Tunable diode laser absorption spectroscopy (TDLAS) is a powerful technology to measure the temperature of burners, such as industrial boilers, scramjet engines, gas turbines, and aero engines. Currently, in TDLAS-based thermometers, two-line thermometry is a common method for calculating the temperature of the target devices based on the monotone variation relation between the ratio of two selected absorption linestrengths and the corresponding temperature. However, the range and accuracy of temperature measurements strongly depend on the absorption linestrengths and the corresponding low-energy levels. Thus, improper selection of the two absorption lines may limit the temperature-measurement range and induce a significant measurement error. In addition, the error is influenced by the uncertainty of the line parameters. Recently, a multiline temperature-measurement method has been developed to improve the tolerance of the measurement error induced by the uncertainty of the line parameters, which can be achieved by scanning with an external cavity laser (ECL) with a wide wavelength region to cover several more absorption lines. However, using ECL may result in problems, such as the difficult baseline retrieval caused by the irregular variation of the emission intensity in the wavelength scanning region and the low temporal resolution limited by the response speed of the ECL internal mechanical motor.
Considering the above ECL limitations, a multiple distributed feedback (DFB) laser-based TDLAS system was developed to measure water vapor (H2O) absorption lines near 1339, 1343, 1392, 1393, and 1469 nm. The wavelengths of the lasers can be tuned to match the absorption region of interest by adjusting the lasers’ temperatures and injection currents with homemade laser temperature and current controllers. Five time-division multiplexing ramp signals were generated using an FPGA-based function generator to control the injection currents of the laser. The laser beams were collimated and passed through the gas medium in tube furnaces and scramjet engines. The transmitted beams were collected on photodetectors, and the corresponding H2O absorptions were recorded using an FPGA-based data acquisition circuit. In this study, we applied a typical multiline temperature-measurement method using Boltzmann diagrams to calculate the temperature of the target species. Temperatures ranging from 1000 K to 1600 K were measured in the high-temperature tube furnace (Fig. 2) to assess the performance of this system. Subsequently, the cross-sectional temperatures of the expansion section of the scramjet engine were measured (Fig. 8), where 16 optical paths were arranged in the wall openings of the expansion section. Two experiments were performed under the same conditions to verify the repeatability of the experiment.
The high-temperature tube furnace measurements show that the temperature-measurement accuracy, as expected, presents a strong dependency on the number of the selected absorption lines (Fig. 5) and is enhanced by increasing the absorption lines. A relative measurement error of less than 1% for the temperature range of 1000?1600 K can be achieved by using five H2O absorption lines; this error is lower than that induced by using a smaller number of absorption lines, indicating that the five wavelengths have good adaptability to the measurement over a wide temperature range (Fig. 6). For the scramjet, the measured hydrogen combustion time and fuel combustion time are approximately 0.2 and 0.5 s, respectively. Moreover, the experimental results show that hydrogen combustion is distributed in the upper layer of the expansion section before and after fuel ignition, and the combustion state remains stable owing to the small temperature difference between the set optical paths when the fuel is burned. To test the repeatability of the hydrogen combustion state, we repeated the same experiment twice for the preset conditions A and B. The final results indicate that the two experiments for the same condition have good repeatability, where the average deviation of the temperature measurement is approximately 16 K for the two experiments under condition A, whereas the average temperature is approximately 7 K for the two experiments under condition B.
In this study, we have developed a miniaturized TDLAS system using multiple DFB lasers as light sources to measure the temperatures of H2O in a high-temperature tube furnace and scramjet. A Boltzmann diagram-based multiline temperature-measurement approach is used to improve temperature-measurement precision. The temperature-measurement performance of the furnace using different numbers of H2O absorption lines is assessed in terms of the measurement error, and the optimized number of H2O absorption lines is determined to be five. A temperature measurement error of 1% is obtained by using five H2O absorption lines to calculate the temperature, which also illustrates that the method used in this study is suitable for temperature measurement with a wide temperature range and high accuracy for complex combustion environments. In addition, the linear average temperature is measured in the expansion section of the scramjet engine at different positions of the cross-section, and the experimental results show that the hydrogen combustion state remains highly consistent according to the average deviations of the two experiments of temperature measurement for different preset conditions A and B. All the above results indicate that the current method based on the Boltzmann diagram shows a high potential for high-precision temperature measurements in the combustion flow field, such as in engines with intense combustion.This method will be further developed to provide more wavelength information for the field distribution reconstruction of combustion and improve accuracy.
