It is an open problem in dynamical systems if every triangle has a periodic billiard orbit. So far it has been proven that equilateral triangles, isosceles triangles, right triangles, and obtuse triangles with no angle greater than 100 degrees have periodic billiard orbits.

I am wondering if there are any other broad or specific cases I am missing? For example, in this paper, it is shown that any triangle with angles $\alpha$, $\beta$ satisfying $k\alpha=l\beta$ have a perpendicular periodic trajectory, where $k, l \in \mathbb{N}$ (page 13). There is a similar finding in this paper stating that every triangle with angles $\alpha$, $\beta$ satisfying $k\alpha + n\beta = \pi$, where $k, n \in \mathbb{N}$ admits a perpendicular periodic trajectory.

Does anyone know of any other general cases? I am particularly interested in the existence of cases where triangles have periodic trajectories when its angles are given by some function, that is if $\alpha$ is an angle in triangle $T$, then the second angle used to define $T$ is $\beta = f(\alpha)$ for example. My department doesn't have people specializing in this area, so I apologize in advance if my question does not belong here or is unclear, give any comments and I will try my best to reciprocate.

obtusetriangles with no angle greater than 100 degrees"? $\endgroup$2more comments