Near Field Scattering For Particle Size Measurement

Particle sizing widely exists in many fields such as energy, chemic engineering, metallurgy, pharmacy and agriculture. It is of great significance to accurately control and measure the particle size, which is useful to enhance the performance and quality of products, to reduce energy consumption, to improve the environment and human health. Light scattering methods have been increasingly attractive due to their advantages of strong applicability, wide measurement range, high repeatability, high automation and intelligence, less interfering factors, non-invasion and on-line measurement. In this paper, the particle sizing based on near-field scattering techniques has been studied to overcome the problems of the traditional low-angle light scattering techniques, such as high intensity of the transmitted light, influence of the stray light and small scattering angle. Moreover, a novel and effective method for measuring scattering intensity is proposed.The basic concept of the near-field scattering is introduced, and then the relationship between the sizes of near-field speckle and the particle is disscussed. From a mathematical point of view, the power spectrum of the near-field speckle image based on the Fast Fourier Transform is equal to the intensity of far-field scattering light. Then the scattering light intensity distributions are analyzed under different particle sizes, refractive indexes and incident wavelengths. The scattering light intensities of different particle sizes and transfer waves are calculated as the estimated values for inversion.The independent mode algorithms are mainly introduced to inverse the particle size distribution. Because the traditional Chahine algorithm is extremely sensitive to noise, prone to generate oscillation, pseudo-peak and instability while inversing unimodal and bimodal sample, an optimized Chahine algorithm is proposed in this thesis. The SIRT algorithm is used to achieve regularization of the non-negativity constraints, and then the non-negative regularized solution is set as the initial value of Chahine algorithm to iterate the regularization equation. The unimodal and bimodal distributions, which obey Johnson-SB distribution function, are verified by simulations. The numerical results show that the optimized Chahine algorithm is better than the traditional Chahine algorithm in noisy circumstance.The particle size measurement system based on near-field scattering technique is designed and built, including a laser, a spatial filter, a microscope objective, CCD, and a computer. The selection criterions of the optical element parameters are given. Compared with the traditional forward small-angle scattering system, this system is more compact and simple. It weakens the absolute alignment of the optical path, and increases the maximum scattering angle up to 40.5 degrees. Experimental results for both unimodal and bimodal samples with known diameters (39.2μm and 67.3μm) are presented. For unimodal samples, measurement error is less than 5%, and for bimodal samples, the measurement error is about 10%.