In order to study the distribution laws and types of plastic zone of surrounding rock in large-span roadway, we analyzed the distribution laws with different spans and lateral pressures using FLAC3D numerical calcu- l...In order to study the distribution laws and types of plastic zone of surrounding rock in large-span roadway, we analyzed the distribution laws with different spans and lateral pressures using FLAC3D numerical calcu- lation software. Based on the roadway support difficulty and distribution laws of the plastic zone of sur- rounding rock, we defined the large-span roadway and classified the types of large-span rectangular roadways. As a result, the distribution laws of the plastic zone on surrounding rock in a rectangular roadway with different spans and lateral pressures were obtained. The results show that the area of the plastic zone on surrounding rock increased with the increase of the spans and lateral pressures, and the plastic zone was symmetrical to the center line of roadway. At λ=0.5, 1.0, 1.5, and 2.0, the plastic zone presented ＂addle- shape＂ distribution, ＂ellipse＂ distribution, ＂thin and high＂ distribution and ＂inverse trapezium＂ distribu- tion, respectively. In addition, we classified the roadways into four types according to the different lateral pressures and spans, including small-span, moderate-span, large-span and extreme-large-span roadways.展开更多
The vertical profiles of longshore currents have been examined experimentally over plane and barred beaches. In most cases, the vertical profiles of longshore currents are expressed by the logarithmic law. The power l...The vertical profiles of longshore currents have been examined experimentally over plane and barred beaches. In most cases, the vertical profiles of longshore currents are expressed by the logarithmic law. The power law is not commonly used to describe the profile of longshore currents. In this paper, however, a power-type formula is proposed to describe the vertical profiles of longshore currents. The formula has two parameters: the power law index (a) and the depth-averaged velocity. Based on previous studies, power law indices were set as a=1/10 and a=1/7. Depth-averaged velocity can be obtained through measurement. The fitting of the measured velocity profiles to a=1/10 and a=1/7 was assessed for the vertical longshore profiles. The vertical profile of longshore currents is well described by the power-type formula with a=1/10 for a plane beach. However, for a barred beach, different values of a needed to be used for different regions. For the region from the bar trough to the offshore side of the bar crest, the vertical profiles of longshore currents given by the power-type formula with a=1/10 and a=1/7 fit the data well. However, the fit was slightly better with a=1/10 than that with a=1/7. For the data over the trough region of cross-shore distribution of the depth-averaged longshore currents, the power formula with a=1/3 provided a good fit. The formulas with a=1/10 and a=1/7 were further examined using published data from four sources covering laboratory and field experiments. The results indicate that the power-type formula fits the data well for the laboratory and field data with a=1/10.展开更多
This paper introduces a deep learning workflow to predict phase distributions within complex geometries during two-phase capillary-dominated drainage.We utilize subsamples from Computerized Tomography(CT)images of roc...This paper introduces a deep learning workflow to predict phase distributions within complex geometries during two-phase capillary-dominated drainage.We utilize subsamples from Computerized Tomography(CT)images of rocks and incorporate pixel size,interfacial tension,contact angle,and pressure as inputs.First,an efficient morphology-based simulator creates a diverse dataset of phase distributions.Then,two commonly used convolutional and recurrent neural networks are explored and their deficiencies are highlighted,particularly in capturing phase connectivity.Subsequently,we develop a Higher-Dimensional Vision Transformer(HD-ViT)that drains pores solely based on their size,with phase connectivity enforced as a post-processing step.This enables inference for images of varying sizes,resolutions,and inlet-outlet setup.After training on a massive dataset of over 9.5 million instances,HD-ViT achieves excellent performance.We demonstrate the accuracy and speed advantage of the model on new and larger sandstone and carbonate images.We further evaluate HD-ViT against experimental fluid distribution images and the corresponding Lattice-Boltzmann simulations,producing similar outcomes in a matter of seconds.In the end,we train and validate a 3D version of the model.展开更多
基金Financial supports are from the National Natural Science Foun-dation of China (No. 50874104)the Scientific Research Indus-trialization Project of Jiangsu Universities (No. JH07-023)
文摘In order to study the distribution laws and types of plastic zone of surrounding rock in large-span roadway, we analyzed the distribution laws with different spans and lateral pressures using FLAC3D numerical calcu- lation software. Based on the roadway support difficulty and distribution laws of the plastic zone of sur- rounding rock, we defined the large-span roadway and classified the types of large-span rectangular roadways. As a result, the distribution laws of the plastic zone on surrounding rock in a rectangular roadway with different spans and lateral pressures were obtained. The results show that the area of the plastic zone on surrounding rock increased with the increase of the spans and lateral pressures, and the plastic zone was symmetrical to the center line of roadway. At λ=0.5, 1.0, 1.5, and 2.0, the plastic zone presented ＂addle- shape＂ distribution, ＂ellipse＂ distribution, ＂thin and high＂ distribution and ＂inverse trapezium＂ distribu- tion, respectively. In addition, we classified the roadways into four types according to the different lateral pressures and spans, including small-span, moderate-span, large-span and extreme-large-span roadways.
基金supported by Science and Technology Support Program of Fujian Province,China(Grant No.2015Y0035)the National Natural Science Foundation of China(Grant No.10672034)
文摘The vertical profiles of longshore currents have been examined experimentally over plane and barred beaches. In most cases, the vertical profiles of longshore currents are expressed by the logarithmic law. The power law is not commonly used to describe the profile of longshore currents. In this paper, however, a power-type formula is proposed to describe the vertical profiles of longshore currents. The formula has two parameters: the power law index (a) and the depth-averaged velocity. Based on previous studies, power law indices were set as a=1/10 and a=1/7. Depth-averaged velocity can be obtained through measurement. The fitting of the measured velocity profiles to a=1/10 and a=1/7 was assessed for the vertical longshore profiles. The vertical profile of longshore currents is well described by the power-type formula with a=1/10 for a plane beach. However, for a barred beach, different values of a needed to be used for different regions. For the region from the bar trough to the offshore side of the bar crest, the vertical profiles of longshore currents given by the power-type formula with a=1/10 and a=1/7 fit the data well. However, the fit was slightly better with a=1/10 than that with a=1/7. For the data over the trough region of cross-shore distribution of the depth-averaged longshore currents, the power formula with a=1/3 provided a good fit. The formulas with a=1/10 and a=1/7 were further examined using published data from four sources covering laboratory and field experiments. The results indicate that the power-type formula fits the data well for the laboratory and field data with a=1/10.
基金supported by the International Cooperation Programme of Chengdu City(No.2020-GH02-00023-HZ)。
文摘This paper introduces a deep learning workflow to predict phase distributions within complex geometries during two-phase capillary-dominated drainage.We utilize subsamples from Computerized Tomography(CT)images of rocks and incorporate pixel size,interfacial tension,contact angle,and pressure as inputs.First,an efficient morphology-based simulator creates a diverse dataset of phase distributions.Then,two commonly used convolutional and recurrent neural networks are explored and their deficiencies are highlighted,particularly in capturing phase connectivity.Subsequently,we develop a Higher-Dimensional Vision Transformer(HD-ViT)that drains pores solely based on their size,with phase connectivity enforced as a post-processing step.This enables inference for images of varying sizes,resolutions,and inlet-outlet setup.After training on a massive dataset of over 9.5 million instances,HD-ViT achieves excellent performance.We demonstrate the accuracy and speed advantage of the model on new and larger sandstone and carbonate images.We further evaluate HD-ViT against experimental fluid distribution images and the corresponding Lattice-Boltzmann simulations,producing similar outcomes in a matter of seconds.In the end,we train and validate a 3D version of the model.
