Measurement of radiation-induced attenuation in optical fibers by optical-time-domain reflectometry

Abstract

ABSTRACT Co radiation-induced attenuation is measured by optical time domain reflectometry (OTDR) techniques and compared toconventional optical throughput measurements. Relative advantages and disadvantages of the OTDR technique are discussed. i_t_ INTRODUCTION Since 1983, the Nuclear Effects Task Group (NETG) formed by NATO RSG.12 (Panel IV, A/C 243) has been utilizing inter-laboratory comparisons of radiation-induced attenuation in optical fibers to develop an international test protocol for thesemeasurements.1'2'3 The protocol was to be based on the combined experience of participating Laboratories within the NATOcountries. Each of these facilities accomplished their previous attenuation measurements by documenting the change in lighttransmission through the fiber during and following irradiation (i.e., by throughput techniques). NETG results through 1989have been published.4'5 These efforts resulted in a finalized test protocol,6 which has been submitted to several certificationorganizations for consideration as an international standard.The NETG protocol defines procedures for measuring attenuation changes due to radiation exposure using throughput tech-niques. The procedure also includes a section suggesting that similar measurements should be possible using optical timedomain reflectometry techniques. However, since no OTDR-based measurements had been completed for radiation-inducedattenuation when the current draft of the protocol was finalized, no detailed recommendations were made regarding the use ofan OTDR.At the 1990 NETG meeting at the Royal Military College of Science (Shrivenham, UK), the absence of any data for evalua-tion of OTDRs for radiation-induced attenuation measurements was cited as a major need within the testing community andseveral of the participating Laboratories were tasked to investigate this area for subsequent evaluation by the NETG. Thispaper reports on the OTDR investigation within the Los Alamos National Laboratory. In the process, a better understandingwas gained for situations wherein the use of OTDR may be critical for enabling a measurement. In general cases, however,we suggest that the use of OTDR is more cumbersome and less accurate than throughput measurements.Significant assistance was received from colleagues who are also pursuing OTDR tests7'8 who have shared their progress withthe present authors. Only one paper in the literature, to our knowledge, has previously appeared reflecting their experience.9