Research On Characterization Of Pore-Fissure Structure And Seepage Characteristics Of Rock And Soil Materials Based On Digital Images

Vigorously developing clean new energy can effectively alleviate the energy pressure in our country and reduce environmental pollution.In the field of unconventional natural gas development,the porosity of the reservoir determines the reserves of this type of natural gas,and the gas permeability determines the level of production capacity.In the field of CO₂ geological storage,the permeability of a reservoir is an important indicator for assessing its CO₂ storage,reserves,and feasibility.In the field of nuclear safety,the permeability of the backfill material bentonite determines the safety and airtightness of the entire repository.Therefore,the physical characteristics of rock and soil materials and the physical index of permeability are of great significance and role to related engineering fields.Through theoretical research,model algorithm derivation and establishment,laboratory experiment and numerical simulation,this paper systematically studied the characterization algorithm of micro pore-crack structure based on digital image,as well as the calculation and prediction model of seepage characteristics.The validity and accuracy of the algorithm and model are verified by laboratory experiments and numerical simulations.Furthermore,by combining digital images and neural networks,a rapid permeability detection method was researched to provide scientific theoretical basis and technical support for engineering practice in related percolation fields.The specific research content is as follows:?1?By using a variety of imaging techniques to image the microstructure of a series of different geotechnical materials,the relevant image preprocessing techniques are studied.Image brightness,contrast,noise,and segmentation threshold will greatly affect the subsequent processing of the image.Through quantitative research,the adjustment method of image brightness and contrast is given.At the same time,a new threshold determination algorithm and program are independently developed,which can effectively divide the pore-fracture structure of the rock and soil material,and lay the foundation for subsequent research.?2?The pore structure in the digital image of GMZ bentonite was quantitatively characterized.The pore size distribution was obtained using DPSD and CPSD algorithms,and the permeability prediction model was developed.The size of the calculation area selected by the image has a great influence on the permeability prediction result,and the determination of the area of the characterization unit?that is,REA?is very important.The magnification of digital images also has a certain effect on the prediction of permeability.The permeability prediction result based on the CPSD algorithm is closer to the actual value,and prediction result has a certain reference value.?3?Through the independently developed THMC multi-field coupling seepage test system and gas transmission system,the gas permeability of various materials was measured experimentally,and the effectiveness of the micro-scale permeability prediction model was verified.Establishing a numerical simulation model based on the actual fracture structure is an effective method for characterizing the non-uniformity of rock and soil materials.Quantitative analysis of the effect of crack opening and tortuosity on gas permeability,while considering the slippage effect,the calculation formula obtained by the fit is in good agreement with the calculation model recognized by the academic community.?4?The appendix gives a neural network-based permeability prediction method,which can quickly and effectively predict the permeability of digital images of rock and soil materials in large quantities,providing a new technical method for related fields.