Daniele Giordano Discusses the Viscosity of Magmatic Liquids

Fast Moving Front Commentary, November 2011

Daniele Giordano

Article: Viscosity of magmatic liquids: A model


Authors: Giordano, D;Russell, JK;Dingwell, DB
Journal: EARTH PLANET SCI LETT, 271 (1-4): 123-134, JUL 15 2008
Addresses: Third Univ Rome, Dept Geol Sci, Largo S Leonardo Murialdo 1, I-00154 Rome, Italy.
Univ British Columbia, Volcanol & Petrol Lab, Vancouver, BC V6T 1Z4, Canada.
Univ Munich, D-80333 Munich, Germany.

Daniele Giordano talks with ScienceWatch.com and answers a few questions about this month's Fast Moving Fronts paper in the field of Geosciences.


SW: Why do you think your paper is highly cited?

Understanding and predicting the physical properties of naturally occurring melts is critical to explaining the volcanic phenomena on terrestrial planets because, ultimately, they control the formation, transport, and eruption of magma. These properties can be divided into thermodynamic and transport properties. Amongst transport properties, viscosity is the most important because it exerts the greatest control on volcanic processes. For example, viscosity is the principle factor in controlling the rise rates of magmas, the efficacy of magmatic differentiation and mixing, and whether volatile-rich magmas will flow (effusive eruption) or fragment (explosive eruption).

Melt viscosity is strongly dependent on temperature and composition and, therein, lies the challenge presented to earth scientists. The melts encountered in earth sciences are diverse in composition and form over a wide range of temperatures. Most magmatic melts are dominantly silica (SiO2) and feature SiO44- tetrahedra as the main structural units. Depending on their geological origins, these "silicate" melts have varying amounts of other oxides, including, in decreasing abundance: Al2O3, FeO(T), MgO, CaO, Na2O, K2O, TiO2, MnO, P2O5.

All terrestrial silicate melts also contain some dissolved volatile species such as H2O, F, Cl, and S species depending on the pressure (P) and temperature (T) conditions at which they form. Furthermore, these properties (i.e., composition and temperature) can be highly transient during magma transport and eruption.

Figure 1:
Main Panel. Interval of viscosity variation for the anhydrous (black circle) and hydrous (gray circles) melts used to calibrate GRD model. Inset. Compositional variation of melts for which viscosity has been measured, as illustrated by a Total Alkali vs. Silica diagram.
Main Panel. Interval of viscosity variation for the anhydrous (black circle) and hydrous (gray circles) melts used to calibrate GRD model. Inset. Compositional variation of melts for which viscosity has been measured, as illustrated by a Total Alkali vs. Silica diagram.

View/download additional figures (3).

Our research addressed this issue by developing a robust model (GRD viscosity calculator) that allows the earth science community to calculate and predict melt viscosity over the full range of temperature-composition conditions found in terrestrial of silicate melts. The model is finding usage by an extremely wide range of geoscientists (e.g., volcanologists, experimentalists, material scientists, geodynamists, petrologists, planetary scientists) and is being used in highly imaginative and unforeseen ways.

We have built a public web-based viscosity calculator which is intended to facilitate usage and diffusion of the GRD model; colleagues can upload melt compositions and the web-tool computes GRD model values of viscosity, glass transition temperature (Tg), and fragility (m) in real-time.

SW: Does it describe a new discovery, methodology, or synthesis of knowledge?

Prior to our work, the prediction of viscosity for geological melts was limited to a model published more than 30 years earlier (Shaw, 1972). The Shaw model ascribed an Arrhenian temperature dependence to melt viscosity described by log n = log no + E/T(K), where both log no and E have compositional dependence. It was calibrated with < 200 experiments spanning a wide range of temperatures (~600 to 1700oC) and viscosities (~10-0.2 to 1012 Pa s). Given the relatively sparse dataset and the remarkably small number of model parameters (N=5) the Shaw model did a superb job and served the community for long time.*

*Due to web formatting constraints, the text "log n = log no" and "log n" from above are shown below as an image:
   

However, our knowledge of the properties of geologically relevant melts has grown immensely over the intervening years mainly because of the concerted efforts and innovations of the experimentalists. These efforts expanded the melt viscometry database by making measurements over a wider range of melt compositions and temperatures and exposed the serious limitations of the Shaw model.

In particular, new experimental strategies allowed for measurement of melt viscosity at lower temperatures and the higher viscosities close to the interval of the glass transition region. These measurements demonstrate that many geologically relevant silicate melts show pronounced non-Arrhenian temperature dependence which the Shaw model is incapable of capturing. It is into this landscape that our research program was launched.

SW: Would you summarize the significance of your paper in layman's terms?

The significance of the paper is the synthesis of knowledge and an approach to the implementation of scientific concepts and progresses obtained by scientists of various disciplines in the materials science and physics relating them to the behavior of natural melts and magmas. We benefitted greatly by inheriting the culmination of over 30 years of experimental viscometry by various well-established research groups in the geosciences which allowed us to exploit ~2,000 experimental data points (Figure 1).

SW: How did you become involved in this research, and how would you describe the particular challenges, setbacks, and successes that you've encountered along the way?

"Understanding and predicting the physical properties of naturally occurring melts is critical to explaining the volcanic phenomena on terrestrial planets because, ultimately, they control the formation, transport, and eruption of magma. These properties can be divided into thermodynamic and transport properties."

The GRD model is the result of a very fruitful long-term collaboration between Italian (Third University of Rome), German (Ludwig Maximilians University and the Bayerisches GeoInstitut), and Canadian (UBC) institutes and scientists initiated during my Ph.D. studies. My current work modeling continues in the Group of Volcanology of Barcelona (GVB, CSIC, ICTJA). The first step toward the realization of this current model  was constituted by the implementation, during my Ph.D., of the first non-Arrhenian model (Giordano and Dingwell, 2003a, b) for predicting the viscosity and the fragility of anhydrous natural silicate melts over a wide range of compositions using the empirical compositional parameter (SM).

Our model adopted the well-established Vogel Fulcher Tammann (VFT) equation (log n = A+ B/(T-C)) which, via three empirical fit-parameters, provided a means of capturing the non-Arrhenian temperature dependence of silicate melts.* In parallel to that, our earliest work (Russell et al. 2002; 2003) showed the high-T limiting behavior (A) of all silicate melts to be operationally identical, implying that all compositional dependencies reside exclusively in B (pseudo-activation energy) and C (VFT temperature; Figure 2a, b, c).

*Due to web formatting constraints, the text "log n" from above are shown below as an image:

The GRD model represents the latest non-Arrhenian multicomponent model for melt viscosity. The large majority of data used to the model calibration and the scientific background derive from the experimental and theoretical research developed during my undergraduate (University of Pisa), Ph.D. (LMU, BGI), and Postdoctoral studies at the Third University of Rome and at the University of British Columbia (Killam PDF).

Our optimization strategy was to use the ~2,000 experimental observations of melt viscosity at controlled temperature and composition to simultaneously solve for a single value of A and 17 other adjustable coefficients that account for the effects of composition on the VFT parameters B and C. Our estimate for the value of A is ~10-5 Pa s which is fully consistent with theoretical estimates.

SW: Where do you see your research leading in the future?

There remain a number of outstanding scientific questions which might be addressed through our computational model. For example, the predicted variations in the model parameters B, C, or properties such as fragility due to composition (e.g., alkali or water content) are a means of testing ideas on melt speciation. What is the structural role of water relative to other structure modifying cations (Figure 2a, b, c)? Does this imply a specific and critical role of water in affecting degassing paths and eruption dynamics processes in nature?

Conversely, fragility expresses the rate at which viscosity varies with temperature and is indicative of the melts capacity for storing energy. Is it also a key parameter that can be used to explain transitions in eruptive style between effusive vs. explosive events (Figure 2d, e)? Such concepts can be tested by coupling our model for melt viscosity to the numerical models currently used to simulate the dynamics of magma ascent and eruption.

*Due to formatting constraints, the text log n = log no and log n from above are shown below as an image:

Page: < 1 | 2 >

 
 

   |   BACK TO TOP