Tunnelling has become a preferred construction as urbanization increases and available land decreases in all over the world. One method that is commonly used in tunneling is so called conventional tunnel method (CTM); despite tunnel boring machines are becoming more popular. One of the main consideration concerning construction of tunnel is the stability of tunnel face during excavation. Scientist and engineer depend on physical modelling to understand different phenomena such as deformation patterns and failure mechanism. Since full-scale physical model is expensive and hard to repeat, it is preferable to use reduced model. There are three categories of reduced physical model: centrifuge, 1-g scale and extrusion tests. The main aim of the present research is to study stability of the deep underground opening of CTM by adapting results of extrusion tests which are obtained numerically thanks to the use of the commercial code Midas GTS NX 2014. The study focuses on rocks or soils whose viscous effect are negligible and behave under undrained condition. Static analyses using an elasto-plastic Tresca constitutive model with axisymmetric model are used to obtain relationship between stresses and displacements or characteristic curve and the development of platic zone surrounding the face. Interpretation of the results is done by considering real behaviour of phenomena which are three-dimensional problems. Therefore, the results have been always compared with those of a 3D tunnel as a reference model. Although the past experiments on extrusion furnished a guide that is sufficiently reliable for qualitative predictions of the phenomena, however, problems that are typical of studies performed in the laboratory do exists. One of the problems associates with an effect of small scale due to geometry. This leads us to investigate any possible effect on boundaries in extrusion tests considering different specimen dimensions. The study indicates that the numerical results of extrusion tests show a strong dependency of the geometry of specimen and the ratio between initial stress and shear strength. For instance, the results reveal that the effect of small scale is less significant when the ratio between the initial stress and the undrained shear strength is equal or lower than 8.5, and is applied in the standard triaxial specimens whose ratio between diameter of extrusion and specimen is 5.4. Another problem of extrusion tests is difficulty in reproducing in situ conditions: distribution of initial stresses and boundary conditions. As consequence, the results give different outcomes with respect to those of the 3D tunnel. The major concern is the evidence of stress evolution stresses in the boundaries close to the face during excavation or reduction of the face pressure. The reduction of stresses in one of the boundaries is evidence in the extrusion tests. In the other hand, the stress evolution is negligible in all the boundaries in the 3D tunnel. This reduction causes, therefore, reduction in destabilizing loads and, hence, reduction of displacement at the face. This is the main reason why the characteristic curve produced by the model extrusion tests is more stable than that of the 3D tunnel. The present study implies also some conclusions that are valid for both models. (i) Pincers effect occurs and is quantified as the stress migrates around the face. (ii) By means of a suitable dimensionless plane, all of the numerical results are collapse into a unique curve (so called front mother curve (FMC)) even though the soil parameters are varied. (iii) Although, the FMC of the extrusion tests is different than that of obtained by the 3D tunnel (due to different boundary conditions), both models give similar responses.
La costruzione di opere sotterranee sta acquisendo una sempre maggiore diffusione per via della sempre crescente urbanizzazione ed occupazione del suolo. Nonostante la sempre crescente popolarità dello scavo meccanizzato, lo scavo in tradizionale è ad oggi ancora molto utilizzato. Un aspetto importante nello scavo di gallerie è l’aspetto della stabilità del cavo. Per valutare il comportamento del sistema, prestando particolare attenzione allo sviluppo di deformazioni e di meccanismi di rottura, la ricerca ha fatto uso di modellazione fisica, in piccola scala: in particolare è possibile eseguire prove in centrifuga, prove a 1g o utilizzare prove di estrusione. Lo scopo principale del presente elaborato è studiare la stabilità di scavi di gallerie profonde adattando i risultati di analisi numeriche, eseguite mediante il codice di calcolo Midas GTS NX 2014, che riproducono prove di estrusione su campioni triassiali. Lo studio si concentra su terreni in condizioni non drenate che hanno un comportamento, in prima approssimazione, indipendente dal tempo. Si sono eseguite analisi statiche in assialsimmetria, utilizzando un legame costitutivo elastico-perfettamente plastico con legge di rottura alla Tresca, per ottenere la relazione fra sforzi e spostamenti del fronte (curva caratteristica) e lo sviluppo di plasticizzazioni intorno al fronte. L’interpretazione dei risultati è stata guidata dall’analisi della risposta di fronti di scavo di gallerie. Nel passato si sono utilizzate le prove di estrusione triassiali per prevedere qualitativamente la risposta di tunnel, tuttavia queste prove sono affette dai problemi legati alla sperimentazione in scala. Questo ha portato alla ricerca dell’influenza delle condizioni al contorno, in particolare la geometria del dominio, sui risultati delle prove di estrusione. Questo studio mostra che c’è una forte dipendenza fra la geometria del sistema e il rapporto fra io stato di sforzo iniziale e la resistenza al taglio. Ad esempio i risultati numerici mostrano che nel caso in cui il rapporto fra sforzo iniziale e coesione non drenata è inferiore a 8.5 è possibile considerare un campione il cui rapporto fra diametro esterno e diametro della camera di estruzsione è pari a 5.4. Un altro problema è riprodurre le condizioni in sito in termini di stato di sforzo e condizioni al contorno. In particolare si nota che per effetto dello scavo (visto come progressiva riduzione della pressione al fronte) nelle prove di estrusione sulla base del campione opposta alla camera di estrusione si assiste ad una progressiva diminuzione dello stato di sforzo, mentre nel caso di scavi di gallerie lo sforzo corrispondente rimane costante. Per tale ragione si ottiene che nel sistema riprodotto da prove di estrusione triassiale la forza instabilizzante si riduce progressivamente e dunque la risposta del sistema sarà più stabile rispetto alla risposta che avrebbe una galleria nelle stesse conclusioni. Nello studio sono presenti delle conclusioni che sonovalide sia per le prove di estrusione sia per lo scavo di gallerie: (i) avviene una migrazione di sforzi definita “effetto tenaglia”. (ii) Definendo un opportuno piano adimensionale tutti i risultati delle analisi numeriche collassano su un'unica curva definita Curva Caratteristica Madre per il Fronte (FMC). (iii) Nonostante le curve madri ottenute dai diversi modelli siano diverse (per effetto delle condizioni al contorno) i modelli forniscono una risposta simile.
Extrusion test on triaxial specimen : numerical investigation and theoretical interpretation
ADI TIRTA, BUDIWAN
2014/2015
Abstract
Tunnelling has become a preferred construction as urbanization increases and available land decreases in all over the world. One method that is commonly used in tunneling is so called conventional tunnel method (CTM); despite tunnel boring machines are becoming more popular. One of the main consideration concerning construction of tunnel is the stability of tunnel face during excavation. Scientist and engineer depend on physical modelling to understand different phenomena such as deformation patterns and failure mechanism. Since full-scale physical model is expensive and hard to repeat, it is preferable to use reduced model. There are three categories of reduced physical model: centrifuge, 1-g scale and extrusion tests. The main aim of the present research is to study stability of the deep underground opening of CTM by adapting results of extrusion tests which are obtained numerically thanks to the use of the commercial code Midas GTS NX 2014. The study focuses on rocks or soils whose viscous effect are negligible and behave under undrained condition. Static analyses using an elasto-plastic Tresca constitutive model with axisymmetric model are used to obtain relationship between stresses and displacements or characteristic curve and the development of platic zone surrounding the face. Interpretation of the results is done by considering real behaviour of phenomena which are three-dimensional problems. Therefore, the results have been always compared with those of a 3D tunnel as a reference model. Although the past experiments on extrusion furnished a guide that is sufficiently reliable for qualitative predictions of the phenomena, however, problems that are typical of studies performed in the laboratory do exists. One of the problems associates with an effect of small scale due to geometry. This leads us to investigate any possible effect on boundaries in extrusion tests considering different specimen dimensions. The study indicates that the numerical results of extrusion tests show a strong dependency of the geometry of specimen and the ratio between initial stress and shear strength. For instance, the results reveal that the effect of small scale is less significant when the ratio between the initial stress and the undrained shear strength is equal or lower than 8.5, and is applied in the standard triaxial specimens whose ratio between diameter of extrusion and specimen is 5.4. Another problem of extrusion tests is difficulty in reproducing in situ conditions: distribution of initial stresses and boundary conditions. As consequence, the results give different outcomes with respect to those of the 3D tunnel. The major concern is the evidence of stress evolution stresses in the boundaries close to the face during excavation or reduction of the face pressure. The reduction of stresses in one of the boundaries is evidence in the extrusion tests. In the other hand, the stress evolution is negligible in all the boundaries in the 3D tunnel. This reduction causes, therefore, reduction in destabilizing loads and, hence, reduction of displacement at the face. This is the main reason why the characteristic curve produced by the model extrusion tests is more stable than that of the 3D tunnel. The present study implies also some conclusions that are valid for both models. (i) Pincers effect occurs and is quantified as the stress migrates around the face. (ii) By means of a suitable dimensionless plane, all of the numerical results are collapse into a unique curve (so called front mother curve (FMC)) even though the soil parameters are varied. (iii) Although, the FMC of the extrusion tests is different than that of obtained by the 3D tunnel (due to different boundary conditions), both models give similar responses.| File | Dimensione | Formato | |
|---|---|---|---|
|
201510_Budiwan_Adi_Tirta_V02.pdf
accessibile in internet per tutti
Descrizione: Thesis in pdf
Dimensione
4.87 MB
Formato
Adobe PDF
|
4.87 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/112893