To overcome the inherently low temperature sensitivity of fiber Bragg gratings (FBGs) in engineering applications under low-temperature conditions, a sensitivity-enhanced FBG temperature sensor based on a polytetrafluoroethylene (PTFE) encapsulation sleeve was developed. Four adhesive materials—silicone thermal grease, polydimethylsiloxane (PDMS), epoxy resin, and modified acrylic ester—were employed to package the FBG within the PTFE sleeve to improve its temperature sensitivity. Thermal stress simulations of the proposed sensor structure were carried out using COMSOL Multiphysics^® 6.2, and the simulation results showed good agreement with the experimental data. Based on the experimental results, the sensitivity-enhancement effects of PTFE combined with different adhesives, as well as the influences of the PTFE sleeve length and wall thickness, were systematically investigated. The results indicate that, within the temperature range of −35 °C to 15 °C, increasing both the length and thickness of the PTFE sleeve can effectively improve the temperature sensitivity of the sensor. When epoxy resin was used as the encapsulating adhesive, the sensor achieved a maximum sensitivity of 117.4 pm/°C, corresponding to a 13.19-fold increase compared with that of a bare FBG sensor. This sensitivity-enhancing packaging structure significantly improves both the temperature sensitivity and linearity of FBG temperature sensors, while also substantially reducing fabrication costs.

Aerospace and power equipment require precise temperature monitoring in extremely high-temperature environments, which often feature narrow and curved installation spaces. Since traditional thermocouples cannot provide both measurement performance and structural flexibility simultaneously, we propose a flexible fiber Bragg grating (FBG) temperature-sensing array. The sensor features a dual-layer composite packaging structure comprising an inner braided quartz-wool layer and an outer nickel-based alloy tube. Combined with a high-temperature annealing process, the packaged array was evaluated against a WRP-191 reference thermocouple. Over the investigated heating and cooling points up to 1000 °C, the maximum absolute indication error, max|T_FBG − T_ref|, was 2.02 °C and the mean absolute indication error was 1.25 °C. These quantities are comparison errors relative to the reference thermocouple and are not expanded uncertainties. The dual-layer design also allowed a minimum demonstrated bending radius of 22 mm, supporting deployment in confined and curved installation paths.