Numerical investigation of ground surface settlement due to circular tunnelling influenced by variations of geometric characteristics of tunnel and mechanical properties of saturated soil and its prediction in the artificial neural network

The settlement of the ground surface due to tunneling are considered as tunnel instability factors, as well as the displacement of the crown of the tunnel. Considering the importance of the subject, numerous research have been discussed but not considered on the effect of shape and type of ground su...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:مجله مدل سازی در مهندسی 2021-04, Vol.19 (64), p.27-39
Hauptverfasser: vahed ghiasi, Mehdi Koushki
Format: Artikel
Sprache:per
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The settlement of the ground surface due to tunneling are considered as tunnel instability factors, as well as the displacement of the crown of the tunnel. Considering the importance of the subject, numerous research have been discussed but not considered on the effect of shape and type of ground surface settlement and its magnitude with respect to different factors. However, there is still no accurate equation to predict ground surface settlement, considering all effective parameters, including geometrical parameters of tunnel section and soil mechanical properties. In this paper, some numerical and parametric analysis of circular tunneling in frictional-cohesive saturated soil has been investigated using 2D FEM by ABAQUS. The behavior of ground surface, considering to change the different values of depth to diameter ratio(H/D), soil cohesion, internal friction angle and permeability coefficient, and the influence of these variables on settlement of surface in each model have been separately evaluated. Then, a multi-layer perception (MLP) artificial neural network is designed to predict the ground surface settlement. MLP is a type of feedforward artificial neural network utilizing back propagation technique for training phase and the Levenberg-Marquardt method are used to reduce the errors and distance between the network outputs and finite element method results. There are four independent variables in the input layer and a dependent variable in the output layer. The middle layer consists of 7 neurons. Finally, the high potential of the artificial neural network with a correlation coefficient of 0.98 is shown in the prediction of ground surface settlement.
ISSN:2008-4854
2783-2538
DOI:10.22075/jme.2019.18022.1735