Influence of particle density on flow behavior and deposit architecture of concentrated pyroclastic density currents over a break in slope: Insights from laboratory experiments Article uri icon

abstract

  • Geological granular flows are highly complex, gravity-driven phenomena whose different behaviors depend on the mechanical properties, density and granulometric distributions of the constituent materials. Years of research have produced significant advances in understanding transport and deposition processes in granular flows. However, the role and effects of clast densities and density contrast in a granular flow are still not fully understood. In this paper we show the effect that pumice has on dry granular flows; specifically on flow velocity and longitudinal segregation of the deposits. Our work confirms, by experimental results, field observations on pumice/lithic segregation and longer pumice runout. We report results of velocity decay and deposit architecture for a granular flow passing over a break in slope (from 38° to 4° inclination). The 30 experimental runs were carried out in a five-meter long laboratory flume equipped with a series of sensors that include laser gates and high-speed cameras (400 fps). We used two polydisperse mixtures of dacitic lithics and rhyolitic pumice in varying amounts, with Weibull and Gaussian particle size distributions. The pumice/lithic ratio changes the flow response passing over a break in slope. This effect is particularly evident starting from 10%25 of pumice volume into the flow mixture, independently of its granulometric distribution. Runout relates to mass following a power law, with an exponent close 0.2. The experiments confirm that pumice segregation affects polydispersed mixtures, similarly to what has been observed in real field deposits, where density decoupling produces lithic-enriched proximal areas and pumice-enriched distal areas. The results obtained prove that the presence of low-density materials in a dense granular flow has a strong influence on its behavior. © 2016 Elsevier B.V.

publication date

  • 2016-01-01