Session 1 “Forecasting landslide displacements” – 28th September 2020
Coordinators: Sabatino Cuomo (UniSa), Jean Vaunat (CIMNE), Núria M. Pinyol (UPC)
Dear Colleagues,
the Workshop is organized in the framework of the agreement between ALERT Geomaterials and the international LARAM School (LAndslide Risk Assessment and Mitigation, University of Salerno).
Geomechanics-driven models will be presented, including issues like incrementally non-linear behaviour of soil, hydro-mechanical coupling, weathering process, rate- and thermal effects. These models are widely applied for slope analysis at local scale, especially for back-analysis of failures, individuation of the key factors for failure, investigation of triggering mechanisms, modelling of post-failure and tempo-spatial evolution of slopes. Contributions about small-strain approaches for cohesive soils, large-displacements approaches for slope failures evolving into flows, or landslide propagation analysis are welcome.
Engineering-oriented procedures will also presented in the context of landslide risk theory to provide relevant information for early-warning systems, urban planning maps, countermeasures against landslide movement and land control tools operational, both at slope scale and over wide areas.
The two classes of tools are really complementary and they should be ideally integrated to effectively tackle the forecasting of landslide displacements. And, in fact the Workshop aims at reinforcing the links between the scholars coming from ALERT and LARAM communities.
DEADLINE: 30th April 2020 to send title and abstract of presentations (only abstract respecting the format suggested by ALERT, can be accepted).
SUBMISSION: please send your abstract to: scuomo@unisa.it
Best regards, see you in Aussois
Sabatino, Jean and Nuria
Session 2 “ Machine Learning and Geomechanics ” – 29th September 2020
Coordinators: Ioannis Stefanou (EC-Nantes) and Felix Darve (3SR)
Machine Learning (ML) is a promising ensemble of mathematical tools and methods that have already lead to astonishing results in science and technology.
The objective of this session is to demonstrate that machine learning can be used to bypass some of the current limitations of several experimental and numerical methods in the field of geomaterials, geomechanics and more generally in solid mechanics.
There is hope that this new approach may lead to more realistic, physics-based models with several applications in geo-energy resources (e.g. geothermal energy, oil and gaz extraction, …), in nuclear waste disposal, in CO2 sequestration, and in the prediction and prevention of natural risks (e.g. earthquakes, volcanoes, landslides, rockfalls, snow avalanches, debris flows, …).
The session will consist of invited talks only. Renowned invited speakers will share their experience on ML showing the large perspectives of the method in solid mechanics and geomechanics.
Session 3 “Bridging the gap between experiments and modelling: from laboratory testing to material models prediction” – 30th September
Coordinators: Béatrice Baudet (b.baudet@ucl.ac.uk) (University College London, U.K.), Federica Cotecchia (federica.cotecchia@poliba.it) (Politecnico di Bari, Italy), Cristina Jommi (cristina.jommi@polimi.it) (Politecnico di Milano, Italy) and (C.Jommi@tudelft.nl) (Delft University of Technology, The Netherlands)
At this point in time, when soil laboratory testing has advanced to the extent that we can determine the contact behaviour between two grains, while soils can be modelled at the microscale with realistic particle properties, it may be a good time to pause and ask: are we really aware of what aspects of the soil internal processes we are modelling? This day session attempts to address this. The morning is organized around sessions on laboratory testing providing evidence of a behavioural framework consolidated, natural and compacted soils, and assess how the different processes determine aspects of the behaviour. In the afternoon, examples of how some of that evidence is integrated in modelling are given. The morning and afternoon sessions will close with a 40-minute discussion.
The sessions will be divided into: 1. Experimental evidence of physical processes in soils and their experimental characterisation, 2. Modelling micromechanical properties and processes with discrete mechanics, and 3. Continuum modelling of micromechanical processes with macromechanics. The modelling sessions should focus on models that actively account for experimental observations.
DEADLINE AND SUBMISSION: Abstracts are invited to be submitted to the organisers by 30th April 2020 (respecting the format suggested by ALERT). Authors of selected abstracts will be invited to give a presentation within the appropriate mini-session.