Study On Angular Distributions Of Light Scattering Matrices For Typical Fire Smoke And Atmospheric Aerosol Particles

In the field of fire detection technology,the identification of fire smoke and non-fire interference particles is helpful to reduce the false alarm rate of smoke detectors based on light scattering,so that they can alarm real fires without affecting normal production and life.In the field of atmospheric aerosol monitoring,the identification of aerosol particle types based on differences in light scattering characteristics is helpful to more accurate assessments of their climate and health effects.Basical characteristics of different kinds of particles like particle size distribution,micro morphology,refractive index,and chemical composition usually have significant differences,resulting in different light scattering characteristics.In-depth study of light scattering characteristics of different kinds of particles is of great significance for the effective identification of typical fire smoke and various atmospheric aerosol particles.Therefore,light scattering characteristics of particles has been a research hotsport in recent years.Compared with scattering light intensity distribution,light scattering matrix provides more abundant information of particle scattering characteristic,which provides more advanced theoretical and technical supports for particle type identification.At present,although there have been measurement and simulation studies on light scattering matrices for several particle samples,angular distributions of light scattering matrix for cement,Asian dust and soot particles are still need to be further studied.Furthermore,the identification method for several kinds of particles needs to be further demonstrated.Therefore,in this thesis,the angular distributions of non-zero elements of scattering matrix for typical fire smoke particles(black carbon particles)and atmospheric aerosol particles(dust particles)as well as the feasibility of identifying particle types based on matrix elements were investigated.The specific research contents are as follows:(1)The particle light scattering matrix measurement apparatus was improved.Two improvements were made based on the original measurement platform.During the first improvement,the ability of measurement of matrix element ratio F33/F11 was supplemented.During the second improvement,the maximum coverage of scattering angle was extended from 160° to 175°,and the experimental apparatus can perform measurements at both 532 nm and 405 nm wavelength.The improved experimental apparatus were verified using water droplets produced by liquid aerosol generator.These water droplets are assumed to be spherical particles whose diameter follow a log-normal distribution.Calculated results based on Lorenz-Mie theory were used to fit measured scattering matrices.The good agreements between measured results of water droplets and the best fitted calculation results indicates that experimental results of improved measurement apparatus are reliable.(2)The similarities of light scattering matrices between anthropogenic origin cement dust and natural mineral dust was revealed.Dust aerosol was generated using a particle disperser,and the angular distribution of scattering matrix for cement dust was measured experimentally.SEM images show that both cement and natural dust have various irregular particle shapes.Angular distribution patterns of matrix elements for cement as well as natural dust such as volcanic ash and desert dust are rather similar.And the values of non-zero matrix elements are confined in relatively narrow regions.The angle range covered by measured scattering matrix for cement dust is only 5-160°.By combining Lorenz-Mie calculation and extrapolation based on measured results,synthetic scattering matrix for cement dust covering 0-1800 full angle range was constructed.(3)The similarities and differences of angular distribution patterns of light scattering matrices for different loess dust samples were investigated.Chinese Loess Plateau is one of the main sources of airborne dust over East Asia.Several loess samples were studied from the perspectives of loess regional variation and dust particle transmission.Measured results show that loess samples from different sampling points,samples with different size distributions,and the same sample at different wavelengths have very similar angular distributions in their light scattering matrices,while there are still some differences that cannot be ignored.Analyses of simulation results of scattering matrices for particles in literatures show that size difference can be used to qualitatively explain the distinctions in measured scattering matrices for loess particles with different size distributions,and Gaussian sphere model has certain feasibility in simulating light scattering of loess dust.Finally,an average light scattering matrix for loess dust was proposed by averaging synthetic matrices for different loess samples.(4)The discrepancies between measured light scattering matrices of soot particles and calculated results of ideal fractal aggregate model are investegated.Soot is an important detection target for photoelectric smoke detectors.Soot particles were generated separately by the combustion of toluene and n-heptane pool fire,and they were collected and dispersed into aerosols again.Measured angular distributions of light scattering matrices for these two soot samples have strong consistencies,while scattering phase functions at 532 nm and 405 nm wavelengths are significantly different.Electro mobility diameter of soot samples were measured and converted into volume equivalent diameter.Morphological parameters such as monomer size distribution,overlap factor,fractal dimension,and fractal prefactor were obtained by analyzing electron microscope images.Ideal fractal aggregate models were built using a particle generation program,and multi-sphere T matrix method was used for optical calculations at 532 nm wavelength.Comparisons between calculated light scattering matrices and experimentally measured results show that ideal fractal aggregated cannot simultaneously reproduce all the measured angular distributions of matrix elements for soot samples.(5)The identification criteria of several kinds of aerosol particles based on light scattering matrix elements was developed.Measured scattering matrices for several samples of dust,soot,smoldering smoke and water droplets at 532 nm wavelength were selected for comparisons.It can be found that these four kinds of particles can be effectively identified based on the distinctions in the angular distribution of matrix elements.The following three kinds of characteristic parameters are proposed:(a)single matrix element ratio at single scattering angle,that is,F33/F11 at 1600 angle;(b)the same matrix element ratio at different scattering angles,such as the combination of measured results of element ratio-F12/F11 at both 90° and 160°angles;(c)different matrix element ratios at the same scattering angle,such as the combination of-F12/F11 and F44/F11 at 1600 angle or the combination of-F12/F11 and F22/F11 at 1600 angle.These three kinds of characteristic parameters can also be used to identify above four kinds of particles at 405 nm wavelength,while the characteristic backscattering angle at 405 nm is 1500 when multi-angles or multiple matrix element ratios were selected.Furthermore,the values of identification criteria at 532 and 405 nm are also slightly different.The results obtained in this thesis can provide instructions and data supports for the further development of complex models for dust and soot particles in the field of fire smoke detection and atmospheric aerosol monitoration.The results can also provide characteristic parameters for the further development of identification methods for different kinds of particles.