Fiber Bragg Grating (FBG) temperature sensors are optical devices that encode temperature as a shift in the reflected Bragg wavelength, and they are widely embedded in concrete structures, bridges, tunnels, and dams for long-term monitoring. Testing these sensors before deployment is a distinct discipline from electrical sensor calibration, because the measurement chain includes the interrogator, the optical fiber path, and any strain-coupling effects from the host structure. This article describes a complete Fiber Bragg Grating temperature sensor testing procedure for civil engineering use. It covers the measurement scope and physical principle, specimen preparation and calibration setup, wavelength-shift measurement with temperature calibration, evaluation of sensitivity, accuracy, and repeatability, co-tested parameters such as strain compensation and durability, and the acceptance criteria and engineering scenarios in which validated sensors are applied.
Scope and principle
The test scope covers bare FBG temperature elements and packaged FBG temperature probes intended for embedment in or surface mounting on civil structures, together with their配套 optical interrogator. The measurand is the Bragg wavelength of the grating, defined as twice the grating period multiplied by the effective refractive index of the fiber core. Temperature changes this wavelength through two coupled effects: the thermo-optic coefficient alters the refractive index, and thermal expansion alters the grating period. Because the two mechanisms act simultaneously, the sensor is treated as a single-input, single-output device whose wavelength shift is proportional to temperature change over its rated range. Strain applied to the fiber produces an additional wavelength shift of the same order as the thermal effect, so the test program must isolate temperature from mechanical loading. For this reason, temperature calibration is performed on strain-decoupled specimens, and strain sensitivity is characterized separately as a co-test parameter. The output of the principle-level verification is a wavelength-versus-temperature relationship suitable for linear regression.
Sensor
specimen preparation and calibration setup
Specimen preparation begins with visual and optical inspection of each sensor: the fiber, connector, and package are examined for coating damage, and optical time-domain reflectometry may be used to locate breaks or excessive attenuation along the path. Each sensor is labeled with its grating serial number and reference wavelength recorded at a stated reference temperature. For calibration, the sensor is installed in a thermostatic bath or a precision dry-block calibrator whose medium covers the rated temperature range of the device; stirred liquid baths are preferred for their uniformity, while dry-blocks suit packaged probes. A reference platinum resistance thermometer, itself traceable to national standards, is placed at the same depth and radial position as the sensor under test. The interrogator is warmed up per the manufacturer's instruction, and its wavelength reference is verified against a known reference before data collection. Care is taken to keep the fiber outside the temperature zone free of sharp bends and to route it so that no axial load reaches the grating during the soak periods.
Test
methods: wavelength shift measurement and temperature calibration
The calibration follows an up-down cycling procedure across at least five evenly spaced temperature points spanning the rated range, for example from below-zero conditions to the upper service limit. At each point, the bath is allowed to stabilize until the reference thermometer indicates a steady reading, then a soak interval is observed so that the sensor and the reference reach thermal equilibrium. Wavelength data are then logged continuously, and the mean of a stable acquisition window is taken as the observed value. The sequence is repeated in the descending direction so that hysteresis can be assessed from the difference between the two branches. A first-order linear regression of wavelength against reference temperature yields the temperature sensitivity coefficient and the intercept; where curvature is observed, a second-order fit is reported alongside. Residuals of the fit are examined for systematic deviation, which often reveals inadequate soaking, connector instability, or strain coupling. The calibrated coefficients are written into the sensor record and used by the monitoring software to convert wavelength to temperature.
Performance
metrics: sensitivity, accuracy, and repeatability
Three metrics govern whether a calibrated sensor is fit for structural monitoring. Sensitivity is the slope of the wavelength-temperature regression, typically on the order of ten picometers per degree Celsius for silica-based gratings; it is reported with its confidence interval, and the value is checked against the sensor specification. Accuracy is evaluated as the deviation between the temperature computed from the calibrated coefficients and the reference temperature at each calibration point; the maximum absolute deviation and the root-mean-square deviation are both reported. Repeatability is assessed by repeating at least three full calibration cycles on the same specimen and computing the spread of the wavelength readings at each fixed point; a narrow spread confirms that the grating and its package return to a consistent state. Hysteresis between ascending and descending branches is quantified and added to the uncertainty budget. Interchangeability between nominally identical sensors is also reviewed, since civil monitoring networks often replace a single sensor without recalibrating the entire array.
Co-test
parameters: strain compensation and durability testing
Because embedded gratings respond to strain as well as temperature, the test program includes a strain-sensitivity measurement. A grating specimen is fixed to a calibrated tensile stage or beam fixture, and wavelength shift is recorded under controlled displacement steps at constant temperature. The resulting strain coefficient is used either to design a strain-isolating package or to pair a loose-tube temperature sensor with a strain-measuring grating for compensation. Durability testing addresses the civil service environment: packaged sensors are subjected to temperature cycling beyond the rated range to check package integrity, to immersion testing for water resistance, and where relevant to alkaline-solution exposure simulating pore water in concrete. Optical continuity, sensitivity drift, and calibration offset are re-measured after each exposure, and any change beyond the stated tolerance triggers investigation. Connector and cable tensile and bend performance are verified against the installation method, since most field failures occur at the lead fiber rather than at the grating itself.
Acceptance
criteria and engineering application scenarios
Acceptance is judged against the project specification and the sensor datasheet jointly. Typical requirements state that the fitted sensitivity shall fall within the datasheet tolerance band, that maximum calibration deviation shall not exceed the accuracy class required by the monitoring design, and that repeatability and hysteresis shall each remain below defined limits over three cycles. Post-durability drift in sensitivity and offset must stay within stated allowances, and every accepted sensor receives a calibration certificate listing its coefficients, reference wavelength, test range, and uncertainty. Validated sensors are deployed in mass-concrete temperature monitoring during curing, where internal thermal gradients guide cooling and formwork removal; in long-term health monitoring of bridges and tunnels, where slow temperature drifts are separated from structural response; and in dam and embankment instrumentation, where multi-point FBG chains measure internal temperature fields. In each scenario, the calibration record allows measured wavelengths to be converted to engineering units with a traceable and documented uncertainty.
Frequently Asked Questions
Which
FBG temperature sensor types fall within this testing scope?
The scope covers bare grating elements, packaged probes, and multi-point grating chains intended for civil structures, whether embedded in concrete or surface-mounted, together with their optical interrogators. Sensitivity coefficients differ between package designs, so each configuration is calibrated individually rather than assumed from the fiber alone.
What
does an FBG temperature sensor test report contain and how is it used?
The report lists the grating serial number and reference wavelength, the calibration range and points, the regression coefficients with uncertainty, sensitivity, maximum deviation, repeatability, and hysteresis results, plus durability re-check data. Engineers use these coefficients and limits to convert field wavelength readings into temperature values and to judge whether the sensor meets the monitoring design.
What
should be communicated with the laboratory before submitting FBG sensors for testing?
Clients should state the rated temperature range, the packaging and installation method, whether strain compensation is required, and the interrogator used on site. Agreeing on calibration points, soak criteria, and any durability exposures in advance prevents retesting and keeps the calibration coefficients consistent with the actual monitoring configuration.