Endgame

To achieve the main objective of ENDGAME, a series of fluid-dynamics experiments will be performed. We will conduct jet flow and decompression experiments, investigating the fragmentation of the magma analogue within a fast-moving gas stream and during rapid decompression, respectively. 

We will also deploy a well-established visual technique, which has been rarely used in volcanology: Schlieren shadow photography. This technique is used to capture the flow of fluids with varying density, allowing us to visualize shock-waves generated during decompression and to better distinguish bubbles and particles within the fluid. There are several setups of the Schlieren shadow photography. For our experiments we adopted the single mirror setup. A point light source is located close to the camera, so that light is collimated back as a point in front of the camera lens. A knife edge is then used to block part of the light going back to the camera. When light passes through the fluid, it is deviated according to the refractive index and the density of the fluid, and different percentages of the light is blocked by the knife edge relative to the surrounding air. This results in brighter or darker areas in the images, showing where the density contrasts are located. This setup, in combination with a high speed camera, will allow us to study the fragmentation process at a very high temporal and spatial resolution.

We performed preliminary tests aimed to prove out the configuration of the large 2D apparatus in combination with the Schlieren shadow-photography and high-speed cameras. We created a small 2D setup using either 2 parallel glass sheets separated by rubber seals and filled with a viscous liquid. The liquid was obtained by mixing hair gel and distilled water with different proportions. We also used a small spherical mirror and a high-speed camera. Air was injected into the 2D setup through a capillary tube connected either to a syringe or to a continuous gas supply and a flowmeter to control the air flow.

Preliminary results show that we can see density perturbations within both water and a viscous liquid. We can observe shock waves propagating inside large bubbles when coalescence is occurring, likely due to the rupture of the bubble wall. Our results demonstrate that we can properly set up the Schlieren shadow photography in combination with a high-speed camera and see the density perturbation within a viscous fluid and shock waves propagating within large gas bubbles. 

Starting from this setup, we implemented the new apparatus for both 3D and 2D shock-tube experiments. The shock-tube apparatus consists of a high pressure reservoir, pressurized using compressed air at different pressures (ΔP = 2, 4, 6, 8 bar), an electrovalve for rapid decompression (<0.005 s), and an open cylindrical conduit. Schlieren shadow photography was achieved with a parabolic mirror (40 cm diameter, 180 cm focal length), a 150 W fiber optic illuminator, a razor blade and a high speed camera, allowing > 30000 fps. Acoustic signals were recorded at 200 KHz using two microphones, a PCB Piezotronic microphone and a GRAS 46 DP-1 microphone above the vent.

To perform the experiments we prepared an analogue material, a shear-thinning fluid (hydroxyethyl cellulose, HEC) that is viscoelastic, transparent and it has already been used as analogue material for volcanological fluid dynamics experiments. As preliminary experiments, we also adopted a mixture of hair gel and distilled water, with different proportions to simulate viscous fluid with different viscosities. Viscosities have been also measured with a rheometer at the host institution.

We performed a series of experiments using the 3D shock-tube apparatus in combination with the high speed Schlieren shadow photography, investigating different pressure reservoir, different length and diameter of the pipe, different volumes of pressurized gas. We collected data from different acoustic sensors deployed near the shock tube, and for some experiments we also deployed a pressure sensor at the vent of the pipe. We conducted experiments with pressurized gas only, adding particles (with different shape and volume), distilled water, or different analogue materials. Below you can see an example of the results from an experiment with the 3D shock-tube apparatus using a viscous analogue material.

To better understand the dynamics of jet flow occurring during our laboratory experiments, and more in general during explosive eruptions, we performed numerical simulations with a magma ascent model. The model is implemented through the OpenFOAM framework (https://openfoam.org/), a free open source C++ platform for Computational Fluid Dynamics which also allows the use of a High Performance Computing (HPC) facility. 

The numerical model has been used to replicate some of the laboratory experiments performed with the 3D shock-tube setup. 

Overall, there is an agreement between jet flow dynamics observed from the experiments compared with that simulated with the numerical code. We can reproduce the propagation of the shock wave in the atmosphere when the pressurized gas exits the tube, and the formation of the vortex ring and of the shock cells due to supersonic jet flow.

 

We also performed numerical simulations with a new numerical model of magma ascent, investigating how different grain size distribution, particle flatness and elongation affect magma ascent dynamics. As test case scenarios, we studied explosive activity at Etna and Stromboli, and preliminary results for the Etna explosive scenario are illustrated below.

To collect data from an active volcano during explosive activity, two field campaigns at Stromboli volcano were conducted. Stromboli is a persistently active volcano producing mild explosive activity approximately every 15-30 minutes. Together with colleagues from other INGV projects (UNO and DYNAMO), we collected multiparametric data during Strombolian activity.




We simultaneously collected images of the explosive activities from a high speed camera and from a thermal camera, as well as acoustic signals from a microphone. Analyses and interpretation of this data have started but they are still in progress. Having multiparametric datasets of real explosive activities which can be compared with laboratory experiments and numerical modeling will provide a great opportunity to better understand explosive dynamics at Stromboli volcano, and in general at basaltic volcanoes.