Multiscale analysis of physical rock properties at Stromboli Volcano: what controls the frictional properties?

Stromboli volcano, located in the north-easternmost island of the Aeolian archipelago (Southern Italy) and well known for its persistent volcanic activity, has experienced at least four sector collapses over the past 13 thousand years. The most recent activity resulted in the formation of the Sciara del Fuoco (SDF) horseshoe-shaped depression and a tectonic strain field believed to have promoted flank collapses and formed a NE / SW trending weakness zone across the SDF and the western sector of the island. The tectonic strain field interplayed with dyking and fracturing appears to control the episodes of instability and the onset of slip surfaces. This study presents new data identifying areas of damage that could promote fracturing via remote sensing and rock friction measurements taken on rocks around the SDF and the coupled “weak” zone. We have carried out a multiscale approach by integrating satellite and microscale observations with frictional tests carried out in triaxial configuration on cm scale slabs.

 

Key units have been sampled on the field (Paleostromboli, Vancori and Neostromboli) with reference to SDF and the weak zone. Direct-shear tests in triaxial configuration were carried out to explore the frictional and seismic properties using rectangular basalt slabs at 5 – 15 MPa confining pressure in dry and saturated conditions, while recording acoustic emissions (AE) via two Piezo-Electric Transducers. The sliding velocity was changed to acquire rate and state friction parameters (RSF). Preliminary results show a variation in the friction coefficient (m) between 0.55 and 0.9 with a general m decrease with increasing confining pressure and saturation. RSF parameters a-b (0.1 < a-b < 0.1) and steady state friction coefficient (m_ss) (0.6 < m_ss < 0.9) are controlled by changing sliding velocity, confinement and by the physical properties of each unit, in particular the porosity.  AE key attributes, such amplitude, frequency and duration and their evolution confirm the relation to sliding velocity, confinement and porosity. Ongoing post mortem SEM analysis are aiming to assess the impact that textural features, such porosity, crystal distribution and glass groundmass for the different units have on the evolution of crack damage and their control on the frictional properties. Quantitative crack density analysis will be carried out using the Matlab tool box FracPaQ on the microstructures to quantify fractures properties and highlight which mechanical features (for example crystals or pores) control the development of asperities/stress concentration. This finding can be related to the field scale fracture density analysis, providing quantitative support for the identification of structurally weak zones across the SDF and constraint the mechanical behaviour of the fractured zones prone to instability.