Temperature Sensor

In this case study, the performance of a Temperature Sensor based on FSI technology is compared against measurements obtained through a Pico PT-104 Logger using up to 4 RS PRO Type PT 1000 4mm diameter Platinum Resistance Thermometer.
The performed tests show that we can track time-varying temperatures with a maximum average error across all FSI segments of 0.9 degC with respect to the baseline sensor, obtained with a resolution of 0.005 degC.
 

Nine FSI sensors written on a standard single mode optical fibre SMF28 with Acrylate coating are bonded to a PEEK tube to protect them from external factors and allow a complete strain isolation. This defines what we are calling FSI Temperature Sensor in the present case study. The FSI Temperature Sensor is then bonded with tape on an Aluminium plate. On the same Aluminium plate, a Pico PT-104 Logger using up to 4 RS PRO Type PT 1000 4mm diameter Platinum Resistance Thermometer and a TE Connectivity HPP809A031 Humidity Sensor are installed to validate the FSI-based measurements.
The Aluminium plate is then placed in a temperature chamber, to induce a time-varying temperature profile and perform measurements in a controlled environment.

The test starts from a controlled temperature of 25 degC, then ramps up to 70 degC and then down again to 25 degC. The chamber temperature is then lowered to -35 degC, and then up again to the initial temperature value.

The measured temperature profiles are displayed in Figure 1. Data is logged with a frequency of 25 Hz in this case. The FSI-based measurements follow closely the measurements provided by the PT100 sensor, not showing a time-delay when the temperature changes, even if the fibre optics is protected by the PEEK tube. Influence of humidity is not considered in this case, given the strong performance of the FSI temperature sensor across the range of temperatures tested.

When looking at steady-state performance only, a calibration graph of the FSI-based measurements against the PT1000 measurements is shown in Figure 2. The FSI-based measurements are within a 1.5 degC error band with respect to the values measured by the PT1000 sensor, in fact the average difference between the two measurements is 0.9 degC.

Finally, in Figure 3 it is shown a comparison of the resolution from the two different measurement technologies used. The FSI-based measurements have a resolution of 0.005 degC, measured as standard deviation of the signal over 300 seconds, against a resolution of 0.05 degC from the PT1000 sensor.
 

A sampling frequency of 25 Hz is used in this application, given the long timeframe spanned by the test and the slow dynamics involved, but sampling frequencies up to 205 kHz can be obtained with our technology.

The use of the PEEK tube to protect the optical fibre from external factors and achieve strain isolation allows a large range of applications and fibre layouts to be covered.

Additionally, if higher temperatures need to be reached, the currently used Acrylate coating can be replaced with a Polyamide coating, allowing to safely reach maximum temperatures up to 350 degC.