Periodic Reporting for period 1 - STRAIN (fault STRength breAkdown and Implications for earthquake Nucleation)
Berichtszeitraum: 2017-09-14 bis 2019-09-13
Most earthquakes are generated along pre-existing faults that suddenly fail after prolonged periods of tectonic stressing. Indeed, an earthquake is generated by the imbalance between the elastic energy provided by the rocks surrounding the fault and the strength drop of the fault itself, which is degraded by progressive slip. A large number of experimental and geophysical data well characterize the first-order resistance of the rocks either before (“static” strength) or after the initiation of seismic slip (“dynamic” strength). However, there is a fundamental lack of understanding about how exactly the fault strength decrease slip (STRAIN weakening) and about the real elasticity of the rock masses around the fault, i.e. the “spring” that triggers the earthquakes. Our ability to limit the societal and economic impact of the earthquakes is strongly tied to our understanding of the physics lying behind the earthquake phenomenology.
The objective of this project is the systematic study of the changes in rock strength with progressive strain using two world-class deformation apparatuses hosted at Durham University, integrating a wide range of microstructural observations and potentially lab seismology to have insights into the microphysics of deformation. This multidisciplinary study will provide laboratory datasets to formulate new strain-dependent rheological laws applicable to the upscaling of brittle rock deformation from the laboratory to the field and, for the first time, will provide a dataset on real fault rocks elasticity at earthquake nucleation conditions. In the process, we will improve the present conceptual framework to understand the dynamics of earthquake nucleation, including the potential study of seismic precursors.
The objective of this project is the systematic study of the changes in rock strength with progressive strain using two world-class deformation apparatuses hosted at Durham University, integrating a wide range of microstructural observations and potentially lab seismology to have insights into the microphysics of deformation. This multidisciplinary study will provide laboratory datasets to formulate new strain-dependent rheological laws applicable to the upscaling of brittle rock deformation from the laboratory to the field and, for the first time, will provide a dataset on real fault rocks elasticity at earthquake nucleation conditions. In the process, we will improve the present conceptual framework to understand the dynamics of earthquake nucleation, including the potential study of seismic precursors.
The research within the STRAIN project was conducted along two main lines of research: laboratory experiments & geological fieldwork, both complemented by extensive microscopy studies.
I designed state-of-the-art friction experiments to study the role of mechanical strength evolution, i.e. the resistance to sliding in faults smulated in the laboratory. From these experiments we retrieved the sheared materials that were analyzed under optical and electron microscopes to study the evolution of the microstructures in response to the deformation (e.g. deformation/breaking of crystals, particle size evolution, extent and geometrical arrangement of deformed zones in the samples etc.).
We documented the (low) friction of important rocks commonly found in faults of the Earth's Crust, such as serpentines and foliated clays, which control the strength if fault interfaces a many levels of faults at plate boundarires. In addition, in the laboratory we simulated two fundamental modes of fault slip: i.e. sliding within rock gouges (powders) and sliding along solid bare surfaces, which are commonly observed in nature. With sliding experiments on bare rocks, we discovered that macroscopic frictional strength is extremely sensitive to surface roughness: we observed that low roughness surfaces have extremely low strength and the slip behaviour is unstable giving rise to laboratory earthquakes. This and other peculiar processes of the sliding surfaces may be strongly akin to processes of slip in shallow natural faults and open new possibilities to study the stability and regularity of slip in the rocks. On the other hand, experiments made sliding rock powders documented how friction evolves in rocks when we have progressive comminution of the grain size in the powders. We found that when powders are extremely fine, frictional behaviour becomes unstable or irregularly unstable, depending on the conditions surrounding the fault (e.g. elasticity, pressure). This allowed us to study what happens at the end of the natural processes of grinding and comminution in natural faults. We demonstrated, with the aid of microstructures, that the stable or unstable behaviour is due to the trade-off between the efficiency of brittle and viscous processes. Brittle and viscous processes are both active at the comminution limit of the rocks, even if the low ambient temperature of the fault would result, in principle, into sole brittle deformation.
In the field, we investigated several faults exhumed to the Earth's surface, including shallow faults in the NE Apennines of Italy, intermediate depth faults in the W Apennines, and deep faults in the Caledonides of NW Scotland. We described in detail the architecture of fault zones, showing the extent of damage imposed by the sliding and flowing of rocks. We documented also the deformation processes in the fault zones and the evidence of fluid pressure and migration ad depth. In particular, we recognize that dissolution-precipitation processes have a paramount role in controlling the strength of faults, the cycle of fluid pressure and the overall structure of the fault zone, at all crustal depths.
The integration of mechanical data and field observation led us to formulate a model to put forward the genesis of slow slip and arrest in tectonic faults, one of the promising avenues to study the genesis of earthquakes in the last decades. In our model, slow slip is generated by brittle shear in conjunction with dissolution-precipitation forming a network of mechanically weak platy minerals and the slowness and arrest of the process is due to the geometrical complexity of the fault zone we observe in the field.
I designed state-of-the-art friction experiments to study the role of mechanical strength evolution, i.e. the resistance to sliding in faults smulated in the laboratory. From these experiments we retrieved the sheared materials that were analyzed under optical and electron microscopes to study the evolution of the microstructures in response to the deformation (e.g. deformation/breaking of crystals, particle size evolution, extent and geometrical arrangement of deformed zones in the samples etc.).
We documented the (low) friction of important rocks commonly found in faults of the Earth's Crust, such as serpentines and foliated clays, which control the strength if fault interfaces a many levels of faults at plate boundarires. In addition, in the laboratory we simulated two fundamental modes of fault slip: i.e. sliding within rock gouges (powders) and sliding along solid bare surfaces, which are commonly observed in nature. With sliding experiments on bare rocks, we discovered that macroscopic frictional strength is extremely sensitive to surface roughness: we observed that low roughness surfaces have extremely low strength and the slip behaviour is unstable giving rise to laboratory earthquakes. This and other peculiar processes of the sliding surfaces may be strongly akin to processes of slip in shallow natural faults and open new possibilities to study the stability and regularity of slip in the rocks. On the other hand, experiments made sliding rock powders documented how friction evolves in rocks when we have progressive comminution of the grain size in the powders. We found that when powders are extremely fine, frictional behaviour becomes unstable or irregularly unstable, depending on the conditions surrounding the fault (e.g. elasticity, pressure). This allowed us to study what happens at the end of the natural processes of grinding and comminution in natural faults. We demonstrated, with the aid of microstructures, that the stable or unstable behaviour is due to the trade-off between the efficiency of brittle and viscous processes. Brittle and viscous processes are both active at the comminution limit of the rocks, even if the low ambient temperature of the fault would result, in principle, into sole brittle deformation.
In the field, we investigated several faults exhumed to the Earth's surface, including shallow faults in the NE Apennines of Italy, intermediate depth faults in the W Apennines, and deep faults in the Caledonides of NW Scotland. We described in detail the architecture of fault zones, showing the extent of damage imposed by the sliding and flowing of rocks. We documented also the deformation processes in the fault zones and the evidence of fluid pressure and migration ad depth. In particular, we recognize that dissolution-precipitation processes have a paramount role in controlling the strength of faults, the cycle of fluid pressure and the overall structure of the fault zone, at all crustal depths.
The integration of mechanical data and field observation led us to formulate a model to put forward the genesis of slow slip and arrest in tectonic faults, one of the promising avenues to study the genesis of earthquakes in the last decades. In our model, slow slip is generated by brittle shear in conjunction with dissolution-precipitation forming a network of mechanically weak platy minerals and the slowness and arrest of the process is due to the geometrical complexity of the fault zone we observe in the field.
The advances carried out with the STRAIN project mostly concern our understanding of the microphysics of brittle deformation, and in particular on the processes of localization and grain size reduction occurring during friction of rocks. We documented the friction of important rocks that were poorly (or not yet) investigated, drawing important informations about fault rheology. Moreover we made new discoveries on the frictional properties of sliding surfaces. Some of these processes (e.g. healing/restrengthening behavior) were found to control the stability of sliding (e.g. stable sliding vs. unstable slip resulting in earthquakes). These discoveries have both an impact on our understanding of fault rheology and have a technical impact on the our ability to make new experiments to study the conditions slip stability. In particular, allowed us to design new experiments to study the factors that influence slip stability (e.g. stability vs. P-T conditions or strain rate) and potentially study slip precursors in the laboratory.
On the other side of the project we provided an important wealth of data on fault structure, slip localization and processes. Together with mechanical data, all field and microstructural data derived from the study of ancient natural faults and shear zones are of paramount importance to produce physics-based numerical models. Since these models allow us to try and predict the behaviour of modern active faults, it is crucial to reduce our uncertainties identifying the correct geological models.
On the other side of the project we provided an important wealth of data on fault structure, slip localization and processes. Together with mechanical data, all field and microstructural data derived from the study of ancient natural faults and shear zones are of paramount importance to produce physics-based numerical models. Since these models allow us to try and predict the behaviour of modern active faults, it is crucial to reduce our uncertainties identifying the correct geological models.