Pier Paolo Comida
1
, Pierre-Simon Ross
1
, Bernd Zimanowski
2
, Ralf Büttner
2
0.0% 8.7% 2.4% 78.1% 2.3% 4.0% 4.0% 0.4% 0.0% 0 100 200 300 400 500 600 700N
um
be
r o
f p
ar
tic
le
s
Sommata-dry - n = 744
Juvenile pyroclasts come in a range of sizes, shapes, surface features and internal textures (e.g., for basalt, sideromelane versus tachylite). These parameters are influenced by how magma deforms, fragments and cools, which is controlled by factors such as magma viscosity, surface tension, crystallinity, volatile content, and interaction with external water, including Molten Fuel Coolant Interactions (MFCI) and less explosive in-teractions.
To understand how both silica content and interaction with external water impact on magma fragmentation and the resulting juvenile
pyro-clasts, a series of laboratory scale experiments were performed at the Physikalisch Vulkanologisches Labor in Würzburg (Germany). In each run, 200 g of volcanic rock was re-melted to 1200°C within 1 hour using an induction furnace. The melt was fragmented and expelled from the steel crucible through the use of compressed argon injected from the base; this is known informally as a “dry blowout”. In wet blowouts, a layer of liquid water was added on top of the magma just before the start of deformation. Dry and wet blowouts were performed on three melt compositions ran-ging from olivine-melilitite (ultramafic) to basaltic trachy-andesite (intermediate); these three compositions have approximately the same
equili-brium viscosity for low shear rates at 1200°C. Dry blowouts are somewhat comparable to fire fountains in nature.
The artificial pyroclasts were collected and hand-sieved in order to obtain grain-size distribution, then different size fractions examined under the binocular microscope. Particles in the 0ϕ size fraction were assigned to different classes based on shapes and other visual features. Major differences in the particle shapes were observed between the three dry blowouts, despite using the same experimental conditions, comparable magma surface tensions, and equilibrium viscosities. Therefore, instantaneous viscosity and non-Newtonian behaviour probably plays a role in controlling particle shapes in lava fountains and other eruptive styles.
Exp. parameters
• Mass of volcanic material: 200 g, granulated
• Water volume (wet blow-out only): 100 ml • Driving pressure: 500 cm3 Argon at 10 MPa • Magma temperature: 1200°C
Spheres
Tears
Lobate/Ovoid
Elongate
Angular/Blocky,
sharp edges
Irregulars,
smooth edges
Platy/"Wings"/
"Folded glass"
Bubbles
Crystals
Legend
23.5% 10.9% 26.3% 10.4% 0.4% 14.7% 13.4% 0.2% 0.2% 0 50 100 150 200 250 300 350 400 N um be r o f p ar tic le s 25.5% 13.8% 28.9% 6.6% 0.9% 10.3% 12.9% 1.0% 0.1% 0 50 100 150 200 250 300 350 400 N um be r o f p ar tic le s 0.8% 28.0% 9.7% 50.2% 2.8% 8.3% 0.1% 0.1% 0.0% 0 100 200 300 400 500 600 N um be r o f p ar tic le s 0.1% 16.4% 4.8% 49.2% 2.1% 10.9% 16.1% 0.4% 0.0% 0 50 100 150 200 250 300 350 400 N um be r o f p ar tic le sBilstein-wet - n = 752
0.0% 4.0% 1.1% 56.6% 1.5% 15.8% 20.1% 0.7% 0.0% 0 100 200 300 400 500 600 700 800 N um be r o f p ar tic le sSommata-wet - n = 1241
Hohenstoffeln-dry - n = 1321
Hohenstoffeln-wet - n = 1294
Bilstein-dry - n = 1028
Foidite Ba-salticande-site Ande-site
Dacite Basalt Pcb Trachyte/ trachydacite Rhyolite TrBas Trachy-andesite Teph/bas PhT TPh Phonolite BTrA 52 41 57 63 0 5 15 10 35 45 55 65 75
SiO
2(wt %)
Na
2O + K
2O (wt %)
Composition Area Information and age SiO2
(wt %) Na2O + K2O (wt %) Equilibrium viscosity (Pa∙s)
Baden-Württemberg, Germany Hohenstoffeln lava lake (Miocene, 15 Ma) 38.1 3.89 73
Olivine-melilitite
Alkaline basalt Bilstein, Brilon, Germany Continental lava flows (mid-Devonian, ??) 45.3 4.52 50
Basaltic
trachy-andesite Vulcano, Italy La Sommata scoria cone (Pleistocene, 99 - 20 ka) 54.4 6.16 64
100 10 0.01 0.1 1 10 [Pa ∙s ] [S-1] Hohenstoffelna Bilsteinb La Sommataa
Equilibrium viscosities of the three tested magma compo-sitions at 1200°C, showing the strong dependence on the shear rate. (a) data from Hobiger et al., 2011. (b) data from
Sonder et al., 2006. 0 10 20 30 40 -5 -4 -3 -2 -1 0 1 2 3 4 Pan
No
rm
al
ize
d
W
t %
PhiHohenstoffeln-dry
0 10 20 30 40 -5 -4 -3 -2 -1 0 1 2 3 4 PanNo
rm
al
ize
d
W
t %
PhiHohenstoffeln-wet
0 10 20 30 40 -5 -4 -3 -2 -1 0 1 2 3 4 PanNo
rm
al
ize
d
W
t %
Phi 0 10 20 30 40 50 -5 -4 -3 -2 -1 0 1 2 3 4 PanNo
rm
al
ize
d
W
t %
Phi 0 10 20 30 40 -5 -4 -3 -2 -1 0 1 2 3 4 PanNo
rm
al
ize
d
W
t %
Phi 0 10 20 30 40 -5 -4 -3 -2 -1 0 1 2 3 4 PanNo
rm
al
ize
d
W
t %
PhiBilstein-dry
Bilstein-wet
Sommata-wet
Sommata-dry
Geochemistry and equivalent viscosity
Experiment set-up
Abstract
Discussion and preliminary conclusions
References
Particle shape proportions (0ɸ)
Grain-size distribution of the artificial pyroclasts
Hobiger, M., Sonder, I., Büttner, R. and Zimanowski, B., 2011. Viscosity characteristics of selected volcanic rock melts. Journal of Volcanolo-gy and Geothermal Research, 200(1-2): 27-34.
Shimozuru, D., 1994. Physical parameters governing the formation of Pele's hair and tears. Bulletin of Volcanology, 56(3): 217-219.
Sonder, I., Zimanowski, B. and Büttner, R., 2006. Non-Newtonian visco-sity of basaltic magma. Geophysical Research Letters, 33(2).
Walker, D. and Mullins, O., 1981. Surface tension of natural silicate melts from 1,200°–1,500° C and implications for melt structure. Contributions to Mineralogy and Petrology, 76(4): 455-462.
950 cm
50 cm
8 mm
Argon
gas
tank
Ar solenoidPressure
transducer
Force transducerWater-cooled
base support
under the
crucible
Metal plateCrucible
Base plate
Melt Water solenoid Water 5 cm internal diameterDrawing not to scale 0
5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm 0 5 10 15 20 25 30 35 40 cm
Hohenstoffeln
Bilstein
Sommata
72.5 ms
after opening of the Ar solenoid75 ms
after77.5 ms
after70.5 ms
after opening of the Ar solenoid73 ms
after75.5 ms
after74.5 ms
after opening of the Ar solenoid77 ms
after79.5 ms
afterParticle shape - SEM (4ϕ)
2 mm 2 mm 2 mm 2 mm 2 mm 2 mm
Particle shape - binocular (0ɸ)
Hohenstoffeln - dry
Bilstein - dry
Sommata - dry
Hohenstoffeln - wet
Bilstein - wet
Sommata - wet
• Dry hand-sieving of
each sample,
separat-ing the sseparat-ingle size
frac-tions up to 4ɸ (63 μm)
• Observation of all
size fractions at the
bin-ocular microscope
• 0ɸ (1 mm) fraction
selected for i) particle
shape analysis and
counting at the
binocu-lar; ii) internal features
in thin sections
• 4ɸ (63 μm) fraction
selected for surface
features study and
geochemistry at
SEM-EDS
Methods
Sphere Lobate Tear Angular Elongate Platy Irregular, smooth edges Bubble• Three magma compositions with similar equilibrium viscosities were tested under the same experimental conditions
• Important differences in particle shapes were observed, for the dry blowouts: ◦ The ultramafic magma produced mostly spheres, lobate clasts and tears ◦ The intermediate magma produced mostly elongate particles
◦ The mafic magma produced particles with transitional characteristics
• Experimental results on ink jets (Shimozuru, 1994) linked the production of Pele’s hairs versus Pele’s tears as dependent on several parameters such as magma viscosity (ƞ),
magma and air density (ρ0, ρ), droplet velocity (ν) and surface tension (σ), all represented by
Pele number (Pe), defined by Pe = ρ∙ν∙ƞ / ρ0∙σ. Pele's hair are produced by larger (Pe) and
Pele's tears for smaller (Pe) (Shimozuru, 1994)
• The parameters defining Pele number initially appear to be the same for all dry blowouts, so identical particles would be expected regardless of magma composition. But this erroneously assumes that equilibrium viscosity of the magma is the relevant parameter
• Yet instantaneous viscosity, which is directly applicable to what happens in the crucible and air during the experiments, is thought to be much lower for ultramafic melts due to their
non-Newtonian behavior (shear thinning). This allows the reshaping of the ultramafic melt drops to configurations such as spheres and tears
• In contrast, Newtonian behavior shown by our intermediate magma leads to a clear preva-lence of elongated fragments under the same experimental conditions
• Finally, fast interaction of the melt with water during the wet runs seems to have a role in the formation of platy particles, which appear for the mafic and intermediate compositions.
• In summary, non-Newtonian behavior probably has a major influence in controlling particle shapes in lava fountains and other eruptive styles
Crystal
Hohenstoffeln - dry
Hohenstoffeln - dry
1
Institut National de la Recherche Scientifique, 490 rue de la Couronne, Québec (Qc), G1K 9A9, Canada;
2Physikalisch-Vulkanologisches Labor, Universität Würzburg, Pleicherwall 1, Würzburg, 97070, Germany
Artificial juvenile pyroclasts from wet and dry “eruptions”: impact of magma composition on
grain sizes and particle shapes
PHYSIKALISCH VULKANOLOGISCHES LABOR Universität Würzburg Germany