Dynamic characterization of Fiber Bragg Grating temperature sensors

Abstract

To reliably characterize fast dynamic heat transfer mechanisms, fast-response temperature sensors are crucial, including knowledge about the temporal response. In this paper, the dynamic behavior of a Fiber Bragg Grating temperature sensor is investigated and compared to different types of fast-response thermocouples using two different experimental dynamic characterization methods. A temperature step is generated by either plunging the sensor into a fluid or exposing it to a fluid droplet at different temperatures. The step response is evaluated to determine the sensor response time. Calibration runs are performed for a silica-based 0.125mm FBG sensor, as well as for 0.16mm and 0.8mm exposed tip and 0.25mm sheathed tip type K thermocouples. Water, glycerin, oil and GaInSn were used to cover a broad range of applications regarding different thermal diffusivities and viscosities. The FBG sensor showed the shortest response times compared to the thermocouples, ranging from 60ms in oil down to 3ms in liquid metal, which is 20% up to 70% faster compared to a 0.25mm sheathed tip type K thermocouple. Additional plunging calibration runs of the FBG sensor were performed in a ternary nitrate molten salt mixture (HITEC) to determine its overall and dynamic behavior in corrosive fluids at elevated temperatures. It turns out that the FBG sensor is not affected by the molten salt and shows similar response times to those measured in water. Regarding the characterization methods, both techniques show reproducible results, even though the droplet method is inapplicable for sensors with higher heat capacity or lower thermal conductivity than the calibration fluid. Furthermore, splashing effects for fluids with low viscosity reduce the reliability of the droplet method. The results also show that a dynamic characterization is indispensable for temperature measurements with high temporal resolution because the response time depends on the sensor size and the heat transfer coefficient between sensor and surrounding, which in turn depends on the sensor type, fluid properties and the flow parameters.