Key metrology considerations for riber Bragg grating sensors

We discuss the current status of fiber Bragg grating (FBG) sensor metrology. High-accuracy wavelength measurements are critical for FBG strain sensors, because wavelength measurement uncertainties even as small as I pm lead to an uncertainty of nearly 1 microstrain. We administered an FBG wavelength measurements round robin in which twelve participants measured the spectral reflectance of four FBGs; we found that the measured peak wavelengths of a single FBG varied by as much as 35 pm. We have determined that this variation results from two factors: the uncertainty of the measurement systems and the wavelength instability of the athermally packaged FBGs. We found that the wavelengths of athermally packaged gratings will drift with time and can undergo large jumps and that it is difficult to achieve stability better than 4 pm/year. We discuss the measurement uncertainties for several common wavelength measurement systems, including tunable laser, optical spectrum analyzer (OSA), and interferometric measurements. We show that when using an OSA, as many of the round-robin participants did, it is difficult to achieve a measurement uncertainty better than 10 pm, and if the OSA is not accurately calibrated to a known wavelength reference; then wavelength measurement uncertainty can be as large as 1 nm. Wavelength references that are based on molecular absorption lines are ideal for calibrating OSAs, but they are not optimal for calibrating FBG interrogation units where a reference based on reflection lines rather than absorption lines is often preferred. We discuss other possible candidates for wavelength references for FBG sensor interrogation units.