Evaluation of the methods of particle morphology characterization: CT scanning, digital imaging and light microscopy

The morphology of granular materials can be characterized two or three-dimensionally by different measurement techniques such as micro-computed tomography (μCT), static and dynamic image analysis (SIA and DIA). Because there are fundamental differences in the working principles among these techniques, their efficacy in characterizing the morphology of granular particles needs to be precisely examined but has not yet been scientifically evaluated. In this study, a systematic evaluation of the efficacy of four DIA devices and a light microscope (LM) with reference to μCT was made by selecting statistically representative samples of three standard sands which covered a wide range of particle form and angularity. Morphological data of these sands were used and probability distributions obtained from the different imaging techniques were compared with μCT results using a precise quantitative measure based on the information theory. The results revealed that most of the DIA devices chosen were capable of providing accurate measurements for the coarse descriptors such as aspect ratio in sands across a broad particle size range, whereas some devices require improvement in their camera resolution to produce accurate measurements for fine sands. Modifying device configuration and/or computational algorithm to either use the captured image with the largest area or obtain an average from multiple random projections in the order of 100 can improve the accuracy of morphology characterization in DIA devices. LM with proper imaging and image processing can provide results comparable to μCT.