Han 2005 paper coefficient of restitution between ball and cushion

[tailuge]

Han 2005 suggests e can be modelled as

e = 0.39 + 0.257*Vx-0.044*Vx^2  (eqn. 26)

https://www.wolframalpha.com/input?i=plot%280.39%2B0.257v-0.044v%5E2%2C+v+%3D+0+to+5%29+

In practice I found this to give unrealistic results and use e=0.86 as a constant.
How does this compare to https://github.com/ekiefl/pooltool implementation?

In this project Vx has maximum of 4.5