1. Introduction

The fiber fuse effect, named due to its similarity with a burning fuse, is a phenomenon that occurs in optical fibers in the presence of high power optical signals. Although this phenomenon is known since 1988 [1], only in recent years it became an important issue. The evolution of optical communications led to a significant increase of the optical power level present in the optical fiber infrastructure, hence bringing attention to the fuse effect as a serious potential problem for telecommunication networks [2].

The fiber fuse is ignited on localized high attenuation points, usually, on damaged or dirty connectors or in tight bent fibers [2–4]. Once initiated, the optical fuse zone propagates towards the signal source, accompanied by a characteristic emission of white light, known as optical fuse discharge. The fuse effect propagation continues until the optical intensity becomes lower than the fiber-dependent threshold value [2]. For silica based fibers, the reported optical intensity thresholds are in the interval ~1-5 MWcm⁻², being also dependent on the fiber type and signal wavelength [4–6]. The fuse effect phenomenon leads to the destruction of the optical fiber since after the fuse propagation the fiber core exhibits a string of voids, being unable to guide the optical signal [1, 2]. Besides this damage in the fiber, the optical fuse discharge can reach, and destroy, the optical emitters or other network active equipment.

Several fuse discharge detection methods have been reported. The measurement of the electrical spectrum of the back-reflected signal was proposed in [7]. The same authors also proposed the monitoring of the fuse discharge using an optical time-domain reflectometer [8]. A method based on the use of fiber Bragg gratings, as temperature sensors, was also demonstrated [9].

In addition to detection methods, it will be also desire to develop one safety devices that are able to halt the fuse propagation before it reaches the network active equipments. The use of a tapered fiber to stop the fiber fuse propagation was first reported in 1989 for a 450 nm signal and a single value of optical power [10]. However, that signal wavelength was very short, when compared to the ones used nowadays in commercial optical fibers, and the process that led to the fuse discharge extinction has not been properly explained. According to [10], the optical fuse discharge extinction on the tapered fiber is explained by the reduction of the fraction of the guided power that actually propagates on germanium (Ge) doped regions. This explanation relies on the assumption that the fuse effect is only driven by Ge related effects, however, it is well known that the fuse effect also occurs on fibers without germanium [5]. More recently, a fiber fuse terminator based on thermally-diffused expanded core (TEC) fiber was proposed [11]. This device reduces the power intensity in the core region and was used to stop the fuse discharge propagation generated with an optical power of 2 W, but no further results have been reported. The halting of the fiber fuse propagation in a fiber section with reduced cladding thickness (etched fiber tapers) was also reported [12].

In this paper, we report and thoroughly analyze the extinction of the optical fuse discharge propagation in optical fiber microwires (OFMs). The microwires act as a mode field expander, decreasing the local optical intensity to a value below the fuse effect intensity threshold. The proposed device is suitable for being introduced into the network infrastructure ahead of the optical active equipments acting as a passive protecting system since OFMs allow a low loss connection and are easily spliced to standard single mode fibers [13].

This paper is organized as follows: section 2 gives a brief introduction to the OFMs main features, in section 3 we present the setup that was implemented for this study, in section 4 the results are discussed and finally the conclusions are presented in section 5.

2. Optical fiber microwires

Optical fiber microwires are used in several applications and components [13, 14]. The OFM characteristics, namely the mode field diameter (MFD), can be tuned by choosing the appropriate diameter of the microwire and the geometry of the two conically tapered transition regions [13, 15].

The mode field diameters at a wavelength of 1480 nm for a step index fiber were estimated numerically considering a three cylindrical layers model, where the fiber core and cladding regions are surrounded by a large layer of air. The transverse distribution of the fundamental mode electric field of the OFM was obtained using a finite element method through Comsol^® Multiphysics software package. From these results, we obtained the fiber effective area and then the mode field effective diameter. Figure 1 displays the variation of the MFD of the fundamental mode as the cladding diameter ranges from the single mode fiber (SMF) initial value of 125 µm to a diameter of 12 µm. The MFD local minimum that corresponds to the maximum con□nement in the core region occurs for a taper diameter around 100 µm and the local maximum occurs for a taper diameter around 40 µm, just before the guiding is transferred to the interface silica/air. In this range, the MFD increases almost exponentially with the decrease of the fiber diameter, leading to lower values of optical intensity as the injected power is maintained.

 

Fig. 1 Fiber microwire MFD as function of the cladding diameter, obtained numerically for a wavelength of 1480 nm and parameters that are consistent with the technical data of the G.652D fiber.

Download Full Size | PDF

3. Experimental setup

A setup for the controllable ignition of the fuse effect was implemented as in [9], based on a short length (~3 m) test fiber connected to a Raman fiber laser (IPG, model RLR-10-1480), emitting at 1480 nm (Fig. 2 ). To promote the fuse effect ignition, the fiber end was put in contact with a metallic foil, causing a localized heating. As safety precaution and to protect the optical source, an optical isolator and a 20 m dummy fiber were also used. The fiber under test was the G.652.D, from Corning, and the OFM was spliced between the signal source and the test fiber where the fuse effect was ignited.

 

Fig. 2 Experimental setup. The fuse effect is triggered in the fiber end and is halted by the OFM.

Download Full Size | PDF

The OFMs were produced from the G.652D fiber by the flame brushing technique, where the fiber is locally heated above its softening point (~1500°C) by an acetylene flame with 2 mm width. The fiber is stretched mechanically by two linear stages in several oscillatory “brushing” movements, controlled by precision stepper motors (Newport UTS-150-pp), with a predefined velocity sequence. This tapering procedure allows the control of the OFM waist and transition region profile [15]. The transition region profiles were chosen to minimize optical loss and to ensure that only the fundamental mode is propagating (adiabatic situation). Two symmetric OFMs were produced and tested, they have identical waist regions but two different shapes of the piecewise-linear transition regions, as shown in Fig. 3 . Both OFMs have a waist region with a length of 35 mm, a diameter of 17 μm and insertion losses of ~0.2 dB. Hereafter, these samples will be denominated as OFM A and OFM B.

 

Fig. 3 OFM measured profiles. Radius values were estimated using microscopic images obtained with an optical microscope (Olympus BH-2).

Download Full Size | PDF

The fiber fuse discharge zone was initiated with different values of injected optical power. For each value, the experimental procedure was repeated four times using the two different OFM profiles.

4. Results and discussion

We have observed the extinction of optical fuse discharge in the fiber tapered zones, for the two different OFM profiles and for all the tested optical power values. For some of the highest optical power values, we observed the OFMs breakup as the fuse discharge was propagating through it. The region around the fuse discharge halting point was observed using an optical microscope (Olympus BH-2). Figure 4(a) and 4(b) shows two of the obtained images where a string of voids is headed by a large void. Recall that this void configuration is typical of fuse effect stopping points and the large void divides the regions of damaged and undamaged fiber [16]. Moreover, the string of voids becomes irregular as it approaches the stop point, which suggests that the optical intensity is closer to the threshold value for the fuse effect propagation [16]. Using the microscopic images, the OFM diameter in the halting zone was measured. These diameters are displayed in Fig. 4(c), as function of the injected optical power. The obtained values for the two different OFM adiabatic profiles are, within experimental uncertainty, coincident, which indicates that the taper diameter decrease rate is not relevant.

 

Fig. 4 Left - Microscopic images (magnification of × 50) of the fiber fuse discharge halting zone in OFM A, for an optical power of 3.0 W (a), and in OFM B for an optical power of 2W (b).The arrow represents the optical signal propagation direction. Right - OFMs diameter in the optical discharge halting zone (the line is a visual guide) (c) and corresponding MFD (d), both as function of the injected optical power.

Download Full Size | PDF

Figure 4(d) shows the correspondent MFD in the halting point as estimated using the data in Fig. 1. Since the MFD values increase with the injected power, it suggests that the MFD on the halting point is such that the intensity there is equal for all optical power. However, our calculations showed that the average intensity on these halting points is not constant but has a decreasing tendency with the increase of the MFD (Fig. 5 ). For completeness, we have added the threshold intensity for this SMF that was reported in [17] as 1.8 MWcm⁻², which is higher than any other value in the graph. The fiber fuse effect has complex dynamics that could depends on several factors, such as the fiber core composition and profile, MFD and the thermal losses. On one hand, a possible explanation for this behavior could be the change of the germanium doped region profile in the tapered zone due to the distribution of the optical power over an increasing area of silica cladding, as the MFD increases. On the other hand, as referred in the introduction, the fuse effect occurs in silica core fibers and the intensity threshold is equivalent to the one in germanium doped core fibers [5]. In [18], was proposed a one-dimensional thermal model where the thermal loss decreases as the mode field radius increases. A decrease on the thermal losses will benefit the fuse propagation and so tends to decrease the threshold intensity. However, the decrease of the threshold intensity can also be a consequence of the conjugation of several factors.

 

Fig. 5 Optical intensity in the optical discharge halting zone as function of the MFD also in the halting zone

Download Full Size | PDF

Nevertheless, the main mechanism responsible for the fuse discharge extinction in the OFMs is the decrease of the optical intensity to values below the fuse effect threshold. Using the threshold intensity for this SMF [17] and the maximum MFD value (Fig. 1), we have determined the upper optical power limit at which this device will theoretically fail to halt the propagation of the fiber fuse effect that is ~7.5 W. However, we believe that for higher optical power values the OFM will effectively halt the fuse discharge propagation, but then due a mechanism similar to the one reported in [12]. In this latter work, only the cladding was etched such that the MFD was maintained approximately equal, however, as the cladding thickness was reduced it was not able to contain the high temperature and pressure of the fuse zone within the core. The resulting expansion of the fuse zone changes the temperature and eventually halts the fuse propagation or even breaks the fiber. In our device, if the optical power was higher than our theoretically calculated value of 7.5 W, the fuse could in fact not halt due to MFD expansion since it cannot expand further than ~23 µm but it would eventually halt due to reduced cladding strength at the OFM regions of diameters under 40 µm as in [12]. Note that even with lowers optical powers, we have observed the destruction of some of ours OFMs.

5. Conclusion

A method to halt the fiber fuse discharge propagation, based on OFMs produced from SMF, was demonstrated. The fundamental mechanism responsible for the fuse propagation halting is the expansion of the MFD that is occurring as the taper diameter is reduced, which leads to the decrease of the optical intensity down to values below the fuse effect threshold. However, the results also showed that the average optical intensity at the halting point is not constant for all OFMs but was lower as the local mode field diameter was larger. Nevertheless, this behavior does not interfere with the ability of the OFM to halt the fuse effect. Hence, this taper with low insertion loss can be introduced ahead of the transmission lasers and other active network components, acting as a passive protection device.