Seepage erosion in layered stream bank material

Current stream restoration practices often require anthropogenic manipulation of natural field soils to reconstruct stream banks in the absence of stabilizing vegetation. For this study, researchers conducted laboratory experiments on reconstructed, non‐vegetated stream banks with layered soils experiencing seepage. The objective of the study was to determine the effect of seepage, pore water pressure, and bank geometry on erosion and bank stability of layered streambanks. The experimental design consisted of an intermediate‐size soil lysimeter packed with a sandy clay loam top soil and an underlying fine sand layer at three bank slopes (90°, 45° and 26°). Shallow groundwater flow and seepage resulted in bank failure of geometrically stable banks. Pop out failures, liquid deformation, and piping were all observed failure mechanisms in the underlying sand material, dependent on the bank angle. Groundwater seepage processes created small‐scale failures of the underlying sand leading to larger‐scale failures of the overlying sandy clay loam. The underlying sand layer eroded according to the initial bank angle and change in overburden loading. The overlying loam layer failed along linear failure planes. The gradually sloped bank (i.e. 26° slope) failed faster, hypothesized to be due to less confining pressure and greater vertical seepage forces. Researchers analyzed the laboratory experiments using the Bank Stability and Toe Erosion Model, version 4·1. The model calculated an accurate shear surface angle similar to the failure angle observed in the lysimeter tests. The model predicted failure only for the undercut 90° bank slope, and indicated stable conditions for the other geometries. Steeper initial bank slopes and undercut banks decreased the bank factor of safety. The observed failure mechanisms and measured saturation data indicated an interaction between overburden pressure, seepage forces, and bank slope on bank stability. Future bank stability modeling would benefit by incorporating lateral seepage erosion and soil liquefaction prediction calculations. Copyright © 2009 John Wiley & Sons, Ltd.     

Flow energy and channel adjustments in rills developed in loamy sand and sandy loam soils

The storms usually associated with rill development in nature are seldom prolonged, so development is often interrupted by interstorm disturbances, e.g. weathering or tillage. In laboratory simulated rainfall experiments, active rill development can be prolonged, and under these conditions typically passes through a period of intense incision, channel extension and bifurcation before reaching quasi‐stable conditions in which little form change occurs. This paper presents laboratory experiments with coarse textured soils under simulated rainfall which show how channel adjustment processes contribute to the evolution of quasi‐stability. Newly incised rills were stabilized for detailed study of links between rill configuration and flow energy. On a loamy sand, adjustment towards equilibrium occurred due to channel widening and meandering, whereas on a sandy loam, mobile knickpoints and chutes, pulsations in flow width and flow depth and changes in stream power and sediment discharge occurred as the channel adjusted towards equilibrium. The tendency of rill systems towards quasi‐stability is shown by changes in stream power values which show short‐lived minima. Differences in energy dissipation in stabilized rills indicate that minimization of energy dissipation was reached locally between knickpoints and at the downstream ends of rills. In the absence of energy gradients in knickpoints and chutes, stabilized rill sections tended toward equilibrium by establishing uniform energy expenditure. The study confirmed that energy dissipation increased with flow aspect ratio. In stabilized rills, flow acceleration reduced energy dissipation on the loamy sand but not on the sandy loam. On both soils flow deceleration tended to increase energy dissipation. Understanding how rill systems evolve towards stability is essential in order to predict how interruptions between storms may affect long‐term rill dynamics. This is essential if event‐based physical models are to become effective in predicting sediment transport on rilled hillslopes under changing weather and climatic conditions. Copyright © 2008 John Wiley and Sons, Ltd.     

The Effects of Hydrophilic Polymer and Soil Salinity on Corn Growth in Sandy and Loamy Soils

The interaction effects of different applied ratios of a hydrophilic polymer (Superab A200) (0, 0.2, 0.6% w/w) under various soil salinity levels (initial salinity, 4 and 8 ms/cm) were evaluated on available water content (AWC), biomass, and water use efficiency for corn grown in loamy sand and sandy clay loam soils. The results showed that the highest AWC was measured at the lowest soil salinity. The application of 0.6% w/w of the polymer at the lowest salinity level increased the AWC by 2.2 and 1.2 times greater than those of control in the loamy sand and sandy clay loam soils, respectively. The analysis of variance of data showed that the effect of salinity was significant on biomass and water use efficiency of corn in the loamy sand and sandy clay loam soils. The highest amounts of these traits were measured in soils with the lowest salinity level. Application of polymer at the rate of 0.6% in the loamy sand soil and at the rate of 0.2% in the sandy clay loam soil resulted in the highest aerial and root biomass and water use efficiency for corn. At these polymer rates the amounts of water use efficiency for corn were 2.6 and 1.7 times greater than those of control in the loamy sand and sandy clay loam soils, respectively. Thus, the use of hydrophilic polymer in soils especially in the sandy soils increases soil water holding capacity, yield, and water use efficiency of plant. On the other hand, decreases the negative effect of soil salinity on plant and helps for irrigation projects to succeed in arid and semi‐arid areas.     

Hydraulic conditions for rill incision under simulated rainfall: A laboratory experiment

A series of controlled laboratory experiments were conducted in order to obtain precise data on the hydraulic and sediment transport conditions during rill formation. Tests were carried out using a crusting-prone binary mixed soil in a 15 m long flume at an average slope of 0·087 under simulated rainfall. Rainfall intensities varied from 30–35 mm h^?1 and developed about 70 per cent of the kinetic energy of natural rainfall of similar intensity. Runoff and sediment discharge measured at the downstream weir were strongly influenced by rill forming processes. Essentially, rill incision reduced runoff discharge as a result of increased percolation in rill channels but increased sediment discharge. Secondary entrainment processes, such as bank collapse, also increased sediment discharge at the weir. Knickpoint bifurcation and colluvial deposition, however, decreased sediment discharge. Rills always developed through the formation of a knickpoint. The critical condition for knickpoint initiation was the development of supercritical flow and waves which mould and incise the bed. Prior smoothing of the soil surface by entrainment and redistribution of sediment facilitated supercritical flow. Statistical analysis showed that hydraulic and sediment transport conditions differed significantly in rilled and unrilled flows. The relationship between sediment discharge, rill erosion, and flow hydraulics was found to be nonlinear, conforming to a standard power function in the form y = ax^b. Rills were also associated with significantly increased sediment transport capacities. However, rill initiation was not clearly defined by any specific hydraulic threshold. Instead, rilled and unrilled flows were separated by zones of transition within which both types of flow occur.     

Spatial and temporal variations in bank erosion on sand‐bed streams in the seasonally wet tropics of northern Australia

Bank erosion rates and processes across a range of spatial scales are poorly understood in most environments, especially in the seasonally wet tropics of northern Australia where sediment yields are among global minima. A total of 177 erosion pins was installed at 45 sites on four sand‐bed streams (Tributaries North and Central, East Tributary and Ngarradj) in the Ngarradj catchment in the Alligator Rivers Region. Bank erosion was measured for up to 3·5 years (start of 1998/99 wet season to end of 2001/02 wet season) at three spatial scales, namely a discontinuous gully (0·6 km²) that was initiated by erosion of a grass swale between 1975 and 1981, a small continuous channel (2·5 km²) on an alluvial fan that was formed by incision of a formerly discontinuous channel between 1964 and 1978, and three medium‐sized, continuous channels (8·5–43·6 km²) with riparian vegetation. The bank erosion measurements during a period of average to above‐average rainfall established that substantial bank erosion occurred during the wet season on the two smaller channels by rapid lateral migration (Tributary Central) and by erosion of gully sidewalls due to a combination of within‐gully flows and overland flow plunging over the sidewalls (Tributary North). Minor bank erosion also occurred during the dry season by faunal activity, by desiccation and loss of cohesion of the sandy bank sediments and by dry flow processes. The larger channels with riparian vegetation (East Tributary and Ngarradj) did not generate significant amounts of sediment by bank erosion. Deposition (i.e. negative pin values) was locally significant at all scales. Bank profile form and channel planform exert a strong control on erosion rates during the wet season but not during the dry season. Copyright © 2006 Commonwealth Government of Australia.     

Influence of seepage undercutting on the stability of root‐reinforced streambanks

Several mechanisms contribute to streambank failure including fluvial toe undercutting, reduced soil shear strength by increased soil pore‐water pressure, and seepage erosion. Recent research has suggested that seepage erosion of noncohesive soil layers undercutting the banks may play an equivalent role in streambank failure to increased soil pore‐water pressure. However, this past research has primarily been limited to laboratory studies of non‐vegetated banks. The objective of this research was to utilize the Bank Stability and Toe Erosion Model (BSTEM) in order to determine the importance of seepage undercutting relative to bank shear strength, bank angle, soil pore‐water pressure, and root reinforcement. The BSTEM simulated two streambanks: Little Topashaw Creek and Goodwin Creek in northern Mississippi. Simulations included three bank angles (70° to 90°), four pore‐water pressure distributions (unsaturated, two partially saturated cases, and fully saturated), six distances of undercutting (0 to 40 cm), and 13 different vegetation conditions (root cohesions from 0·0 to 15·0 kPa). A relative sensitivity analysis suggested that BSTEM was approximately three to four times more sensitive to water table position than root cohesion or depth of seepage undercutting. Seepage undercutting becomes a prominent bank failure mechanism on unsaturated to partially saturated streambanks with root reinforcement, even with undercutting distances as small as 20 cm. Consideration of seepage undercutting is less important under conditions of partially to fully saturated soil pore‐water conditions. The distance at which instability by undercutting became equivalent to instability by increased soil pore‐water pressure decreased as root reinforcement increased, with values typically ranging between 20 and 40 cm at Little Topashaw Creek and between 20 and 55 cm at Goodwin Creek. This research depicts the baseline conditions at which seepage undercutting of vegetated streambanks needs to be considered for bank stability analyses. Copyright © 2008 John Wiley & Sons, Ltd.     

Modelling of the retreat process of composite riverbank in the Jingjiang Reach using the improved BSTEM

Bank retreat in the Jingjiang Reach is closely related not only to the near‐bank intensity of fluvial erosion but also to the composition and mechanical properties of bank soils. Therefore, it is necessary to correctly simulate bank retreat to determine the characteristics of fluvial processes in the Jingjiang Reach. The current version of bank stability and toe erosion model (5.4) was improved to predict riverbank retreat, by inputting a dynamic water table, and calculating the approximation of the distribution of dynamic pore water pressure in the soil near the river bank face, and considering the depositional form of the failed blocks, which is assumedly based on a triangular distribution, with the slope approximately equalling the stable submerged bank slope and half of collapsed volume deposited in the bank‐toe region. The degrees of riverbank stability at Jing34 were calculated using the improved bank stability and toe erosion model. The results indicate the following trends: (a) the degrees of riverbank stability were high during the dry season and the rising stage, which led to minimal bank failure, and (b) the stability degrees were low during the flood season and the recession stage, with the events of bank collapse occurring frequently, which belonged to a stage of intensive bank erosion. Considering the effects of bank‐toe erosion, water table lag, and the depositional form of the collapsed bank soil, the bank‐retreat process was simulated at the right riverbank of Jing34. The model‐predicted results exhibit close agreement with the measured data, including the total bank‐retreat width and the collapsed bank profile. A sensitivity analysis was conducted to determine the quantitative effects of toe erosion and water table lag on the degree of bank stability. The calculated results for toe erosion indicate that the amount of toe erosion was largest during the flood season, which was a primary reason for bank failure. The influence of water table lag on the degree of stability demonstrates that water table lag was an important cause of bank failure during the recession stage.     

Streambank dewatering for increased stability

Streambank erosion is often the dominant source of sediment leaving modified watersheds. Mass failure of high, steep banks is one of the most serious forms of streambank erosion. The risk of a given bank experiencing mass failure is a function of bank height, angle, and soil strength, which is governed by soil moisture. Two methods for bank dewatering were tested in adjacent sections of streambank bordering a deeply incised channel in northern Mississippi: a low‐cost pump system and subsurface horizontal drains. Pore water pressures (both positive and negative pressures, or matric suction) were continuously monitored for 2 years at the pumped site, at an adjacent untreated control section, and for 1 year at the site stabilized with horizontal drains. Resulting data were used to calculate a time series of the factor of safety using a computer model. Over the course of two wet seasons, average bank retreats for the control and pumped plots were 0·43 and 0·21 m, respectively. More limited monitoring revealed that the site with passive drains retreated about 0·23 m. At the pumped site pore water pressure was 3–4 kPa lower than at the control site during the most critical periods. Accordingly, computed factors of safety were above the failure threshold at the pumped site, but fell below unity at the control site on 11 occasions over the period of observation. Similarly, the drained site displayed generally lower pore water pressure and higher safety factors except for two events when drains were evidently overwhelmed with the volume of local surface and subsurface flows. These results suggest, but do not prove, that bank dewatering promoted lower rates of bank retreat and higher levels of stability since the three sites had slight differences in soils, geometry and boundary conditions. Initial cost of the dewatering treatments were significantly less than orthodox bank stabilization measures, but operation and maintenance requirements may be greater. Published in 2009 by John Wiley & Sons, Ltd.     

Measuring streambank erosion due to ground water seepage: correlation to bank pore water pressure,precipitation and stream stage

Limited information exists on one of the mechanisms governing sediment input to streams: streambank erosion by ground water seepage. The objective of this research was to demonstrate the importance of streambank composition and stratigraphy in controlling seepage flow and to quantify correlation of seepage flow/erosion with precipitation, stream stage and soil pore water pressure. The streambank site was located in Northern Mississippi in the Goodwin Creek watershed. Soil samples from layers on the streambank face suggested less than an order of magnitude difference in vertical hydraulic conductivity (K_s) with depth, but differences between lateral K_s of a concretion layer and the vertical K_s of the underlying layers contributed to the propensity for lateral flow. Goodwin Creek seeps were not similar to other seeps reported in the literature, in that eroded sediment originated from layers underneath the primary seepage layer. Subsurface flow and sediment load, quantified using 50 cm wide collection pans, were dependent on the type of seep: intermittent low‐flow (LF) seeps (flow rates typically less than 0·05 L min^?1), persistent high‐flow (HF) seeps (average flow rate of 0·39 L min^?1) and buried seeps, which eroded unconsolidated bank material from previous bank failures. The timing of LF seeps correlated to river stage and precipitation. The HF seeps at Goodwin Creek began after rainfall events resulted in the adjacent streambank reaching near saturation (i.e. soil pore water pressures greater than ?5 kPa). Seep discharge from HF seeps reached a maximum of 1·0 L min^?1 and sediment concentrations commonly approached 100 g L^?1. Buried seeps were intermittent but exhibited the most significant erosion rates (738 g min^?1) and sediment concentrations (989 g L^?1). In cases where perched water table conditions exist and persistent HF seeps occur, seepage erosion and bank collapse of streambank sediment may be significant. Copyright © 2007 John Wiley & Sons, Ltd.     

The use of small flumes for the determination of soil erodibility

The objective of this study was to examine the possibility of determining soil erodibility of loamy soils with small flumes. This was done by comparing the classification of soil erodibility obtained in the field with that obtained in the laboratory. Therefore twenty soils with a texture varying from silty loam to sandy loam were selected from the Leuven region. The erosion in the field was determined by measuring the volumetric evolution of the rill pattern. In the laboratory the soils were tested with a rainfall simulator and small flumes. The conclusion was that for loamy soils the flume experiments are a quick, simple, and reliable method for the determination of the relative soil erodibility.     

Influence of cryptogamic crust disturbance to wind erosion on sand and loam rangeland soils

A portable field wind tunnel was used to assess the sediment flux rates of loam and sand textured soils in the Mallee region of southeastern Australia. Three levels of crust disturbance (nil, moderate and severe) simulating stock trampling were investigated. The results demonstrated the importance of cryptogamic crusts in binding the soil surface and providing roughness after the soil was moderately disturbed. On the loamy soil, the crust helped maintain sediment flux rates below the erosion control target to 5 g m⁻¹ s⁻¹ for a 65 km h⁻¹ wind measured at 10 m height. Once the crust was severely disturbed, sediment fluxes increased to 1·6 times the erosion target. On the sandy soil, even with no crust disturbance the sediment flux was 1·6 times the erosion control target. Disturbing the crust increased sediment fluxes to a maximum of 6·7 times the erosion control target. Removal of the crust also decreased the threshold wind velocity that resulted in an increase to the risk of erosion from <5 per cent to 20 per cent. © 1998 John Wiley & Sons, Ltd.     

Entrainment of E. coli and Listeria monocytogenes from sediment in irrigation canal

E. coli and Listeria monocytogenes (or L. monocytogenes) are bacteria affecting fresh produce that is harmful for health of humans and animals. If these bacteria are present in surface waterbody (e.g., irrigation canals), they will impair irrigation water quality and threaten produce safety. This paper studied the resuspension of E. coli and Listeria from bed sediment into irrigation water through several sets of laboratory experiments in an open channel flume. We studied three types of sediments using several flow rates in different velocities and shear stress. Bacteria's concentration in water increases with the bed shear stress. Two empirical relations were derived to correlate the concentration of E. coli and L. monocytogenes with the dimensionless bed shear stress. The experimental data favorably verified the relationships for sandy loam, loamy sand, and loam. The results showed that both bacteria could entrain from sand more efficiently compared to other sediments (i.e., sandy loam or loam). These relationships can be applied to water quality models for simulating E. coli and L. monocytogenes transport in irrigation canals for better managing irrigation water quality.     

Effects of wet meadow riparian vegetation on streambank erosion. 1. Remote sensing measurements of streambank migration and erodibility

We quantified how rates of stream channel migration in a montane meadow vary as a function of the riparian vegetation community. The South Fork of the Kern River at Monache Meadow, located in California's southern Sierra Nevada range, supports two distinct types of vegetation: a dry meadow community dominated by sagebrush and non‐native grasses (xeric scrub and meadow), and a wet meadow community dominated by rushes and sedges (hydric graminoids). We measured rates of lateral stream migration for dry versus wet meadow reaches from aerial photographs spanning a 40‐year period (1955–1995). While stream migration rates averaged only 0·24 ± 0·02 m a^?1 in the wet meadow, the dry meadow channel migrated an average of 1·4 ± 0·3 m a^?1. We used a linear model of meander migration to calculate coefficients that characterize bank migration potential, or bank erodibility, independent of channel curvature. These calculations demonstrate that, at Monache Meadow, banks without wet meadow vegetation are roughly ten times more susceptible to erosion than banks with wet meadow vegetation. Where stream bank heights consistently exceed 1 m, low water availability creates riparian habitats dominated by dry meadow vegetation. Thus, channel incision may reduce bank stability not only by increasing bank height, but also by converting banks from wet meadow to dry meadow vegetation. Copyright © 2002 John Wiley & Sons, Ltd.     

A large weighing lysimeter for evapotranspiration and soil‐water–groundwater exchange studies

A large weighing lysimeter was installed at Yucheng Comprehensive Experimental Station, north China, for evapotranspiration and soil‐water–groundwater exchange studies. Features of the lysimeter include the following: (i) mass resolution equivalent to 0·016 mm of water to accurately and simultaneously determine hourly evapotranspiration, surface evaporation and groundwater recharge; (ii) a surface area of 3·14 m² and a soil profile depth of 5·0 m to permit normal plant development, soil‐water extraction, soil‐water–groundwater exchanges, and fluctuations of groundwater level; (iii) a special supply–drainage system to simulate field conditions of groundwater within the lysimeter; (iv) a soil mass of about 30 Mg, including both unsaturated and saturated loam. The soil consists mainly of mealy sand and light loam. Monitoring the vegetated lysimeter during the growing period of winter wheat, from October 1998 through to June 1999, indicated that during the period groundwater evaporation contributed 16·6% of total evapotranspiration for a water‐table depth from 1·6 m to 2·4 m below ground surface. Too much irrigation reduced the amount of upward water flow from the groundwater table, and caused deep percolation to the groundwater. Data from neutron probe and tensiometers suggest that soil‐water‐content profiles and soil‐water‐potential profiles were strongly affected by shallow groundwater. Copyright © 2000 John Wiley & Sons, Ltd.     

Effect of bed topography on soil aggregates transport by rill flow

After its formation, a rill may remain in the field for months, often receiving lower flow rates than the formative discharge. The objective of this work was to evaluate the rill flow transport capacity of soil aggregates at discharges unable to erode the rill, and to analyse the influence of the rill macro‐roughness on this transport process. A non‐erodible rill was built in which roughness was reproduced in detail. In order to assess only the rill macro‐roughness, a flat channel with a similar micro‐roughness to that in the rill replica was built. Rill and channel experiments were carried out at a slope of 8 and at six discharges (8·3 × 10^?5 to 5·2 × 10^?4 m³ s^?1) in the rill, and eight discharges (1·6 × 10^?5 to 5·2 × 10^?4 m³ s^?1) in the channel. Non‐erodible aggregates of three sizes (1–2, 3–5 and 5–10 mm) were released at the inlet of the rill/channel. The number of aggregates received at the outlet was registered. The number and position of the remaining aggregates along the rill/channel were also determined. The rill flow was a major sediment transport mechanism only during the formation of the rill, as during that period the power of the flow was great enough to overcome the influence of the macro‐roughness of the rill bed. At lower discharges the transport capacity in the previously formed rill was significantly less than that in the flat channel under similar slope and discharge. This was determined to be due to local slowing of flow velocities at the exit of rill pools. Copyright © 2006 John Wiley & Sons, Ltd.     

Processes and forms of an unstable alluvial system with resistant,cohesive streambeds

As a response to channelization projects undertaken near the turn of the 20th century and in the late 1960s, upstream reaches and tributaries of the Yalobusha River, Mississippi, USA, have been rejuvenated by upstream‐migrating knickpoints. Sediment and woody vegetation delivered to the channels by mass failure of streambanks has been transported downstream to form a large sediment/debris plug where the downstream end of the channelized reach joins an unmodified sinuous reach. Classification within a model of channel evolution and analysis of thalweg elevations and channel slopes indicates that downstream reaches have equilibrated but that upstream reaches are actively degrading. The beds of degrading reaches are characterized by firm, cohesive clays of two formations of Palaeocene age. The erodibility of these clay beds was determined with a jet‐test device and related to critical shear stresses and erosion rates. Repeated surveys indicated that knickpoint migration rates in these clays varied from 0·7 to 12 m a^?1, and that these rates and migration processes are highly dependent upon the bed substrate. Resistant clay beds of the Porters Creek Clay formation have restricted advancement of knickpoints in certain reaches and have caused a shift in channel adjustment processes towards bank failures and channel widening. Channel bank material accounts for at least 85 per cent of the material derived from the channel boundaries of the Yalobusha River system. Strategies to reduce downstream flooding problems while preventing upstream erosion and land loss are being contemplated by action agencies. One such proposal involves removal of the sediment/debris plug. Bank stability analyses that account for pore‐water and confining pressures have been conducted for a range of hydrologic conditions to aid in predicting future channel response. If the sediment/debris plug is removed to improve downstream drainage, care should be taken to provide sufficient time for drainage of groundwater from the channel banks so as not to induce accelerated bank failures. Published in 2002 John Wiley & Sons, Ltd.     

Hydraulics of runoff and loess loam deposition

On many more or less loamy soils, rill erosion is reported to start on slopes that are equal to or steeper than 2–3°; critical Froude numbers for the start of rill wash on these slopes vary between 2·0 and 3·0. This explains why colluvial deposition often occurs on slopes below 2–3°, when water spreads out at the downslope extremities of the rills. The critical hydraulic conditions for loess loam deposition were tested in the laboratory for slopes of 0·5° and 2°, applying unit-discharges (q) up to 10 cm²/s. It appeared from these experiments that for afterflow, without raindrop impact, deposition starts for critical load concentrations (c_cr) varying between several g/1 and about 60 g/l. Under rain c_cr amounts to a minimum value of 100–125 g/l and it increases when the runoff film becomes thinner. Nevertheless, deposition in pluvial runoff is also possible, as was the case during the Weichselian, according to data from quarries in Belgium and in The Netherlands. A modified Kalinske equation is proposed for c_cr prediction, with the introduction of a typical empirical coefficient C_r and considering such factors as shear stress and mean particle size. Massive sedimentation may occur when it stops raining and afterflow starts, since c_cr values are then much lower. It is shown from the Shields' diagram that loamy suspensions are more sensitive to sedimentation than sands in clear water.     

Bank undercutting and tension failure by groundwater seepage: predicting failure mechanisms

Groundwater seepage can lead to the erosion and failure of streambanks and hillslopes. Two groundwater instability mechanisms include (i) tension failure due to the seepage force exceeding the soil shear strength or (ii) undercutting by seepage erosion and eventual mass failure. Previous research on these mechanisms has been limited to non‐cohesive and low cohesion soils. This study utilized a constant‐head, seepage soil box packed with more cohesive (6% and 15% clay) sandy loam soils at prescribed bulk densities (1.30 to 1.70 Mg m^?3) and with a bank angle of 90° to investigate the controls on failure mechanisms due to seepage forces. A dimensionless seepage mechanism (SM) number was derived and evaluated based on the ratio of resistive cohesion forces to the driving forces leading to instability including seepage gradients with an assumed steady‐state seepage angle. Tension failures and undercutting were both observed dependent primarily on the saturated hydraulic conductivity, effective cohesion, and seepage gradient. Also, shapes of seepage undercuts for these more cohesive soils were wider and less deep compared to undercuts in sand and loamy sand soils. Direct shear tests were used to quantify the geotechnical properties of the soils packed at the various bulk densities. The SM number reasonably predicted the seepage failure mechanism (tension failure versus undercutting) based on the geotechnical properties and assumed steady‐state seepage gradients of the physical‐scale laboratory experiments, with some uncertainty due to measurement of geotechnical parameters, assumed seepage gradient direction, and the expected width of the failure block. It is hypothesized that the SM number can be used to evaluate seepage failure mechanisms when a streambank or hillslope experiences steady‐state seepage forces. When prevalent, seepage gradient forces should be considered when analyzing bank stability, and therefore should be incorporated into commonly used stability models. Copyright © 2013 John Wiley & Sons, Ltd.     

Experimental investigation on cohesionless sandy bank failure resulting from water level rising

In the last decade, sediment replenishment forming cohesionless sandy banks below dams has become an increasingly common practice in Japan to compensate for sediment deficits downstream. The erosion process of the placed cohesionless sediment is a combination of lateral toe-erosion and the following mass failure. To explore cohesionless bank failure mechanisms, a series of experiments was done in a soil tank using a compacted sandy soil mass exposed to an increasing water level. Two types of uniform sand(D_(50) = 0.40 mm and 0.17 mm) and two bank heights(50 cm and 25 cm) were used under the condition of a constant bank slope of 75°. The three dimensional(3D) geometry of the bank after failure was measured using a handheld 3D scanner. The motion of bank failure was captured using the particle image velocimetry(PIV) technique, and the matric suction was measured by tensiometers. The compacted sandy soil was eroded by loss of matric suction accompanying the rise in water level which subsequently caused rotational slide and cantilever toppling failure due to destabilization of the bank. The effect of erosion protection resulting from the slumped blocks after these failures is discussed in the light of different failure mechanisms. Tensile strength is analyzed by inverse calculation of cantilever toppling failure events. The tensile strength had non-linear relation with degree of saturation and showed a peak.The findings of the study show that it is important to incorporate the non-linear relation of tensile strength into stability analysis of cantilever toppling failure and prediction of tension crack depth within unsaturated cohesionless banks.