OSA's Digital Library

Optics Express

Optics Express

  • Editor: Michael Duncan
  • Vol. 14, Iss. 16 — Aug. 7, 2006
  • pp: 7312–7318
« Show journal navigation

A mechanical criterion for the design of readily cleavable microstructured optical fibers

Véronique François and Seyed Sadreddin Aboutorabi  »View Author Affiliations


Optics Express, Vol. 14, Issue 16, pp. 7312-7318 (2006)
http://dx.doi.org/10.1364/OE.14.007312


View Full Text Article

Acrobat PDF (1795 KB)





Browse Journals / Lookup Meetings

Browse by Journal and Year


   


Lookup Conference Papers

Close Browse Journals / Lookup Meetings

Article Tools

Share
Citations

Abstract

Some complex microstructured fibers (MSFs) are well known to produce poor-quality cleaves or even to break at cleavage. But to find widespread use in photonics technology, MSFs will have to be easily cleavable using mechanical cleavers, since more sophisticated techniques add complexity. In this paper, the very different, yet reproducible cleavage patterns of three high air-fraction, double-clad microstructured fibers are analyzed. Fracture faces reveal the fracture propagation paths and provide measurements of the fracture lengths in the intercapillary bridges. These lengths prove to be always shorter than the critical fracture length predicted by fracture mechanics. A criterion based on critical fracture length is thus proposed to design cleavage-robust MSFs.

© 2006 Optical Society of America

1. Introduction

During the past few years, there has been some investigation into the damage to microstructured fibers (MSFs) caused during conventional mechanical cleavage [1

1. S. Huntington, K. Lyytikainen, and J. Canning “Analysis and removal of fracture damage during and subsequent to holey fiber cleaving,” Opt. Express 11, 535–540 (2003). [CrossRef] [PubMed]

]. MSFs are optical fibers with air capillaries running throughout their length and many novel designs are being explored for promising applications. For handling-cost reduction and simplicity, they are preferably cleaved using conventional mechanical cleavers. However, using this technique, the fracture faces of even the highest-quality fibers frequently show damage around the capillary holes [1

1. S. Huntington, K. Lyytikainen, and J. Canning “Analysis and removal of fracture damage during and subsequent to holey fiber cleaving,” Opt. Express 11, 535–540 (2003). [CrossRef] [PubMed]

–3], which accelerates fiber degradation and degrades splice quality [4–6

4. R. O. Ritchie, “Mechanics of fatigue-crack propagation in ductile and brittle solids,” Int. J. Fract. 100, 55–83 (1999). [CrossRef]

]. In this paper, we demonstrate that fracture mechanics of brittle materials can explain various MSF damage patterns at cleavage and can be applied at the design stage to improve MSF robustness.

2. Experiments

Three samples of high-air-fraction, double-clad MSFs, prototyped by OFS Fitel Denmark ApS, were studied. These fibers are currently the subject of intense research efforts because of their ability to amplify signals to very high powers, thanks to their unsurpassed high numerical aperture, not achievable with conventional double-clad fibers. All three fibers are made of the same materials under the same manufacturing conditions. The diameters of the inner cladding, ring of capillaries, outer cladding and fiber with its polymer coating are given in Table 1 for the three fibers together with their mean bridge thicknesses. A “bridge” refers to the wall between two adjacent capillaries. Although all three fibers feature similar geometries, they reproducibly produce the vastly different cleavage patterns shown in Fig. 1.

Table 1. MSF dimensions.

table-icon
View This Table
| View All Tables
Fig. 1. Cleaved surfaces of the MSF samples, obtained with an EFC-11 cleaver; a) out-of-focus optical micrograph of Fiber 1 revealing the fracture propagation path, from the blade-impact point to the final fracture ridge; b), c) and d) scanning-electron micrographs of Fibers 1, 2 and 3 respectively.

An Ericsson EFC-11 mechanical fiber cleaver was used, which initiates a crack that spreads throughout the fiber cross-section, leaving a mirror-like fracture face in its wake if appropriate stress is applied [7

7. A. D. Yablon, Optical Fiber Fusion Splicing (Springer, Germany,2005).

, 8

8. T. Haibara, M. Matsumoto, and M. Miyauchi, “Design and developpement of an automatic cutting tool for optical fibers,” J. Lightwave Technol. LT-4, 1434–1439 (1986). [CrossRef]

]. Cleavage features were measured using both optical and scanning electron microscopes (SEM) at magnifications up to 300,000X. Fig. 1(a) is a slightly out-of-focus optical micrograph that reveals the fracture propagation path in Fiber 1: the shock wave creating the fracture induces wave-like imperfections at the surface, which in turn cause optical interference in the micrograph. Fracture propagates from the blade impact point to the final fracture ridge in the outer cladding in both the clockwise and the counterclockwise directions. The ridge is formed as the clockwise and the counter-clockwise fractures propagate in slightly different planes and tear the material off at their junction. Each fracture front induces independent residual fractures in the bridges while passing them, as evidenced by the wave-like interference in the inner cladding of Fig. 1(a) and previously reported using atomic-force microscopy [1

1. S. Huntington, K. Lyytikainen, and J. Canning “Analysis and removal of fracture damage during and subsequent to holey fiber cleaving,” Opt. Express 11, 535–540 (2003). [CrossRef] [PubMed]

]. Good fracture face quality can be achieved if residual fractures are able to propagate all the way through the bridges and recombine in the same plane to cleave the inner cladding with a mirror-like quality. This occurs with Fiber 1, as shown in the SEM image of Fig. 1(b). The roughness of the resulting surface is less than 1 μm, which is smooth enough for splicing.

In contrast, mechanical cleavage of Fibers 2 and 3 leads to the catastrophic patterns shown in Fig. 1(c)and 1(d). The inner cladding of Fiber 2 is pulled out, leaving the outer cladding as a hollow tube. The majority of the bridges break near the outer-cladding, residual fractures apparently stopping at the bridge entrance. The inner cladding of Fiber 3 recedes inside the outer cladding, the majority of the bridges breaking near their junction with the inner cladding and interrupting the mirror-like surface at this point. These observations, reproducible under the entire range of cleaving force, indicate that most residual fractures in Fiber 2 and 3 bridges stop propagating somewhere on the bridges, preventing high-quality cleavage of the inner cladding.

3. Analysis

Fracture mechanics of glass fiber is described by the classic Griffith model of brittle fracture further developed by a number of authors [9

9. D. Glodge, P. W. Smith, D. L. Bisbee, and E. L. Chinnock, “Optical fiber end preparation for low-loss splices,” Bell Syst. Tech. J. 52, 1579–1587 (1973).

]. In this model, the fracture does not open while propagating, but the two newly created surfaces rather remain parallel to each other and the fracture features an infinitely small radius of curvature at the tip as shown in Fig. 2. When such a fracture propagates, the released strain energy is the sum of the surface energy of the two newly created surfaces and the kinetic energy associated with the moving fracture. The theoretical model shows that the energy balance equation for a fracture of length a propagating in a brittle isotropic material subject to a locally applied tensile stress σapp is a function of the density, the Young’s modulus and the surface tension of the material [9

9. D. Glodge, P. W. Smith, D. L. Bisbee, and E. L. Chinnock, “Optical fiber end preparation for low-loss splices,” Bell Syst. Tech. J. 52, 1579–1587 (1973).

]. The general expression of the theoretical model is bounded by two limiting cases of special interest.

Fig. 2. Ingliss model of fracture propagation in a bar.

The first case of interest is the minimum, or threshold, condition for the fracture to be able to propagate. It is reached when the released strain energy equals the surface energy and is just sufficient to make up for the creation of the two new surfaces. In this case, the kinetic energy associated with the moving fracture is essentially zero. This condition is generally written as [7

7. A. D. Yablon, Optical Fiber Fusion Splicing (Springer, Germany,2005).

]:

σappacY=KIC,
(1)

where ac is the minimum or critical fracture length that can induce rupture of the material under the given applied stress σ app, Y is a geometrical configuration factor and KIC is the material toughness. KIC is an experimentally measurable material constant, which is a simple function of the Young’s modulus and the surface tension used in the general theory [9

9. D. Glodge, P. W. Smith, D. L. Bisbee, and E. L. Chinnock, “Optical fiber end preparation for low-loss splices,” Bell Syst. Tech. J. 52, 1579–1587 (1973).

].

The second case of interest occurs when the propagating fracture fails creating only two, optically smooth surfaces, known as the mirror zone, but rather creates a mist zone and a hackle zone on the fracture face. As either the fracture length a or the applied tensile stress σ app increase, more and more strain energy is converted into kinetic energy, until the fracture reaches a limiting velocity. At this point, the excess energy begins to be taken up by the creation of additional surfaces in the mist zone. This condition is generally written as [7

7. A. D. Yablon, Optical Fiber Fusion Splicing (Springer, Germany,2005).

]:

σappDmist=Kfract,
(2)

where Dmist is the distance from the crack initiation site to the mist boundary and Kfrac is another material constant, which is a function of the material toughness KIC, the material density and the fracture limiting velocity [9

9. D. Glodge, P. W. Smith, D. L. Bisbee, and E. L. Chinnock, “Optical fiber end preparation for low-loss splices,” Bell Syst. Tech. J. 52, 1579–1587 (1973).

].

In between these two extreme cases, a fracture of length a subject to a locally applied tensile stress σapp produces a mirror like fracture face. There, we can write:

KICYσappaKfract
(3)

and the released strain energy is sufficient to create two optically smooth surfaces. Some of the released energy is transformed into kinetic energy and the fracture speeds up, but the fracture speed never reaches the limiting velocity. We call such a fracture front autonomous.

The material toughness KIC of pure silicate optical fibers is 0.73 MPa.m1/2 [10

10. NIST, “SiO2 Base Glasses: S100a”, retrieved May 31, 2006, http://www.ceramics.nist.gov/srd/summary/glss100a.htm.

]. The constant Kfrac was experimentally measured to be 2.18 MPa.m1/2 using SMF-28 fracture faces cleaved under various tensile strengths [11

11. S. S. Aboutorabi, “Clivage mécanique des fibres microstructurées,” M. Eng. Thesis (École de Technologie supérieure, Montreal, QC, Canada,2006).

]. This value is in reasonable agreement with the value of ~1.8 MPa.m1/2 extracted from a plot in [7

7. A. D. Yablon, Optical Fiber Fusion Splicing (Springer, Germany,2005).

] and the value of 2.32 MPa.m1/2 reported in [9

9. D. Glodge, P. W. Smith, D. L. Bisbee, and E. L. Chinnock, “Optical fiber end preparation for low-loss splices,” Bell Syst. Tech. J. 52, 1579–1587 (1973).

] for fused silica glass. Both KIC and Kfrac are expected to be the same for our MSF samples as they are made of usual fiber glass. Y is a dimensionless geometrical factor equal to √π·f(a/W), where f(a/W) accounts for the finite with W of the material compared to the fracture length a, which intensifies the local stress at the fracture tip as compared to the remote applied stress σapp. In the case of a fracture propagating from the material boundary in a plane perpendicular to the stress direction [12

12. H. Tada, P. C. Paris, and G. R. Irwin, The Stress Analysis of Cracks Handbook, 3rd ed. (American Society of Mechanical Engineers Press, New York,2000). [CrossRef]

],

f(a,W)=1.120.231(aW)+10.55(aW)221.72(aW)3+30.39(aW)4.
(4)

In the case of the double-clad MSFs, W is the outer cladding diameter and, when the fracture reaches the intercapillary bridges, the fracture length is already equal to the outer cladding thickness, which can be computed from the data of Table 1 as a = (W - Dc) /2, where Dc is the diameter of the ring of capillaries. The corresponding values of f(a/W) are tabulated in Table 2. For any given applied stress, the critical length ac required to cleave a bridge can then be computed using Eq. (1) and could be made as short as necessary by increasing the applied stress. However, Eq. (2) provides a maximum value for the applied stress, as the minimum mist diameter should be equal to the fiber outer diameter to ensure that the entire cross sectional area of the fiber falls within the mirror zone.

Table 2. Cleaving conditions using the EFC-11 cleaver.

table-icon
View This Table
| View All Tables

Table 2 summarizes our cleavage data for the various MSFs. The cleavage tensile force used in the experiments is the one specified by the cleaver manufacturer for each respective fiber diameter, which provides a mist diameter always slightly larger than the fiber diameter. The corresponding applied tensile stress was computed as the ratio of the force to the total fiber cross sectional area.

Fig. 3. Measurement of fracture length in a typical broken bridge.

The experimental fracture length is defined in the propagation plane and on the line normal to the fracture front, from the fracture front to the nearest air boundary. Fig. 3 depicts the measurement AB of the fracture length in a typical broken bridge. Fracture depths AB on the bridges of Fibers 2 and 3 measured using dozens of high-resolution SEM micrographs, are shown in Fig. 4. Heavily damaged bridges in Fiber 2 decreased the number of possible measurements as well as their accuracy compared to Fiber 3. Indeed, 94% of the fracture depths measured in Fiber 2 and 100 % of the fracture depths measured in Fiber 3 turn out to be shorter than the respective computed values of ac.

Fig. 4. Maximum fracture length in bridges for Fibers 2 and 3.

Once the residual fractures have stopped in the bridges, the inner cladding and most of the bridges length remain attached in the case of Fiber 2, while only the inner cladding remains attached in the case of Fiber 3. These remaining attached parts give way under the stress, after a short, yet noticeable delay. In Fiber 2, where fractures stop at the bridge entrances close to the outer cladding, the bridges give way first, stripping the outer cladding off and providing the cleavage pattern of Fig. 1(c). In Fiber 3, where fractures stop at the bridge ends close to the inner cladding, the inner cladding itself gives way under the stress, providing the cleavage pattern of Fig. 1(d).

4. Design criterion

To achieve a smoothly cleaved surface, the main fracture in the outer cladding and the residual fractures in the bridges should propagate without any significant delay. This implies that all fractures should be autonomous. Fig. 5 illustrates the essential aspects of the proposed criterion. A is the impact point. AC is a propagation path tangential to air capillary c1 at point B. From point C on capillary c2, two distinct and independent fracture fronts emerge, one penetrating the bridge and the other continuing through the outer cladding. Both fronts must be autonomous in order to propagate through the bridges and reach the inner cladding imposing the condition BC > ac. In general, a fracture will remain autonomous in a bridge if each line tangential to c1 from the points on the perimeter of c2 has a length at least equal to ac (i.e. ED > ac). This condition guarantees fracture propagation in those bridges that are not directly in line with the initial fracture point. The same condition should be satisfied when the fracture travels from the inner cladding to the outer cladding. The criterion to determine the suitability of capillary shape and position within the fiber cross-section can be summarized as follows: 1) For each point on a capillary perimeter, draw a tangential line to its neighbouring capillaries. 2) Place adjacent capillaries so that the length of those tangential lines is slightly longer than ac. This guarantees the existence of at least one autonomous fracture point and is verified for Fiber 1.

Fig. 5. Design criterion for robust capillary bridges: any BC and DE segments should be larger than the critical fracture length ac.

5. Conclusion

Concepts of fracture propagation stop and critical fracture depth were put forward in the context of complex microstructured fibers. Measured fracture depths in broken bridges were found to be in complete agreement with these concepts. Based on these results, we proposed to use the critical fracture depth as a novel criterion to design complex MSFs capable of readily sustaining mechanical cleavage. In particular, we believe that this criterion permits an appropriate selection of bridges’ lengths and widths, hence of capillary shapes, which will allow to design very large NA double clad MSFs that both efficiently confine pump light and are easily cleavable.

Acknowledgments

The authors would like to thank Dr. Torben Veng, from OFS Fitel Denmark ApS, who kindly provided the fiber samples and professors Jacques Masounave and Philippe Bocher for expert advice. They also would like to acknowledge the financial support of the National Science and Engineering Research Council of Canada.

References and links

1.

S. Huntington, K. Lyytikainen, and J. Canning “Analysis and removal of fracture damage during and subsequent to holey fiber cleaving,” Opt. Express 11, 535–540 (2003). [CrossRef] [PubMed]

2.

C. Simonneau, P. Bousselet, G. Melin, L. Provost, C. Moreau, X. Rejeaunier, A.Le Sauze, L. Gassa, and D. Bayart, “High-power air-clad photonic crystal fiber cladding-pumped EDFA for WDM applications in the C-band,” presented at the Europeen Conference on Optical Communications (ECOC), PD57, (2003).

3.

W. J. Wadsworth, M. R. Percival, G. Bouwmans, J. C. Knight, T. A. Birks, T. D. Hedley, and P. S. J. Russel, “Very high numerical aperture fibers,” IEEE Photon. Technol. Lett. 16, 843–845 (2004). [CrossRef]

4.

R. O. Ritchie, “Mechanics of fatigue-crack propagation in ductile and brittle solids,” Int. J. Fract. 100, 55–83 (1999). [CrossRef]

5.

D. G. Holloway, “The fracture of glass,” in Physics Education (1968), pp. 317–322. [CrossRef]

6.

T. Kuwabara, Y. Mitsunaga, and H. Koga “Calculation method of failure probabilities of optical fibers,” J. Lightwave Technol. 11, 1132–1138 (1993). [CrossRef]

7.

A. D. Yablon, Optical Fiber Fusion Splicing (Springer, Germany,2005).

8.

T. Haibara, M. Matsumoto, and M. Miyauchi, “Design and developpement of an automatic cutting tool for optical fibers,” J. Lightwave Technol. LT-4, 1434–1439 (1986). [CrossRef]

9.

D. Glodge, P. W. Smith, D. L. Bisbee, and E. L. Chinnock, “Optical fiber end preparation for low-loss splices,” Bell Syst. Tech. J. 52, 1579–1587 (1973).

10.

NIST, “SiO2 Base Glasses: S100a”, retrieved May 31, 2006, http://www.ceramics.nist.gov/srd/summary/glss100a.htm.

11.

S. S. Aboutorabi, “Clivage mécanique des fibres microstructurées,” M. Eng. Thesis (École de Technologie supérieure, Montreal, QC, Canada,2006).

12.

H. Tada, P. C. Paris, and G. R. Irwin, The Stress Analysis of Cracks Handbook, 3rd ed. (American Society of Mechanical Engineers Press, New York,2000). [CrossRef]

OCIS Codes
(060.0060) Fiber optics and optical communications : Fiber optics and optical communications
(060.2280) Fiber optics and optical communications : Fiber design and fabrication
(160.2290) Materials : Fiber materials

ToC Category:
Photonic Crystal Fibers

History
Original Manuscript: June 13, 2006
Revised Manuscript: July 27, 2006
Manuscript Accepted: July 27, 2006
Published: August 7, 2006

Citation
Véronique François and Seyed Sadreddin Aboutorabi, "A mechanical criterion for the design of readily cleavable microstructured optical fibers," Opt. Express 14, 7312-7318 (2006)
http://www.opticsinfobase.org/oe/abstract.cfm?URI=oe-14-16-7312


Sort:  Author  |  Year  |  Journal  |  Reset  

References

  1. S. Huntington, K. Lyytikainen and J. Canning "Analysis and removal of fracture damage during and subsequent to holey fiber cleaving," Opt. Express 11, 535-540 (2003). [CrossRef] [PubMed]
  2. C. Simonneau, P. Bousselet, G. Melin, L. Provost, C. Moreau, X. Rejeaunier, A. Le Sauze, L. Gassa and D. Bayart, "High-power air-clad photonic crystal fiber cladding-pumped EDFA for WDM applications in the C-band," presented at the Europeen Conference on Optical Communications (ECOC), PD57, (2003).
  3. W. J. Wadsworth, M. R. Percival, G. Bouwmans, J. C. Knight, T. A. Birks, T. D. Hedley and P. S. J. Russel, "Very high numerical aperture fibers," IEEE Photon. Technol. Lett. 16, 843-845 (2004). [CrossRef]
  4. R. O. Ritchie, "Mechanics of fatigue-crack propagation in ductile and brittle solids," Int. J. Fract. 100, 55-83 (1999). [CrossRef]
  5. D. G. Holloway, "The fracture of glass," inPhysics Education (1968), pp. 317-322. [CrossRef]
  6. T. Kuwabara, Y. Mitsunaga and H. Koga "Calculation method of failure probabilities of optical fibers," J. Lightwave Technol. 11, 1132-1138 (1993). [CrossRef]
  7. A. D. Yablon, Optical Fiber Fusion Splicing (Springer, Germany, 2005).
  8. T. Haibara, M. Matsumoto and M. Miyauchi, "Design and developpement of an automatic cutting tool for optical fibers," J. Lightwave Technol. LT-4, 1434-1439 (1986). [CrossRef]
  9. D. Glodge, P. W. Smith, D. L. Bisbee and E. L. Chinnock, "Optical fiber end preparation for low-loss splices," Bell Syst. Tech. J. 52, 1579-1587 (1973).
  10. NIST, "SiO2 Base Glasses: S100a," retrieved May 31, 2006, http://www.ceramics.nist.gov/srd/summary/glss100a.htm.
  11. S. S. Aboutorabi, "Clivage mécanique des fibres microstructurées," M. Eng. Thesis (École de Technologie supérieure, Montreal, QC, Canada, 2006).
  12. H. Tada, P. C. Paris and G. R. Irwin, The Stress Analysis of Cracks Handbook, 3rd ed. (American Society of Mechanical Engineers Press, New York, 2000). [CrossRef]

Cited By

Alert me when this paper is cited

OSA is able to provide readers links to articles that cite this paper by participating in CrossRef's Cited-By Linking service. CrossRef includes content from more than 3000 publishers and societies. In addition to listing OSA journal articles that cite this paper, citing articles from other participating publishers will also be listed.


« Previous Article  |  Next Article »

OSA is a member of CrossRef.

CrossCheck Deposited