Optical Birefringence Fiber Temperature Sensors in the Visible Spectrum of Light

This article describes experimental tests to determine PM fibers Panda style responses to a thermal source with different initial temperature. The aim of this study was to determine the sensitivity of a polarization maintaining fiber to the radiating heat, and to upgrade the space configuration and time response when using the 635 nm light. The sensitivity of the polarization maintaining fiber during excitation of both polarization modes is the principle of this sensor function. This excitation is caused by temperature change and by absorption of thermal radiation. This mechanism is used as an indicator for detection of temperature field disturbance. This article also provides links to previously published results and compares them to the results in this article.


Introduction
Polarization Maintaining Fibers (PMF) work on the basis of creating mechanical strain that causes artificial birefringence.Typical fiber applications include cases, where state polarization maintenance along the light propagation through the fiber is required, and where the excitation of one polarization axis occurs.One effective way of creating birefringence is arranging structures with different thermal expansion into a fiber cladding.At an excitation of both polarization axes, when the light propagates through the fiber, the phase shift occurs between these axes.The measurement site and set of measurement results were published in [1] and [2].Comparison works describing highly birefringence fibers, temperatures and strains, measurements, dependences and other new generation of optical fiber sensors are described in [3] and [4].
The phase shift is dependent on temperature and therefore also on incident thermal radiation.This phase shift can be evaluated by polarizer.Previous experiments [1] and [2] showed that the most suitable fiber, from the sensitivity point of view, is PANDA type fiber.The achieved results suggest to apply such sensor in cases, where the temperature field was disturbed by human body proximity, i.e. in the range of specific temperatures.This paper deals with the same site and arrangement as the previous experiments, but components for wavelength of 635 nm were used.Suitable application could be, for example, for property protection against illegal manipulation, and thus, some requirements for the sensor sensitivity, system configuration and time response were established.The aim of this article is to analyze the influence of polarization angle setup on the input and output polarizer.And finally, to examine this problem in spread range temperature too.

Model of Fiber Thermal Segment Exposition
A thermal source was simulated by a plastic basin with various water temperatures.This arrangement enabled changes to the test water temperature, to the distance from basin bottom to the fiber and also to the number of exposed fiber segments.An ideal model of a fiber sensor, for detection of temperature field disturbance caused by various water temperature in the basin, was set to simulate realistic conditions and also to define analysis and subsequent conditions.Thermal exposition was performed by defined temperature of water with the basin bottom placed 6 cm from the exposed fiber segment.For shorter distances (up to 1 cm), the response of PMF had linear characteristic, and for longer distances, it had non-linear characteristic [2].Water temperature of 0  For maximum response of the sensor element, it was necessary to excite both polarization axes uniformly.This could be achieved by circular polarization excitation or by the axes excitation caused by linear polarization angle of π/4.The advantage of the first solution is that the polarization axes of the fiber might have an orientation relative to the rotation; the disadvantage is the need to include a phase retarder of π/4 behind the LD output with an expected linear polarization character.Preferably, the second solution introduces the optical power of the laser diode directly into the fiber.However, the disadvantage is the need of the exact orientation of the optical fiber axes relative to the laser polarization.Given that the sensor should form a compact unit, the second variant seems advantageous, where the mutual rotation of the input lin-ear polarization towards the polarization axes can be achieved by appropriate orientation of the connector and optical fiber.The system can be described as Jones vectors and matrixes (Eq.( 1)) [7], [8] and [9].Where β is the angle between input linear polarization and axis x, θ is the angle between axis of linear polarizer and axis x, δ is the phase shift between the polarization axes.
The coherent matrix behind the optical fiber (before the polarizer) is: The values of Stokes parameters are: When the fiber rotation is exactly set at π/4 and the angle of input intensity β = 0, Stokes parameters are cos δ, 0, sin δ.Responses to the arrangement of optical fiber polarimeter are shown in [9].
The resulting intensity of the optical wave output was: By substituting numerical values, it can be shown that in the range of reasonable error settings, approximately ±5 • , the deviation of the mean value was approximately 3 %.In terms of the evaluation phase, it was not substantial [7], [8] and [9].

Mechanism of Heat Transfer
Transferred heat Q T from water of temperature T e to a fiber of temperature T was calculated as follows [6]: where λ T is a coefficient of thermal conductivity, ∆l is the heat source (basin bottom) distance from the fiber segment, S is the surface area, ∆T e is temperature difference between the source and ambient temperature, ∆T is the temperature difference between the water and ambient temperature and t is time.As the phase shift was calculated with respect to the change of temperature, all the temperatures were relative to the initial temperature of the fiber, i.e. to the difference between the given and the ambient temperature.
As the transfer of heat depended on the difference between the temperatures, we substituted the difference of absolute temperatures T by the difference of temperatures ∆ϑ in • C, in which the temperature of the applied water was measured.The following formula was obtained:

Experimental Results
Firstly, we prepared the measurement equipment as a preliminary workplace.We took advantage of the benefits of our new polarimeter equipment.We could comfortably observe online the following: the development of SOP changes at outputs from the light source, linear in-fiber polarizer and PM fiber sensor.In this phase, we dismounted the linear polarizer on output and the photodiode, too.Then it was possible to observe the immediate polarization state on polarimeter.
Our measured values are the demonstration of functionality of our temperature sensor (Fig. 2).The measurement workplace arrangement: Temperature controller TED200C, power current supply LDC202C -mounting TCLDM9 -optical source Thorlabs LPS-PM635-FC pigtail LD -connector FC/FC (ADAFC3) -fiber sensor = PM 635 fiber (length = 2 m, Lightcom/Safibra) -connector FC/FC (ADAFC3) -PM patch cord (PMJP-FC-FC-635-900-5-1) -Polarimeter PAX5720IR1-T.A demonstration of the time response for this set of components is depicted in Fig. 3 and Fig. 4. Firstly, (Fig. 3) shows the on-line exhibition from the polarimeter.We can see clearly one part of the ray on the Poincare sphere (approximately 10 s).Some difference in the output polarization is evident from this graph.From the viewpoint of clearness, the most important is Fig. 3, as it shows long periods, approximately 15 minutes.It represents the measurement for 0 • C. Thanks to the reference (25 • C) measurement, this can be considered authoritative.A demonstration of the time response for this set of components is depicted in Fig. 3 and Fig. 4. Firstly, (Fig. 3) shows the on-line exhibition from the polarimeter.We can see clearly one part of the ray on the Poincaré sphere (approximately 10 s).Some difference in the output polarization is evident from this graph.From the viewpoint of clearness, the most important is Fig. 3, as it shows long periods, approximately 15 minutes.It represents the measurement for 0 • C. Thanks to the reference (25 • C) measurement, this can be considered authoritative.By comparing the results, the conclusion was different responses of PMF sensor to minus ∆ and plus ∆ temperature from human body temperature.This was caused by different mechanisms of the heat transfer.SOP development in polarization maintaining fiber sensor.The track of the light polarization state on the Poincare sphere was very similar in each deflection caused by temperature change.When the heat source was removed, the track in reverse direction was clearly observed.

Experimental Evaluation of the Time Response
By comparing with the previously published results [1] it seems that the response in the visible spectrum of light is more massive.It can be due to beat length, which is double for 635 nm.
As shown in Fig. 4, the temperatures influence is minimal when the measured medium temperature is approaching to ambient temperature.The delta of temperatures have to be the same response but it is not due to different heat transfer.
A typical progress on regular circle was unfortunately disturbed by some inaccuracies, inhomogeneity and reflections, especially on the connectors.The amplitude fluctuation was apparently caused by sampling.There were also some other affecting factors: imperfect excitement of both axes, low degree of coherence, but also a noncircular LD beam profile.From practical realization point of view, it is also necessary to take into considerations the excitation by LD, with regard to the compatibility of individual sensor components.

Conclusion
The described temperature sensor at 635 nm can be used in many different ways.The above described arrangement, all component exactly as in-line fiber construction, can be used with advantages in many industrial sectors.This construction largely eliminates the sensitivity to other factors, e.g.mechanical vibration, radiation or pressure changes.For significant exposed length, there was an appropriate response.The option of adjusting sensitivity by changing the exposed length is also advantageous.A fast sensor response was achieved by simultaneous exposition of short segments.When compared to the previous measurement [1], we can clearly state that both wavelengths are suitable for this type of sensor.The degree of sensitivity will be the subject of further research.
Future work will be aimed at the studying of 1310 nm wavelength and comparing the results with previous studies.The trend of development will be directed to replace the connector by splicing.Gradual replacement of polarimeter may allow to use linear in-line fiber polarizer and photodiode.

Fig. 1 :
Fig. 1: Floor projection of measuring workplace with exposed length depiction and its sectional view.

Fig. 2 :
Fig. 2: The arrangement of fiber sensor model.Disposed optical fiber sensor was on the surface of the polystyrene plate.Cork stoppers were used as spacers.Measuring station was inclosed by polystyrene plates.The entire workstation was wrapped in polyethylene foam to avoid external thermal influences.Sensory fibers were accessible by upper slot for heat only.
Exciting two lengths (25 cm per length) of fiber 6 cm from heat source (basin of water), temperature at the time of attaching was 35 • C (human body), 0 • C and 53 • C. Heat source was located at time 2:20 min and removed at 7:50 min.

Fig. 4 :
Fig. 4: The evolution of the output Stokes element S 3 .