tailuge / billiards Goto Github PK

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

How do the following quantities compare to pooltool and Han2005

Constants and derived quantities

Mx,My    = 0.000720
Mz       = 0.000370
I        = 0.000099
Max vel  = 4.912500
Max rvel = 187.5000
g = 9.8
mu = 0.00985
muS = 0.105
rho = 0.025
m = 0.23
R = 0.03275
e = 0.86

in constants.ts

export const mu = 0.8

But, Moving friction is much larger than Rolling friction

It should not be one variable.

In the description of Key equations,

rolling motion <-> sliding motion

Hey @tailuge, I just wanted to say thank you for building this project. I’ve long been wanting to build an algorithm to suggest routes given the current set of balls. I found this project to implement the physics engine and debugging tools to be super convenient to play with my ideas.

Current status:
I’m able to place two balls randomly on the table and shot the other one in the pocket. Demo: https://vjeux.github.io/billiards/dist/diagrams.html

Next is to start building the route planning. I have an idea that I believe will work.

Some of the tweets from my hack session today:

• Capture and normalize table from my garage: https://twitter.com/vjeux/status/1337885313113358336?s=21
• Make a shot: https://twitter.com/vjeux/status/1337991561401884672?s=21
• Find the best hole: https://twitter.com/vjeux/status/1338009999239311361?s=21

So far it’s been dead simple to run and haven’t hit any bugs.
The only thing is that the spins don’t behave quite like the physical world. If I get far enough in my prototype I’ll give some more concrete repro, my hunch is that it’s going to be about tweaking the constants.