More and more telecommunication and fibre optic measuring systems refer to devices that analyse the interference of two optical waves. The information given by the interferences cannot be used unless the combined amplitude is stable in time, which means, that the waves are in the same state of polarization. In those cases it is necessary to use fibres that transmit a stable state of polarization.

As light passes through a point in space, the direction and amplitude of the vibrating electric field traces out a path in time. Polarized light can be classified as linearly, elliptically or circularly polarized, in them the linearly polarized is the simplest.

Polarization maintaining fibre optic systems requires specialised fibre and connectors and careful assembly as well as alignment to achieve optical performances, which are affected by the polarization of the light travelling through the fibre. Many systems, such as fibre interferometers and sensors, fibre laser and electro-optic modulators, also suffer from polarization dependent losses that can affect system performance.

To solve this problem, several manufacturers have developed polarization-maintaining fibres (PM fibres). These fibres work by inducing a birefringence within the fibre core, which refers to a difference in the propagation - constant of light travelling through the fibre for two perpendicular polarizations. Birefringence is created within a PM fibre either by forming a non-circular fibre core (shape induced birefringence), or by inducing constant stresses within the fibre with stress applying parts (SAP) (stress induced birefringence). This birefringence breaks the circular symmetry in an optical fibre, creating two principal transmission axes within the fibre, known respectively as the fast and slow axes of the fibre.

At present the most popular fibre type in the industry is the circular SAP type (stress induced birefringence), or PANDA fibre. One advantage of PANDA fibre over most other fibre types is that the fibre core size and numerical aperture is compatible with regular single mode fibre. This ensures minimum losses in devices using both types of fibres.

Provided the input light into a PM fibre is linearly polarized and orientated along one of these axes, then the output light from the fibre will remain linearly polarized and aligned with the principal axis, even when subjected to external stresses.

While in theory one can produce perfectly linearly polarized light, in practice, this is not the case. Instead there is always some residual polarization, random or elliptical, present in the output light. To measure the quality of the polarized beam, one must measure its polarization extinction ratio (ER). Naturally, how well a fibre maintains polarization depends on the input launch conditions into the fibre. Perhaps the most important factor is the angular alignment between the polarization axis of the light with the slow axis of the fibre. Assume that we have a perfectly polarized input beam into an ideal fibre, misaligned by an angle f with respect to the slow axis of the fibre. Because of this alignment, a small amount of light will be transmitted along the fastest axis of the fibre. This will degrade the ER of the output beam. The maximum possible ratio is thus limited by: ER = 10 log (tan² F)

Thus to achieve output extiction rations greater than 20dB, the angular misalignment must be less than 6 degrees. For 30dB extinction ratios, the angular misalignment must be less than 1.8 degrees.

Many things, including reflection from surfaces, stresses within the transmitting media, magnetic fields, and others, can affect the polarization of a light beam. Thus, analysing the polarization of light travelling through optical fibre has uses in a wide range of applications. At the same time, controlling and manipulating the polarization-state of light is also highly desirable.

Understanding how to control polarization of light in a fibre optic system and how to properly use polarization-maintaining (PM) components is vital for successful results. PM Fibre generally resembles ordinary single-mode fibre in core and cladding diameters. It differs in the degree and consistency of optical birefringence, that is, the difference in index of refraction between the fast and slow polarization modes. All single-mode fibres exhibit some degree of birefringence. It is a weak effect that can be modelled as a stack of many, randomly oriented retarder plates.

The limiting impact on telecommunications, polarization-mode dispersion, comes about because of the great lengths of fibre involved. In contrast, PM fibre is designed to exhibit a consistently high value of linear birefringence. For comparison, in single-mode fibre, the fast and slow polarization modes experience 360 degrees of phases shift over many meters. In PM fibre, this distance, called the beat length, is typically a few millimetres.

There are several strategies for creating this consistent, higher linear birefringence in PM fibre. In each case, the core is placed in a uniform transverse mechanical stress field to induce a difference in index between orthogonal axes. The level of built-in stress is intended to be much higher than stresses produced by the application environment, so that the electric field of light launched along one of the principal (fast or slow) axes will remain aligned with that physical axis as it propagates. This is key to most PM fibre applications.

As mentioned earlier, the polarization-state of light travelling through a medium can be influenced by stress within the medium. This can present problems with ordinary single mode fibre. When a normal fibre is bent or twisted, stresses are induced in the fibre. These stresses in turn will change the polarization-state f light travelling through the fibre. Furthermore, if the fibre is subjected to any external perturbations, that due to changes in the fibre's position or temperature, then the final output polarization will vary with time.

The polarization maintaining fibres have a strong linear birefringence, which creates PM fibre applications.

PM Fibre Applications

PM fibre is typically used to guide linearly polarized light from point to point. PM fibre also finds many specialised applications in optical sensors and in both telecommunications and sensor research. In some applications, light is purposely launched in such a way that the electric field projects equally onto fast and slow axes. In this condition, the output polarization-state is highly sensitive to wavelength, temperature and mechanical stress.

What Limits PM Fibre Performance?

In the most common optical fibre telecommunications applications, PM fibre is used to guide light in a linearly polarised state from one place to another. To achieve this result, several conditions must be met. Input light must be highly polarised to avoid launching both slow and fast axis modes, a condition in which the output polarization state is unpredictable.

The electric field of the input light must be accurately aligned with a principal axis (the slow axis by industry convention) of the fibre for the same reason. If the PM fibre path consists of segments of fibre joined by optical connectors or splices, rotational alignment of the mating fibres is critical. In addition, connectors must have been installed on the PM fibres in such a way that internal stresses do not cause the electric field to be projected onto the unintended axis of the fibre.