Idris El Bakri and Dr. Richard H. Selfridge, Electrical and Computer Engineering,
The project was an experimental and theoretical research effort to determine the effect of temperature on the birefringence of optical D fibers. Such fibers have a D shaped cross section, hence their name, and consist of a three major regions: the light guiding core, the cladding and the super cladding. Different components of light waves traveling in this kind of fiber travel at different velocities, and such a fiber is said to have high birefringence (HB). HB fibers are very useful as temperature sensors.
The experimental aspect of this project involved applying extreme temperatures to a short section of a light guiding D fiber and observing how the propagation of light was altered. The theoretical part consisted of the development of a new mathematically simple model to describe the observations. The model was based upon the index of refraction of the core material in the fiber and its relation to temperature changes. The model greatly simplifies the mathematical analysis and will facilitate the use of HB fibers as temperature sensors
To analyze the propagation of light waves in the optical fiber the polarimetric method was used. In such a method, laser light with a definite state of polarization is focused into the tip of the fiber using a precision coupler. Propagating light is analyzed at the output of the fiber using a linear polarizer and a silicon detector that sends its readings to a computer program. The program stores and plots the result. When heat is applied to the fiber using a micro burner, the polarization state of propagating light changes and affects the observed output intensity. The output intensity goes through a series of maxima and minima that are directly related to the changes in the index of refraction of the fiber core.
The mathematical approach used to explain the intensity changes observed relates the change in the core index of refraction with temperature (which can be calculated) with birefringence change. This in turn gives the pattern of intensity changes that are observed.
Good agreement between experimental results and theoretical predictions was obtained. The measured values of the birefringence change were within 20% of the predicted results. This error percentage is acceptable given the inaccuracies inherent in the experiment arrangement. Also, the dimensions of the fiber provided by the fiber manufacturer, which are key to performing the calculations, are only approximate values and extremely sophisticated and expensive methods must be used to determine them exactly.