Closed orejolaphet closed 8 years ago
Weirdness. If you leave the page for a period of time the friction is no longer displayed.
Same behavior you can see for the original Java simulation. It's a feature of selected mathematical model.
@ariel-phet what do you think about this behavior?
Seems like something we will need to address internally
Right now the "thermal energy" is calculated as "total energy - kinetic and potential energy", which causes any inaccuracies and cumulative errors to be added to the thermal energy.
http://www2.phy.ilstu.edu/~rfm/380s13/chapters/CH2.4_Pendulum.pdf notes a differential equation for the change in kinetic/potential energy, which could be approximated on its own at every step.
@ariel-phet, does this seem like a good approach?
Reassigning for a response to the above comment.
In our discussion I think we agreed on the above approach.
Forgot this on my list of things that were fixed by the model updates.
Noted in Issue #82
If you zoom in the the energy graph 50+ times you will eventually see some friction in the system during oscillation even though friction is set to none. (Note the smaller the angle the more you must zoom in)
To reproduce: Open sim Select seen w/ energy graph Open energy graph Click zoom 50+ times Move mass to large angle and begin oscillation
At small angles
At Large angles