基于集成树算法的岩石黏聚力和内摩擦角预测方法

岩石的黏聚力(c)和内摩擦角(φ)是岩石工程设计及稳定性评价的重要参数,其直接测量需通过多组三轴或剪切试验,耗时多且成本高。基于4个易获取的岩石物理力学参数(纵波波速VP、密度ρ、单轴抗压强度UCS和巴西抗拉强度BTS),构建了用于预测c和φ值的智能模型。共收集了199组含不同岩石类型的数据,采用5种集成树算法开发预测模型,使用贝叶斯优化算法对模型的超参数进行优化。模型评估结果表明:构建的模型均具有较好的预测性能,其中极端随机树模型表现最佳(测试R^(2)>0.97)。敏感性分析表明:VP、UCS和BTS对c值的预测结果影响较大,ρ对φ值的预测结果影响较大。研究成果已成功应用于金川矿区,验证了模型的实用性,开发的图形用户界面便于工程技术人员使用。

Intelligent method to experimentally identify the fracture mechanism of red sandstone

Tensile and shear fractures are significant mechanisms for rock failure.Understanding the fractures that occur in rock can reveal rock failure mechanisms.Scanning electron microscopy(SEM)has been widely used to analyze tensile and shear fractures of rock on a mesoscopic scale.To quantify tensile and shear fractures,this study proposed an innovative method composed of SEM images and deep learning techniques to identify tensile and shear fractures in red sandstone.First,direct tensile and preset angle shear tests were performed for red sandstone to produce representative tensile and shear fracture surfaces,which were then observed by SEM.Second,these obtained SEM images were applied to develop deep learning models(AlexNet,VGG13,and SqueezeNet).Model evaluation showed that VGG13 was the best model,with a testing accuracy of 0.985.Third,the features of tensile and shear fractures of red sandstone learned by VGG13 were analyzed by the integrated gradient algorithm.VGG13 was then implemented to identify the distribution and proportion of tensile and shear fractures on the failure surfaces of rock fragments caused by uniaxial compression and Brazilian splitting tests.Results demonstrated the model feasibility and suggested that the proposed method can reveal rock failure mechanisms.

Numerical simulation of microwave-induced cracking and melting of granite based on mineral microscopic models

This study introduces a coupled electromagnetic–thermal–mechanical model to reveal the mechanisms of microcracking and mineral melting of polymineralic rocks under microwave radiation.Experimental tests validate the rationality of the proposed model.Embedding microscopic mineral sections into the granite model for simulation shows that uneven temperature gradients create distinct molten,porous,and nonmolten zones on the fracture surface.Moreover,the varying thermal expansion coefficients and Young's moduli among the minerals induce significant thermal stress at the mineral boundaries.Quartz and biotite with higher thermal expansion coefficients are subjected to compression,whereas plagioclase with smaller coefficients experiences tensile stress.In the molten zone,quartz undergoes transgranular cracking due to theα–βphase transition.The local high temperatures also induce melting phase transitions in biotite and feldspar.This numerical study provides new insights into the distribution of thermal stress and mineral phase changes in rocks under microwave irradiation.

Uniaxial Compressive Strength Prediction for Rock Material in Deep Mine Using Boosting-Based Machine Learning Methods and Optimization Algorithms

Traditional laboratory tests for measuring rock uniaxial compressive strength(UCS)are tedious and timeconsuming.There is a pressing need for more effective methods to determine rock UCS,especially in deep mining environments under high in-situ stress.Thus,this study aims to develop an advanced model for predicting the UCS of rockmaterial in deepmining environments by combining three boosting-basedmachine learning methods with four optimization algorithms.For this purpose,the Lead-Zinc mine in Southwest China is considered as the case study.Rock density,P-wave velocity,and point load strength index are used as input variables,and UCS is regarded as the output.Subsequently,twelve hybrid predictive models are obtained.Root mean square error(RMSE),mean absolute error(MAE),coefficient of determination(R2),and the proportion of the mean absolute percentage error less than 20%(A-20)are selected as the evaluation metrics.Experimental results showed that the hybridmodel consisting of the extreme gradient boostingmethod and the artificial bee colony algorithm(XGBoost-ABC)achieved satisfactory results on the training dataset and exhibited the best generalization performance on the testing dataset.The values of R2,A-20,RMSE,and MAE on the training dataset are 0.98,1.0,3.11 MPa,and 2.23MPa,respectively.The highest values of R2 and A-20(0.93 and 0.96),and the smallest RMSE and MAE values of 4.78 MPa and 3.76MPa,are observed on the testing dataset.The proposed hybrid model can be considered a reliable and effective method for predicting rock UCS in deep mines.

Dynamic mechanical properties and wave propagation of composite rock-mortar specimens based on SHPB tests

Filled inclusions in rock discontinuities play a key role in the mechanical characteristics of the rock and thereby influence the stability of rock engineering. In this study, a series of impact tests were performed using a split Hopkinson pressure bar system with high-speed photography to investigate the effect of interlayer strength on the wave propagation and fracturing process in composite rock-mortar specimens.The results indicate that the transmission coefficient, nominal dynamic strength, interlayer closure, and specific normal stiffness generally increase linearly with increasing interlayer stiffness. The cement mortar layer can serve as a buffer during the deformation of composite specimens. The digital images show that tensile cracks are typically initiated at the rock-mortar interface, propagate along the loading direction, and eventually result in a tensile failure regardless of the interlayer properties. However, when a relatively weaker layer is sandwiched between the rock matrix, an increasing amount of cement mortar is violently ejected and slight slabbing occurs near the rock-mortar interface.

Numerical investigation on the tensile fracturing behavior of rock-shotcrete interface based on discrete element method

Four groups of numerical models of Brazilian tests on rock-shotcrete interfaces were successfully conducted by PFC2D. The tensile strength and Young’s modulus of shotcrete were considered. Six different undulations of rock-shotcrete interface were set up. The influences of multiple parameters on the bearing characteristics of the rock-shotcrete interface were studied. The results showed that a better support performance can be obtained by increasing the Young’s modulus of shotcrete rather than the tensile strength of shotcrete. For different tensile strength and Young’s modulus, the increase of sawtooth height has different effects on the support performance. The failure mechanism of the rock-shotcrete interfaces was analysed in detail. The stress shielding effect and stress concentration effect caused by the shape characteristics of rock-shotcrete interface were observed. The influence of these parameters on the overall support performance should be fully considered in a reasonable support design.

Effects of external dynamic disturbances and structural plane on rock fracturing around deep underground cavern

The occurrence of disasters in deep mining engineering has been confirmed to be closely related to the external dynamic disturbances and geological discontinuities.Thus,a combined finite-element method was employed to simulate the failure process of an underground cavern,which provided insights into the failure mechanism of deep hard rock affected by factors such as the dynamic stress-wave amplitudes,disturbance direction,and dip angles of the structural plane.The crack-propagation process,stress-field distribution,displacement,velocity of failed rock,and failure zone around the circular cavern were analyzed to identify the dynamic response and failure properties of the underground structures.The simulation results indicate that the dynamic disturbance direction had less influence on the dynamic response for the constant in situ stress state,while the failure intensity and damage range around the cavern always exhibited a monotonically increasing trend with an increase in the dynamic load.The crack distribution around the circular cavern exhibited an asymmetric pattern,possibly owing to the stress-wave reflection behavior and attenuation effect along the propagation route.Geological discontinuities significantly affected the stability of nearby caverns subjected to dynamic disturbances,during which the failure intensity exhibited the pattern of an initial increase followed by a decrease with an increase in the dip angle of the structural plane.Additionally,the dynamic disturbance direction led to variations in the crack distribution for specific structural planes and stress states.These results indicate that the failure behavior should be the integrated response of the excavation unloading effect,geological conditions,and external dynamic disturbances.

Rock Strength Estimation Using Several Tree-Based ML Techniques

The uniaxial compressive strength(UCS)of rock is an essential property of rock material in different relevant applications,such as rock slope,tunnel construction,and foundation.It takes enormous time and effort to obtain the UCS values directly in the laboratory.Accordingly,an indirect determination of UCS through conducting several rock index tests that are easy and fast to carry out is of interest and importance.This study presents powerful boosting trees evaluation framework,i.e.,adaptive boosting machine,extreme gradient boosting machine(XGBoost),and category gradient boosting machine,for estimating the UCS of sandstone.Schmidt hammer rebound number,P-wave velocity,and point load index were chosen as considered factors to forecast UCS values of sandstone samples.Taylor diagrams and five regression metrics,including coefficient of determination(R2),root mean square error,mean absolute error,variance account for,and A-20 index,were used to evaluate and compare the performance of these boosting trees.The results showed that the proposed boosting trees are able to provide a high level of prediction capacity for the prepared database.In particular,itwas worth noting that XGBoost is the best model to predict sandstone strength and it achieved 0.999 training R^(2) and 0.958 testing R^(2).The proposed model had more outstanding capability than neural network with optimization techniques during training and testing phases.The performed variable importance analysis reveals that the point load index has a significant influence on predicting UCS of sandstone.

Prediction of uniaxial compressive strength of rock based on lithology using stacking models

Uniaxial compressive strength(UCS)of rock is an essential parameter in geotechnical engineering.Point load strength(PLS),P-wave velocity,and Schmidt hammer rebound number(SH)are more easily obtained than UCS and are extensively applied for the indirect estimation of UCS.This study collected 1080 datasets consisting of SH,P-wave velocity,PLS,and UCS.All datasets were integrated into three categories(sedimentary,igneous,and metamorphic rocks)according to lithology.Stacking models combined with tree-based models and linear regression were developed based on the datasets of three rock types.Model evaluation showed that the stacking model combined with random forest and linear regression was the optimal model for three rock types.UCS of metamorphic rocks was less predictable than that of sedimentary and igneous rocks.Nonetheless,the proposed stacking models can improve the predictive performance for UCS of metamorphic rocks.The developed predictive models can be applied to quickly predict UCS at engineering sites,which benefits the rapid and intelligent classification of rock masses.Moreover,the importance of SH,P-wave velocity,and PLS were analyzed for the estimation of UCS.SH was a reliable indicator for UCS evaluation across various rock types.P-wave velocity was a valid parameter for evaluating the UCS of igneous rocks,but it was not reliable for assessing the UCS of metamorphic rocks.

Characteristic stress and strain precursor information for fine-grained granite during failure process under triaxial loading and unloading conditions

The unloading effect by excavation may cause irreversible and severe damage to the surrounding rock masses in underground engineering.In this paper,both conventional triaxial compression(CTC)tests and triaxial unloading confining pressure(TUCP)tests were conducted on fine-grained granite to study its triaxial compression failure processes due to unloading.Based on the crack volumetric strain(CVS)method,the crack axial strain(CAS)method and crack radial area strain(CRAS)method were proposed to identify the failure precursor information(including stress thresholds and axial strain at the initiation point of crack connectivity stage)during the rock failure processes.The results of the CTC tests show that the stable crack development stressσsd,unstable crack development stressσusd,and crack connectivity stressσct identified by the CAS method are 6%,74%–84%,and 86%–97%of the peak stress,respectively.For the TUCP cases,as the confining pressure increases,the stress thresholds,axial pressure at failure and axial strain at the start of the crack connectivity stage increase,while the time ratio of the crack connectivity stage to the entire unloading stage decreases.This indicates that fine-grained granite is prone to generate more cracks and leads to fail suddenly under high confining pressure.Furthermore,this new method demonstrates that the point at which the derivative of the radial crack area strain transitions from stable to a sudden increase or decrease is defined as the precursor point of rock failure.The results of axial strain at the starting point of the crack connectivity stage are very close to those predicted by the AE method,withβ1 no more than 11%.

Intelligent rockburst prediction model with sample category balance using feedforward neural network and Bayesian optimization

The rockburst prediction becomes more and more challenging due to the development of deep underground projects and constructions.Increasing numbers of intelligent algorithms are used to predict and prevent rockburst.This paper investigated the drawbacks of neural networks in rockburst prediction,and aimed at these shortcomings,Bayesian optimization and the synthetic minority oversampling technique+Tomek Link(SMOTETomek)were applied to efficiently develop the feedforward neural network(FNN)model for rockburst prediction.In this regard,314 real rockburst cases were collected to establish a database for modeling.The database was divided into a training set(80%)and a test set(20%).The maximum tangential stress,uniaxial compressive strength,tensile strength,stress ratio,brittleness ratio,and elastic strain energy were selected as input parameters.Bayesian optimization was implemented to find the optimal hyperparameters in FNN.To eliminate the effects of imbalanced category,SMOTETomek was adopted to process the training set to obtain a balanced training set.The FNN developed by the balanced training set received 90.48% accuracy in the test set,and the accuracy improved 12.7% compared to the imbalanced training set.For interpreting the FNN model,the permutation importance algorithm was introduced to analyze the relative importance of input variables.The elastic strain energy was the most essential variable,and some measures were proposed to prevent rockburst.To validate the practicability,the FNN developed by the balanced training set was utilized to predict rockburst in Sanshandao Gold Mine,China,and it had outstanding performance(accuracy 100%).

Tree-Based Solution Frameworks for Predicting Tunnel Boring Machine Performance Using Rock Mass and Material Properties

Tunnel Boring Machines(TBMs)are vital for tunnel and underground construction due to their high safety and efficiency.Accurately predicting TBM operational parameters based on the surrounding environment is crucial for planning schedules and managing costs.This study investigates the effectiveness of tree-based machine learning models,including Random Forest,Extremely Randomized Trees,Adaptive Boosting Machine,Gradient Boosting Machine,Extreme Gradient Boosting Machine(XGBoost),Light Gradient Boosting Machine,and CatBoost,in predicting the Penetration Rate(PR)of TBMs by considering rock mass and material characteristics.These techniques are able to provide a good relationship between input(s)and output parameters;hence,obtaining a high level of accuracy.To do that,a comprehensive database comprising various rock mass and material parameters,including Rock Mass Rating,Brazilian Tensile Strength,and Weathering Zone,was utilized for model development.The practical application of these models was assessed with a new dataset representing diverse rock mass and material properties.To evaluate model performance,ranking systems and Taylor diagrams were employed.CatBoost emerged as the most accurate model during training and testing,with R2 scores of 0.927 and 0.861,respectively.However,during validation,XGBoost demonstrated superior performance with an R2 of 0.713.Despite these variations,all tree-based models showed promising accuracy in predicting TBM performance,providing valuable insights for similar projects in the future.