Development and Application of a Method for Gas-phase Temperature Measurements in Particle-laden Flows

Abstract

Suspensions of particles in a carrier flow of gas are utilised in, or being developed for, several high-temperature industrial processes. These include for material transformations in calciners and kilns, as fuel in particulate burners and as the medium for radiation absorption in concentrated solar thermal receivers. The efficiency, stability, and emissions from such systems is strongly dependent on the temperature distribution of both the particle and fluid phases, each of which can be highly variable both spatially and temporally. While these systems are widely utilised, there is still a lack of fundamental understanding of the heat transfer processes due to the complexity of turbulent particle-laden flows with a high particle volume fraction. Therefore, this work aims to provide insight into these processes for future optimisation of non-isothermal particle-based systems. This is performed by adapting and applying the technique of laser induced fluorescence (LIF) to measure the gas-phase temperature in a particle-laden flow that is heated using high-flux radiation. This thesis presents the first demonstration of LIF in the densely loaded conditions present in particle-laden flows relevant to industrial application, with the potential for strong optical interference from elastic scattering of radiation from the excitation laser by particles. The two-colour method for thermometry, with toluene as the fluorescent tracer, was used to provide spatially resolved measurements from a < 1 mm thick planar cross-section of the flow. The particle distribution was measured simultaneously with the temperature by imaging the laser light scattered by particles (particle nephelometry). The accuracy and precision of the two-colour LIF method was assessed for a series of particle materials and diameters, including materials that luminesce following absorption of the excitation laser light. The results show that optical filters effectively suppress the detection of elastically scattered light, with other sources of measurement uncertainty including particle luminescence, laser attenuation, and signal trapping identified and assessed. The systematic error in the measurement from these combined sources was shown to increase with local particle loading, but be independent of particle diameter. The two-colour LIF and particle nephelometry methods were applied to simultaneously measure the gas-phase temperature and particle distributions in a particle-laden flow heated using high-flux radiation, evaluated for systematically varied series of particle diameter, particle volumetric loading, and heating power. The measurements were recorded in a particle-laden jet flow issuing from a long, straight pipe with well-defined inlet and co-flow conditions, with the particles heated using an axisymmetric, well-characterised infra-red radiative source generating a beam with a peak flux of up to 42.8 MW/m2 on the axis. The resulting gas-phase temperature profile increased monotonically with distance down-stream from the start of the heating region, at up to 2,200 ◦C/m on the jet centreline. Additionally, attenuation of the heating beam was shown to lead to an asymmetric temperature profile in the jet flow. The rate of increase of the gas temperature was shown to be directly proportional to both the heating flux and the time-averaged particle volumetric loading, within the range of conditions investigated. The temperature decreased significantly with an increase in particle diameter, due to the dependence of radiative and convective heat transfer processes to different exponents of the diameter. The experimental results for the temperature rise on the jet centreline were shown to match the trends from a simplified analytical model. Importantly, the model also predicts that the particle temperature is significantly greater than the gas, from the heating region to the edge of the measurement region investigated. The asymmetry of the flow temperature due to attenuation of the heating beam is also shown to increase with an increase in the particle loading and a decrease in the particle diameter (i.e., an increase in the total cross-sectional area of particles in the flow). The instantaneous distributions of both the gas-phase temperature and particle locations were demonstrated to be highly non-uniform in the radiatively heated particle-laden flow. The particle distributions were analysed using Voronoi diagrams to determine the locations of particle clusters. Void regions (i.e., with no nearby particles) were also identified. The gas-phase temperature around particles was shown to be dependent on the local particle loading, with the measured temperature inside of clusters also greater than that outside of clusters. Localised regions of relatively high or low temperature compared to their surroundings were also identified from the instantaneous images, with these regions shown to remain coherent to the downstream edge of the measurement region. The high temperature regions are shown to be typically associated with regions of high local particle-loading, while regions with low temperature are shown to be in the void regions or with a low particle loading. These results suggest that the structures in the flow are long-lived with a sufficient particle-gas temperature difference, both within the heating region and in the near-field downstream, for convection between the particles and gas to influence the gas-phase temperature field more significantly than entrainment, mixing, and convection within the gas flows.