This study presents a classification for subaqueous clay-laden sediment gravity flows. A series of laboratory flume experiments were performed using 9%, 15%, and 21% sediment mixture concentrations composed of sand, silt, clay, and tap water, on varying bed slopes of 6°, 8°, and 9.5°, and with discharge rates of 10 and 15 m3/hr. In addition to the characteristics of the boundary and plug layers, which have been previously used for the classification of open-channel clay-laden flows, the newly presented classification also incorporates the treatment of the free shear layer. The flow states within the boundary and free shear layers were established using calculation of the inner variable, self-similarity considerations, and the magnitude of the apparent viscosity. Based on the experimental observations four flow types were recognized: (1) a clay-rich plug flow with a laminar free shear layer, a plug layer, and a laminar boundary layer, (2) a top transitional plug flow containing a turbulent free shear layer, a plug layer, and a laminar boundary layer, (3) a transitional turbidity current with a turbulent free shear layer, no plug layer, and a laminar boundary layer, and (4) a fully turbulent turbidity current. A connection between the emplaced deposits and the relevant flow types is drawn and it is shown that a Froude number, two Reynolds numbers, and a dimensionless yield stress parameter are sufficient to associate an experimental flow type with a natural large-scale density flow.@en

A new viscoelastic constitutive model for subaqueous clay-rich gravity flows is presented. It is explained that for the materials which exhibit a minimum in their strain controlled flow curves the structure parameter must be a symmetric function of the strain rate and the stress. Therefore, the destruction of structure within the material is modeled using the dissipation energy. An expression for the elastic strain of the flowing structure is derived. The final set of equations can reproduce the viscosity bifurcation that clay suspensions may exhibit under a given load. This is explained to be important for the prediction of the run-out distance of clay-rich gravity flows. The ability of the model to reproduce the general response of pasty materials to step stress and step shear rate tests is examined. The model requires four empirical parameters. A methodology is presented for obtaining these parameters and power law functions are given for their calculations for a limited rest time of 3000 s. The ability of the model to reproduce the rheological behavior that clay-rich suspensions exhibit under both stress and strain controlled conditions is examined using rheometry tests.@en

Debris flows are driven by gravity, which in the tail region is overcome by the yield strength of the flow, forcing it to freeze. These flows are capable of achieving staggeringly large run-out distances on low gradients. The case in point, described in previous publications, is the flow which resulted in the deposit of Bed 5 of the Agadir megaslide on the north-west African margin. Debrites of this flow have been recorded several hundred kilometres away from the original landslide. Previous studies have attributed such long run-out distances to hydroplaning, low yield strength, and flow transformation. It is known that the net force acting on a volume of fluid in equilibrium is zero. In this work we show that clayladen flows are capable of approaching equilibrium. The flows which can achieve the maximum run-out distance are cohesive enough to resist some of the surrounding disturbances, that can upset the equilibrium, and reach close to equilibrium conditions, yet are dilute enough to have low viscous stress, and relatively low yield strength and lose little sediment due to deposition. A flow that is not in equilibrium will always seek to approach equilibrium conditions by speeding up or slowing down, depositing sediment, eroding the substrate, contracting in the form of the tail approaching the head, stretching, entraining water and growing in height, or dewatering and collapsing. Here we present a theory that shows that two dimensional (2D) flows in equilibrium do not grow in height. 2D flume experiments were conducted on different mixtures of kaolinite, sand, silt, and water, on varying slopes and a transitionally rough bed (sand glued), and using various discharge rates, in order to map out different stages in the evolution of a density flow from a cohesive plug flow into a turbidity current. The following flow types were observed: high density turbidity currents, plug flows, and no flow. From the velocity profiles, certain runs demonstrated close to equilibrium behaviour. For these flows, very little flow height growth and velocity variation was observed over the length of the flume. In all cases the flow appeared to be laminar within the boundary layer with Kelvin-Helmholtz instabilities at the top which were suppressed to a large extent for higher sediment concentrations. A deposit consisting of thick muddy sand, with approximately uniform thickness, was observed for higher sediment concentrations, indicating relatively higher yield strength values, while a thinner more sandy deposit was observed for more dilute flows. It was concluded that high sediment concentrations on more moderate slopes result in slower moving plug flows which are capable of suppressing turbulence at the top, while lower sediment concentrations on steeper slopes result in faster moving, more turbulent currents. The flows which can achieve the largest run-out distance are located between these two extremes. @en

Debris flows are driven by gravity, which in the tail region is overcome by the yield strength of the flow, forcing it to freeze. These flows are capable of achieving staggeringly large run-out distances on low gradients. The case in point, described in previous publications, is the flow which resulted in the deposit of Bed 5 of the Agadir megaslide on the north-west African margin. Debrites of this flow have been recorded several hundred kilometres away from the original landslide. Previous studies have attributed such long run-out distances to hydroplaning, low yield strength, and flow transformation. It is known that the net force acting on a volume of fluid in equilibrium is zero. In this work we show that clayladen flows are capable of approaching equilibrium. The flows which can achieve the maximum run-out distance are cohesive enough to resist some of the surrounding disturbances, that can upset the equilibrium, and reach close to equilibrium conditions, yet are dilute enough to have low viscous stress, and relatively low yield strength and lose little sediment due to deposition. A flow that is not in equilibrium will always seek to approach equilibrium conditions by speeding up or slowing down, depositing sediment, eroding the substrate, contracting in the form of the tail approaching the head, stretching, entraining water and growing in height, or dewatering and collapsing. Here we present a theory that shows that two dimensional (2D) flows in equilibrium do not grow in height. 2D flume experiments were conducted on different mixtures of kaolinite, sand, silt, and water, on varying slopes and a transitionally rough bed (sand glued), and using various discharge rates, in order to map out different stages in the evolution of a density flow from a cohesive plug flow into a turbidity current. The following flow types were observed: high density turbidity currents, plug flows, and no flow. From the velocity profiles, certain runs demonstrated close to equilibrium behaviour. For these flows, very little flow height growth and velocity variation was observed over the length of the flume. In all cases the flow appeared to be laminar within the boundary layer with Kelvin-Helmholtz instabilities at the top which were suppressed to a large extent for higher sediment concentrations. A deposit consisting of thick muddy sand, with approximately uniform thickness, was observed for higher sediment concentrations, indicating relatively higher yield strength values, while a thinner more sandy deposit was observed for more dilute flows. It was concluded that high sediment concentrations on more moderate slopes result in slower moving plug flows which are capable of suppressing turbulence at the top, while lower sediment concentrations on steeper slopes result in faster moving, more turbulent currents. The flows which can achieve the largest run-out distance are located between these two extremes. @en

Debris flows are driven by gravity, which in the tail region is overcome by the yield strength of the flow, forcing it to freeze. These flows are capable of achieving staggeringly large run-out distances on low gradients. The case in point, described in previous publications, is the flow which resulted in the deposit of Bed 5 of the Agadir megaslide on the north-west African margin. Debrites of this flow have been recorded several hundred kilometres away from the original landslide. Previous studies have attributed such long run-out distances to hydroplaning, low yield strength, and flow transformation. It is known that the net force acting on a volume of fluid in equilibrium is zero. In this work we show that clayladen flows are capable of approaching equilibrium. The flows which can achieve the maximum run-out distance are cohesive enough to resist some of the surrounding disturbances, that can upset the equilibrium, and reach close to equilibrium conditions, yet are dilute enough to have low viscous stress, and relatively low yield strength and lose little sediment due to deposition. A flow that is not in equilibrium will always seek to approach equilibrium conditions by speeding up or slowing down, depositing sediment, eroding the substrate, contracting in the form of the tail approaching the head, stretching, entraining water and growing in height, or dewatering and collapsing. Here we present a theory that shows that two dimensional (2D) flows in equilibrium do not grow in height. 2D flume experiments were conducted on different mixtures of kaolinite, sand, silt, and water, on varying slopes and a transitionally rough bed (sand glued), and using various discharge rates, in order to map out different stages in the evolution of a density flow from a cohesive plug flow into a turbidity current. The following flow types were observed: high density turbidity currents, plug flows, and no flow. From the velocity profiles, certain runs demonstrated close to equilibrium behaviour. For these flows, very little flow height growth and velocity variation was observed over the length of the flume. In all cases the flow appeared to be laminar within the boundary layer with Kelvin-Helmholtz instabilities at the top which were suppressed to a large extent for higher sediment concentrations. A deposit consisting of thick muddy sand, with approximately uniform thickness, was observed for higher sediment concentrations, indicating relatively higher yield strength values, while a thinner more sandy deposit was observed for more dilute flows. It was concluded that high sediment concentrations on more moderate slopes result in slower moving plug flows which are capable of suppressing turbulence at the top, while lower sediment concentrations on steeper slopes result in faster moving, more turbulent currents. The flows which can achieve the largest run-out distance are located between these two extremes. @en

Transformations of a subaqueous density flow from proximal to distal regions are investigated. A classification of these transformations based on the state of the free shear and boundary layers and existence of a plug layer during transition from a debris flow to a turbidity current is presented. A connection between the emplaced deposit by the flow and the relevant flow type is drawn through the results obtained from a series of laboratory flume experiments. These were performed using 9%, 15%, and 21% sediment mixture concentrations composed of sand, silt, clay, and tap water, on varying bed slopes of 6, 8, and 9:5, and with discharge rates of 10[m3=h] and 15[m3=h]. Stress-controlled rheometry experiments were performed on the mixtures to obtain apparent viscosity data. A classification was developed based on the imposed flow conditions, where a cohesive flow may fall within one of five distinct flow types: 1) a cohesive plug flow (PF) with a laminar free shear and boundary layers, 2) a top transitional plug flow (TTPF) containing a turbulent free shear layer, a plug layer, and a laminar boundary layer, 3) a complete transitional plug flow (CTPF) consisting of a turbulent free shear and boundary layers and a plug, 4) a transitional turbidity current (TTC) with a turbulent free shear layer and a laminar boundary layer, and, 5) a completely turbulent turbidity current (TC). During the experiments, flow type PF resulted in en masse deposition of a thick uniform ungraded muddy sand mixture, which was emplaced once the yield stress overcame the gravitational forces within the tail region of the flow. Flow type TTPF resulted in deposition of a thin ungraded basal clean sand layer during the run. This layer was covered by a muddy sand deposit from the tail. Flow type TTC did not deposit any sediment during the run. A uniform muddy sand mixture was emplaced by the tail of the flow. Flow type TC resulted in deposition of poorly sorted massive bottom sand layer. This layer was overlain by either a muddy sand mixture or a sand and silt planar lamination. Flow type CTPF was not observed during the experiments. Furthermore, it was observed that flows which are in transition from a TTC to a TTPF result in a thin bottom clean sand layer covered by a banded transitional interval. This was overlain by a muddy sand layer and a very thin clean sand layer, resulting from traction by dilute turbulent wake. In all cases a mud cap was emplaced on top of the deposit after the runs were terminated.@en

Transformations of a subaqueous density flow from proximal to distal regions are investigated. A classification of these transformations based on the state of the free shear and boundary layers and existence of a plug layer during transition from a debris flow to a turbidity current is presented. A connection between the emplaced deposit by the flow and the relevant flow type is drawn through the results obtained from a series of laboratory flume experiments. These were performed using 9%, 15%, and 21% sediment mixture concentrations composed of sand, silt, clay, and tap water, on varying bed slopes of 6, 8, and 9:5, and with discharge rates of 10[m3=h] and 15[m3=h]. Stress-controlled rheometry experiments were performed on the mixtures to obtain apparent viscosity data. A classification was developed based on the imposed flow conditions, where a cohesive flow may fall within one of five distinct flow types: 1) a cohesive plug flow (PF) with a laminar free shear and boundary layers, 2) a top transitional plug flow (TTPF) containing a turbulent free shear layer, a plug layer, and a laminar boundary layer, 3) a complete transitional plug flow (CTPF) consisting of a turbulent free shear and boundary layers and a plug, 4) a transitional turbidity current (TTC) with a turbulent free shear layer and a laminar boundary layer, and, 5) a completely turbulent turbidity current (TC). During the experiments, flow type PF resulted in en masse deposition of a thick uniform ungraded muddy sand mixture, which was emplaced once the yield stress overcame the gravitational forces within the tail region of the flow. Flow type TTPF resulted in deposition of a thin ungraded basal clean sand layer during the run. This layer was covered by a muddy sand deposit from the tail. Flow type TTC did not deposit any sediment during the run. A uniform muddy sand mixture was emplaced by the tail of the flow. Flow type TC resulted in deposition of poorly sorted massive bottom sand layer. This layer was overlain by either a muddy sand mixture or a sand and silt planar lamination. Flow type CTPF was not observed during the experiments. Furthermore, it was observed that flows which are in transition from a TTC to a TTPF result in a thin bottom clean sand layer covered by a banded transitional interval. This was overlain by a muddy sand layer and a very thin clean sand layer, resulting from traction by dilute turbulent wake. In all cases a mud cap was emplaced on top of the deposit after the runs were terminated.@en