I take some time to model a toy that most of you have seen: the old Swinging Balls Apparatus, and them run a Motion study to see if I can get the actual results... however, when I place one ball on an elevated position and let it swing under gravity and hit the other six ball, after the impact all the balls show a swing motion, which is not the real effect you get with the actual apparatus.

the swinging balls apparatus must satisfy all the energy conservation laws, not just one or two. The only way to satisfy both momentum and energy conservation in this case is for one ball to go in and one ball to go out, or two balls in, two balls out, but I am not able to get the proper results.

What is happening? Do I have something wrong on the model? or is that Motion can not perform the correct calculations and simulate the behavior of the actual apparatus?

Maybe I am not using the proper parameters for contact, particularly to define the impact parameters (damping, penetration, etc.) or is that the software perform an incorrect calculation?

It look to me like the software is just taking care of the conservation of energy not the momentum where the effect to the mass and velocity of the ball being lifted and released under gravity force to collide with the other ball result in:

1m v1 = 7m v1/7

Since after impact all the balls swing at a lower momentum. This may satisfy the energy conservation law but is not correct, since both energy and momentum should be conserved.

I believe that the problem is that I am not using the proper parameters when I define the contact, particularly to define the friction and impact parameters (damping, penetration, etc.), or maybe I am using the wrong "integrator"

Maybe someone out there (a GURU on Motion Simulation), can take a look at the model and make it work... entering the correct parameters, I will like to see those, once the model is working properly.

Attached is a zip file containing the assembly and the Motion Study (Motion Analysis Add-In need to be active). The assembly is SW 2009 sp 4.0

Hello,

To get this model working, I did the following:

Edit the 3D contact to use the Restitution coefficient (I used 1 for this)

In the study Properties, changed Accuracy to 0.001

In the study Properties, used precise contact

In the Advanced properties, reduced the Maximum Integrator Step Size to 0.001

The Poisson model (Restitution Coefficient) does not require specification of damping coefficients and does correctly account for energy dissipation. This model is therefore most attractive in situations where energy dissipation is of great importance. The determination of the Poisson model parameters (restitution coefficient) is more straightforward than that of the impact force model. In many instances, you can measure it using standardized testing methods (ASTM F 1887-89, for example) or find it in various tables. The model is not suitable for persistent impacts (where contact is developed for a prolonged period of time) and you should use the impact force model in these cases. Be aware that the properties in the Impact Force Model can be significantly more difficult to compute.

Also, attached is the model that is used in the 2009 Motion Simulation Training Course (Lesson 8, Exercise 16) which is very similar.

Hope this helps.

-Matt