Ancient Collapse Research Articles

Analysis on water-inrush the process of deep excavation in karst area caused by soil internal erosion

Non-negligible internal erosion threatens deep foundation pit excavation safety and stability. A water-inrush incident in Guilin city, a typical karst area, is thoroughly analysed theoretically after revealing groundwater seepage effects on hydraulic conductivity, internal erosion degree and water-flooding probability. The distribution characteristics of the main granular soils at the site are examined. Internal erosion by suffusion of readily erodible gap-graded soil, widely distributed in situ, is specifically studied. The hydraulic conductivity of soil can be determined by the internal erosion extent, S (%), directly calculated by the effective particle size. Comparing the influences of different parameters on permeability, it is found that the formula with uniform coefficient Cu is more practical, because it is more applicable in gap-graded soil with a wide range of hydraulic conductivity, such as the permeability coefficient range of 10−6 to 101 cm/s. The permeability coefficient-based relationship between seepage velocity and hydraulic gradient is determined. By establishing and implementing a numerical model, the pore-pressure history in each excavation phase reveals that the pore pressure notably increases during excavation, especially in areas adjacent to ancient collapse bodies. Seepage flow driven by excess pore pressure potentially generates a penetrating flow upward through the overlaying soil, possibly triggering a water-inrush incident at a maximum seepage velocity of 8.0×10−1 cm/s. Several suggestions are proposed to prevent water-inrush incidents during deep foundation pit excavation in similar geological setting areas.

Three-dimensional stability analysis of a deep excavation with seepage in Karst deposit

Groundwater seepage is an important factor threatening the safety of excavation. In order to reduce the threatening, a simulation of excavation and the seepage process in Guilin city is carried out by FLAC3D. The paper establishes a three-dimensional numerical model through spline interpolation method. The model built by the system of CAD can show the scope of an ancient collapse and the topological relationship among the soil layers in small scale. After that, the fluid-solid coupling technology is used to the pore pressure calculation, which adjusts the parameter by comparing the model simulation results and the site observation results. The regular pore pressure in the excavation is obtained. The results show that the pore pressure fluctuates greatly in excavation area nearby the ancient collapse during the excavation process, the local upward seepage can be driven and flow up under the reverse hydraulic power. It provides suggestions for the excavation construction and as a reference for analyzing stable seepage of groundwater in deep excavation.

Detection of pre-eruptive seismic velocity variations at an andesitic volcano using ambient noise correlation on 3-component stations: Ubinas volcano, Peru, 2014

Volcano monitoring and eruption forecasting are based on the observation and joined interpretation of several precursory phenomena. It is thus important to detect new types of precursor and to study their relationship with forthcoming eruptions. In the last years, variations of seismic velocity have been observed in some volcanoes, mainly basaltic, before eruptions. In this paper, we look for velocity variations and waveform decorrelations before the 2014 eruptive sequence of the andesitic Ubinas volcano in Peru. We compute velocity changes by using seismic ambient noise cross-correlation (between pairs of stations) and cross-components correlation (between vertical and horizontal components of single stations), as well as coda wave interferometry of seismic multiplets. With these different approaches, we show that the major explosions that occurred from 13 to 19 April were preceded by a clear velocity decrease and waveform decorrelation. The amplitude of velocity change is generally larger on single-station cross-components correlation than on two-station cross-correlation in all the frequency ranges tested (between 0.1 and 8 Hz). We highlight an apparent anisotropy of velocity change in single-station cross-components correlation, with larger amplitudes when correlating vertical and tangential components than using vertical and radial components with respect to the crater. The Mw = 8.1 Iquique earthquake on 1 April 2014 produced also a marked co-seismic velocity drop detected in a high frequency range(3–5 Hz) in both single-station cross-components correlation and cross-correlations. We locate in the horizontal plane and in depth the velocity perturbation and the structural change related with decorrelation. During the main phase of eruptive activity, the velocity decrease at low frequency (0.1–1 Hz) appears to affect the whole edifice mainly at depth of about 1 to 3 km below the surface. The structural perturbation is more concentrated on the south flank of the volcano, a zone that corresponds to an ancient collapse. We suggest that the observed velocity variations are due to the dilatation of the edifice and to microfracturation induced by magma pressurization. The structural change may be locally enhanced by a possible zone of material weakness in the southern sector. The co-seismic velocity perturbation is located mostly in the southeast flank, at depth smaller than 0.5 to 1 km, and may be related to the presence of the hydrothermal system of the volcano.