# LAME

### MECHANISMS OF CATASTROPHIC LANDSLIDES

The Environment and Climate programme of the European Commission Area 2.3 Technologies to forecast, prevent and reduce natural risks Research task 2.3.1 Hydrological and hydrogeological risk Project Reference : ENV4970619**Objectives****Work Content****Project Milestones and Deliverables****Benefits****Economic and Social Impact****Project Management Structure****The Partnership****Relevant Links**

## 1. OBJECTIVES

#### Background:

Soils are granular materials and occupy an intermediate position between solids and liquids. So, they can withstand shear stresses at rest and, thus, inclined soil surfaces - also called slopes - can remain stable. However, every now and then catastrophic slides occur at slopes, and the geotechnical engineer is faced with the problem how to predict and how to improve the stability of slopes. The first who analysed the stability of slopes was COLLIN in 1846. On the occasion of several slope failures which occurred along the tracks of the Swedish railways and in Gothenburg Harbour in the beginning of our century, FELLENIUS introduced his method of slope stability analysis in soil mechanics based on slip circles. This method is still under use together with some more advanced methods which have been later introduced by BISHOP, JANBU and others. However, there are many cases where these methods prove to be inadequate; thus slopes have failed catastrophically although they were shown to be stable according to these methods . Such collapse events are dramatic failures without warning and without any obvious trigger. CASAGRANDE e.g. writes that äthe overlying mass can slide out as if suddenly placed on roller bearingsô. The deeper reason for the shortcoming of the æclassicalÆ methods is that they are based on an oversimplified understanding of the phenomenon of failure (i.e. the inability to stand the applied loads) of granular bodies. This understanding goes back to COULOMB and MOHR and assumes that failure occurs only when the stress obliquity (i.e. the ratio of shear to normal stresses) reaches a critical value. An early attempt to explain the discrepancy between the conventional stability analysis and the occurrence of landslides is the decrease of soil strength from a peak to a residual value at continued deformation . However, this is not satisfactory as this is the fact to be explained rather than its explanation. In the meanwhile, our understanding of catastrophic failure has been substantially improved. Today we know that the aforementioned critical stress obliquity (or critical shear stress, if we refer to cases without drainage) is not a soil constant but depends on density, effective pressure level and mode of deformation. Moreover, we learned that even before the onset of æclassicalÆ shear failure a series of other peculiar events may occur which, eventually, can lead to a catastrophic failure of the considered soil body. Such events are related with the phenomena of spontaneous loss of stability and frequently also shear localisation, and are highly influenced by the interaction between grains and pore fluid. In several cases this interaction is responsible for a reduction or even a complete loss of shear strength, a phenomenon which is known as liquefaction or fluidisation. It should be stressed that our knowledge on the above phenomena is very new and has not yet been systematically incorporated into the analysis of slope stability for practical purposes. This is exactly the aim of the present proposal. In order to assess the role of the pore fluid we intend to investigate both types of landslides, the ones determined by the interaction of the grains with the porefluid and the ones for which the porefluid is immaterial. In order to have a fresh approach to the problem we intentionally omit the several proposed classifications of landslides. We wish to clarify that our scientific viewpoint has its origin in geomechanics and geotechnical engineering, i.e. we intend to focus on the conditions for the transition from a stable configuration of soil and rock masses to an unstable one with subsequent catastrophic motion.

## 2. WORK CONTENT

**Work package 1: Material behaviour**-
*Activity 1: Constitutive models**Activity 2: Laboratory tests**Activity 3: Material Stability**Activity 4: Phase Transitions and Percolation Theory*

**Work package 2: System behaviour**-
*Activity 5: FEM simulations**Activity 6: Numerical simulations with distinct elements**Activity 7: System stability considerations*

**Work package 3: Mitigation**-
*Activity 8: Material properties during and after the slide (post-failure behaviour)**Activity 9: Loads on shelters**Activity 10:Mitigation by monitoring, drainage or densification*

#### Activity 1 (Constitutive models)

An ideal constitutive model is expected to describe all types of soils realistically and to be easily calibrated. As such a mathematical model cannot be achieved, we have to restrict our attention to particular requirements. Thus, the envisaged models should be able to particularly describe density dependence of the mechanical behaviour of granular materials. This in view of the fact that loose soils are more susceptible to volumetric collapse and therefore to spontaneous liquefaction. Furthermore, the constitutive model should properly take into account the effective pressure level dependence of the material behaviour. It is realised that the attributes ædenseÆ or ælooseÆ can only be used with reference to particular pressure levels of the grain skeleton. Thus, even dense materials can exhibit collapse at high pressure levels due to grain crushing. This fact can also lead to liquefaction . Further important features of constitutive models are their ability to exhibit limits in stress space with respect to HILLs stability criterion, the controllability of element tests, and the possibility of shear banding. Of particular importance for landslides seems to be their ability to describe the so-called static (more properly spontaneous) and cyclic (in particular earthquake-induced) liquefaction. An important possible source of stability loss is the swelling occurring with rainfalls after long dry periods. This volumetric increase leads to softening and can be responsible for a series of catastrophic landslides occurring not only in arid zones (e.g. the huaycos in Peru) but also in Mediterranean countries. We intend to critically examine relevant constitutive laws and, to a feasible extent, also to improve them. This comprises elastoplastic and other constitutive models of the newest generation. A point of concern of this activity will be the investigation of the role of the grain rotations and the energy which can be acquired by their moment of inertia. As far as this is of importance for the initiation and catastrophic evolution of landslides, the constitutive laws have to be enlarged in order to properly incorporate polar effects .

#### Activity 2 (Laboratory tests)

In order to highlight the mechanical behaviour of materials involved in catastrophic landslides and the mechanisms which cause collapse, a series of triaxial, biaxial (plane strain) and ring shear tests on loose sand and morainic material specimens will be performed. Two different classes of tests have been chosen: load controlled tests in drained conditions strain controlled tests in drained and undrained conditions The following parameters will be considered: relative density for granular specimens and water content for cohesive materials; degree of cementation; initial saturation index; type of consolidation (isotropic or anisotropic); stress history (different values of over-consolidation ratio); time factor and imposed strain rate values. In particular, load controlled tests will be performed by imposing finite load increments or by continuously increasing the stress applied on the specimens. The strain rate will be recorded, by paying particular attention to the dynamic evolution of the mechanical response. In fact, according to recent experimental evidences and theoretical interpretations, the dynamic collapse of the micro-structure plays an important role in global failures both of laboratory specimens and in large scale problems.

#### Activity 3 (Material Stability)

While methods of structural stability have been extensively applied in civil engineering practice, the study of material stability (i.e. the stability in relation with the constitutive behaviour of the material as it can be considered for homogeneous stress and strain fields) is one of the clues to better understand apparently incomprehensible landslides. This activity will be organised into three successive steps. The first step consists to apply LYAPUNOV's definition of stability to experimental results obtained with homogeneous deformations in order to put the stability analyses of collapsible slopes on a rational base. From this definition it will be possible to determine certain stress states which are unstable although strictly inside a plastic limit condition. Thus some material instabilities exist from a physical point of view while a conventional shear failure condition is not yet reached. Since some mechanisms of loss of stability have been experimentally clarified, we propose in a second step to analyse them by means of HILL's condition of mechanical stability. For that we will consider suitable constitutive relations and a comparison between some of them will be carried out including elastoplasticity, hypoplasticity and other incrementally non-linear models. The comparison is based on a graphical representation of the second order work plotted at certain stress- strain states after a given strain history. The first stress-strain states where the second order work is vanishing in at least one stress direction are points of a stability limit surface. The stress directions where this second order work is vanishing constitute unstable directions and will be analysed and discussed in the light of experimental observations.

#### Activity 4 (Phase Transitions and Percolation Theory)

The subsequent step must be considered as an open but fascinating question: At the liquefaction state the intergranular stresses are vanishing as the contacts between grains are lost. Therefore such a state can be also understood as a percolation threshold since precisely at this point there is no more connection between grains (the percolation limit can be viewed e.g. as the vanishing possibility to transmit electrical current through an assembly of conductive grains). Now let us consider granular materials on a slope just at the collapse threshold. The assembly of grains may be viewed as such that between two values of the slope angle (this is exactly the mass of soil which will slide) the density is very low, i.e. the contacts between grains are rare. Thus the liquefaction condition may be approximately verified and therefore also the percolation limit. As a possible conclusion, liquefaction state, percolation limit and unstable condition might be proven to coincide on a slope just before collapse. Such an analysis should give realistic mechanical criteria for possible landslides and for the mass of soil which would be involved in such a case. The analysis of collapsible slopes with concepts of percolation will be combined with modern concepts of material instability. A first attempt for slopes has been recently proposed . CASAGRANDEÆs effective-pressure dependent critical void ratio can be understood and used as a percolation threshold: For a higher void ratio with the same effective pressure, only contractant deformation is possible, leading to volumetric collapse and spontaneous liquefaction without drainage. For lower void ratios with constant effective pressure rearrangements of the skeleton lead to shear localisation. With the dead load of the granular mass void ratios above critical cause chain reactions leading to catastrophic landslides. This is a percolation on the macro-scale comparable to forest fires. Details of work program: Laboratory tests: Biaxial and shear tests for precise determination of pressure dependent percolation thresholds of materials occurring in pilot field projects, evaluation with constitutive relations implying percolation thresholds for state limits, deliver element test data and correlation to the partners working on numerical aspects, 1 g model tests with ca. 1 mþ samples for verifying mechanisms, adapting monitoring systems and improving stabilisation methods, deliver model test data to the partners doing numerical work for verification, analysis of autogeneous dynamic effects in critical states (spontaneous emission of shear and pressure waves, the latter in skeleton and porewater), evaluation of field test data from catastrophic events (we have access to such data from loose mining deposits), proposals for field investigation, monitoring with early warning, stepwise prediction via observational method.

#### Activity 5 (FEM simulations)

The purpose of simulations with Finite Element codes will be (i) to reproduce the conditions under which known landslides have taken place, and (ii) to assess the risk of a failure taking place at a particular place. There exist a large number of Finite Element codes available in the market. However, they have to be scrutinised whether they are able to simulate key problems of landslides. The first problem is the coupling with the interstitial pore pressure, and the second the implementation of suitable constitutive models. This restrict the choices to codes such as ABAQUS, GEFDYN, CESAR, PLAXIS and CRISP (among the commercial codes), and SWANDYNE as a representative of research codes developed at universities. Simulations of landslides by the partner will be made with whichever suitable code is available to them, provided it fulfil the conditions mentioned above. In fact, it will be fundamental to implement the advanced constitutive models prepared by the partners. When a landslide is triggered, failure quite often develops as localisation of deformation in a narrow zone. Therefore, it will be necessary to use meshes in which this narrow zone is modelled with small elements. Here, it will be necessary to use adaptive remeshing combined with constitutive regularisation techniques. The goal here is to assess the relative performance of existing Finite Element codes in modelling landslides which have occurred in the past.

#### Activity 6 (Numerical simulations with distinct elements)

Landslides and rockfalls involve a widespread type of geomaterials, ranging from clays to rocks. Numerical simulations of the related failure phenomena must take into account the different nature of these materials, which can be considered continuous for fine grained soils, but are discrete when dealing with failures in coarse grained soils, debris-flows or rockfalls. The distinct element method is certainly the most adequate to reproduce failures in discrete materials, and is able to catch dynamic effects, due to the explicit scheme it is built on. In the beginning, 2D codes will be used. The possibility to conveniently use 3D codes to model large-scale failures will be checked in a second step. Distinct elements will be used to model the material behaviour at two different scale levels, i.e. volume element behaviour and the real-scale failures. In the first case numerical simulations are to be devoted to the understanding of basic properties of granular materials leading to local instabilities; in the latter one, to the modelling of large scale problems, with particular attention to the assessment of design-loads for rockfall shelters and to the evaluation of debris-flows run length. Comparisons with FEM simulations will be made for those materials whose nature is difficult to label as continuous or discrete.

#### Activity 7 (System stability considerations)

A great class of soils, including loose granular deposits and anisotropically normally to slightly overconsolidated cohesive soils show unstable behaviour expressed in terms of deviatoric softening and/or frictional strain-rate softening. In all cases the action of pore-fluid has additional destabilising effect leading to soil liquefaction due to contractancy . According to our understanding, a catastrophic liquefaction is a chain reaction due to spontaneous liquefactions with release of kinetic energy). Strain softening in turn can have the tendency to localise in narrow zones of few grain diameters in form of practically fully drained shear bands, that manifest themselves macroscopically as global failure surfaces . However shear bands do not form at once everywhere in the failed zone, but progress gradually in space following the redistribution of dead loads of the corresponding geostructure (the sloping ground here). The original Palmer and Rice conceptual model of progressive localisation accounted only for the dead loads necessary to produce slope instability. This model was based on a simple strain-softening elastoplastic model for clay, and could not provide any information about lifetime. In addition to the retardation effect due to internal fluid flow and the stabilising effect of inertia , loose granular soils and soft clays show pronounced frictional strain-rate sensitivity, which, depending on initial conditions, may lead to unstable behaviour in the sense of accelerated strain evolution. Here the initial-boundary value problem of a creeping infinite slope of thickness H will be considered. The slope is creeping under the action of dead loads (self weight) and follower loads (seepage). Time may be introduced through: a) Darcian fluid flow, b) strain-rate sensitivity of the soil itself, and c) macro and micro-inertia of the granulate. Here the factor (b) will be emphasised, and accordingly the soil will be modelled in terms of effective stresses and a elastic-viscoplastic constitutive model that accounts for strain-rate friction softening. In particular 1D and 2D numerical stability and bifurcation analyses 13, 14, 15 will be performed which will provide a better understanding of tertiary creep phenomena and crude life-time estimates of the slope. It should be emphasised that tertiary creep is modelled here on a constitutive level through the internal non-linear dynamics of evolution of the various internal variables.

#### Activity 8 (Post-failure behaviour)

An important parameter to be assessed when evaluating the risk-of-failure of a slope, especially in the case of rock-falls, debris-flows (Muren), liquefaction failures, is the run-out distance of the slide. The awareness of the risk and relocation of public as well as private buildings could have saved many lives in Europe (e.g. Aberfan 1966, Randa 1991). The knowledge of the pressure exerted on pipelines by a liquefied submarine flow has also an important impact from many viewpoints: technical economical, environmental and social. The research programme will be developed by performing numerical and small-scale 1 g model tests (which are promising due to the low stress level) in order to arrive at an understanding of the phenomena. Mathematical models for different flow types will be formulated and the predictions one will be able to obtain by means of them will be compared to available experimental data. Simplified methods will be then developed with the aim of arriving at engineering formulae that will help the designer of an engineering work or the architect planning a new urban or industrial development to estimate the run-out distance of a flow and/or the pressure exerted on underground lifelines. The models developed in this activity will be also important for the following one devoted to the prediction of the load on shelters and the interaction between soil and structures.

#### Activity 9 (Loads on shelters)

Once a catastrophic landslide has been triggered, a mass of soil starts moving downhill, and deformations cannot be considered any longer as small. As one of the aims of the project is to provide an estimation of the loads of landslides on shelters, the problem has to be posed in a different way, much closer to the formulations which are used in Computational Fluid Dynamics. The efforts will be directed along two lines (i) Developing simplified assumptions, such as assuming for a certain shape and foundation that it is immersed in a viscous fluid, and (ii) developing a suitable CFD code.

#### Activity 10 (Mitigation by monitoring, drainage or densification)

Mitigation means low cost control based on a better understanding of mechanisms possibly leading to catastrophic landslides. Recommendations for geotechnical monitoring and stabilisation will be worked out in co-operation with the partners in the group. Monitoring: With the improved understanding of the kinematics of creep and its possible transition to spontaneous flow, displacement measurements can more efficiently be deployed and evaluated. Measurements are made by surface geodesy and photogrammetry, and also in boreholes of suitable size and shape. The use of modern receivers for borehole deformations (inclinometers, deflectometers etc.) will be incorporated. Measurements of porewater pressure are included for the unsaturated zone with gas channels (tensiometers), the saturated or nearly saturated zone (pressure receivers), and for low-amplitude P-waves (hydrophones). The transition from suction to positive porewater pressure by closure of gas channels, caused by flooding of the ground, is a percolation which can indicate the onset of spontaneous liquefaction in case of void ratios higher than critical. Spontaneous increase of porewater pressure in the saturated zone can indicate the transition from creep to catastrophic flow. Spontaneous emission of P-waves with certain spectral properties (typically 1/f spectrum) appears to be a precursor of catastrophic flow, comparable with precursors of earthquakes. Stabilisation: Geotechnical stabilisation methods with comparably low costs will be recommended in combination with monitoring which requires an improved understanding of mechanisms to be worked out by the partners. Minor earth moving, i.e. resloping, covering and the like, can have a substantial effect when designed on the base of adequate mechanisms and related numerical studies. Drainage at the surface and inside the ground with ditches, wells, boreholes and geotextiles can likewise more effectively be tailored on the base of an improved understanding of failure mechanisms. Novel technologies will be initiated using controlled moderate pulsations and shocks introduced into the ground via boreholes. An improved understanding of the propagation of mechanical waves in the grain skeleton and the porewater will enhance the development of low-weight devices for this task. Combinations with the aforementioned monitoring systems will enable to work in this way even at near- critical states of the slope.

## 3. PROJECT MILESTONES AND DELIVERABLES:

The deliverable items are listed below with reference to the proposed Activities and Work Packages.

### WORK PACKAGE 1 (Material behaviour):

#### Activity 1: Constitutive models

Suitable constitutive (i.e. mathematical) models for soils (most of them developed by the partners) will be selected and presented (together with their calibration procedures). It will be explained, why these models are appropriate to describe the material instability which is one of the key issues to catastrophic landslides. Possibilities to detect the material properties in situ will be pointed out.

#### Activity 2: Laboratory tests

Necessary special laboratory test procedures will be compiled from references and/or worked out, which are needed to determine the relevant material properties. Based on own experiments and on references, a set of typical values of the relevant parameters will be presented.

#### Activity 3: Material stability

Although stability is recognised as a system property, recent research points to the fact that several important sources of instability do reside in the material behaviour. We intend to present a critical evaluation of the existing evidence in this field. This evaluation will be illustrated with numerical and experimental examples.

#### Activity 4: Phase Transitions and Percolation Theory

A more physical approach to liquefaction phenomena is accomplished by the Percolation Theory, which is inherently related to the theory of Phase Transitions. We intend to incorporate this concept into the problem of slope failure. In supplement with Activity 1 we will estimate critical concentrations of relevant parameters which can be decisive for the release of catastrophic events.

### WORK PACKAGE 2 (System behaviour):

#### Activity 5: FEM simulations

Based on appropriate constitutive models a Finite Element implementation will be presented to assess the stability of slopes under various actions (water infiltration, loading, etc.).

#### Activity 6: Numerical simulations with distinct elements

By use of distinct element codes simulations of rockfalls and debris-flows will be made. Based on the distinct element method results load estimations for the design of shelters will be worked out.

#### Activity 7: System Stability Considerations

The results of a system stability analysis will be used: as guideline for the identification of the proper set of variables that are meaningful for the description of slope progressive failure with full use of scaling laws crude lifetime estimates based on parameter analysis, which are typical for various geometrical layouts and soil types.

### WORK PACKAGE 3 (Mitigation):

#### Activity 8: Material properties during and after the slide

Post failure properties are decisive for the design of shelters and measures for rescue. A simplified mathematical model describing the mechanical properties of Muren (debris avalanches) and other liquefied soil/rock masses will be worked out.

#### Activity 9: Loads on shelters

The mechanical properties of failed and loosened soil/rock masses (as delivered by activity 8) will serve as basis to develop simplified design forces for shelters.

#### Activity 10: Mitigation by monitoring, drainage or densification

Recommendations will be worked out for the compaction and drainage procedure necessary to mitigate the danger of landslides.

### MILESTONES:

- Selection of suitable constitutive models and specification of relevant soil parameters (11th month)
- Specification of relevant stability criteria for slopes (12th month)
- Validation of numerical and analytical simulations of landslides in the light of field experience. Reevaluation of existing evidence (24th month)
- Recommendations to mitigate the danger of landslides (24th month).

## 4. BENEFITS

The European countries are particularly vulnerable by catastrophic landslides for several reasons:

- high density of population
- Europe is transversed by many mountainous regions (Alps, Pyreneans, Appennines, Pindos) crossed by important traffic and other life-lines
- Europe has a particularly long shoreline, the economical exploitation and environmental protection of which is deteriorated by on-shore and off-shore landslides
- dump slopes have been created over centuries by mining activities such that huge areas are still inaccessible due to their high risk of sudden landslides and constitute a very important environmental handicap. In this context it should be mentioned the reclamation project of Lausitz in Eastern Germany and the landslide of a hydraulic fill tailings dam in Stava (Northern Italy) . Even flat European countries such as the Netherlands are faced with such problems.

The formulation of risk-of-catastrophy criteria and design of retaining and/or sheltering structures is expected to contribute saving lifes and economic resources. The social benefits of carrying out a policy of prevention in sites where landslides can happen will increase the quality of life in endangered regions. The project is expected to assist engineers in a well-documented design when dealing with highly dangerous slopes. Up to now they must feel understandably uneasy and uncertain when they are facing the task to assess the safety of a slope in a region where catastrophic slope failures may occur. The expected benefits from our research proposal can be illustrated by the fact that after catastrophic landslides vivid discussions arose as to whether the location of the destructed structure was properly chosen. The results of this project will be made available to the concerned European Authorities and they will contribute to a better assessment of safety and risk with respect to catastrophic landslides.

## 5. ECONOMIC AND SOCIAL IMPACT

Rock-falls and flow-slides are the cause of many casualties and huge damages in all mountainous regions e.g. in Huascaran with 4.000 and (eight years later) 18.000 casualties, in mining regions and along shorelines. Sizes up to some million cubic meters, more than 100 casualties and more than 10 events per annum are reported. Of course, landslides catastrophes are not limited to the EU-states. The latter are, however, particularly vulnerable in view of their dense population and the dense network of lifelines. Particularly the alpine regions are very often hit by landslides, e.g. the ones in Vajont (in 1963 with 3.000 casualties) or in Randa (near Zermatt), where in April and May 1991 two landslides moved 15 millions of mþ soil and rock each, and Val Pola in 1987 with 30 casualties. Even in the days of the formulation of the present proposal (November 1996) a rock-slide south of Brenner interrupted the main rail connection between Germany and Austria with Italy for several days. A better understanding of the phenomena and a more appropriate modelling could lead to substantial savings of lives and properties. A better estimation of risks will help to lower the insurance costs. Another impact of landslides prevention will be the conservation of environment. It is remarkable that landslides threaten more pronouncedly those regions which rely their economic life and welfare from the beauty of environment (e.g. the Alps).

- Landslides endanger shorelines

- Debris from Huascaran catastrophy

- Landslides occur also in flat areas

- Landslides occur also in absence of water. (This landslide occurred on the planet Mars. The role of water is here probably played by the electrical interaction between the individual grains.)

## 6. PROJECT MANAGEMENT STRUCTURE

Each of the partners will be responsible for one or more activities:

Activity |
Responsible |

1 Constitutive models | Kolymbas |

2 Laboratory tests | Nova |

3 Material stability | Darve |

4 Phase Transitions and Percolation Theory | Gudehus |

5 FEM simulations | Pastor |

6 Numeric simulation with distinct elements | Nova |

7 System stability considerations | Vardoulakis |

8 Material properties during and after the slide | Nova |

9 Loads on shelters | Pastor |

10 Mitigation by monitoring, drainage or densification | Gudehus |

Twice a year there will be meetings in order to exchange the results and discuss the further steps. The last meeting will take place before the end of the project and will serve as a coordination for the publication of our results. The Project Coordinator will be responsible for the coordination of the project. He will also be the main interface between the research team and the European Commission. He will consolidate the project planning, progress reports, milestone reports, cost statements and budgetary overviews etc. using the inputs from the other partners (see below). The communication strategy aims to keep all the partners fully informed about the project status, the planning and all the other issues which are important to the partners in order to obtain maximum transparency for all involved and increase the synergy of the cooperation. Interactive management meetings and technical meetings take an important role in the communication strategy. All information (like minutes of meetings, visit reports, task reports, relevant publications, etc.) will be communicated to the Project Coordinator, who will be responsible for channelling this information to the other partners, where appropriate. The communication strategy also aims to effectively communicate with authorities responsible for the prevention of catastrophes. The communication strategy - which includes a planning for publications to be made, presentations to be given and conferences to be attended on behalf of the team - will be a leading topic at each meeting. Each partner and Activity Responsible will formally report every 6 months to the Project Coordinator about the progress of the work, on the basis of a regularly updated detailed planning. The reporting includes information about the technical progress, results obtained (e.g. deliverables) and compliance with the workprogramme. The progress status of the activity will also be reported in terms of percentage of completion, estimated time to completion, actual man-months spent and man-months needed to complete the activity. The Coordinator will summarise the overall project status and planning. To this end he will also regularly update the time schedule and the manpower matrix using the data he receives from the partners. After the first 12 months, the Coordinator will prepare a consolidated overview of the budgetary situation of the project ( in ECU), on the basis of the cost statements he has received from the partners for submission to the Commission and of the payments that have been made. The budgetary situation will also be compared with the initial costs-per-year planning which is to be made at the kick-off phase of the project. In addition to these activities we plan to organise a workshop, where selected scientists from neighbour fields such as hydrodynamics, geology and others will be invited and report on observations and problems related to catastrophic landslides.

### Coordinator:

The project coordinator, Prof. Kolymbas, is head of the Institute of Geotechnical Engineering and Tunnelling of the University of Innsbruck, which has been involved with the analysis of several slopes and slides in Tyrol. Its staff of 15 people contains a secretary and a bookkeeper. At present three EU-projects are running under coordination of Prof. Kolymbas.

- ERBCHBGCT 940554
- ERBFMMACT 960110
- IN/RU-95-0742

Prof. Kolymbas has carried out several research projects in Germany and Austria. He also participated in two joint research projects (SFB) in Germany. In 1992 he successfully organised the workshop äModern Approaches to Plasticityô with participation of 60 scientists from all over the world.

## 7. THE PARTNERSHIP

Overview of the consortium

Organisation |
Abbreviation |
Country |
Responsible |

Institut für Geotechnik und Tunnelbau, Universität Innsbruck | IGT | AT | Prof. D. Kolymbas |

Institut für Boden- und Felsmechanik, Universität Karlsruhe | IBF | DE | Prof. G. Gudehus |

Laboratoire Sols Solides Structures, Institut National Polytechnique de Grenoble | SSS | FR | Prof. F. Darve |

Dipartimento di Ingegneria Strutturale, Politecnico di Milano | PdM | IT | Prof. R. Nova |

Escuela de Caminos, Fundacion Agustin de Bethencourt | FAB | ES | Prof. M. Pastor |

National Technical University of Athens | NTUA | GR | Prof. I. Vardoulakis |

All organisations are universities and belong to the type EDU.

### Cooperation:

It should be stressed that the six partners come from six different European countries. They know each other from close collaboration in the past including common projects and publications. So, they know their mutual capabilities and esteem each other, a fact which guarantees a successful cooperation. More importantly, the thematic area of the project will allow to expose younger engineering scientists of the different university environments to the European dimension of such catastrophic events.

### Profile of the individual partners:

#### 1. Dimitrios KOLYMBAS (IGT):

**Title:**

Prof. Dr. techn. habil.

**Affiliation:** Institut f?r Geotechnik und Tunnelbau, Universit?t Innsbruck

**Address:** Techniker Str. 13, A-6020 Innsbruck, Austria

**Tel.**: +43/512/507-6670

**Fax:** +43/512/507-2996

**e-mail:** dimitrios.kolymbas@uibk.ac.at

**Expertise:**

Introduction of the theory of hypoplasticity as framework of constitutive models. Experimental, theoretical and numerical investigations of the behaviour of granular materials. Multiphase media.

**Relevant publications:**

- Modern Approaches to Plasticity, Elsevier, 1993.
- Kolymbas, D., Herle, I., v. Wolffersdorff, P.-A.: Hypoplastic constitutive equation with internal variables. International Jounal for Numerical and Analytical Methods in Geomechanics, Vol. 19, 1995.
- Compaction waves as phase transitions, Acta Mechanica, 107, No. 1-4, 1994, 171-182.

**Personnel:**

5 scientists, 5 technicians and administrative staff. Doz. Dr. B. Lackinger is a worldwide known expert on snow avalanches, Mr. Th. Wilhelm is a specialist for the mechanics of two-phase media (sand and porewater) and Mr. W. Fellin works in the field of compaction of loose soils.

#### 2. Gerd GUDEHUS (IBF):

**Title:**

Prof. Dr. Ing. habil.

**Affiliation:** Institute of Soil and Rock Mechanics, University of Karlsruhe

**Address:** Postfach 6980, D-76128 Karlsruhe, Germany

**Tel.:** +49/721/608-2221

**Fax:** +49/721/69 60 96

**e-mail**: gudehus@ibf-tiger.bau-verm.uni-karlsruhe.de

**Expertise:**

Constitutive relations, element tests. Predictor to Easter Sheldt Barrier 1975 - 6 concerning liquefaction. Various patents on in-situ soil stabilisation. Large scale field tests with soil nailing, vibration in soft soil, strutted excavations blasting, slopes, retaining walls, basic research on capillarity, electrocapillarity, thermal activation, dynamic phase transitions.

**Relevant publications:**

- A Comprehensive Constitutive Equation for Granular Materials. Soils and Foundations, 36, 1, 1996, 1-12.
- Spontaneous Liquefaction of Saturated Granular Bodies. In äModern Approaches to Plasticityô (ed. D. Kolymbas), 691-714, Elsevier, 1993.
- Sprengversuche zur Bodenverdichtung. Vortr?ge zur Baugrundtagung Berlin 1996, 523-536.
- State limits and percolation thresholds of granular bodies. Proceedings of Powders and Grains 1997 (under preparation).

**Personnel:**

Group of ca. 20 scientists and 20 technicians and administrative staff.

#### 3. Felix DARVE (SSS):

**Title:**

Prof. Dr.

**Affiliation:** Laboratoire Sols Solides Structures , Institut National Polytechnique de Grenoble

**Address:** B.P. 53, FR-38041 GRENOBLE cedex 9, France

**Tel.:** +33/476 82 52 76

**Fax**: +33/476 82 70 00

**e-mail:** felix.darve@img.fr

**Expertise:**

Incrementally non-linear constitutive relations for soils. Analysis and modelling of static and cyclic liquefaction. Material stability and uniqueness. Bifurcations in relation with the mechanical behaviour of a granular material. Applications of percolation theory.

**Relevant publications:**

- Geomaterials. Constitutive Equations and Modelling, F. Darve ed., Elsevier Applied Science, 418 pages, 1990
- Recent Advances in Geomechanical, Geotechnical and Geoenvironmental Engineering, F. Darve, Y. Meimon, J. Benoit, R. Borden eds., Technip, 184 pages, 1993
- Les G?omat?riaux. Th?ories, Exp?riences, Mod^les, F. Darve, P.Y. Hicher, J.M. Reynouard eds., Herm^s, 208 pages, 1995
- Mecanique des G?omat?riaux, F. Darve, P.Y. Hicher, J.M. Reynouard eds., Hermes, 562 pages, 1995
- Les G?omateriaux. Avanc?es r?centes en Calcul dÆ Ouvrages, F. Darve, P.Y. Hicher, J.M. Reynouard eds., Herm^s, 332 pages, 1995.

#### 4.Roberto NOVA (PdM):

**Title:**

prof. ing. Ph.D.

**Affiliation:** Dipartimento di Ingegneria Strutturale, Politecnico di Milano

**Address:** piazza Leonardo da Vinci 32, I-20133 Milano, Italia

**Tel.**: ++39/2/2399-4232

**Fax:** ++39/2/2399-4220

**e-mail:** calvetti@giuditta.stru.polimi.it

**Expertise:**

the main research activity was devoted to the mathematical modelling of the mechanical behaviour of soils and soft rocks, with special emphasis on material stability, liquefaction of loose sands, inherent and strain induced anisotropy of rocks and soils. He was involved also in many practice oriented research projects dealing with stability of subacqueous slopes, analyses of slope failures triggered by rainstorms, predictions of foundations and sheet-piles movements, design of tunnels and of earth reinforced walls.

**Relevant recent publications:**

- di Prisco C., Nova R., Lanier J. (1993) "A mixed isotropic-kinematic hardening constitutive law for sand" in Modern Approaches to Plasticity, D. Kolymbas ed., Elsevier, 83-124
- Nova R., (1994) "Controllability of the incremental response of soil specimens subjected to an arbitrary loading programme" J. Mech. Behav. Mater. ,5,2, 193-201
- di Prisco C., Nova R., (1994) "Stability problems related to static liquefaction of loose sand" Proc. I.C. Localisation and Bifurcation Theory for Soils and Rocks, Chambon, Desrues, Vardoulakis eds., Aussois, 59-72
- di Prisco C., Matiotti R., Nova R., (1995) "Theoretical investigation of the undrained stability of shallow submerged slopes" G?otechnique, 45,3, 479- 496
- Lagioia R., Nova R., (1995) "An experimental and theoretical study of the behaviour of a calcarenite in triaxial compression" G?otechnique, 45,4, 633- 648

**Scientific personnel:**

C. di Prisco, C. Jommi; 1 research fellow; 3 doctoral students. The Milan group has experience on experimental tests on loose specimens leading to liquefaction, constitutive modelling, numerical and analytical analysis of flow-slides and rock-falls. It was already involved in an EPOCH project on the same premises. It was also involved in the analysis of some actual failures occurred in the Italian Alps.

#### 5. Manuel PASTOR (FAB):

**Title:**

Prof. Dr.

**Affiliation:** Escuela de Caminos, Fundacion Agustin de Bethencourt

**Address:** Alfonso XII 3, 3 y 5, E-28014 Madrid, Spain

**Tel.:** +34/1/335-7226

**Fax:** +34/1/335-7249

**e-mail:** manuel.pastor@cedex.es

**Expertise:**

During the past years, M. Pastor has been involved in research projects concerning both Fluid Dynamics and Soil Dynamics problems. In the first area, the effort has been directed to produce optimal algorithms for convection-dominated flows, and in the second, to produce (i) constitutive equations suitable for geomaterials and (ii) numerical methods for coupled soil dynamics problems.

**Relevant publications:**

- Pastor and M. Quecedo: äA patch test for mesh alignment effects in localised failureô, Comm. Appl. Num. Meth., 11, 1015-1024, 1995.
- Pastor, M. Quecedo and O.C. Zienkiewicz: äA mixed displacement-ressure formulation for numerical analysis of plastic failureô, Comp. and Structures, 1996 (Accepted for publication)
- Zienkiewicz, M. Huang and M. Pastor: äComputational soil dynamics - Anew algorithm for drained and undrained conditionsô, Comp. Meth. Adv. Geomechanics, 47-59. H.J. Siriwardne and M.M. Zaman (Eds.), Balkema, 1994.

#### 6. Ioannis VARDOULAKIS (NTUA):

**Title:**

Prof. Dr. Ing.

**Affiliation:** National Technical University of Athens

**Address:** Iroon Polytechneiou 5, GR-15773 Zografou

**Tel.:** 0030/1/772-1217

**Fax:** 0030/1/772-1302

**e-mail:** i.vardoulakis@mechan.ntua.gr

**Expertise:**

Plasticity and visco-plasticity and fracture-mechanics theories for porous-frictional materials (soils, rocks, salts, ice, concretes, ceramis) with emphasis on microstructure. Bifurcation and localisation phenomena (shear-banding, splitting, surface instabilities) in geomaterials; folding and faulting of geologic structures. Elastodynamics of inhomogeneous media. Static and dynamic poroelasticity of fully and partially saturated porous media. Hydromechanical coupling. Development of new experimental devices and techniques. Non-linear Finite Element analysis with emphasis on post-critical computations and regularization techniques. The Laboratory of Testing Materials in NTUA is well equipped with a variety of experimental measuring devices and computer facilities.

**Relevant publications:**

- Stability and Bifurcation of Undrained, Plane Rectilinear Deformation on Water-saturated Granular Soils, Intern. J. Numerical and Analytical Methods in Geomechanics Vol. 9, 399-414, 1985
- Dynamic Stability Analysis of Undrained Simple Shear on Water-Saturated Granular Soils, Intern. J. Numerical and Analytical Methods in Geomechanics Vol. 10, 177-190 (1986)
- Plane strain compression experiments on water - saturated fine-grained; Han, C. & Vardoulakis I.G. (1991) G?otechnique 41 No. 1, 49-78
- Deformation of Water-Saturated Sand: I. Uniform undrained deformation and shear banding, G?otechnique 46, No. 2, 1-16 (1996) 5. Deformation of Water-Saturated Sand: II. The effect of pore water flow and shear banding, G?otechnique 46, No. 2, (1996)

**Scientific personnel:**

One doctoral student

### Qualification of the partners for this project:

All partners have contributed substantial impacts to the modern geomechanics. Their scientific profile covers the experimental exploration, the mathematical modelling and the numerical simulation of important aspects of the behaviour of soils and geomaterials under the aspect of mechanics of catastrophic landslides and is documented in numerous publications. The expertise and roles of each partner in this project can be outlined as follows:

**D. Kolymbas**is the initiator of the theory of hypoplasticity which constitutes a new framework for mathematical models of soil behaviour. He also worked in the field of stability of 3D geotechnical structures and in the field of mechanics of multiphase materials. Being the new director of the Institute of Geotechnical Engineering and Tunnelling of the University of Innsbruck he will be the coordinator of this project.**G. Gudehus**is a world-wide known and leading scientist in geomechanics. He has shaped related interdisciplinary research world-wide and is now focusing in integrating physical insights into geotechnical engineering. In recognition of his impact he is the main scientific consultant for the large reclamation works in Eastern Germany open pit mining fields. In this project he will bring in his expertise from advanced physics as it can be applied to geotechnical problems and he will be responsible for mitigation and monitoring measures.**F. Darve**Æs approach to constitutive modelling of soils had a paramount impact to geomechanics in France. Being the coordinator of the research groups GRECO and ALERT for a series of years he increased his scientific and science-management expertise. He is now editor-in-chief of the Journal äMechanics of Cohesive-Frictional Materialsô. In this project he will represent Grenoble as one of the centres of excellence for geomechanics in France and he will be responsible for material stability.**R. Nova**combines expertises from soil mechanics testing with experiences from engineering practice in his countryÆs environment. He has being studying static liquefaction of loose sands for the last 20 years and was involved in the analysis of rainstorm-triggered landslides in Valtellina.**M. Pastor**is expert in large-scale computer aided geoengineering modelling with particular emphasis in adaptive and mixed Eulerian-Lagrangean finite element coding.**I. Vardoulakis**is a world-wide recognised expert on bifurcation analysis and micromechanical modelling of geomaterials and has published numerous often-cited papers as well as a monography on this subject. He has designed and constructed a number of soil testing devices, including the biaxial apparatus to study shearbanding in soils.

## 8. RELEVANT WEB PAGES

- European Commission
- Directorates-General and Services of the European Commission
- DG VI Agriculture
- DG XI Environment, Nuclear Safety & Civil Protection
- DG XII Science, Research & Development
- DG XVI Regional Policies & Cohesion
- Joint Research Centre
- EUROPA Server
- CORDIS Server
- INFOREGIO Information on the activities of the European Regional Development Fund (ERDF) and the Cohesion Fund
- LEADER (with DG VI)
- Competence Center on Natural Hazards in Switzerland
- Swiss extra-parlementary commission on natural hazards
- Statistics on natural hazards and catastrophes

by the "SwissRe" reinsurrance company - Informations on 1999 natural catastrophes

by "SwissRe"