## C09 - Dynamics of rock dehydration on multiple scales

**Head(s):** Dr. Marita Thomas (WIAS), Prof. Dr. Timm John (FU Berlin)

**Project member(s):** Konstantin Huber, Lisa Kaatz, Dr. Dirk Peschka, Dr. Johannes Vrijmoed, Andrea Zafferi

**Participating institution(s):** FU Berlin, WIAS

### Project Summary

This project deals with the dynamic formation of dehydration-related fluid flow structures within rocks. Field observations of natural occurrences along with thermodynamic calculations reveal that rock dehydration is characterised by three stages:

(i) the initial formation of porosity caused by fluid liberation during dehydration reactions,

(ii) the intermediate stages of fluid pooling and vein network formation, and

(iii) the final stages of fluid release from the dehydrating system.

While the initial stage is primarily induced by chemical processes, the later stages are dominated by mechanical interactions of solid and fluid. In particular, an increase of fluid pressure causes mechanical stresses that ultimately lead to fracturing. While dehydration-associated mineral reactions are mainly driven by slow conductive heat transfer and occur on grain boundaries, hence on µm-scales, fracture-related fluid release may occur within the time scale of seismic events and up to km-scales. The goal of this project is to decipher the hierarchical structure of the interacting processes on multiple time and length scales. We will develop a thermomechanical multiscale model and validate it with geological field observations. Our approach joins mathematical analysis and fieldand laboratory-based data coupled via numerical simulations. To achieve this goal we will first focus on the scales where the transition from a chemically controlled to a deformation-controlled system can be observed. The research is structured in four workpackages (WP 1 to 4):

**WP1: “From Observation to Model”** will set up a prototype model, based on published data and concepts, featuring the coupling of fluid and heat transport with chemical reactions and with deformations and fracturing of the solid in a thermodynamically consistent way.

**WP2: “Measurements and Field Data”** aims to gather comprehensive field data on multiple scales from natural observations on serpentinite, a hydrous rock that has not experienced dehydration. This dataset will cover domain sizes ranging from µm^{2}, m^{2} up to tens of m^{2}.

**WP3: “Mathematical Analysis”** studies the well-posedness of the prototype model. Modifications of this model will be obtained by incorporating scaling parameters or by restricting certain processes to spatial, scale-dependent subdomains. Via Γ- convergence and homogenisation methods the scaled prototype model will be mathematically rigorously carried over to different scales. This will ultimately lead from the (modified) prototype model (root model) to a hierarchical multiscale model that encodes the aggregated dynamics.

**WP4: “From Model to Observation”** aims to implement the prototype model and its extensions in a numerical code. We will apply the numerical model to simulate the dehydration network pattern formation using the natural data. To validate our model, simulation output will be compared to observed network patterns developed within the same rock type that has undergone dehydration.

### Project publications

Peschka, D. and Thomas, M. and Ahnert, T. and Münch, A. and Wagner, B.
(2018)
*Gradient structures for flows of concentrated suspensions.*
SFB 1114 Preprint at WIAS 10/2018
.
(Unpublished)