National Technical University of Athens (NTUA)
Detailed stochastic finite element simulation in geotechnics
Abstract
dc:descriptionWe know more about the movement of celestial bodies than we know about the soil underfoot” claimed once Leonardo da Vinci; a clever quotation that certainly gives us food for thought. Soils in their natural state are by far among the most variable materials. Unfortunately, lack of available data on their properties has led engineers to seek new ways of dealing with this uncertainty. Let us consider, for example, the adoption of safety factors; the astounding computing power of today’s parallel computers has opened completely new paths for engineers, especially on how to incorporate uncertain soil parameters directly in their numerical analyses. Up to now, some of the most widely used methods lay under the umbrella of stochastic finite element methods (SFEM), consisting of the alternative extension of the classical deterministic finite element method(FEM). This thesis, in particular, deals with the application of SFEM for the simulation of geotechnical problems with uncertain soil parameters, while the spatial variability of material properties such as elasticity or permeability is taken into account via the use of homogeneous random fields. In the first part, essential theoretical concepts are established; to this purpose, basic set and probability theory, together with random processes and fields, are analyzed. Furthermore, the basic variants of SFEM are discussed, followed by a brief introduction to variability response functions (VRF) as an alternative, trustworthy technique for quantifying response uncertainty. The second part consists of numerical case studies of the SFEM methodology with a special regard to problems of a clearly geotechnical nature. Specifically, the first applications consist of settlement of a footing resting on a consolidating soil layer with stochastic Young’s modulus and permeability. Next, the footing problem is revisited on the basis of dynamic variability response functions (DVRF), in order to prove the potential of this alternative methodology. It has to be noted though that, in the second application, the problem of shear wave propagation in a soil layer with spatially variable shear-modulus is studied; this is tackled with the DVRF methodology using the fast Monte Carlo simulation (FMCS), whereas again the accuracy of the FMCS is compared with the results obtained via the MCS. To conclude, throughout the present work, the potential of SFEM is fully demonstrated, something which will definitely prove to be of immense value for risk assessment and uncertainty quantification in the hands of modern engineers. After all, we must never forget what Horace Mann wrote in 1855: “Every addition to true knowledge is an addition to human power” (Lectures on Education, Lecture 1).
Degree
thesis:*- Grantor dc:publisher
- National Technical University of Athens (NTUA)
- Year dc:date
- 2017
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Mnitaras, Theofilos-Ioannis
- Μανιταράς, Θεόφιλος-Ιωάννης
Subjects
dc:subject × 16- Στοχαστικά πεπερασμένα στοιχεία
- Ομογενή τυχαία πεδία
- Δυναμικές συναρτήσεις διακύμανσης της απόκρισης;
- Εδαφική στερεοποίηση
- Διάδοση διατμητικού κύματος
- Stochastic finite element method
- Homogeneous random fields
- Dynamic variability response functions
- Soil consolidation
- Shear wave propagation
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Πολιτικού Μηχανικού
- Νανοτεχνολογία
- Engineering and Technology
- Civil Engineering
- Nano-Technology
Rights
dc:rights- Statement dc:rights
-
- BY
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/44594