(SCHREFLER) INTL. SYMP. ON GEOMECHANICS AND APPLICATIONS FOR SUSTAINABLE DEVELOPMENT

The prediction of the behavior of cementitious materials and concrete structures under severe conditions and/or for long time spans is of paramount importance in civil, environmental and nuclear engineering. Often, commercial tools do not provide a sufficiently accurate response, so it is necessary to use more sophisticated approaches.
In this work, a general framework for the simulation of the non-linear behavior of concrete is shown and described. It is based on the mechanics of multiphase porous media. The mathematical model is developed by writing the relevant balance equations for the constituents at the pore scale, i.e. the local form of governing equations formulated at micro-scale, and by upscaling these equations to the macroscopic scale, taking into account thermodynamic constraints according to the so-called TCAT (Thermodynamics Constrained Averaging Theory) which assures that all the thermodynamics are properly up scaled from the micro to the macro level. Thanks to this approach, all the relevant quantities involved are thermodynamically correct, no unwanted dissipations are generated, and both the bulk phases and interfaces are taken into account. This procedure does not exclude, however, the use of a numerical multiscale approach in the formulation of the material properties. The numerical solution is obtained directly at the macro level by discretizing the governing equations in their final form.
The resulting model can be usefully applied to several practical cases: evaluation of the concrete's performance at early stages of maturing massive structures [1-3], structural repair works [2,3], exposure of concrete to high temperatures, e.g. during fire [4,5], cementitious materials subject to freezing/thawing cycles [6], etc.
In this work, the general model focuses on the specific situations described above and several examples are shown.

Superconductivity is currently exploited in several technological applications, from small-scale electronic devices to large-scale particle accelerators and fusion reactors. High field magnet technologies are still based on the use of low temperature superconducting (LTS) materials, either NbTi or Nb₃Sn. Nb₃Sn cables are brittle and strain sensitive [1]-[3], but they perform better than NbTi ones. Actual challenges given by nuclear fusion and high-energy physics require more and more performing materials, capable to transport high current densities at high temperature and at very high magnetic fields [4]. High Temperature Superconducting (HTS) materials are nowadays considered as possible candidates for such demanding conditions. These materials are thus named because they exhibit superconducting behavior at much higher temperatures than NbTi and Nb₃Sn. Among the various HTS concepts, the coated conductors, also referred to as rare-earth-barium-copper-oxide (Rare-Earth1Ba2Cu3O7-x) or REBCO tapes, are promising competitors. The coated conductor tapes have exhibited high current carrying capability under high magnetic field and good mechanical properties that meet the specific requirements in the superconducting motors and coils [5] – [7]. HTS magnets, however, are a new technology. If the LTS technology is well established, a robust R&D is needed to explore the possible use of high-temperature superconductors in high field magnets, as the superconductors performances are not only influenced by the magnetic field and the operating temperature, but also by the mechanical strain [8]–[9]. As a consequence, the development of these new generations of conductors requires extensive investigation about the impact of the main characteristics of the cable architecture on the electrical performances of a single superconducting tape. In particular, for a proper conductor design, it is important to fully characterize the single tape in its working conditions. In this work, a coupled thermo-electro-mechanical model is developed, suitable to analyze the behavior of HTS tapes and predict their performances inside the coil. The drop of electrical performances at the yield strength of the tapes under different loading conditions is evaluated. The multiphysics model is going to be validated against experimental measurements of the critical current now being performed on REBCO tapes by SuNAM Co., immersed in liquid nitrogen.

Concrete structures are widely used in civil, environmental, and nuclear engineering. During their lifetime, the structures are often exposed to severe environmental conditions, causing various deterioration processes, which might significantly reduce their durability.
The aim of this contribution is to present a general approach [1] to modelling various chemical deterioration processes in concrete, due to combined action of variable hygro-thermal, chemical and mechanical loads. For this purpose, mechanics of multiphase porous media and delayed damage theory are applied.
The mass, energy, and momentum balances, as well as the evolution equations describing the progress of chemical reactions and the constitutive and physical relations are briefly summarized. The kinetics of physicochemical deterioration processes like calcium leaching [2], Alkali Silica Reaction (ASR) [3], and salt crystallization-dissolution [4, 5], are described by means of evolution equations based on linear thermodynamics of chemical reactions. The mutual couplings between the chemical, hygral, thermal and mechanical processes are presented and discussed, both from the viewpoint of physicochemical mechanisms and mathematical modelling. Numerical methods used for solution of the model governing equations are presented. For this purpose, the finite element method is applied for space discretization and the finite difference method for integration in the time domain.
Some examples of the model application for analysis of transient chemo-hygro-thermo-mechanical processes in porous construction materials are presented and discussed. The first example deals with calcium leaching from a concrete wall due to chemical attack of pure water exposed to gradients of temperature and pressure. The second one describes cracking of a concrete element, caused by development of expanding products of ASR. The third example concerns salt crystallization during the drying of a wall made of concrete and ceramic brick, causing degradation of the surface layer due to development of crystallization pressure.

The mechanics of porous media, or briefly poromechanics, sometimes is directly equated with Geomechanics. And this is probably also no surprise as many problems relevant for Geomaterials can be described with theories of poromechanics and also many of the involved interesting effects can be present and relevant for these applications, like large deformations of the solid skeleton, transport of different substances, interface effects on different scales etc. But it is important to note that more or less the same theory, if formulated in a sound and consistent way, can also be used – and actually is often THE appropriate theory – for many other interesting and important applications that also are essential for sustainable development.
With his strong and unique background and expertise in the Mechanics of Porous Media and Geomechanics, Bernhard Schrefler also was a pioneer in transferring this expertise to other interesting areas in different fields of Engineering and the Applied Sciences. This serves as an motivation for me to cover advanced computational models and approaches for porous media for interesting problems beyond Geomaterials.
In this talk we will present some of our recent work on advanced modeling and development of novel computational approaches in important applications, all of which also play an important role in sustainable development, namely energy storage systems as well as highly relevant biomedical and biophysical applications.

The title of this work refers to a fundamental scientific research about bridging the gap between the microstructure of concrete and the studied traffic infrastructure. It has been carried out in the framework of a joint research project of the Institute for Mechanics of Materials and Structures of Vienna University of Technology and the Department of Geotechnical Engineering of Tongji University, Shanghai. It reads as: bridging the gap by means of multiscale structural analysis. Four different topics are treated in the framework of the general subject of this report. The first two topics deal with the analysis of traffic infrastructure, subjected to exceptional load cases. The first topic is on multiscale analysis of thermal stresses in concrete pavements due to sudden temperature changes and the second one is on microstructural analysis of the impact and blast loading in tunnel linings. The last two topics deal with reinforced concrete hinges in different types of traffic infrastructure. One of them is on multiscale structural analysis of segmented tunnel rings used in mechanized tunneling and the other one on experiments and Finite Element modeling of reinforced concrete hinges used in integral bridge construction.

Simulations using a history-dependent and physically-motivated Internal State Variable (ISV) constitutive model implemented into a spherical Finite Element code, TERRA3D, for the entire Earth’s mantle is used to illustrate the catastrophic plate tectonics event in our Earth’s history. We investigate the kinetics of dynamic recrystallization, grain size, and their influences on the mantle dynamics during its convection. The unique aspect of this study was that an explicit recrystallization variable was introduced and connected with the grain size kinetics, thus unifying static (grain size increases) and dynamic recrystallization (grain size decreases). We found that significant dynamic recrystallization (grain size reduction) occurred in the descending slabs and adjacent mantle, thus weakening its strength. Due to the rheological weakening, frequent episodic overturns and mantle avalanches were observed. Furthermore, strongly heterogeneous microstructures and associated viscosities were predicted in the entire mantle, because of the competition between the dynamic recrystallization, grain refinement, and grain growth under the geological setting. The grain size tended to be larger (~106 µm) in the upper mantle (below the lithospheric mantle) as the grain growth rate overtook the grain size reduction rate, while relatively small grain sizes (102 ~ 103 µm) were observed in the lower mantle as dynamically recrystallized downwelling slabs continuously flowed. In particular, exceptional heterogeneity of microstructure and rheology was observed nearby the core-mantle boundary depending on the kinetics of the dynamic recrystallization and grain size. Amazingly, this high rate event appears to align with worldwide flood stories as documented in the bible as well as other ancient manuscripts such as the Epic of Gilgamesh.

Robust models are required in geomechanics to make reliable predictions of engineering applications close to collapse both in the field of cementitious materials and soils.
The lecture is aimed at giving some highlights on possible strategies to the numerical modelling of concrete materials, in particular proving the soundness of FEM numerical models for the correct simulation of its mechanical behaviour, when close to failure.
The suggested approach is an elasto-plastic-damaged formulation in function of two invariants of the deviatoric stress tensor and in line with non-associated plasticity, i.e. the hardening, non-associated model by Menétrey and Willam [1], enriched with the potential function proposed by Grassl et al. [2] with a reformulation of what proposed in [3] for the damage effects; in this case a non-local integral type regularization technique is included [4] to avoid mesh-dependency in the results. In addition to this, to overcome the limitations in the iterative return-mapping scheme of the elasto-plastic model for the cement paste under tensile regime, due to the presence of singularities, or apex points in the adopted non-smooth yield surface, an improved return-mapping procedure is numerically implemented, able to catch locally the optimal return point on the active yield surface [5]. When dealing with concrete, ITZ can be modelled or not, and so characterized mechanically or not, depending if it is expected to have an important role in the failure mechanism; this may happen when maybe the collapse is interface-driven. In case ITZ is mechanically characterized, a possible cohesive formulation is illustrated, accounting for its lower stiffness and different degree of compactness than the surrounding cement paste.
On the other hand, the predictive simulation of damage triggering and evolution in concrete under generic 3D stress states requires the definition of the continuum at a meso-scale level, i.e. as a heterogeneous material, where coarse aggregates are explicitly modeled together with the cement paste, while the finer aggregates are included in the latter component, which is treated as homogeneous. At this purpose, special procedures to conduct meso-scale FE analyses on ordinary concrete made with calcareous aggregates, as well as sustainable concrete made with recycled aggregates from Electric-Arc Furnace (EAF) steel slag, is proposed, based on 3D X-ray computed tomography (X-ray CT) for the digitalization of the outer geometry of the aggregates and for the definition of their orientation in the matrix.

Crack propagation is most frequently implemented on the basis of so-called extrinsic models in which discontinuity surfaces (cracks) are introduced upon satisfaction of an external stress criterion. Often, an implicit time marching scheme is employed in which the crack is kept fixed within the computations of the iterative solver. The crack is advanced to a pre-determined length on the basis of a pre-determined propagation law at the end of the load step. This approach has been shown to lack mathematical soundness and is especially problematic in the context of hydraulic fracturing. The sequential solution of the displacement and crack surface in unknown fields leads to crack propagation velocities that do not converge with time step and mesh size refinement. A consequence of this issue is that the hydraulic fracturing model cannot properly capture the step-wise crack advancement and pressure oscillations in saturated porous media. This is not a coincidence but a manifestation of robustness issues with extrinsic crack propagation algorithms. We propose a hydraulic fracturing model with non-differentiable energy minimization for cohesive fracture in which formation and propagation of cracks are direct outcomes of the computations within the time step. The method allows advancement for any length of crack within a time step given the applied loads without need to introduce crack nucleation and crack increment length criteria. Numerical results show step-wise behavior which also exhibit convergence with time step and mesh size refinement.

For the extraction of heat from deep geothermal layers, the creation of fracture networks in these layers is needed. Water is injected under high pressure and fracture initiation and growth is induced. For modeling fracture growth, the Extended Finite Element Model (XFEM) has proven to be a powerful numerical tool. For hydraulic fracturing in low permeable rocks the so-called Enhanced Local Pressure (ELP) model was recently introduced. In this approach, the fluid pressure within the fracture is included as an additional degree of freedom with respect to the original XFEM displacement field, which greatly increases the applicability of the model. The fluid pressures in the fracture and the surrounding porous material, however, are still only coupled by means of a simplified, analytical Terzaghi relation.
In order to further improve the model, the interaction of the fracture fluid flow and the deformable porous medium is studied. We developed a coupled model in which the free flow is described by the Stokes equations and the fluid-saturated porous medium by Biot’s equations. We solve the coupled problem using a staggered FEA approach. The numerical model is shown to fully couple the free flow and the fluid flow in the saturated poro-elastic medium, taking into account the slippage effect and surface flow impedance.
We now combine the coupled free flow model and the ELP method to better predict fracture propagation in fluid-saturated poro-elastic materials and corresponding fracture leak-off rates.

Geotechnical engineers have been concerned for many years to determine the conditions under which a geostructure would fail. In order to determine the failure load and the mechanism type, mathematical, constitutive and numerical models have been used. As an example, we can consider the case of a slope subjected to seismic loading. Here, the mathematical model has to describe the coupling between the solid skeleton and the pore fluids. The contributions of Olek Zienkiewicz and Bernardo Schrefler have been of paramount importance for both saturated and unsaturated soils. To describe soil behaviour, constitutive relations are used. We will consider here Generalized Plasticity models for both saturated and unsaturated soils where we have included a state parameter. Regarding numerical models, most of practical cases have been modelled using coupled finite elements. Special techniques have been proposed in the past years to improve the accuracy of the models.
Once failure has been triggered, large deformations and displacements can occur. New mathematical models enlarging the domain of application of the classical pre-failure models have been derived, taking into account large relative displacements between phases. Regarding the rheological behaviour of fluidized soil, the progress has been much slower, and much work lies ahead of the concerned researchers. During last years, we have explored the similarity between rheological models and viscoplastic constitutive equations of Perzyna's type, which seem to provide a suitable bridge between solid and fluidized geomaterials behaviour.
Regarding numerical modelling, lagrangian meshless techniques such as SPH provide a suitable framework. In cases of landslides propagating distances much larger than their initial length, thin layer approximations provide suitable compromises between accuracy and cost of computation.

In geotechnics we rarely directly measure the parameters of the soils we need for engineering computations. Thus, the inverse problem that allows the needed values from directly measured data is frequently solved. Although the four presented strategies can be used to elaborate results of many geotechnical tests (such as CPTU or Marchetti dilatometer), in this summary we focus on one example only, namely the Falling Weight Deflectometer (FWD). This test is used to evaluate mechanical parameters of layered structures of road and pavements. Deflections due to an impulse load from the falling weight are measured in several points using geophones aligned on a rigid support. Determination of the mechanical parameters of the pavement layers is done by minimization of a mean square difference between the measured and a theoretical deflection. In the kernel of minimization procedures, there are costly computations of the theoretical deflections and their gradients (see [1]). The simplest use of Artificial Neural Network (ANN) in this context is the application of the well trained ANN as a surrogate of costly FEM computations in the minimization process. Once trained with limited number of results of the direct FE solutions the ANN gives the deflections for trial set of searched parameters. We use the ANN that approximates also the gradients with respect to its input. Such an ANN used it in frame of Truncated Newton method assures acceleration of the procedure. As a second strategy, we approximate directly the inverse relation between the set of parameters of the FE model of the layered structure and the deflection of its surface (see [2]). The input of the ANN is valued with the deflections and the output with the corresponding set of the model parameters. The ANN acts here as a universal approximate of an unknown functional relationship among the observed deflections and the searched parameters. Using various qualitative FE models, we can also train the ANN to discover qualitative properties of the structure which seems to be novel in the field of the inverse problem. Another possibility is the use of the ANN to directly approximate the inverse relation trained with the laboratory data collected from numerous real tests. We show that the necessary number of the laboratory test to train the ANN is reasonably small. The ANN trained with direct laboratory data acts here a special model-less form of phenomenological representation of constitutive relationships, based on observations (as in [3]).

One of the still persisting problems in both solid and fluids mechanics is to deal with singularities in the stress/strain fields or the flow field. An associated problem is the mesh-size dependence and non-convergence in finite element calculations in the material softening regime. A simple regularization method can be used to overcome both difficulties. This is shown for both elastic and plastic solids, as well as for both Newtonian and complex fluids.

Landslides of the flow-type are dangerous and also challenging to study [1]. A wide literature has been investigating the principal mechanisms governing each stage in which these phenomena can be ideally subdivided: namely, triggering [2], post-failure [3] and propagation [4]. However, holistic contributions and general overviews are very rare. In addition, a number of numerical methods have been issued and validated so that new chances exist to efficiently model those threats. One main limitation has been represented by the tremendous gap among those contributions based on the effective stress principle in soil mechanics and other studies conceived in fluid mechanics. The former ones have been applied to slope stability while the latter to landslide evolution, including propagation, deposition and even impact/interaction with structures and protective measurements. As emblematic cases, two classes of rainfall-induced landslides of the flow-type namely debris flows and debris avalanches could be mentioned. The principal numerical methods are reviewed for modeling the landslide initiation and propagation and are later used for analyzing a series of benchmark slopes and real case histories which are successfully simulated.

The mechanics of porous media is among the most fascinating and interesting branches of continuum mechanics since it can be applied to extremely broad fields of science. From the beginning of the XX century, when Karl von Terzaghi postulated the effective stress principle in soils mechanics, the theory of porous media advanced substantially, particularly thanks to the contribution of Maurice Anthony Biot. He introduced the general concept of the poroelastic medium and developed the theory of dynamic poroelasticity (now known as Biot’s theory) which is the basis of porous media mechanics. More recent developments consider averaging procedures which also include interface mechanics [1].
Porous media mechanics is ordinarily used for geomechanical problems at large, but nowadays it is also applied to model biomechanical ones. Teeth and bone decalcification, herniation of intervertebral discs and glaucoma and tumor growth are examples of clinical pathologies which can be modeled using mathematical approaches based on porous media mechanics.
Two very different applications are briefly presented to highlight the flexibility of this theory.
The first one is a thermo-hygro-chemo-mechanical (THCM) model for concrete. The presented approach is inspired by the theoretical framework of Gawin, Pesavento and Schrefler [2]; the reference model has been further improved for structural application by accounting for shrinkage, as well as creep and mechanical damage in a fully coupled fashion.
The second one is a multiphase model of tumor growth. The tumor is modeled as a four-phase system which consists of a solid phase, an extracellular matrix, and three fluid phases. The fluid phases are the interstitial fluid, tumor cells and healthy cells, with the latter two phases modeled as adhesive fluids. Since tumor growth is strongly influenced by nutrient availability, the diffusion of oxygen coming from the nearby existing vessels is also considered. Examples of biological interest will be presented.

Shock waves generated in water by Pulsed Arc Electrohydraulic Discharges (PAED) have, over the past years, offered new perspectives for the stimulation of hydrocarbon reservoirs, aimed at increasing their production. This contribution addresses the implementation of PAED techniques with two objectives: first, the development of an alternative to classical hydraulic fracturing, and second, the stimulation of existing fractures, aimed at removing debris and particles that may clog drains and decrease the production of oil or gas.
With regard to the development of an alternative to hydraulic fracturing that would be more effective for tight formations and potentially less dangerous for the environment, the principle is to induce a shock wave in rock masses generated by PAED. The dynamic load generates distributed damage instead of a localized fracture whose propagation may be difficult to control. In the context of tight rock masses with gas contained in the occluded porosity, it is expected that distributed cracking will be more efficient at connecting these pores, although the volume of rock affected remains confined nearby the well. Experiments reproducing PAED fracturing in reservoir conditions are described. The influences of the amplitude of the shock wave and of the number of shocks applied to laboratory specimens on damage and on the intrinsic permeability of the material are illustrated. Then, a computational model that simulates the entire process is discussed. We focus on the constitutive model for rock masses, based on orthotropic damage with crack closure effects, coupled to an orthotropic description of the evolution of permeability under loads. Rate effects on the damage growth are also included. Finally, numerical simulations of laboratory experiments and PAED fracturing under in situ conditions are discussed.
As far as the stimulation of existing fractures is concerned, the issue is to flush out the various particles that may be packed within the propped fracture and may induce a decrease of permeability of the fracture viewed as a drain for hydrocarbon. Applications go beyond unconventional gas production and cover conventional oil and gas production as well, thus making hydrocarbon production more effective. The shock wave generated inside the borehole by PAED is converted into surface waves travelling on the fracture surface. These waves induce fast variations of pressure that may potentially destabilize flocculates and put in suspension particles that have clogged the drain. In order to check this basic principle, an experimental set up has been developed in which a small portion of fracture is clogged and then unclogged by applying a dynamic load. Experiments illustrate the clogging-unclogging effects as a function of the opening of the propped fracture and of the density of fine that are introduced in order to promote clogging. These results should further help at understanding the basic parameters that govern the clogging-unclogging processes and therefore understanding what could be the best conditions of applicability of the method.

Recent years have shown the growth of a great number of recycled building materials, among which green concretes, made with recycled aggregates coming from construction and demolition waste (C&DW), metallurgical slags, plastic and other man-made materials, are being increasingly studied and used in several real-scale applications. Such a scenario aims to achieve challenging goals in terms of reduction of carbon emissions, reuse of waste materials promoting a circular economy, and avoidance of the use of bulk natural resources. This work shows the recent research carried out at the University of Padova on this topic. Experimental work, aimed at proving the suitability of using recycled aggregates and mainly slag from the Electric Arc Furnace in structural concrete, is shown. Particularly, results both about small-scale specimens for the analysis of mechanical and durability properties, and also about large-scale applications are presented. These latter applications allow the study of the structural performance of reinforced concrete elements under gravity and seismic-like actions. Lastly, some real construction projects are discussed, highlighting how environmental impacts can be effectively mitigated through the use of such materials.

Geomaterials and biological tissues have numerous properties in common. Both are naturally evolving porous media, the fluid of both is water, both exhibit large specimen-to-specimen variability of material properties, both are anisotropic, both have microstructure evolving from an intricate communication between environmental conditions and internal physics, both have ionisation along fluid-solid interfaces, and both have numerous ionic species dissolved in the fluid. It is no wonder that myriads of applications on geomechanical models have been found in biology. Many applications include herniation of the intervertebral disc [1,2], tissue differentiation driven by ion-exchange [3-4], swelling at extremely large deformation [5], osteoporosis, osteoarthritis [6], coronary vascular disease as a multiporosity problem [7], mechanotransduction in gel-like tissues [8-9], diffusiophoresis-driven propulsion [10-11]. These applications have led to novel numerical techniques, novel design of prostheses and better understanding of tissue engineering constructs [9].

In the previous industrial revolution, virtual twins (emulating a physical system) were major protagonists. Usually, numerical models (virtual twins), however, are static, that is, they are used in the design of complex systems and their components, but they are not expected to accommodate or assimilate data. The reason is that the characteristic time of standard simulation strategies is not compatible with the real-time constraints which are mandatory for control purposes. Model Order Reduction techniques opened new possibilities for more efficient simulations.
The next generation of twins, the so-called digital twins, allowed for assimilating data collected from sensors with the main aim of identifying parameters involved in the model as well as their time evolution in real time, anticipating actions using their predictive capabilities. Thus, simulation-based control was envisaged and successfully accomplished in many applications. Despite an initial euphoric and jubilant period, unexpected difficulties appeared immediately. Namely, in practice, significant deviations between the predicted and observed responses were noticed, limiting or abandoning their use in many applications.
In that framework of multi-uncertainty evolving environments, Hybrid Twins we proposed, consisting of three main ingredients: (i) a simulation core able to solve complex mathematical problems representing physical models under real-time constraints, (ii) advanced strategies able to proceed with data-assimilation, data-curation, data-driven modelling and finally data-fusion when using compatible descriptions for the physical and data-based models, and (iii) a mechanism to adapt the model online to evolving environments (control).

References:
Virtual, Digital and Hybrid Twins. A new paradigm in data-based engineering and engineered data. F. Chinesta, E. Cueto, E. Abisset, J.L. Duval, F. El Khaldi. Archives of Computational Methods in Engineering, https:\doi.org10.1007s11831-018-9301-4, 2018.

Natural phenomena have mostly multiphysics and multiscale character. The same is true for many manmade materials. Multiphysics and multiscale models, if based on sound physical principles and satisfying the numerical requirements such a consistency and stability, can give much more insight in the phenomena under investigation than simpler models and can even allow for discoveries. I shall address a few of these cases where complex models made the difference. The use of an appropriate u-p model allowed to find that cavitation is needed for the onset of localization in dilatant porous media [1]. It also allowed to identify internal length scales governing the phenomenon. A three-fluids model for concrete, which considers a chemo-thermo-hydro-mechanical analysis taking into account of dry air, capillary and adsorbed water and water vapour in the pores allowed for a unified treatment of concrete under very high temperatures with particular regard to thermal spalling, tunnel fires and reactor vessels, concrete at early ages and beyond, leaching, Alkali-Aggregate reactions, freezing and thawing [2,3]. A model for fracturing in saturated porous media based on Biot’s theory, standard Galerkin Finite Elements, cohesive fracture and remeshing enabled to discover at least for the mechanics community that fracturing in saturated porous media is not continuous but stepwise with ensuing pressure oscillations [4,5]. This has important implications both in hydraulic fracturing operations and in geophysics, among others to reproduce the en echelon structure observed in nature. In material mechanics a truly multiscale model allowed to predict correctly the residual strains after cool down of Nb3 Sn superconducting strands [6]. Finally a multiphase model based on an transport oncophysics framework enabled to identify for instance a way to obtain less dense tumors which are desirable for drug delivery [7]. Other examples will be shown in the presentation.

In this study, we shall investigate the effect of plastic deformation and inherent anisotropy of rocks on the hydraulic fracturing process. For this purpose, we propose an efficient numerical solution by combining the extended finite element method (XFEM) and the embedded discrete fracture model (EDFM) for studying hydraulic fracturing under coupled thermal-hydraulic conditions in an elastic-plastic porous medium. We consider both the fluid flow (and heat transfer) through the porous medium and the exchange between the medium and fracture. An efficient iterative scheme is developed to deal with the interaction between rock deformation, fracture propagation, fluid flow and heat transfer. The proposed method is assessed through comparisons with analytical solutions for a number of well-established problems. A series of numerical calculations are performed in order to investigate the effect of plastic deformation and inherent anisotropy of rocks on the process of hydraulic fracture propagation. Particular attention will be paid to the analysis of the process zone ahead of fracture in the context of anisotropic and plastic rocks.

Environmental, economic and safety concerns require more and more precise capabilities to perform the life cycle assessment of engineering structures. Therefore, structural engineers should be capable of describing all stages of the structural life even that involving the propagation of cracks and branching under complex loading conditions. The description of the propagation of cracks in structural materials, however, is still an open problem. The unavoidable presence of discontinuities prevents a direct application of the methods based on Classical Continuum Mechanics (CCM). Recently, Peridynamics (PD) has been proposed [1, 2] as a theory in which cracks are not part of the problem but part of the solution; PD is based on integral equations that do not make strong assumptions on the continuity of the displacement field. The integrals of the peridynamic theory are computed on a neighborhood of each material point, which is affected, in a nonlocal way, even by points that are not in direct contact with it. As a consequence of such a nonlocality, computational methods based on PD are usually more computationally expensive than those based on CCM. Several researchers are trying to couple computational methods based on CCM with those based on PD to obtain a computational tool able to simulate crack propagation in an efficient way [3-5]. Coupling two different continuum theories is not straightforward. In our presentation, coupling is realized at the discrete level between the standard displacement version of the Finite Element Method and a meshless version of the Ordinary State based PD. The domain is divided in two portions, one discretized with FEM and the other with OSBPD. If a perfect bonding between the displacements of the two portions is imposed, some out of balance forces are generated. The paper evaluates the magnitude of the out of balance forces and discusses some ways to reduce them.

With a total length of 64km, the Brenner Base Tunnel will be the world's longest tunnel.
The northernmost construction lot, Tulfes-Pfons, comprises 41.5 km of tunnel. Conventional tunnelling methods were used for 26.5 km of this length, and an open gripper tunnel boring machine for the remaining 15 km. The area extends from the Innsbruck Quartz Phyllite zone to the Penninic Bündnerschiefer group. An open tunnel boring machine made it possible to investigate the rock mass in greater detail.
Several fault zones and 10 large overbreaks in the different shapes and geotechnical rock mass behaviours were encountered in the first 13 km. One of them is presented more in detail in this contribution: the overbreak “San Francisco”, which was investigated with two-dimensional finite element back analyses which were including also the interaction effects with the ongoing advance of the main tunnels. The determination of the material parameters for the material laws for the rock mass and sprayed concrete is described and the finite element model including the concrete shells is explained. The calculations show that advanced material models are appropriate and loading/unloading effects can be successfully simulated. Purpose-built heavy steel segments, injected infill concrete and a cement/water suspension for binding the collapsed rock made it possible to bridge these cavities.
Bergmeister, K.; Reinhold, C. (2017): Learning and optimization from the exploratory tunnel - Lernen und Optimieren vom Erkundungsstollen – Brenner Basistunnel. In: Geomechanics and Tunneling. Berlin, 05/2017 Österreichische Gesellschaft für Geomechanik, Ernst&Sohn, Berlin
Bergmeister, K. (2011): Brenner Basistunnel – Der Tunnel kommt. Tappeinerverlag – Lana
Bergmeister, K. (2012): Life Cycle Design for the world longest tunnel project. In: IAALCE (Editors: Strauss, Frangopol, Bergmeister), Vienna

The tunnel track of the Brenner Base Tunnel crosses the base of the Brenner Massif passing through an important aquifer in the Hochstegen zone, which shows an essential influence on the water balance on the ground surface. In order to investigate the hydrological situation before and after the tunnel construction, two-dimensional models of the groundwater flow system with the finite difference code MODFLOW were produced. The scope of these models was on one hand to obtain information about the effects of the drawdown of the groundwater table by the draining effect during the tunnel excavation. On the other hand, an intention of the model was to investigate whether the drawdown of the groundwater table could be kept to a small and acceptable extent through suitable grouting measures and thus keep the influence on the groundwater balance in the model area low. The calculations show that the drawdown of the groundwater table in the zone of the aquifer of the Hochstegen Zone can be reduced to a small, acceptable extent. A drilling campaign with pumping tests was in execution and gave already additional input data. The paper deals with the model, the hydraulic parameters, the data from grouting measures obtained in other deep tunnel projects and from the drilling campaign and their effects on the modelling parameters and describes in detail the knowledge gained.

Polyvinyl chloride (PVC) membranes are used in geoenvironmental endeavours to prevent groundwater contamination due to leakage of leachates from landfill and other hazardous waste sites. PVC membranes constitute an important component of multi-barrier containment systems that also include layers of impermeable clay and PVC leachate collection systems. Geosynthetic membranes used as landfill liners can be exposed to adverse environments, including heat, exposure to ultra-violet light during construction, bacteria, and chemicals [1]. Despite their widespread use, their long-term effectiveness under exposure to chemicals, such as ultra-violet light, radiation, etc., are poorly understood. A primary requirement of a geosynthetic membrane relates to its ability to undergo large deformations and maintain its integrity, thereby impeding the migration of hazardous chemicals and contaminants to the environment. Experiments conducted in connection with this research indicate that the interaction of the geosynthetics with chemicals such as acetone and ethanol leads to loss of plasticizers that are necessary to maintain the hyperelasticity of geosynthetics [2, 3]. The longevity of the containment provided by PVC geosynthetics can be influenced by these factors, specifically in situations involving the thermal desiccation of clay. Desiccation cracking can be caused by moisture depletion in the clay barrier following exothermic processes associated with the decay of organic matter in a landfill. A cracked clay barrier provides a pathway for contaminants to come into direct contact with a geosynthetic barrier.
In contrast to rubber-like elastic materials [4-7], glassy polymeric materials exhibit appreciable irreversible effects, including development of permanent strains during loading-unloading cycles and strain-rate effects. This paper presents constitutive models that first describe the hyperelastic behaviour of a geosynthetic material in its untreated condition. The modelling accounts for both reversible and irreversible components of hyperelastic behaviour and incorporates strain-rate dependency in the constitutive response. The constitutive modelling is then extended to include the long-term loss of hyperelasticity because of exposure to pure ethanol. The constitutive parameters were determined from uniaxial tests and constrained tests conducted at different strain rates. The constitutive models were implemented in a general-purpose finite element code to examine the mechanics of a membrane fixed along a circular boundary and loaded by a hemi-spherical indenter. The comparison between the experimental results and the computational estimates were used for the purposes of validating the modelling approach.

Drop size distribution of aerosols controls the efficiency of crucially important environmental processes, e.g. transfer fluxes greenhouse gases at air-sea interface, and industrial processes, e.g. the energetic and environmental efficiency of energy production from liquid fuels.
In these processes, all the relevant momentum, heat and mass transfer fluxes occur across the tiny interfaces, separating the drops by the carrier fluid.
Interfaces are an inherently hard to define non-place! The accurate determination of interface position, shape and interaction with the fluid turbulence, however, is crucial to predict the overall behavior of the involved macroscopic physical phenomena.
Although considerable observational and theoretical attention is being focused on this topic, a clear and comprehensive picture on formation and properties of the dispersed phases is not currently available yet.
Our effort is to provide a general theoretical framework to describe the evolution of dispersed multiphase systems in turbulent flows. To this aim, Direct Numerical Simulation (DNS) of turbulence and accurate tracking of the interface are required, but the range of scales involved for most of practical environmental and industrial applications is so wide that performing this task is a formidable challenge for present day computers. These challenges include the grid resolution for DNS of turbulence which is of the order of the Kolmogorov scale, but of course, physical interfaces have a much smaller scale (order of few molecules) making the direct resolution unfeasible.
In this talk, we will briefly review the historical pioneering studies and current experimental findings and computational methodologies used to describe interfaces and then, we will focus on the phase-field approach in turbulent flows. In this Eulerian approach, the phase distribution is described by the order parameter ϕ. We will examine several flow instances and phenomena ranging from turbulent stratified flows to turbulent dispersion of drops and bubbles to reveal potentials and limitations of the phase-field method. Interface interactions with turbulence, coalescence and breakup phenomena for different important physical phenomena involving changes of fluids density and viscosity and presence of surfactants (Marangoni effects) will be discussed in connection with the characteristics of turbulence. Finally, potentials to upscale current results and comparison with current theoretical and experimental findings will be presented.

Multiscale Modeling of Metals from Atoms to Components
Prof. Dr. Dr. h. c. Siegfried Schmauder
Institute for Materials Testing, Materials Science and Strength of Materials (IMWF), University of Stuttgart, Pfaffenwaldring 32, D-70569 Stuttgart, Germany
siegfried.schmauder@imwf.uni-stuttgart.de
In this overview the first successful examples of real multiscaling from atoms to macroscale for different applications of metals will be presented.
In this context, multiscale simulation comprises all length scales from atomistics via microme¬cha-nical contributions to macroscopic materials behavior and further up to applications for compo-nents, nowadays called multiscale materials modelling (MMM).
A main focus of the presentation will be put on new developments with special emphasis on MD-simulations and other methods involved and how they interact within the present approach. It will be shown that each method is superior on the respective length scale. Furthermore, the parameters which transport the relevant information from one length scale to the next one are decisive for performing physically based multiscale simulations [1].
While in the past, different methods were tried to be combined into one simulation, it is nowadays obvious in many fields of research that the only way to succeed in understanding the mechanical behavior of materials is to apply scale bridging techniques in sequential multiscale simulations to achieve phy¬sically based practically relevant material solutions without adjustment to any experiment. This has opened the door to real virtual material design strategies.
In a final step it will be shown that the approach is not limited to metals but can be extended to other material classes and can be also applied for composites [2] as well as to many aspects of material problems in modern technical applications where the different disciplines meet, from physics to materials science and further on to en¬gineering applications.
In the second part of the presentation, emphasis will be put on the problem of fatigue of metals where multiscale materials modeling can answer a number of questions such as the influence of the lattice type or the relevance of mate¬ri¬als properties.

The introduction of composite materials in industry forced the development of a specific branch of computational science known as mechanics of composite materials. Homogenization, a basic technique of this theory, has been known for about a hundred years, counting from the pioneering work of Voigt [1] and Reuss [2]. The development of advanced composite structures, however, is still pushing the analysis tools and techniques for multiscale simulations, e.g. [3-6] forward. The paper is aimed to review challenges in the design of complex composite structures and ways to overcome them. The following problems are considered as examples for application of the multiscale simulations: prediction of micro-strains in ITER superconducting cables, design of novel composite materials for 3D printing, and design of a new carbon fiber high-speed catamaran. Application of both conventional and novel techniques for homogenization and heterogenization (unsmearing) in multiscale analysis is discussed and pros and cons of various approaches are highlighted.

Underground gas storage (UGS) is a practice that is becoming widely implemented to cope with seasonal peaks of gas consumption. When the target reservoir is located in a faulted basin, a major safety issue is the re-activation of pre-existing faults, possibly inducing (micro-) seismicity. Faults are reactivated when the shear stress exceeds the limiting acceptable strength. It has been observed that this occurrence can happen unexpectedly during the life of a UGS reservoir, i.e. when the actual stress regime is not expected to reach the failure condition. A numerical analysis has been carried out to cast light in this respect, by investigating the mechanisms and the critical factors that can be responsible for fault activation during the various UGS stages. The reservoir’s geomechanical behavior is simulated by an original elasto-plastic 3D Finite Element (FE) approach where the fault strength is taken into account by means of Lagrange multipliers [1]. The fault geometry is reproduced using special zero-thickness Interface Elements (IE), and the possible activation is controlled by the Mohr-Coulomb failure criterion. The simulations are carried out on a 3D regional-size geological setting, which requires the use of advanced numerical techniques for the solution of the resulting discrete problem [2]. The model is applied in a physical context representative of the typical UGS reservoirs located in the Netherlands, in terms of reservoir properties, fault geometry and pressure history. The Norg and Bergermeer UGS fields represent the reference for this modelling application. The analysis addresses the role of: (i) the space and time pore pressure gradients in the UGS formation, within the faults bounding/compartmentalizing the reservoir; (b) the poroelastic stress changes with respect to the natural stress regime; (c) the specific geological settings, such as the geometric configuration and the hydro-geomechanical properties of the faults and reservoir. The numerical results show that "unexpected" fault re-activations are likely to occur during UGS when micro-seismicity had been already experienced in the primary reservoir exploitation, even if the pore pressure does not exceed the initial undisturbed conditions.

Mining is essential to economy of the nations as well as technological development and human industry. Until we can stop using some of the material mined or replace the process with a more environmental friendly one, reduction of environmental impact of existing processes would be the best way forward. Monitoring and laboratory experiments have provided very good data concerning mining activities but it is usually expensive and time-consuming to perform field studies or laboratory experiments. Numerical modelling is one way to assess the process and to extract useful information from field observations and experimental data. Numerical modelling can also provide predictions for future operations or new mining process. In this paper, the finite discrete element method [1] will first be introduced and the development of the method to model blasting and rock cutting [2] will be explained. With a better understanding of the method involved, its environmental impact can be minimised.

Recently, non-continuous (step-wise) crack propagation has been observed in the numerical and experimental analysis concerning hydraulic fracture propagation in porous rock. It has been shown that the phenomenon is of a fundamental nature in the HF process [1,2]. Simultaneously, various regular crack propagation regimes (e.g. crack speed oscillation) have been discovered in the dynamic processes: in discrete elastic structures (splitting chain strips, crack propagation in lattices with different links [3-5]), and in continuous media (delamination of a flexural elastic beam rested on the Winkler foundation [6]).
The latter case has much in common with the phenomenon discussed in [1,2]. Among others, it was shown in [6] that, under the action of an incident sinusoidal wave, the steady-state mode exists only in a bounded domain of the wave amplitude. For higher amplitudes, local separation segments periodically emerge at a distance ahead of the main transition front. The analytical solution obtained allows analysis of this effect in detail and allows identification of a boundary between the steady-state and forerunning modes into the parametric space.
In a structured material (even of a simplest regular structure), depending on the applied load and the material properties, the following basic established (regular) dynamic fault / fracture propagation regimes can be identified [3-5]: fully open (classic) crack, bridge crack, and cluster-type propagation and forerunning. For more complex materials, all of those modes can appear together as a specific combination organized into rather complex, but still regular, regimes. We do not include the so-called branching crack propagation regime into this classification. This regime may also be very regular but it is not supported in one-dimensional front propagation as the previous modes.

The aim of this lecture is to discuss the applications of surface elasticity determination for effective properties of materials, and for some related phenomena as surface wave propagation. Here, in addition to the constitutive relations in the bulk, constitutive relations at the surface are independently introduced. Nowadays the most popular models of surface elasticity relates to the models by Gurtin and Murdoch [1, 2] and by Steigmann and Ogden [3, 4]. Some other models are also known in the literature, which can describe surface/interface related phenomena, see e.g. [5-7].
First, we discuss some useful surface elasticity models. From a physical point of view, surface elasticity models correspond to an elastic solid with an elastic membrane or shell or another 2D continuum attached to its boundary. The corresponding boundary dynamic boundary conditions are derived at the smooth parts of the boundary as well as its edges and corner points. Let us underline that these conditions also include dynamic terms. As a result, we have here a dynamic generalization of the Laplace-Young equation as known from the theory of capillarity. Second, we discuss the influence of the surface stresses at the effective stiffness parameters of layered plates and shallow shells. For small deformations, we derived the exact formulae for modified tangent and bending stiffness parameters of the plates and shells. The influence of residual surface stresses is also discussed. Unlike the previous case, where surface stresses are slightly changing the material properties, there is another example of the essential influence of surface properties. This example relates to the propagation of anti-plane surface waves. We discuss some peculiarities of the wave propagation.

Porous media such as soil, rocks and concrete are of great importance in the context of civil engineering and environmental geomechanics. They consist of a solid skeleton and pores filled with fluids, e.g. air and water. Complex mechanisms of flow and transport take place within the pore network and can lead to deformation of the solid skeleton and eventually to fracture phenomena [1].
Phase-field modeling of fracture has recently emerged as an alternative to conventional approaches such as remeshing, extended finite element methods or cohesive zone modeling. The phase-field framework can be considered a special type of gradient damage modeling approach, where a diffusive approximation of the crack is taken into account and the continuous phase-field parameter is used to describe the material integrity. The essential advantages are the possibility to describe arbitrarily complicated fracture patterns such as nucleation, branching and merging, without ad-hoc criteria on a fixed mesh, through the solution of partial differential equations derived from variational principles [2-5].
Phase-field modeling of fracture in porous media has been addressed in some recent publications [6-7], which however have only focused on the fully saturated case. Objective of this contribution is to describe fracture in partially saturated porous media using a phase-field approach [8]. In this study the air phase is assumed at constant atmospheric pressure with negligible density (passive air phase assumption) and the solid skeleton is described by its linear-elastic properties. Quasi-statics processes are studied. The equilibrium equations of the porous media, the mass balance equation of the liquid water and the phase-field evolution equation constitute a nonlinear coupled and time-dependent system of equations, which needs to be discretized and linearized. We formulate the coupled non-linear system of partial differential equations governing the problem with displacements, capillary pressure and crack phase-field as unknowns. The spatial discretization is carried out with finite elements of appropriate order for the different unknowns. We discuss its solution and present some relevant examples on desiccation tests [8].
The previous model has recently been extended taking into account the contribution of the air phase and the dynamics (u-p approach). The first preliminary numerical results will be shown and discussed.

Constant growth in world population and intense industrial development inevitably cause degradation of the environment and induce scarcity, sometimes also causing conflicts and tension regarding georesources. Predictive geosciences gather knowledge from fields such as mathematics, physics, and material sciences in order to develop geo-technologies based on Big Data assessment and predictive/inverse numerical simulations. Also necessary is the development of theoretical concepts and continuously improving methods of application derived from recent advancements in scientific and technological innovation in related domains.
Land surfaces and subsurfaces are increasingly solicited for different uses including exploitation of water resources for farming, agriculture, and other uses. In addition, exploration and exploitation of energy resources, useful substances, and storage of undesirable substances in the underground require knowledge of the dynamics of multiphase hydrosystems, from the surface to depths of several kilometers. The fundamental issue is transfer of mass and heat between phases (water - rock - gas - "microorganisms") on different scales of time and space, from soil to groundwater and deep aquifers. Such knowledge and application demands multidisciplinary competencies and complementary methods to enrich and extend both theoretical approaches and databases to build advanced concepts as well as interpretive and predictive numerical models [1]. These models must be continuously fed, updated, improved, tested, and validated on potentially similar and analogous systems on large scales.
The advanced concepts are based on field observations highlighting the expression and footprints of several complex and coupled physico-chemical and biological processes, such as those for reactive facies, redox zone, pH-buffering zone, capillary fringe, and mixing zone, that can be nested in generic and generalizable patterns (i.e., Critical Zone, etc.). These concepts can be used as guides to explore and manage geosystems to exploit resources (water, oil, gas, heat, recyclable materials, etc.) as well as to store some resources temporarily (water, heat, gas, etc.) and store undesirable substances permanently (CO₂, industrial water, acid gases, radioactive waste, etc.) [2]. Several platforms for the environment, artificial recharge of aquifers, mineral extraction, and recycling materials, were constructed at different scales to sustain these scientific developments [3]. Biodegradation processes based on the triggering of biogeochemical reactions result in the installation of active redox zones with relatively variable spatial extension and lifespan [4].
It is clear that future industrial and socio-economic developments must be rethought and improved by moving closer to an ideal of "zero rejection" and optimal and efficient use of natural resources. A comprehensive review of the state of knowledge, success stories, challenges, and innovation potentials will be presented. In addition, the areas of development in geosciences at BRGM (French Geological Survey) and ISTO (Earth Sciences Institutes of Orléans), institutes that share observation sites, research infrastructures, and pilot sites for developing environmental monitoring tools, and validation models supports will be presented.

The present work models problems in which the initiation and propagation of cracks in porous materials represents a key issue and is strongly influenced by the interaction between the solid matrix and the fluid in pores. The methodology is based on a work by the authors already published [1].
In this case, discontinuities are modelled by means of quasi-zero-thickness interface elements using an FEM-based approach. These special elements, which can be used to define either pre-existing or propagating cracks, act as joints that allow representation of the jump in the displacement field and the directional preferences in the fluid flow.
To ensure that the direction of the crack growth is not heavily influenced by the mesh, a non-local damage model is used to predict the degradation pattern of the domain and the interface elements are then inserted, followed by a remeshing.
FIC-stabilized elements of equal order interpolation in the displacement and the pore pressure have been successfully used under complex conditions near the undrained-incompressible limit [2]. A bilinear cohesive fracture model describes the mechanical behaviour of the joints. A formulation derived from the cubic law models the fluid flow through the crack.
Examples in 2-D and 3-D, using 3-noded triangles and 4-noded tetrahedra respectively, are presented to illustrate the features of the proposed methodology in hydraulic fracture processes. Other examples solved by the authors using joint elements in dam engineering [3] will be shown to introduce some of their alternative applications.

In this study, an energy based hydro-mechanical model and computational algorithm for the problem of hydraulically driven fracture networks developing in naturally fractured impermeable media is developed. The model is based on non-differentiable energy minimization for the dynamic deformation and fracture of the body coupled with mass balance of fluid flow within the hydro-fractures. Time-discontinuity induces spurious crack-opening velocity fields which lead to nonphysical solutions for the coupled fluid pressure field defined locally along the crack faces. The use of a time-continuous fracture model, such as the present non-differentiable energy minimization approach, is crucial for the numerical soundness and stability of the hydraulic fracture propagation algorithm. A discontinuous Galerkin finite element formulation is implemented, in which every element edge in the mesh is a potential site of hydro-fracture initiation and propagation. The presence of pre-existing natural fractures, as a common challenge in nearly all geological formations, are modelled with desirable edibility by simply assigning different fracture properties to the element edges defining the natural fractures. Using the graph theory principals, a search algorithm is proposed to identify, among all, the sub-set of cracked interfaces that form the interconnected hydraulically loaded fracture network. Robustness of the proposed computational algorithm and its versatility in the study of hydraulic fracturing is shown through several numerical simulations.

The mechanical behaviour of complex materials, characterized by complex non-linear behavior and complex internal sub-structure (micro), strongly depends on their microstructural features. In particular, in the modelling of these materials, such as particle composites that are polycrystals with interfaces or with thin or thick interfaces, as well as rock or masonry-like materials, the discrete and heterogeneous nature of the matter must be taken into account. This is because interfaces and/or material internal phases dominate the gross behaviour. This is definitely ascertained. What is not completely recognized instead is the possibility of preserving the memory of the microstructure, and of the presence of material length scales, without resorting to the discrete modelling which can often be cumbersome in terms of non-local continuum descriptions. In the possibility of accounting for non-symmetries in strains and stresses, the classical Cauchy continuum (Grade1) does not always seem appropriate for describing the macroscopic behaviour while taking into account the size, the orientation and the disposition of the micro-heterogeneities. This occurs in the case of materials made of particles of prominent size and/or strong anisotropy anisotropic media which lack in-material internal scale parameters. This calls for the need of non-classical continuum descriptions [1, 2], that can be obtained through multiscale approaches, aimed at deducing properties and relations by bridging information at proper underlying sub-levels via energy equivalence criteria. In the framework of such a multiscale modelling, the non-local character of the description is then crucial for avoiding physical inadequacies and theoretical computational problems. In particular, there are problems in which a characteristic internal (material) length, l, is comparable to the macroscopic (structural) length, L [3]. Among non-local theories, it is useful to distinguish between 'explicit' or 'strong' and 'implicit' or 'weak' non-locality [4]. Implicit non-locality concerns generalize continua with extra degrees of freedom, such as micromorphic continua [1] or continua with configurational forces [2].
This talk wants firstly to focus on the origins of multiscale modelling, related to the corpuscular(molecular)-continuous models developed in the 19th century and to give explanations 'per causas' of elasticity (Cauchy, Voigt, Poincare). This is in order to find conceptual guidelines for deriving discrete-to scale-dependent continua that are essentially non-local models with internal length and dispersive properties [4, 5]. Then, a discrete-to-scale dependent continuous formulation, developed for particle composite materials, based on a generalized version of Voigt's molecular/continuum approach is proposed [6]. Finally, with the aid of some numerical simulations concerning ceramic matrix composites (CMC), and microcracked media and masonry assemblies, the focus will be on the advantages of micropolar modelling with respect to other generalized continuum formulations [7-9].

References:
[1] Eringen, A.C. (1999), Microcontinuum Field Theories, Springer-Verlag, New York
[2] Gurtin, M. E. (2000), Configurational Forces as Basis Concept of Continuum Physics, Springer-Verlag, Berlin.
[3] Trovalusci P., Ed. (2016), Materials with Internal Structure. Multiscale and Multifield Modeling and Simulation, P. Trovalusci (Ed.), Springer Tracts in Mechanical Engineering, Vol.18:109-131, Springer.
[4] Trovalusci, P. (2014), Molecular approaches for multifield continua: origins and current developments. CISM (Int. Centre for Mechanical Sciences) Series, 556: 211-278, Springer.
[5] Trovalusci, P., Capecchi, D., Ruta, G. (2009), Genesis of the multiscale approach for materials with microstructure, Archive of Applied Mechanics, 79(11): 981-997.
[6] Trovalusci, P., Varano, V., Rega, G. (2010), A generalized continuum formulation for composite materials and wave propagation in a microcracked bar, Journal of Applied Mechanics, 77(6):061002/1-11.
[7] Trovalusci, P., Pau, A. (2014), Derivation of microstructured continua from lattice systems via principle of virtual
works. The case of masonry-like materials as micropolar, second gradient and classical continua. Acta Mechanica,
225(1):157-177
[8] Fantuzzi, N., Trovalusci P., Dharasura S. (2019), Mechanical behaviour of anisotropic composite materials as micropolar continua, Frontiers, 59 (6):1-11 (https://doi.org/10.3389/fmats.2019.00059).
[9] Settimi, V., Trovalusci, P., Rega, G. (2019), Dynamical properties of a composite microcracked bar based on a generalized continuum formulation, Continuum Mechanics and Thermodynamics, 1-18. (DOI: 10.1007/s00161-019-00761-7)

In this work, a fully coupled hydromechanical formulation for unsaturated 2nd gradient elastoplastic porous media is presented and applied to the numerical modeling of some geomechanics IBVP characterized by strain localization into shear bands. The introduction of internal length scales associated to the weakly non-local character of the constitutive equations effectively regularizes the numerical solutions.
The 2nd gradient elastoplastic model adopted is based on two independent plastic mechanisms. The first one is provided by a three-invariant isotropic--hardening elastoplastic model similar to the one presented by Nova et al. [1], extended to unsaturated soils. In lack of sufficient experimental evidence, the second-gradient mechanism is based on a simple elastic-perfectly plastic formulation.
Foe the numerical solution of the governing system of non-linear PDEs, the Isogeometric (IGA) Finite Element Method [2] has been adopted. When applied to constrained micromorphic media such as second-gradient materials, IGA offers the advantage of providing higher-order continuity of the approximating functions across element boundaries, which allows a more efficient and straightforward implementation of the discrete equilibrium problem, as compared to existing mixed FE formulations based on conventional polynomial shape functions, see [3]. This feature is also very important in coupled hydromechanical problems. In fact, the smoothness of the approximated displacements and pore pressure fields can mitigate significantly the requirements for minimum time steps.
The simulation of some relevant consolidation problems demonstrates the good performance of the IGA implementation, and shows its effectiveness in regularizing the FE solutions when localization patterns occur in the strain field.

The wind climate of Europe and of many other countries in the world is dominated by synoptic extra-tropical cyclones and by mesoscale thunderstorm downbursts. Thunderstorms are frequent phenomena that cause wind speeds and wind-induced damage often greater than those due to synoptic cyclones. This new paradigm of wind engineering has given rise to a recent burst of research despite which there is not yet a model of thunderstorms and their loading of structures like that developed for cyclones in the early 1960s which is still in use in engineering practice (1).
This shortcoming leads to unsafe and expensive structures. The insufficient safety of low- and medium-rise structures is testified by frequent damage and collapse in thunderstorm days. The excessive cost of tall buildings is apparent due to the absence of critical situations due to the wind. Both these aspects are derived from the fact that thunderstorm outflows intensify close to the terrain and reduce their speed while increasing the height, whereas traditional wind speeds and loading increases with height.
THUNDERR (www.thunderr.eu) is a project funded by the European Research Council aiming to pursue three objectives: 1) to formulate a novel, interdisciplinary and unitary model of thunderstorm outflows (2, 3); 2) to assess a wind loading model of structures due to thunderstorm outflows (4, 5) and to encapsulate this and the classical method for cyclones into a novel wind loading format easily transferable to engineering and codification; 3) to spread the results throughout the international community, to strengthen a renewed culture on wind actions on structures.
This paper provides the general framework of the THUNDERR project, illustrates the results obtained in this phase of the research, describes the perspectives of the studies currently undertaken under the scientific and technical viewpoints, and discusses their potential impact on civil and structural engineering, as well as their consequences on building safety and sustainability

Studying the freezing process in water-saturated soils is of great interest in many engineering fields. During the freezing of fine-grained soils, volume expansion known as frost heave is usually observed. Such phenomenon can cause detrimental deformation and damage to highways, building foundations and pipelines in cold regions subjected to seasonal freezing or during mining and tunneling where artificial freezing is adopted. The frost heave is attributed to the formation and growth of ice lenses associated with water migration to the freezing front. Based on several experimental works and previous studies, it is found that the ice lens formation is related to cracking of the soil in the frozen fringe [1]. Therefore, according to the stress criterion, the soil skeleton separates and a new ice lens forms when the pore pressure exceeds the sum of the overburden stress and the separation strength of the freezing soil. Moreover, suction occurs when the pore water solidifies into ice, facilitating water migration from the unfrozen zone to feed the growth of the ice lens [2].
Here we propose a thermo-hydro-mechanical model consisting of a porous solid matrix and a pore-fluid phase representing solid ice and liquid water based on a coupled phase-field-porous media approach. The model accounts for the phase change, water migration, ice lens formation and soil deformation due to the freezing process. We use the macroscopic theory of porous media (TPM) for the description of the deformable, heterogeneous porous solid [3] with the phase changing fluid constituent described by a unified formulation employing a phase-field model (PFM) [4]. The diffusive interface treatment of the freezing front can be easily implemented numerically as no explicit front tracking and application boundary conditions at the interface is required. The ice lens formation is modeled based on the rigid ice model adopting the stress criterion. In this context, a fracture related PFM is used to describe the crack formation preceding the ice lens initiation [5]. Additionally, the Clapeyron equation is employed to calculate the suction pressure at the water–ice interface. Finally, numerical examples are presented to demonstrate the ability of the proposed model in describing the freezing process in fluid-saturated porous media.

Many biological systems integrate inorganic, mineral components which are also used by nature in a geological context. This provides a fascinating avenue for interdisciplinary work at the “bio/geo-interface”. The present lecture will cover 20 years of research concerning works that translate and adopt concepts from poro-micromechanics, a theory having originally arisen in geoengineering, to move towards the deciphering of the mechanics of complex biological systems. In this context, particular emphasis will be laid on the use of the Mohr-Coulomb failure criterion, one of the most fundamental theoretical concepts in geomechanics. This criterion is within the framework of multiscale continuum micromechanics, so it allows prediction of the strength of bone from elastic and strength properties of the material’s elementary components: hydroxyapatite crytals, type I collagen, and water with non-collageneous organics.
In this context, the following theoretical steps, all carefully implemented numerically and tested by very many biomechanical, biophysical, and biochemical experiments, are the following: the hierarchical sequencing of traditional homogenization schemes such as the two-phase Mori-Tanaka and self-consistent scheme [1,2,3], the extension from two-phase to multi-phase systems [4,5], and the consideration of eigenstrains and their upscaling characteristics [6], which paved the way towards a unified vision of bone multiscale biomechanics, encompassing poro-elasticity, poro-plasticity [7,8], and creep [9].