DEM Study On The Reliability Of Image-based Measuring Method For Granular Material

Granular material exists widely in cement, metallurgical, pharmaceutical andchemical industries. It shows special properties between fluid and solid, and has greatimpact on the product quality and energy consumption. It is thus necessary to detectthe state of the material. Image-based measuring method has been widely applied inpractice due to its simple hardware configuration and abundant processing algorithms.However, what the camera can capture is actually images of a certain surface of thematerial because granular material is generally confined in rotating processequipments with high-temperature and heavy dusts. The image-based measuringtechnique assumes that the material state measured from the surface is equivalent tothat of the interior. The reliability and reasonability of such an assumption has notbeen studied and there has been little literature report on this topic.There are some methods available to study the inner state of the material such asfrozen cutting method and probe method. However, they belong to invasive anddestructive measuring methods and are not able to capture the dynamic changes of thematerial state. With the rapid development of high-performance computingtechnology, Discrete Element Method (DEM) becomes a powerful tool for the studyof granular system in confined process equipments. It has the advantages ofsimulating the track of each particle and thus provides lots of information about thematerial state. In the present work, the reliability of the image-based measuringmethod is studied by means of DEM analysis, which is great theoretical and practicalimportance. The research done in the present work and the main conclusions are:(1) Experimental set-up was established to mimic the material movement inrotating process equipment. Image processing algorithms were proposed to obtain thetwo feature variables (mixing time and repose angle). Series of image-basedmeasurement were performed and features of the material surface were obtained.(2) A DEM model for the motion process of the material was established inPFC3D and FISH environment on microscopic level (particle-level). Simulated resultsbased on the DEM model, including surface/inter images of the material, the reposeangel and the dynamic profile of the particle contacts, were verified by experimentaldata.(3) Based on the above DEM model, the motion behavior of the material wassimulated for different operating conditions and the state of the inner material were obtained. The reliability of the image-based measuring method was studied bycomparing the differences of the material surface and its interior in aspects ofcross-section images, mixing curve, mixing time and repose angles(4) It was found in the present work that the image-based measuring method formaterial state is reliable and the features measured can reflect the state of the innermaterial, under the condition that the material is slowly rotated and the motion is atsteady state. However, in the cases of high rotating speeds and dynamic stage of themotion, the measured features show great discrepancy from the state of the innermaterial (about20%). Several suggestions are then given for the improvement of thereliability of the image-based measuring method, and a perspective in this field ismade at the end of the thesis.