[max_likelihood]

I did a project during my undergraduate degree in physics which involved interfering three planar waves at 60 degrees from one another to create a hexagonal intensity pattern. The interesting thing was that at each of the 6 corners of the hexagon was a singularity (optical vortex) where the phase was undefined. At these points, the phase space was shaped like a spiral staircase (screw dislocation) and particles suspended there could actually be rotated. It was like an “optical wrench” if you will.

On a small scale, planar waves can be modeled like flat sheets of paper traveling through space without any angular momentum (no twisting motion). Yet when these sheets hit an object from multiple angles with the right timing, they can actually cause the object to twist.

[mNovak]

Very interesting. To clarify, 60deg phase offset or azimuth (e.g. converging at a single point)? Now I really want to plug this into an EM simulator

[max_likelihood]

Sorry, I misspoke about the angle of separation (it's actually 120 degrees) and did a poor job of describing the orientation. If you imagine a vector perpendicular to the sheet indicating the direction of travel, then these three vectors would intersect at a single point and lie along the surface of a cone. The angle between the axis of the cone and each vector, the azimuth, was 30 degrees.

https://i.ibb.co/VDgS8yr/3beam.png

[kbwt]

> On a small scale, planar waves can be modeled like flat sheets of paper traveling through space without any angular momentum (no twisting motion).

They certainly have angular momentum, it just depends on the choice of origin. If you pick an origin along the peak ray of the plane wave, there will be no twist around that point. Just like with a particle traveling in free space.

[max_likelihood]

Unfortunately, I don’t know enough to intelligently comment on this. I was largely under the guidance of my professor. However, I can tell you that in my modeling I used the paraxial approximation and that the light was linearly polarized. It was my understanding that only circularly polarized light carried intrinsic angular momentum (https://en.wikipedia.org/wiki/Spin_angular_momentum_of_light...).

Also, I would agree with you that in a uniform electric field, a single E vector in isolation would appear to produce a torque on a point some distance away. But if the rest of the field is considered, the net torque at that point would be zero, right?

[afturner]

This is unbelievably fascinating. Can anyone go into more detail?