Consequences of Stress Optical Communications Systems Stress Bending Loss and Reliability in Optical Fibres Increased Loss in the Fibre Increased Probability of Failure Bending Loss in Fibres At a bend the propagation conditions alter and light rays which would propagate in a straight fibre are lost in the cladding. Macrobending, for example due to tight bends Attenuation: Bending Loss Microbending, due to microscopic fibre deformation, commonly caused by poor cable design Microbending is commonly caused by poor cable design Macrobending is commonly caused by poor installation or handling
Ray Diagram View of Macrobending Recall that macrobending is caused typically by poor handling or installation. Ray diagram view used with multimode fibre provides approximate explanation. At a sharp bends light rays which propagate by TIR on straight fibre are lost into the cladding. Result is optical power loss and thus attenuation. Mode Field View of Macrobending Mode field view is more accurate but harder to visualise, a must for singlemode In a fibre a wavefront perpendicular to the direction of travel must be maintained At a sharp bend the outer part of the mode field must travel faster than the inner part to maintain the wavefront Thus outer part of mode field may be forced to travel faster than the velocity of light in the material As this is not possible the energy in the outer part of the mode field is lost through radiation Cladding At a bend loss occurs where TIR fails Loss of a portion of the mode field at a sharp bend Core Power lost via radiation from cladding Mode field Macrobending in Multimode Fibre Critical radius is the bend radius below which loss increases rapidly Critical radius of curvature R c for multimode fibre is given approximately by: R c = 4π 2 3 n 1 λ 2 2 n 1 n 2 32 / Macrobending in Singlemode Fibre In a singlemode fibre as the spot size or mode field radius (MFR) increases the loss at a bend increases Qualitatively this is because a greater proportion of the mode field is lost if the MFR is larger Full analysis of loss is complex and beyond the scope of current discussions Low MFR = Lower Loss Larger MFR = Higher Loss Cladding Cladding Loss can be reduced by using larger refractive index differences For a given bend radius a larger NA will result in a lower R c and thus lower loss Core Power lost via radiation from cladding Core More power lost via radiation from cladding While R c is influenced by wavelength it is found that above R c the loss is not a a strong function of wavelength (multimode fibre only) Mode field Mode field
Quantifying Macrobending in Singlemode Fibre (I) Macrobending can be characterised in SM fibres by the empirical formula: Quantifying Macrobending in Singlemode Fibre (II) Influence of Mac# on loss in db/m at 1320 nm Loss = exp 8.5-519 x D mm 1 λ x Mac# 3 db/m The Mac# (Macrobending Number) is a function of the MFR and the "effective fibre cutoff wavelength λ ce ": Mac# = 2 x MFR λ ce Quantifying Macrobending in Singlemode Fibre (III) The higher the operating wavelength above the cutoff wavelength the lower the V-value Quantifying Macrobending in Singlemode Fibre (IV) Influence of wavelength on loss in db/m for a Mac# of 9 A lower V-value means a larger MFR So for longer wavelengths the MFR and thus the loss increases Thus the loss due to bending can be expected to increase at 1550 nm relative to 1330 nm Typical Mac#'s in singlemode fibre are 8-9 and >10 in so called weakly guiding fibres
Microbending in Fibres Bending Loss Tests for Cables Minimum bend radius for a cable is typically 10 to 20 times the outer diameter of the cable. Microbending in Fibres More critical than macrobending Due to processing rather than mishandling. Loss can occur due to distortion of the core cladding interface, induced by manufacture or poor cable design Common value used in Cabling Standards is 15 times the cable diameter Fibre Reliability Fibre Reliability Fibre is intrinsically very reliable in a benign environment Few documented failure mechanisms Most failures are caused by poor cable choice, poor installation or accidental damage Intrinsic tensile fibre strength exceeds that of an equivalent steel wire Theoretical strength is 20 GPa (2,900,000 Psi) Due to surface defects such as cracks strength in practice is much lower, typically 5 GPa (725 kpsi) Fibre showing surface cracks and flaws (exaggerated) 1kPa = 0.145 Psi
Fibre Proof Testing Crack and Flaw Growth Weak fibres are those with large surface defects after production All produced fibres are proof tested after production Typical proof test stress is three times normal service maximum Failure occurs when under stress a crack grows to some critical dimension Crack growth is depended on the so-called fatigue susceptibility parameter, "n" Larger values of n mean faster crack growth, shorter lifetime Stress accelerates crack growth Simplified proof test apparatus Moisture and high temperatures also accelerate crack growth and reduce lifetime Fibre Failure Examples (I) Minimum Time to Failure Photo shows an end view of a failed fibre Magnification is 2000x Failure caused by small flaw on the fibre surface Two distinct areas visible: Smooth area near flaw were crack propagated quickly but cleanly Jagged area were fibre failed completely Most important parameter for cable designers Assume cable is under a constant stress "s" The time to failure t f is given by: t = A s f A is constant and n is the fatigue susceptibility parameter (15 to 50 for glass, typically 20) As stress grows the time to failure drops rapidly Problem: For n = 20 develop an argument to show that a stress "s" applied for 1 second is equivalent to a stress of 0.35s applied for 40 years -n
Typical industry test Higher reliability tests Proof test stress 50 kpsi (0.35 GPa) 100 kpsi (0.7 GPa) Proof Testing Results Maximum flaw size Predicted lifetime at maximum service stress 2.3 micron 30 Years 0.7 micron >>100 Years Higher proof test stress means longer lifetime But higher stress means more fibres are rejected, lower yield/higher cost Lifetimes assume no moisture ingress and normal temperatures Effect of Moisture Effect of Temperature Effect of Moisture and Temperature Moisture does not penetrate silica glass, so it does not affect propagation Presence of water as OH ions on the fibre surface accelerates crack growth This process is called stress corrosion Moisture protection is important in fibre cables At 90 degrees centigrade the fatigue susceptibility parameter is significantly higher than that at 25 degrees Fibre strength decreases by 25% at 90 degrees compared to 25 degrees High tensile strength and zero moisture ingress cables are essential at elevated temperatures