{"id":17717,"date":"2024-10-22T14:51:59","date_gmt":"2024-10-22T12:51:59","guid":{"rendered":"https:\/\/navier-lab.fr\/?p=17717"},"modified":"2024-10-23T14:13:14","modified_gmt":"2024-10-23T12:13:14","slug":"soutenance-alec-tristani","status":"publish","type":"post","link":"https:\/\/navier-lab.fr\/en\/soutenance-alec-tristani\/","title":{"rendered":"PhD Defense \u2013 Alec Tristani"},"content":{"rendered":"<section class=\"kc-elm kc-css-813790 kc_row\"><div class=\"kc-row-container  kc-container\"><div class=\"kc-wrap-columns\"><div class=\"kc-elm kc-css-634306 kc_col-sm-12 kc_column kc_col-sm-12\"><div class=\"kc-col-container\"><div class=\"kc-elm kc-css-826321 kc_text_block\"><\/p>\n<p>\nAlec Tristani, PhD student in the Geotechnics team, will defend his PhD thesis entitled <span lang=\"EN-GB\">\u201c<em>Enriching tunnel behavior analysis with artificial intelligence tools<\/em>\u201d on <\/span><span lang=\"EN-GB\" data-olk-copy-source=\"MessageBody\">Friday the 25<sup>th<\/sup> of October starting at 2pm in the Navier amphitheatre. <\/span><\/p>\n<p class=\"x_MsoNormal\"><span lang=\"EN-GB\" data-olk-copy-source=\"MessageBody\"><strong>Composition of the jury:<\/strong><br \/>\n<\/span><\/p>\n<ul>\n<li class=\"x_MsoNoSpacing\"><span lang=\"EN-GB\">Chlo\u00e9 Arson \u2013 Cornell University \u2013 Reporter<\/span><\/li>\n<li class=\"x_MsoNoSpacing\"><span lang=\"EN-GB\">Daniela Boldini \u2013 Sapienza University of Rome \u2013 Reporter<\/span><\/li>\n<li class=\"x_MsoNoSpacing\"><span lang=\"EN-GB\">Georg Anagnostou \u2013 ETH Zurich \u2013 Examiner<\/span><\/li>\n<li class=\"x_MsoNoSpacing\">Jean-Michel Pereira \u2013 \u00c9cole des Ponts<span lang=\"EN-GB\"> \u2013 <\/span>Examiner<\/li>\n<li class=\"x_MsoNoSpacing\">Jean Sulem \u2013 \u00c9cole des Ponts<span lang=\"EN-GB\"> \u2013 <\/span>Thesis Director<\/li>\n<li class=\"x_MsoNoSpacing\">Lina-Mar\u00eda Guayac\u00e1n-Carrillo \u2013 \u00c9cole des Ponts<span lang=\"EN-GB\"> \u2013 <\/span>Co-Supervisor<\/li>\n<li class=\"x_MsoNoSpacing\">Didier Subrin \u2013 Centre d\u2019\u00c9tudes des Tunnels<span lang=\"EN-GB\"> \u2013 <\/span> Guest<\/li>\n<li class=\"x_MsoNoSpacing\"><span lang=\"EN-GB\">Huy Tran-Manh \u2013 Itasca Consulting \u2013 Guest<\/span><\/li>\n<\/ul>\n<p class=\"x_MsoNoSpacing\"><span lang=\"EN-GB\">\u00a0<\/span><\/p>\n<p>\n<\/div><div data-closeall=\"true\" class=\"kc-elm kc-css-115515 kc_accordion_wrapper\">\n    <h3 class=\"kc-accordion-title\">Abstract<\/h3>\n<div class=\"kc-elm kc-css-341302 kc_accordion_section group \"><h3 class=\"kc_accordion_header ui-accordion-header\"><span class=\"ui-accordion-header-icon ui-icon\"><\/span><a href=\"#show-the-abstract\" data-prevent=\"scroll\"><i class=\"\"><\/i> Show the abstract<\/a><\/h3><div class=\"kc_accordion_content ui-accordion-content kc_clearfix\"><div class=\"kc-panel-body\"><div class=\"kc-elm kc-css-212972 kc_text_block\"><\/p>\n<p style=\"text-align: justify;\">Tunnel design relies on multiple approaches. Analytical and empirical formulations are reliable and useful for rapid estimations. Numerical modeling enables to overcome too restrictive assumptions and may be required in complex situations. However, additional efforts are needed to improve the robustness of these approaches. Especially, the time-dependent effects need to be better described to evaluate the long-term behavior of tunnels and there is a need for new constitutive models. Furthermore, to analyze the increasing amount of measurements collected during tunnelling, new techniques must be developed. Machine learning appears to be a promising field as previously unknown relationships between observations and measurements can be established. Thus, in this thesis, analytical, numerical, and data-driven approaches are emphasized to analyze the behavior of tunnels, both in the short term and in the long term. By combining these methods, a global framework for tunnel design is developed.<\/p>\n<p style=\"text-align: justify;\">On the one hand, an original approach is presented for the back-analysis of convergence measurements in deep tunnels to determine the constitutive parameters of the surrounding rock mass. Since time-dependent deformations can result from both the advance of the face and the delayed response of the ground, the two effects must be considered during excavations. To that end, new analytical solutions are derived by assuming an unlined circular tunnel excavated in a homogeneous isotropic ground under an initial isotropic stress field and assuming a fractional viscoelastic plastic behavior. Combining the closed-form solution that takes into account the progressive face advance and an empirical approach, convergences are back-analyzed based on a least-squares optimization method to calibrate the constitutive parameters of the ground. The presented process aims to characterize the short-term and long-term behavior of tunnels and offers the advantage of being applicable during the excavation phase as soon as convergence measurements are available. The method is illustrated by two case studies related to the Fr\u00e9jus road tunnel and the Saint-Martin-la-Porte access gallery (SMP2). Additionally, the ground-lining interaction is studied through the development of an analytical solution for a lined tunnel excavated in a fractional viscoelastic ground.<\/p>\n<p style=\"text-align: justify;\">On the other hand, the potential use of artificial intelligence is explored. To this end, three machine learning tools are built based on synthetic data to predict both the short-term and the long-term behavior of the ground. The first model aims to assess the support pressure and the radial displacement at equilibrium in the lining and the radial displacement that occurs close to the face at the installation distance of the support. The second model computes the extrusion occurring at the core of the face considering an unlined gallery. The last model includes the time-dependent effect in the analysis to evaluate the ground-lining interaction in the long term. In particular, the bagging method is applied to neural networks to enhance generalization. Good performance is obtained using relatively scarce datasets. The modeling of the surrogate models is explained from the creation of the synthetic datasets to the evaluation of their performance and their limitations are discussed. In practice, these machine learning tools should be helpful in the field during the excavation phase.<\/p>\n<p>\n<\/div><\/div><\/div><\/div><\/div>\n<\/div><\/div><\/div><\/div><\/section>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":158,"featured_media":17743,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[251,291],"tags":[],"class_list":["post-17717","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-annonces-en","category-annonces-geotech-en"],"acf":[],"_links":{"self":[{"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/posts\/17717","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/users\/158"}],"replies":[{"embeddable":true,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/comments?post=17717"}],"version-history":[{"count":13,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/posts\/17717\/revisions"}],"predecessor-version":[{"id":17756,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/posts\/17717\/revisions\/17756"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/media\/17743"}],"wp:attachment":[{"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/media?parent=17717"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/categories?post=17717"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/navier-lab.fr\/en\/wp-json\/wp\/v2\/tags?post=17717"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}