Periodic Reporting for period 4 - CoQuake (Controlling earthQuakes)
Berichtszeitraum: 2022-09-01 bis 2023-11-30
Earthquakes are among the most deadly and economically devastating natural disasters. Today, we know that humans can cause earthquakes by injecting fluids into the Earth's crust. However, what we don't know is whether humans can control them. CoQuake has focused on addressing this question, demonstrating theoretically, numerically, and experimentally that there are ways for us to prevent earthquake instabilities on large faults. Specifically, CoQuake showed that by adjusting the fluid pressure at injection points near a fault, it is possible to induce aseismic, creep-like slip of a controlled amplitude.
In other words, CoQuake has shown that we can design a "cruise control" system for faults and prevent earthquake-like instabilities. Preventing natural earthquakes can have a huge impact on societies, and CoQuake has taken a significant step in this direction. This finding, together with other innovations in the fields of mechanics of solids, geomechanics, control theory, and experimental research, have a direct impact on both society and science.
The CoQuake project achieved groundbreaking results by developing innovative experimental setups, creating new theoretical frameworks, and advancing knowledge in various interdisciplinary fields (see below for some examples). Based on CoQuake's outcomes, new avenues for research have been opened.
In other words, CoQuake has shown that we can design a "cruise control" system for faults and prevent earthquake-like instabilities. Preventing natural earthquakes can have a huge impact on societies, and CoQuake has taken a significant step in this direction. This finding, together with other innovations in the fields of mechanics of solids, geomechanics, control theory, and experimental research, have a direct impact on both society and science.
The CoQuake project achieved groundbreaking results by developing innovative experimental setups, creating new theoretical frameworks, and advancing knowledge in various interdisciplinary fields (see below for some examples). Based on CoQuake's outcomes, new avenues for research have been opened.
CoQuake has achieved significant results by combining knowledge from various scientific fields to explore innovative ways to understand and control earthquakes. Some notable achievements include:
- Combining control theory with earthquake mechanics: CoQuake successfully merged control theory with earthquake mechanics. This pioneering combination allows us (in theory) to mitigate seismic activity.
- Paradigm shift in geomechanics and geophysics: The project introduced a novel approach, challenging the traditional need for detailed geological data before interventions for mitigating seismicity. CoQuake demonstrated the ability to stabilize and control earthquake faults systems without detailed geological information.
- Controlling self-organized critical systems: CoQuake serves as an additional example, among the limited instances, of successfully controlling self-organized critical systems, particularly those resembling earthquakes.
- Uniting artificial neural networks with thermodynamics: For the first time, CoQuake combined artificial neural networks with thermodynamics for constitutive modeling. This innovative approach accelerated simulations by several orders of magnitude and allowed the automatic identification of intrinsic state variables for complex inelastic materials.
- Reinforcement learning for earthquake control: CoQuake introduced Reinforced Learning to earthquake control, providing a new perspective within geomechanics and geophysics communities for mitigating seismicity.
- 3D-printing rock-like samples with sand: CoQuake 3D-printed rock-like samples with sand, bridging materials engineering with mechanics, geomechanics, civil engineering, geophysics, and geology. These samples were crucial for reproducing earthquake-like instabilities in our experiments, mimicking the behavior of real seismic faults in the laboratory.
- Stochastic DEM analyses for fault gouges: CoQuake used stochastic analyses and Discrete Elements to describe the frictional properties of fault gouges, revealing new emerging features of granular assemblies at the macroscale.
- Mathematical proof on strain localization: Through bifurcation and stability analysis, CoQuake mathematically proved that viscosity and inertia cannot regularize strain localization. This reinforces the necessity of using micromorphic continua in mechanics.
- Discovering new phenomena in fault gouges: CoQuake uncovered new physical phenomena during the shearing of fault gouges, including the emergence of a Portevin–Le Chatelier effect for granular materials, traveling shear bands, frictional strength regain during thermal pressurization, and migrating grain crushing.
- Expanding control theory concepts: CoQuake expanded control theory concepts of passivity to discontinuous and non-autonomous systems, providing valuable insights for the control of a broader range of complex systems.
- Laboratory stabilization of analog faults: CoQuake demonstrated how to stabilize the unstable slip of analog faults, impose controlled slip rates, and avoid earthquake-like instabilities through innovative experimental setups.
- Applying limit analysis for volcanic eruptions: Although less connected to the main objectives, CoQuake applied limit analysis from mechanics to describe instabilities during volcanic eruptions.
- Numerical platform development: The project created a numerical platform, called "Numerical Geolab", enabling fast implementation and solution of multiphysics problems involving micromorphic continua and inelastic materials.
These achievements highlight CoQuake's interdisciplinary approach, pushing the boundaries of scientific understanding and offering new perspectives on earthquake control and related phenomena. Most of these findings have been already published in scientific journals, conferences and public engagement events.
- Combining control theory with earthquake mechanics: CoQuake successfully merged control theory with earthquake mechanics. This pioneering combination allows us (in theory) to mitigate seismic activity.
- Paradigm shift in geomechanics and geophysics: The project introduced a novel approach, challenging the traditional need for detailed geological data before interventions for mitigating seismicity. CoQuake demonstrated the ability to stabilize and control earthquake faults systems without detailed geological information.
- Controlling self-organized critical systems: CoQuake serves as an additional example, among the limited instances, of successfully controlling self-organized critical systems, particularly those resembling earthquakes.
- Uniting artificial neural networks with thermodynamics: For the first time, CoQuake combined artificial neural networks with thermodynamics for constitutive modeling. This innovative approach accelerated simulations by several orders of magnitude and allowed the automatic identification of intrinsic state variables for complex inelastic materials.
- Reinforcement learning for earthquake control: CoQuake introduced Reinforced Learning to earthquake control, providing a new perspective within geomechanics and geophysics communities for mitigating seismicity.
- 3D-printing rock-like samples with sand: CoQuake 3D-printed rock-like samples with sand, bridging materials engineering with mechanics, geomechanics, civil engineering, geophysics, and geology. These samples were crucial for reproducing earthquake-like instabilities in our experiments, mimicking the behavior of real seismic faults in the laboratory.
- Stochastic DEM analyses for fault gouges: CoQuake used stochastic analyses and Discrete Elements to describe the frictional properties of fault gouges, revealing new emerging features of granular assemblies at the macroscale.
- Mathematical proof on strain localization: Through bifurcation and stability analysis, CoQuake mathematically proved that viscosity and inertia cannot regularize strain localization. This reinforces the necessity of using micromorphic continua in mechanics.
- Discovering new phenomena in fault gouges: CoQuake uncovered new physical phenomena during the shearing of fault gouges, including the emergence of a Portevin–Le Chatelier effect for granular materials, traveling shear bands, frictional strength regain during thermal pressurization, and migrating grain crushing.
- Expanding control theory concepts: CoQuake expanded control theory concepts of passivity to discontinuous and non-autonomous systems, providing valuable insights for the control of a broader range of complex systems.
- Laboratory stabilization of analog faults: CoQuake demonstrated how to stabilize the unstable slip of analog faults, impose controlled slip rates, and avoid earthquake-like instabilities through innovative experimental setups.
- Applying limit analysis for volcanic eruptions: Although less connected to the main objectives, CoQuake applied limit analysis from mechanics to describe instabilities during volcanic eruptions.
- Numerical platform development: The project created a numerical platform, called "Numerical Geolab", enabling fast implementation and solution of multiphysics problems involving micromorphic continua and inelastic materials.
These achievements highlight CoQuake's interdisciplinary approach, pushing the boundaries of scientific understanding and offering new perspectives on earthquake control and related phenomena. Most of these findings have been already published in scientific journals, conferences and public engagement events.
The CoQuake project has achieved groundbreaking progress beyond the state of the art in earthquake control and geomechanics. By integrating control theory with earthquake mechanics, the project introduced innovative methodologies and demonstrated the feasibility of seismic control without exhaustive geological data. CoQuake's achievements include controlling self-organized critical systems, integrating artificial neural networks with thermodynamics, accelerating multiscale numerical simulations by several orders of magnitude, applying reinforced learning for earthquake control, 3D-printing rock-like samples with sand, and utilizing multiphysics analyses for fault gouges. The project also made mathematical proofs for earthquake control, discovered new phenomena in fault gouges, expanded control theory concepts, stabilized analog faults in the laboratory, and developed a numerical platform (Numerical Geolab) for multiphysics problems and generalized continua. These accomplishments, disseminated through various channels, set the stage for continued advancements in (geo-)mechanics, machine learning and seismic risk mitigation.