For many industrial and commercial processes, temperature monitoring ensures operational safety and efficacy. Conventional electric temperature sensors are adequate if replaced often and effectively shielded from electromagnetic interference (EMI). However, they all suffer from the same inherent limitation – they can only measure temperature at a single location. In practice, these sensors are often deployed at a handful of locations, so the overall temperature distribution remains unknown.Fiber optic sensing system (FOSS) technology, an alternative method to measure temperature, acquires continuous profiles along the entire length of an optical fiber with millimeter spatial resolution. The temperature at thousands of sensing points can be monitored using a single lead cable. Processes that rely on temperature sensors to maintain ambient temperature uniformity or to detect hot spots can benefit considerably from understanding the overall temperature distribution. When performing distributed temperature sensing, four key factors will determine success: fiber temperature sensitivity, coating effects, sensor preparation, and calibration.
Fiber temperature sensitivity
FOSS interrogators perform two direct measurements – mechanical strain and temperature. Temperature sensitivity stems from two phenomena: changes in the core refractive index with respect to temperature and thermally induced strain. The fundamental objective behind fiber optic temperature sensing is minimizing the mechanical strain component such that the measured apparent strain is only comprised of effects due to temperature.
Optical fiber must be coated to reduce its fragility and to allow it to be handled without breaking. The coating material also determines the fiber’s performance as a sensor. Stiff polymer coatings, such as polyimide and Ormocer, are widely used for strain sensing applications due to their excellent strain transfer properties across a wide operational temperature range. However, like all polymers, these coatings are hygroscopic in nature and will expand volumetrically as they absorb the air’s moisture. Consequently, humidity-driven coating expansion transfers some strain into the fiber optic core, resulting in an additional humidity-dependent hysteresis. Since relative humidity (RH) is intrinsically dependent on ambient temperature, this effect is undesirable for temperature sensing and limits the accuracy of the sensor. Although a stiff coating is essential for strain sensing, a softer coating material, such as Ormocer-T, is suitable for temperature sensing.
Prior to using a FOSS interrogator for distributed temperature sensing, the fiber must be conditioned and configured. Considerations include:
Thermal conditioning – Fiber must be preconditioned to its expected operating temperature range. At a minimum, it is recommended that a single preconditioning cycle is performed where the fiber is subjected to the maximum and minimum operating temperatures for a few hours.
Packaging – The optical fiber must be completely isolated from mechanical strain when performing distributed temperature sensing. This is typically accomplished by packaging the fiber inside a small tube or capillary. The capillary is adhered to the substrate while the fiber housed inside floats freely. If friction effects are small, the measured apparent strain only includes effects due to temperature and a repeatable calibration curve can be generated.
Installation – Depending on how the fiber is packaged and installed, the thermal expansion/contraction of the capillary material can induce mechanical strain within the fiber via friction effects, resulting in incorrect temperature measurements. Depending on several factors, it is possible to package the entire fiber even if the installation requires turns and bends. Factors include fiber length, number of turns required, bend radius of each turn, capillary material, the capillary’s ID, and temperature range. Generally, only a few turns will be able to be accommodated. Installation configurations should be validated before use.
Overall accuracy of a fiber optic temperature sensor is also highly dependent on the calibration quality. Thermocouples and resistance temperature detectors (RTDs) are the two primary devices employed as reference temperature sensors during calibration. Both devices can be used effectively; however, calibrating with thermocouples is typically less cumbersome because you can measure the temperature at a highly localized point.
Fiber optic sensing technology provides a level of insight into surface and ambient temperature distributions that allows users to thermally map areas of interest in real-time with 1.6mm spatial resolution – impractical to achieve using traditional single-point temperature sensors. Due to its small size, chemical inertness, and immunity to electromagnetic interference, optical fiber can be installed in environments where alternative sensors cannot operate. Processes that rely on maintaining temperature uniformity, such as curing composite parts or thermal management of battery packs, stand to benefit greatly from these capabilities.
In the rocket industry, adequate thermal insulation is critical to the survivability of the rocket and overall mission success. Distributed temperature sensing can optimize the thermal insulation design and reduce weight. Other applications include optimizing the performance and effectiveness of heat exchangers, such as a radiator, by thermally mapping the path of the working fluid. Due to the continuous nature of the measurement, temperature gradient distributions are fully captured, providing engineers greater insight into the underlying physics of their device.
Many applications that currently employ traditional single-point temperature sensors, such as thermocouples, would benefit from having thousands of additional measurement points. The primary reasons that thermocouples are currently deployed in limited quantities is the installation time associated with each sensor, the cumbersome wire bundles, and the associated weight penalty. Fiber optic sensing technology overcomes all three of these issues, enabling engineers to capture information that would otherwise be impractical to gather.
About the author: Ryan Scurlock is an application engineer at Sensuron. He can be reached at firstname.lastname@example.org.