Basic ball bounce isbn ebook

No mass m depending.
Runge-Kutta (fix epsilon step) or Dormand-Prince (varying step). But here : auto step LSodar in order to modify absolute and relative tolerance from 10e-6 to 10e-10 as a test, even if not conclusive !

Curve depending from k factor, explaing flat ending

... because to computation demanding !

k=-.80 :

k=-.95 :

So how to correct this problem ?