This paper discuss a set of tools for metrological evaluation of optical fiber grating sensors, including uncertainty analysis of the measurements. Background and definitions about basic specifications of the optical fiber grating sensors, such as response curve, calibration curve, sensitivity, resolution, repeatability, reproducibility and conformity are presented. Besides, the paper proposes the procedures to identify, quantify, and express uncertainties in optical fiber grating sensor measurements based on the International Standard Organization's Guide to the Expression of Uncertainty in Measurement. In order to illustrate the application of the approach, two fiber grating measuring systems were evaluated under the light of the methodology. The proposed route is a pragmatic step towards the standardization of optical fiber sensors.

In this work is evaluated the performance of etched fiber Bragg gratings, assembled in different configurations to operate as a refractometric sensor, in the analysis of water-ethanol mixtures. Two fiber Bragg gratings operating close to 1300 nm and 1500 nm were wet-etched and employed in the sensor design. Four configurations for the refractometric sensor were studied, in dual-wavelength and single-wavelength operation modes, and at two sample temperatures. Calibration curves were determined for the range between 0.0 and 100.0 % v/v of water in ethanol, and the sensor performance for each configuration was analyzed by comparing its sensitivity, conformity, repeatability and combined uncertainty. The best results showed that the sensor can be used to measure the ethanol-water concentration with combined uncertainty of 2.8 % v/v for the range up to 80.0 % v/v of ethanol concentration and 7.0 % v/v of uncertainty for the range above 80.0 % v/v of ethanol concentration for the single-wavelength operation mode at two temperatures.

This work shows prospects of long-period fibre grating applications as transducers for fuel conformity analysis. The proposed long-period grating transducer was employed to assess the gasoline conformity in commercial gas stations. Grating responses were used to train and validate a radial base function topology of an artificial neural network. The obtained results show that fibre optic sensors supervised by artificial neural networks can integrate systems for smart sensing with high applicability in the petrochemical field. The radial base function had reached a correct classification probability of approximately 94%. The device applicability in the analysis of hydrated ethanol fuel was also investigated by measuring the concentration of ethanol in ethanol–water mixtures. The results showed that the developed transducer can be used to infer the ethanol–water concentration with a resolution of up to 0.23%.