1
$\begingroup$

I have a Prism glued to a prism plate. Let's say the method of using optical glue is good enough, and minimal human error is tolerated. What experiment would be a good to check if the prism is not tilted to one side (flatness deviation) beyond certain threshold.

The prism is a retro-reflecting dispersing prism. So my idea is to shoot a collimated laser beam and across the reflective surface and see the deviation of reflected beam.

Any other ideas ?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
2
$\begingroup$

Use a sodium lamp positioned above the prism and slightly to the side to create a reflection at the interface. Look for interference patterns at the prism/prism plate interface. If the spacing between dark fringes is small (and you have many fringes), you have a problem. If you have very few fringes and the spacing is large, the interface is close to being flat.

Improve this answer
$\endgroup$
5
  • $\begingroup$ This would be a good idea for quick estimation. If I want a quantitative result (for eg conduct the same experiment over 100 such cases) then its better to have a experimental data to compare and form a toleration point. $\endgroup$
    – Rene Duchamp
    Commented Aug 27, 2015 at 18:27
  • $\begingroup$ Wouldn't the spacing between the fringes be quantitative? $\endgroup$
    – Paul
    Commented Aug 27, 2015 at 20:19
  • $\begingroup$ Yes, adjacent fringes indicate a spacing change of $\lambda$/4, unless the phase shift at the bottom reflection is different from the top reflection. $\endgroup$
    – Bill N
    Commented Aug 28, 2015 at 17:13
  • $\begingroup$ Can this minimization of phase shift between bottom and top be carried out analytically ? Any references you could provide for the same ? $\endgroup$
    – Rene Duchamp
    Commented Jun 23, 2016 at 23:44
  • $\begingroup$ Any good intermediate optics text should discuss this. Jenkins & White is the classic standard. $\endgroup$
    – Bill N
    Commented Jun 25, 2016 at 18:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.