Research On Multi-scale Structure Of Porous Media Based On Micro X-ray CT And Data-constrained Modeling

The pore and composition distribution form of porous media at different scales are important factors affecting their macroscopic physical properties.X-ray CT imaging has been widely used in the field of porous media structure characterization as a three-dimensional(3D)non-destructive characterization technique.For the 3D structure,the sample size represented by CT technology is about 10~3times the size of the voxel.The mainstream image thresholding segmentation method often lose composition distribution information which is smaller than the CT resolvable scale.This restricts the in-depth application of x-ray CT in porous media.In this paper,CT imaging combined with a data-constrained modeling(DCM)is used to characterize the 3D multi-scale structure of two nano-iron modified biomass carbon samples.The samples are used in waste water treatment and the two samples belong to pre-treatment and end-treatment respectively.Based on the multi-scale structure characterization results,a partial permeability lattice boltzmann method(PP-LBM)was used to numerically simulate the single-phase seepage behavior of water in the sample.The DCM obtains the quantitative distribution information of compositions smaller than the size of the CT voxel,which effectively expands the characterization scale of CT imaging technology.This obtains the multi-scale structure of the sample to some extent.Later,the method is extended to the multi-scale structure characterization of a jaw bones of ancient domestic pigs.The pores and iron-containing minerals in the nano-iron modified biomass carbon sample have obvious multi-scale distribution characteristics.The distribution forms of these two compositions and their changes before and after the reaction have an important effect on the waste water treatment capacity of the sample.The multi-scale structure and composition distribution of ancient biological jaws and teeth is an important basis for understanding the ancient living environment and biological eating habits.The main tasks completed in this paper are as follows:The modified biomass carbon samples before and at the end of treatment and a jaw bone of ancient domestic pigs are scanned by a micro x-ray CT.The data-constrained modeling of the two types of samples are established and the multi-scale structure information of the samples are quantitatively obtained.The 3D distribution characteristics of the composition in the modified biomass carbon samples before and at the end of the reaction are compared and analyzed.The distribution characteristics of each composition at different scales and the 3D distribution of connected clusters are quantitatively analyzed.Based on the partial permeability lattice boltzmann method,the single-phase seepage flow in the sample pores is numerically simulated.The relationship between the velocity distribution and the spatial distribution characteristics of each composition is discussed.3.The effect of different filters with different materials and thickness on the beam hardening correction during the CT imaging of the ancient pig jaw bone sample is compared.The clusters distribution morphology of the high x-ray absorption composition and pores in the sample were calculated.The 3D distribution characteristics of the composition of the sample are analyzed.The results of this paper provide a new method for the multi-scale structure characterization of porous media based on CT imaging technology.The research results on modified biomass carbon samples is helpful to understand the migration mode of each compositions during the waste water treatment process and its relationship with the water flow rate.The results provide data support for sample reaction activity evaluation and materials modification.The multi-scale structure characterization results of the jaw bones and teeth of ancient domestic pigs are a data basis for further analysis and measurement of the sample structure characteristics.For archaeological samples the presented method provide not only a non-destructive method but also a new idea of the digital modeling and archiving.