Bend Constant of 0. With those values, the hand ticks quite normally. But in hand #4, the Motor Torque value is also 1000, but the Bend Constant is 1. With those values, the axle doesn't have enough torque to overcome the force of gravity!!! Can someone please explain what's happening here? Why does motorTorque and bendConstant have such an unusual relationship? And why does motorTorque have any effect at all when using the Bend function? Shouldn't it be disabled?

[Ivan]

I've researched this and have come to the following conclusions:
If bendConstant is 0, then the axle's torque will always be motorTorque, otherwise, the torque is calculated by multiplying bendConstant and angle delta limited to a maximum determined by motorTorque.

[Xray]

I've researched this and have come to the following conclusions:
If bendConstant is 0, then the axle's torque will always be motorTorque, otherwise, the torque is calculated by multiplying bendConstant and angle delta limited to a maximum determined by motorTorque.

I understand the first part of your conclusion (if bendConstant == 0, then axle torque = motorTorque), but I do not understand the second part. What is "angle delta"? And does what you found make sense? I mean, do you believe that the system was designed this way for some good technical reason, or is it simply a bug caused by a design oversight?

[Ivan]

I don't think this is a bug. It's more like a feature that enables you to chose whether you want elastic bending or non-elastic.

"Angle delta" means the difference between the actual angle and the angle that the axle is bend to. If bendConstant is not 0, the closest thing i could compare that behaviour to is a rubber band propeller. The more it is turned away from default angle, the more torque it will have in tendance to get back to the default angle.


Here is the graph that describes that bahaviour.

[Xray]

That's good, useful information! Thanks very much!

[Ivan]

You're welcome!

[Kitesurfer]

Some googling gives that the bend stuff is actually to set up the axle as a servo. Also; good use for Algodoo in some biomechanica research! https://mori0091.github.io/Biped/biped6.0.html