Investigation of a screen structure for mitigating debris-flows along coastal roads

Abstract

Debris-flow is one of the major natural hazards in Norway. The mountainous terrains of Norway are exposed to debris-flow hazards in relation to extreme weather events resulting in intense and prolonged rainfall. This hazard is expected to increase with the changing climate and has been affecting human settlements and damaging infrastructures such as roads, railways, and bridges. The E39 coastal highway is one of the main corridors in the western part of Norway that is exposed to debris-flow hazards. In connection with the long-term goal to upgrade the E39 coastal highway as an improved, continuous, and ferry free coastal highway route, the Norwegian Public Roads Administration (NPRA) sponsored this Ph.D. work to assess countermeasures for the debris-flow mitigation along steep coastal terrains. Based on a broad pilot study, it was decided to focus the work on a screen-type debris-flow countermeasure for its potential to mitigate the debris-flow threats. Moreover, debris-flow mobility and impact behaviors were studied. The study was conducted using a laboratory flume model experiment and a numerical simulation. This work presents the behavior of debris-flow, performance of a screen-type debris-flow countermeasure in reducing debris-flow mobility and impact force, and performance dependency of the screen in debris-flow composition based on tests conducted on laboratory flume models. A numerical model simulation of the laboratory flume model tests and a real debris-flow case are presented. The work also presents a new method to measure impact force using a pillar (passable structure), a pore-water pressure measuring system, and a mass mixing and releasing cylinder, each of which was designed, developed, and implemented during this study. The debris-flow behavior study investigated the mobility and impact force dependency on debris-flow composition. Variations in proportions of water content and fines content, along with total flow volume, are seen to affect its behaviors. High water and fines content facilitate debris-flow mobility and result in a relatively longer run-out distance and a faster flow. The change in debris-flow mobility is influenced by the change in water content,more so than the change in the fines content. In investigating effect of debris-flow composition on the dominating stress, it was found that tests conducted by 5.4% fines content with 55% - 60% solids concentration and tests conducted by 14% fines content with 60% solids concentration exhibited flow regimes dominated by frictional stresses that most real debris-flows probably fall in. In the tests conducted on screens, the results generally showed that a screen could hinder the debris-flow mobility and its impact force. Longer screen lengths retain more debris material than shorter counterparts. In contrast, an optimum opening width of screen grids is seen to be related to the mean diameter of the debris material used, d50. In other tests conducted using this optimum opening width, run-up height on a downstream guide-wall of an underpass is more reduced when longer screen lengths are used. Longer screen also reduces the mobility of debris-flows while accumulating more debris on its surface. The debris accumulation decreases when more water and fines contents are used while debris accumulation increases when the total debris volume is increased. The screen potential is also evaluated using the accumulation process by particle image velocimetry (PIV) analyses. The screen working mechanism is observed to be a progressive layer-bylayer accumulation causing the upward shifting of the shearing layer. The speed of this shearing layer shift can describe the rate of accumulation where it is found faster when total volume and water contents are reduced. However, the slower rate of accumulation from large total volume cases is seen to result in maximum final accumulation thicknesses. The laboratory flume model tests are supplemented with a numerical investigation using a tool called the GeoFlow SPH-FD numerical model. The model is used, in this study, for simulating the debris-flow behavior and the screen performance along with a simulation of screen application to mitigate a real debris-flow event. The simulations show that the numerical model is capable of capturing and replicating the behaviors of the laboratory tests and a real debris-flow case with proper selection of its governing parameters. The laboratory test simulation results show that the turbulence factor of the Voellmy rheology relates to the solids concentration of the debris-flow. In contrast, the consolidation factor, that controls the evolution of pore-water pressure, relates to the fines content of the debris-flow. However, the basal friction angle of the numerical model is found to be low and could not fit with the conventional friction angle. A low friction angle needed to be used, which is believed to be realistic due to turbulence in mass flows. With these back-calculated parameters, the numerical model simulation is shown to reasonably replicate the performance of screens seen during the laboratory tests. The numerical model simulation of the real debris-flow event, from the coastal terrains of Norway, is conducted by incorporating the effects of entrainment (erosion) and porewater pressure evolution. The application of a screen countermeasure is demonstrated by simulating different installation and placement options in the debris-flow channel. Three consecutive screens that are strategically placed in the channel can stop the entire flow according to the numerical simulation. With this the numerical model is seen to be a promising tool in evaluating the performance of the screens in a real debris-flow case. The E39 coastal highway route and similar roads in steep terrain need effective countermeasures for the debris-flow threats. This study contributes to building insight and tools necessary to choose debris-flow countermeasures. Both the screen and the numerical model were shown to be promising tools in mitigating areas prone to debris-flows. The screen is shown to potentially reduce the mobility of the debris-flow through laboratory model investigations and numerical simulations. The knowledge obtained from the laboratory investigations, together with the simulation capability of the GeoFlow SPH-FD numerical model to capture the performance of the screen, is beneficial in debris-flows mitigation works. These knowledge help in scaling-up the size of screens to field level while 50% opening ratio or the d50 of the debris material can be used to determine the screen opening width. With this, strategically selected and placed screen(s), and perhaps combined with other countermeasure types, can serve as debris-flows mitigation method for nearby infrastructures and settlements. In the impact force study, the flow mobility and impact pressure relationship given by the hydrodynamic power-function between the Froude number, Fr = v/(p gh), and the empirical pressure coeffcient (normalized impact pressure, = F/(Apv2)), is found to serve as a bridge between real debris-flows (Fr ≤ 3) and the laboratory flume test results,nwhich typically have Fr > 3. Also, it is shown that increasing fines and water contents is seen to increase the flow mobility and to slightly decrease the normalized impact pressure. The basal pore-water pressure was successfully measured in few tests. The results show that, at peak flow height, the liquefaction level (ratio of pore-water pressure to the total pressure) at basal surface (shearing layer) can reach up to full liquefaction level. Also, the contribution of pore-water pressure to the flow mobility and its prolonged presence in debris deposit long after the debris-flow stops are observed. Moreover, the challenges and limitations of measuring pore-water pressure in an open channel flows are also shown.