Recent advancements in high temperature systems such as gasifiers, molten slag furnaces, super-heated steam turbines, and energy harvesting systems require in-situ temperature monitoring. Conventional temperature sensors such as thermocouples, thermistors, and resistance temperature devices (RTDs) are limited to such applications due to poor reliability. Additionally, these sensors require one end of the device to be maintained in a relatively cold junction which limits the integration with the sophisticated modern high temperature systems. To mitigate these issues, research has been focused on adapting wireless technology for measuring temperature in harsh environments where a conventional temperature sensor cannot be integrated with the system. Initially, research was focused heavily on fiber optics sensor systems, surface acoustic wave (SAW) sensors, laser interferometers, and infra-red sensors. However, these sensor systems have limited lifespan, poor reliability due to reactions taking place at the hot junction, and integration of the sensor with the high temperature system is challenging due to sophisticated design of the furnaces, gasifiers, or turbine blades. In order to mitigate these issues, research has been focused on adapting radio-frequency (RF) technology for high temperature systems. The objective of this work was to design and fabricate inductor-capacitor (LC) resonator based wireless sensors to monitor temperature systems above 500oC.The two-dimensional (planar) LC resonator sensors were designed using the ANSYS Maxwell simulation package. The inductor design was based upon a fifteen-turn coil. The coil geometry was based on a width of 0.15 mm and a spacing of 0.15 mm spacing between the turns. The outer diameter was ~20 mm and the inner diameter was near ~10 mm. The capacitor was based on an interdigitated capacitor (IDC) which could be easily printed onto a ceramic substrate, which would act as the dielectric. The common substrate used in this work was alumina (Al2O3) or yttrium-stabilized zirconia (YSZ). The LC circuit patterns were printed onto alumina or zirconia substrates using inks containing semi-conducting ceramic particles. A few different semi-conducting ceramic compositions were evaluated in this work. One example composition was the La2NiO4 (LN) composition, which displays relatively good high-temperature stability and high electrical conductivity in the range from >40 S/cm at temperatures above 450⁰C. The ceramic powder compositions were synthesized using a conventional solid-state reaction route by mixed the stoichiometric constituents by attrition-milling using zirconia media in ethanol. The resultant slurry was dried and calcined >1000°C for 4 h. X-ray diffraction (XRD) analysis was performed using Panalytical X-ray diffractometer to confirm single phase formation. The circuit patterns were fabricated by using a micro-casting process. Conventional photolithography was used to pattern the photoresist onto the ceramic substrates. The patterned photoresist acted as the molds to deposit the ceramic inks to form the desired LC patterns. The micro-casted patterns were dried and then fired in a muffle furnace to ~1200oC to densify the patterns and bond them to the ceramic substrates. During the firing process, the photoresist would be removed in situ through a basic pyrolysis mechanism. For initial testing in this work, platinum (Pt) interconnects connected to the contact pads were made using Pt ink and further connected to Pt wires to connect to the electrical analysis equipment. Figure 1 shows a magnified optical microscopy picture of a section of the inductor pattern, which shows the level of resolution of the ceramic pattern. The sensors were tested within a high temperature muffle furnace in air. Initial tests used matching sensors placed parallel to each other within the furnace. One sensor was attached using the Pt wires to a signal generator (HP E4433B, USA) to act as the interrogator. The second sensor was hard-wired to a signal analyzer (Tektronix RSA306B, USA). The sensors were characterized at 500 – 1000oC in an ambient atmosphere with an RF signal ranging from 10 – 80 MHz at 2 MHz/sec sweep rate. The sensors showed a sensitivity of 0.2 kHz/oC. Two signal processing methods were implemented in algorithm form (within MATLAB) to estimate the temperature given the frequency response of the sensor. The methods evaluated were the cross-correlation and minimization of absolute error methods. In addition to these two signal processing methods, a split window technique for analyzing the temperature/frequency correlation was used. The cross-correlation signal processing method showed the best fitting for the initial sensor testing. The first set of sensors based on the LN compositions demonstrated sensitivities ~0.2 kHz/oC up to 1000°C. The work will further discuss the results from all sensor compositions and geometries. In addition, the investigation of cyclic stability and repeatability analysis will also be discussed in the presentation. Figure 1
Read full abstract