Příspěvek k modelovému stanovení únosnosti tyčových mikropilot

Abstract

This thesis deals with the determination of the bearing capacity of reinforced micropiles by the FEM. At present the research of geotechnical problems concentrates on the application of the advanced material models which enable to model the behaviour of soil in dependence on the void ratio and the confining stress. For a design of micropiles with bearing capacity determined by load test results, it is advantageous to use the numerical model of the finite element method. After the validation of the numerical model of the load test it is possible to model already easily the micropiles of different geometric parameters in similar geological environment. The bearing capacity of micropiles in non-cohesive soils is significantly influenced by the dilative behaviour that is defined as an enlargement of the soil volume by the shear stress which, however, exists only before the so called critical state has been reached. As soon as the maximal void ratio had been reached the dilative behaviour ceases. It is therefore necessary to use a suitable material model which takes this limitation of the dilatancy into account. These requirements are fulfilled, for instance, by the widely used hypoplastic model which can be implemented in the program PLAXIS 8.2 used for modeling. The Mohr-Coulomb (MC) material model will be used as the second material model. The results of both material models will be presented and analysed. The used finite element model was validated using the full scale load tests on micropiles by FOREVER research program in Saint Rémy-lés-Chevreuse conducted in a fine grained Fontainebleau sand. The MC material model parameters have been taken from literature. The hypoplastic parameters for the sand were not available and have been determined by the author of the thesis from the results of the triaxial and oedometric tests in published in literature. After the calibration of the models, the load bearing capacity has been estimated for a slightly changed pile parameters in a comparable geological environment. In the appendix C and D the estimation of the load-displacement relation of the micropiles subjected to compression and tension have been stated. Micropiles of the diameter 100, 150 and 200 mm with lengths 4.0, 4.5 and 5.0 m have been modeled in different geology with characteristic density of Id = 0.2, 0.6 and 1.0. The effect of the goestatic initial stress and the influence of the micropile installation have been modeled by the Earth pressure at rest value k0 = 0.5, 1.0 and 1.25. The numerical results show that the hypoplastic model models the behaviour of the micropiles in the good agreement with the experimental results of the load test. The model enables to model the non-cohesive soil with a different density Id. The MC model, however, uses the volume weight of soil which does not realistically interpret the deformations. The compared results of the both material models are shown in appendix C and D. The estimated result accuracy of the bearing capacity of the micropiles is influenced by a correct assumption of the geostatic initial stress, a correct estimation of the influence of the micropile installation on this geostatic initial stress and a correct determination of the density of the soil. In the field of micropile installation on the original soil arise many unsolved questions and further investigation is therefore required.