| Rockfalls are always the main security concern of many engineering projects in mountainous areas.Construction of infrastructures in southwestern mountainous area is an important measure to promote the Development of Western Regions forming a new pattern under a new era.Especially,some projects being built are located in the Hengduan Mountains which are characterized by dramatically fluctuating topography,making rockfall hazard mitigation particularly prominent.However,this issue has not been solved,among which the key points include identification of dangerous rock blocks and quantitative assessment of rockfall hazard.Considering this,proposing a new analysis method is meaningful for safety construction and normal operation of infrastructures.Taking a road engineering being built in southwest mountainous area as the background,this study focuses on the development characteristics of high position dangerous rock blocks above the tunnel entrance and quantitatively evaluates the corresponding rockfall hazard based on the digital products generated by modern onsite unmanned aerial vehicle(UAV)digital photogrammetry and by means of multiple advanced algorithms and a numerical simulation method.From this,an integrated analysis framework"rockfall sources delineation–identification of dangerous rock blocks–rockfall hazard assessment"is proposed.The main research contents and findings of this study are summarized as follows:A high-resolution three dimensional(3D)rock slope surface model is established by applying a UAV-based multi-angle nap-of-the-object photography technique.The image quality and model accuracy can facilitate the procedure of rock discontinuities fine collection.A total of 2434 discontinuities are identified from the produced model,and a program named"rock discontinuity geometric parameter acquisition"is developed based on their feature points.Among all geometric parameters,a spatial sampling multiscanline method and a modified agglomerative nesting method are introduced for spacing calculation and automatic identification of discontinuity sets,respectively.The spatial sampling multiscanline method incorporates the real distribution of discontinuity traces and orientations,and reduces the sampling error caused by undulate rock surface.The modified agglomerative nesting method is independent of initial parameters,and shows good performance in partitioning the poles at the boundary of adjacent joint sets.Based on the collected geometric and mechanical information of all identified discontinuities,the general development features of discontinuities in the studied slope are summarized and the rock mass basic quality is also evaluated.Rockfall sources quantitative assessment is conducted based on a preliminary grasp of rock mass structure features,and a morphology–spatial distribution coupled approach for identification of rockfall sources in high steep fractured rock slopes is proposed.In this approach,a threshold slope angle obtained by probabilistic kinematic and block theory analyses is firstly used to delineate rockfall-prone areas.Then,rockfall trajectory simulation with consideration of the mean block size is adopted to analyze the hazard level of each rockfall-prone area.Finally,the risk levels of rockfall sources are assessed by applying a risk matrix.The results show that the planned tunnel entrance is threatened by five sources.Of these,#1,#4 and#5 are sources with moderate risk level,and#6 and#8 are sources with high risk level.A study on location of dangerous rock blocks constituting of multiple free surfaces and discontinuities is performed in the five identified sources.A new systematic procedure"block identification–geometric characterization(complex polyhedral model)–stability analysis(deterministic block theory)"for distinguishing this type block is proposed.It incorporates the actual position and real geometric information of in-situ blocks.A total of 46 blocks with volumes greater than 1 m~3 are determined by applying the proposed framework.They are mainly distributed in the rockfall source#6.Their sizes are less than 30 m~3,and their sliding surfaces belong to a joint set with SE dip direction and middle-steep dip angle.Although their sizes are limited,they are the main rockfall sources due to their high gravitational potential energy.After determination of distribution features and stability states of the former block type,a study on the development features of another block type constituting of a slope surface and multiple line-type discontinuities is performed in the five identified sources.From this,an automatic dangerous rock block searching approach based on a trace map extracted from a rock exposure is proposed.In this approach,a depth-first search algorithm is firstly used to detect cycles.Then,topology and edge vector method are used to perform block finiteness and removability analysis.Finally,limit equilibrium is adopted to conduct block stability analysis.The proposed approach considers the contribution of line-type discontinuities on block formation and integrates the geometric and mechanical modelling procedures in one algorithm,which addresses the limitations of current in-situ rock block automatic extraction methods.A total of 113finite blocks from rockfall sources#4~#6 are obtained by applying this approach.However,only three of them are unstable.Therefore,the probability of rockfall events caused by this type blocks are very low.In other words,this block type is not the main rockfall source.Taking the identified dangerous rock blocks as input parameters,rockfall hazard assessment is conducted by a 3D rockfall simulation considering the real block shapes and sizes.Simulation results of 25 real blocks show that rockfalls will reach tunnel entrance within 20~30 s after initiation.Of these,the velocity of some blocks can reach40 m/s;the general rockfall velocity is between 13 and 30 m/s;the corresponding kinetic energy is in the range of 1300~11500 k J;the jump height of some blocks can reach 15 m;the general rock jump height is in the range of 1 m and 7 m;and the normal impact force is in the range of 4500 and 30000 k N.These rockfall kinematic parameters indicate that they have very high disaster-induced ability.From this,an integrated rockfall prevention measure(clear and protective nets)is suggested for hazard mitigation. |