Les dépôts morainiques holocènes de la zone axiale pyrénéenne: approche déterministe de leur instabilitéà Verdun sur Ariège, (Pyrénées centrales, France)

Résumé

Les résultats de cette note montrent que l'instabilité des moraines d'un secteur des Pyrénées centrales peut être abordée à partir d'un modèle géotechnique simple faisant appel au calcul du coefficient de sécurité F par la méthode des tranches de Bishop. Ce calcul est appliquéà des glissements de terrains morainiques anciens et actuels de 100.000 à 400.000 m³, circulaires, qui sont respectivement classés en glissements rotationnels et glissements translationnels. L'acquisition des paramètres géotechniques du matériau morainique permet de montrer le rôle majeur joué par l'eau dans le déclenchement des glissements de terrain actuels, avec une hauteur d'eau critique, avant rupture, de 60 à 85% de la hauteur totale de la moraine pour F=1,3. Cette approche déterministe de l'instabilité des versants morainiques s'appuie sur des limites géométriques précises des glissements, conformes aux modèles de calcul, avec les mêmes cercles de rupture que ceux reconnus sur le terrain.

Abstract

 This paper reports a study of the historic instability of the till deposits in an area of the central Pyrenees between France and Spain upslope of the town of Verdun. The object of the work was to create a model based on the geomorphological and geological mapping of landslides and a geotechnical survey of the instability of the whole of the mountainous slopes of the Domanial Verdun area. The detailed mapping allows the historic landslides to be distinguished from the more recent/active movements which involve volumes of 100,000 to 400,000 m³ of material. In addition, it was possible to assess the areal percentage of the actively slipping zones (5%) compared to the historic slide zones (12%) and the stable areas (78%). Following the European classification, two types of slides were identified in the Verdun study area: (1) historic rotational slides and (2) active translational slides located in the Verdun area at between 1000 and 1250 m altitude. The stability was calculated using the classic "method of slices", subdividing the slipped zone into vertical slices along a suitable cross section. The so-called factor of safety (F) was determined by dividing the moments of resisting forces (M_R) by the moments of driving forces (M_D). The state of limiting equilibrium has a "factor of safety" of 1. The physical parameters of the till deposits of Verdun were established as: unit weight γ=24.9 kN/m³ (calculated using the percentage of gneissic blocks contained in the tills) and porosity n=0.24. The results of 15 triaxial tests in a gravelly sand matrix were plotted in a Lambe diagram following a linear regression model [x=(σ₁+σ₃)/2 and y=(σ₁–σ₃)/2 with sin φ′ =tan θ]. From this the friction angle φ′=33°±3 and c′=45±5 kPa were established. The factor of safety calculated for the moraine deposits in the historic slides was 1.44 without water (Hw=0). With a height of water of 7 m, representing 85% of the till thickness, this was reduced to F=1. To achieve a factor of safety of 1.3, the maximum water level within the till should not exceed 2.5 m, representing 65% of the till thickness. Similarly, the factor of safety was calculated for the active slides of another area (shown as section 4 in Fig. 3 in the paper). Using slice number 9 from the middle of the slide, the factor of safety was 1.08 when the height of the water was taken as 90% of the till thickness. This high calculated factor of safety for the height of water is consistent with the slow movement of the actual slides. However, a lower internal cohesion of the till deposits or the presence of a weathered zone would decrease the factor of safety from 1 to 0.8. It is also possible that other parameters, such as the regional seismic activity, could have been sufficient to initiate movement (F<1) during the last 50 years. It is of note that the map of seismic activity shows that more than a 100 earthquakes with magnitudes greater than 3 have occurred in the central Pyrenees since 1660. The paper emphasises the importance of high-quality mapping which identifies and classifies areas of historic and recent instability. From this, a single geotechnical model to calculate the stability can be established. The level of the water is shown to be the critical parameter and of more significance than the variations in the effective friction angle, which itself is more important than variations in the effective cohesion. With this information it is possible to determine those areas where some form of stabilisation and/or drainage of the till deposits is necessary.