SCARLET-1.0: SpheriCal Approximation for viRtuaL aggrEgaTes

Abstract

Abstract. Aggregation of particles occurs in a large variety of settings and is therefore the focus of many disciplines, e.g., Earth and environmental sciences, astronomy, meteorology, pharmacy, and the food industry. In particular, in volcanology, ash aggregation deeply influences the sedimentation of volcanic particles in the atmosphere during and after a volcanic eruption, affecting the accuracy of model predictions and the evaluation of hazard and risk assessments. It is thus very important to provide an exhaustive description of the outcome of an aggregation process, starting from its basic geometrical features such as the position in space of its components and the overall porosity of the final object. Here we present SCARLET-1.0, a MATLAB package specifically created to provide a 3D virtual reconstruction for volcanic ash aggregates generated in central collision processes. In centrally oriented collisions, aggregates build up their own structure around the first particle (the core), acting as a seed. This is appropriate for aggregates generated in turbulent flows in which particles show different degrees of coupling with respect to the turbulent eddies. SCARLET-1.0 belongs to the class of sphere-composite algorithms, a family of algorithms that approximate 3D complex shapes in terms of a set of sphere-composite nonoverlapping spheres. The conversion of a 3D surface to its equivalent sphere-composite structure then allows for an analytical detection of the intersections between different objects that aggregate together. Thus, provided a list of colliding sizes and shapes, SCARLET-1.0 places each element in the vector around the core, minimizing the distances between their centers of mass. The user can play with different parameters that control the minimization process. Among them the most important ones are the cone of investigation (Ω), the number of rays per cone (Nr), and the number of orientations of the object (No). All the 3D shapes are described using the Standard Triangulation Language (STL) format, which is the current standard for 3D printing. This is one of the key features of SCARLET-1.0, which results in an unlimited range of applications of the package. The main outcome of the code is the virtual representation of the object, its size, porosity, density, and the associated STL file. In addition, the object can be potentially 3D printed. As an example, SCARLET-1.0 has been applied here to the investigation of ellipsoid–ellipsoid collisions and to a more specific analysis of volcanic ash aggregation. In the first application we show that the final porosity of two colliding ellipsoids is less than 20 % if flatness and elongation are greater than or equal to 0.5. Higher values of porosities (up to 40 %–50 %) can instead be found for ellipsoids with needle-like or extremely flat shapes. In the second application, we reconstruct the evolution in time of the porosity of two different aggregates characterized by different inner structures. We find that aggregates whose population of particles is characterized by a narrow distribution of sizes tend to rapidly reach a plateau in the porosity. In addition, to reproduce the observed densities, almost no compaction is necessary in SCARLET-1.0, which is a result that suggests how ash aggregates are not well described in terms of the maximum packing condition.

First field evidence of the electrical multipolar nature of volcanic aggregates

“…Following a Lagrangian perspective, the porosity of the PCs increases as the collisional processes evolve in time. For example, Rossi and Bonadonna ( 45 ) have recently quantified that PC1 type aggregates can have a porosity that evolves in time from negligible values (e.g., <10%) up to a maximum 40% plateau for a tight packing and a 70% plateau for a loose packing. We thus expect that the electrical shielding is actually promoted in those regions of the volcanic plume far from the vent and the jet region, where the aggregate can be made by sufficiently high values of porosity (e.g., at the NBL and in the umbrella spreading).…”