What would be outcome of experiment?

[zonde]

What would be outcome of such experiment:
We take EPR photon experiment where we direct one entangled photon to Bob's side and the other to Alice's side. We rotate polarization of photon at Alice's side by 22.5°. At both sides we direct photon to PBS and detect them in two detectors.
Only difference is that at each side before detection at one of two detectors we destroy entanglement of photon (if it's can be done).

 

[zonde]

Would it be easier to predict outcome of EPR like experiment if instead of destroying entanglement we would rotate polarization by 45° before all four detectors (but after PBS)?

 

[DrChinese]

It is not clear what you are asking of the experiment.

You can rotate Alice and Bob (using say wave plates) and that will not affect their entanglement. If they are not entangled, you will not expect any special correlation over and above chance. Using a PBS with 2 detectors on each side, entanglement is terminated when you detect them. You could put additional polarizers in front of the detectors and after the PBS, and play with them that way I guess.

 

[zonde]

The question is that: from QM perspective - is it possible to change outcome of experiment by changing polarization properties after PBS?
It seems to me that it is possible because entangled pair of photons is wavefunction over all paths including paths after PBS. And changing wavefunction may change outcome of experiment.

But from local realism perspective it would seem that changing polarization properties after PBS should not change outcome because if photon is there then changing it's properties does not change the fact that it is there. Unless of course we assume unfair sampling at detectors.