Computational Model Of Fiber Optic, Arc Fusion Splicing; Experimental Comparison

Abstract

Acknowledgement: The assistance and support of the MICOM Army Missile Command is gratefully appreciated. An analytical tool to investigate the arc fusion splicing of optical fibers is developed. The physical model incorporates heat transfer and thermal, visco elastic strain. The heat transfer equations governing radiation, conduction and convection during arc heating are formulated. The radiation heat flux impinging on the fiber optics is modeled based on reported experimental analysis of a generic type arc discharge. The fusion process considers deformation of the fiber due to thermal, viscous and elastic strain. A Maxwell stress-strain relationship is assumed. The model assumes an initial gap at the beginning of the arc which is closed by a press-stroke during the heating cycle. All physical properties of the fused silica glass fibers are considered as functions of temperature based on available experimental data. A computer algorithm has been developed to solve the system of governing equations and parametric studies carried out. An experiment using a FSM-20 arc fusion splicer manufactured by Fujikura Ltd. was carried out to provide experimental verification of the analytical model. In the experiment a continuous fiber was positioned in the arc and cyclic heating and cooling was carried out. One end of the fiber was clamped and the other was free to move. The fiber was heated for 6 seconds and cooled for 3 minutes for several cycles. At the end of each cooling process, photographs of the deformation of the fiber were taken. The results showed that the fiber necked down on the free end and buldged up on the fixed end. With repeated heating and cooling cycles, the optical fiber eventually necked down to the point that it melted in two. The analytical model was run for the conditions of the experiment. Comparisons of the predicted deformation of the optical fiber with those measured is given. The analytical model displays all of the physical phenomenon of fiber deformation observed in the experiment. The magnitude of the deformation is over predicted in some instances and underpredicted in others. This effect is explained on the basis of unknowns in the model of radiation heat flux and in the variation of density with temperature (i.e., crystal structure changes during heating). It is concluded, however, that the model simulates all the physics correctly, and therefore, provides a meaningful technique for analyzing the significant parameters involved in arc fusion splicing. Plans for further parametric studies are described.