No abstract

Landslide risks arising from boulder falls and debris flows are commonly mitigated using rigid and flexible barriers. Debris-barrier interaction is a complicated process so current design methods rely on the use of the pseudo-static force approach. In addition to physical testing, numerical simulations can be used to provide insight into the impact mechanism. This paper presents the applications of numerical models to simulate rigid and flexible barriers subjected to rockfall and debris flow impacts respectively. For rigid barriers, rock-filled gabions, recycled glass cullet, cellular glass aggregates and EVA foam were assessed for their performance as cushioning materials. From the results, empirical equations were established for predicting the boulder impact forces and penetration into the cushion layer. Amongst the materials considered in this study, rock-filled gabions appear to be the most promising for use in practice. For flexible barriers, finite-element models, calibrated using documented case histories, were developed to simulate the debris-barrier interaction. The models were used to investigate the barriers' behaviour under debris impacts from both force and energy perspectives. From the results, the hydrodynamic pressure coefficient was found to be lower than the current recommended value whilst only a small amount of debris energy was transferred to the barrier.

Landslides are major natural hazards that can pose significant risks to human lives and properties in mountainous areas. Mobility assessments of landslide debris are integral to the management of landslide risks. Over the past two decades, the Geotechnical Engineering Office (GEO) in Hong Kong has expended considerable effort on the development of debris mobility models, and a suite of advanced numerical tools has been developed using different numerical techniques, including particle-in-cell, smoothed particle hydrodynamics and arbitrary Lagrangian–Eulerian methods. The accuracy of the tools was corroborated against historical landslides in Hong Kong and typical values of model parameters were established for use in routine practice. These numerical tools offer geotechnical practitioners a pragmatic and efficient approach to assess the dynamics of landslide debris in landslide hazard investigations and quantitative risk assessments. Some of the numerical tools recently developed by the GEO can also be used to assess the dynamic interaction between landslide debris and structural countermeasures, thereby allowing design optimisation. This paper presents a holistic review of the technical development and engineering applications of these numerical tools with a focus on their roles in landslide risk management in Hong Kong. It also documents the experience of Hong Kong in steering and promoting the use of numerical modelling of large ground mass deformations for the purposes of slope safety management.

The use of steel flexible barriers to mitigate landslide risk on natural hillsides is becoming common in the last decade in Hong Kong. The current design approach for this kind of barrier structure involves the adoption of the hydrodynamic load model to predict dynamic impact forces, followed by non-linear structural analyses of flexible barriers using numerical programs based on the pseudo-static method. From local guidelines, the dynamic pressure coefficient is taken as 2.0. This empirically considers the effect of impacts from boulders up to 2.0m in diameter. With a view to rationalising the design approach, a series of physical impact tests and numerical analyses was conducted to investigate the dynamic impact on flexible barriers and the resulting barrier response. The tests involved up to 9m³ of debris resisted by a 1.5m high steel barrier, conducted at the 28m long flume facility at the Kadoorie Centre, Hong Kong. Numerical modelling using computer programs LS-DYNA and NIDA-MNN was conducted to analyse the dynamic response. The study aims to evaluate the dynamic pressure coefficient and to verify the current design approach. Results indicate that a coefficient of 1.0 is in general appropriate for design purposes for debris comprising primarily water and fine-grained particles.

No abstract