The measurement of birefringence in optical glass by electronic means
Askowith, Burton Jacob
MetadataShow full item record
This paper first presents a summary of several old and new methods of measuring birefringence, indicating their limitations in respect to large optical glass blanks, and then explains in considerable detail the writers' new radiation method, which is free of many of these limitations. The procedure and theory of this new method are given here briefly. Procedure: A collimated beam of monochromatic (yellow - green) light is passed through a circular polarizer and then through the specimen to be measured, that is, a large optical glass blank, which is held stationary. The beam next passes through a plane-polarized analyzer, which may be rotated to permit maximum and minimum transmissions of the elliptically polarized light emerging from the specimen. The intensity of this transmitted light is indicated on an ammeter by means of a sensitive photocell and amplifier. Using a variac, the maximum intensity is adjusted to indicate a predetermined maximum position on the ammeter scale. The corresponding minimum position will then indicate the birefringence in terms of degrees of phase difference with the help of a special scale to be used in conjunction with the ammeter scale. The circular light emerging from the circular polarizer enables the optical blank under test to be placed in any arbitrary azimuth in respect to its principal directions thus eliminating the necessity for lining up the heavy optical blank. The necessity for lining up is one of the chief objections to previous methods as far as their use with large optical blanks is concerned. Theory: When the circular light from the circular polarizer passes through the specimen, the additional birefringence of the latter causes the beam to become elliptically polarized. Mathematical analysis of this elliptic light shows that the maximum and minimum intensities emerging from the rotatable plane analyzer determine by their ratio the birefringence of the specimen by the simple formula. SinΔφ=(A-B)/(A+B)=((A/B)-1)/((A/B)+1) where A is the maximum intensity, and B is the minimum, and Δφ is the birefringence of the specimen in terms of phase difference. This device is to be used to measure birefringence in optical blanks between 0° and 300° of phase-difference. Therefore, to distinguish between the four possible phase-differences associated with each absolute sine value on the scale, advantage is taken of the fact that the major axis of our birefringence ellipse is orientated in the first 180° (0° - 180°) of specimen phase difference in a direction perpendicular to its orientation in the second 180° (180° - 360°) of specimen phase-difference. Thus, if we note that our maximum intensity position occurs with the analyzers indicator clockwise to the specimen's "fast" axis in the one group (0° - 180°) and anti-clockwise in the other group (180° - 360°), we can distinguish between these two groups. Also, by inserting a rotatable 60° - retardation plate in the beam's path, we can determine the sum of the retardations, and on subsequent subtraction of the known 60° - retardation, this will leave only one sensible answer, thus distinguishing between the (0° - 90°) group and the (90° - 180°) group, or their equivalent groups between 180° and 360°. By dividing the birefringence scale properly into upper and lower double groups of 90° span each, this procedure can be made reasonably simple. The "fast" axis of the optical blank at any point can be determined closely enough for the above purpose by the geometry of the round blank, (since principal directions are almost always radial and tangential), or else by previous qualitative examination. Due to light-leakage through Polaroid plates at "extinction" positions, zeroing of the device may be accomplished by means of a shutter with only fair accuracy (relative inaccuracy around 90° of phase-difference). However, by zeroing the ammeter through crossed plane-Polaroids, we allow for most or all of this leakage at all phase-differences. The principal advantage of the new method is the one previously suggested, namely, that we do not have to line up the heavy, bulk-y optical glass blank at each point to make the principal directions of the specimen coincide with those of the polarizer or analyzer. A second advantage over previous radiation methods is its full wave-length range. Another advantage is its simplicity of operation, at least for birefringence known to lie between 0° and 90° of phase-difference.
Thesis (M.A.)--Boston University