The influence of slope-angle ratio on the dynamics of granular flows: insights from laboratory experiments

Laboratory experiments on granular flows using natural material were carried out in order to investigate the behaviour of granular flows passing over a break in slope. Sensors in the depositional area recorded the flow kinematics, while video footage permitted reconstruction of the deposit formation, which allowed investigation of the deposit shape as a function of the change in slope. We defined the slope-angle ratio as the proportion between slope angle in the depositional area and that of the channel. When the granular flow encounters the break in slope part of the flow front forms a bouncing clast zone due to elastic impact with the expansion box floor. During this process, part of the kinetic energy of the dense granular flow is transferred to elutriating fine ash, which subsequently forms turbulent ash cloud accompanying the granular flow until it comes to rest. Morphometric analysis of the deposits shows that they are all elliptical, with an almost constant minor axis and a variable major axis. The almost constant value of the minor axis relates to the spreading angle of flow at the end of the channel, which resembles the basal friction angle of the material. The variation of the major axis is interpreted to relate to the effect of competing inertial and frictional forces. This effect also reflects the partitioning of centripetal and tangential velocities, which changes as the flow passes over the break in slope. After normalization, morphometric data provided empirical relationships that highlight the dependence of runout from the product of slope-angle ratio and the difference in height between granular material release and deposit. The empirical relationships were tested against the runouts of hot avalanches formed during the 1944 AD eruption at Vesuvius, with differences among actual and calculated values are between 1.7 and 15 %. Velocity measurements of laboratory granular flows record deceleration paths at different breaks in slope. When normalized, the velocity data show third-order polynomial fit, highlighting a complex behaviour involving interplay between inertial and frictional forces. The theoretical velocity decays were tested against the data published for volcaniclastic debris flows of the 5-6 May 1998 event in the Sarno area. The comparison is very good for non-channelized debris flows, with significant differences between actual and calculated velocities for the channelized debris flows.