Bibliographic Metadata

Prediction of tunnel boring machine performance using machine and rock mass data / von Ghulam Dastgir
Additional Titles
Vorhersage der Leistung von Tunnelbohrmaschinen mittels Maschinendaten und Gebirgsparametern
AuthorDastgir, Ghulam
CensorTentschert, Ewald-Hans ; Jodl, Hans-Georg
DescriptionX, 184 Bl. : Ill., graph. Darst., Kt.
Institutional NoteWien, Techn. Univ., Diss., 2012
Zsfassung in dt. Sprache
Bibl. ReferenceOeBB
Document typeDissertation (PhD)
Keywords (EN)TBM, ROP, RMC, RMR, Linear Prediction Modeling
Keywords (GND)Tunnelbohrmaschine / Tunnelvortrieb / Leistungsprognose
URNurn:nbn:at:at-ubtuw:1-60392 Persistent Identifier (URN)
 The work is publicly available
Prediction of tunnel boring machine performance using machine and rock mass data [24.65 mb]
Abstract (English)

Performance of the tunnel boring machine and its prediction by different methods has been a hot issue since the first TBM came into being. For the sake of safe and sound transport, improvement of hydro-power, mining, civil and many other tunneling projects that cannot be driven efficiently and economically by conventional drill and blast, TBMs are quite frequently used. TBM parameters and rock mass properties, which heavily influence machine performance, should be estimated or known before choice of TBM-type and start of excavation. By applying linear regression analysis (SPSS19), fuzzy logic tools and a special Math-Lab code on actual field data collected from seven TBM driven tunnels (Hieflau expansion, Queen water tunnel, Vereina, Hemerwald, Maen, Pieve and Varzo tunnel), an attempt was made to provide prediction of rock mass class (RMC), rock fracture class (RFC), penetration rate (PR) and advance rate (AR). For detailed analysis of TBM performance, machine parameters (thrust, machine rpm, torque, power etc.), machine types & specification and rock mass properties (UCS, discontinuity in rock mass, RMC, RFC, RMR, etc.) were analyzed by 3-D surface plotting using statistical software R. Correlations between machine parameters and rock mass properties which effectively influence prediction models, are presented as well. In Hieflau expansion tunnel AR linearly decreases with increase of thrust due to high dependence of machine advance rate upon rock strength. For Hieflau expansion tunnel three types of data (TBM, rock mass & seismic data e.g. amplitude, pseudo velocity etc.) were coupled and simultaneously analyzed by plotting 3-D surfaces. No appreciable correlation between seismic data (Amplitude & Pseudo velocity) and rock mass properties and machine parameters could be found. Tool wear as a function of TBM operational parameters was analyzed which revealed that tool wear is minimum if applied thrust is moderate and that tool wear is high when thrust is too low or too high. An empirical linear model for advance rate was predicted with a high accuracy. On the other hand, in Hemerwald tunnel thrust and AR have the same correlation as in Hieflau.

A significant correlation between machine parameters and rock mass properties was found. An empirical linear equation with great accuracy was achieved to predict AR as a function of different rock mass properties and machine parameters. After analyzing the data from seven tunnel sites, based on rock strength, fracture class and behavior of thrust versus advance rate (AR), seven case histories have been divided into two major groups.

Group one consists of Hieflau, Hemerwald, Maen and Pieve tunnels. Rock mass strata mainly comprise of Limestone, Schistose-Gneiss, Micaschist and Meta-granite. For group one rock strength (UCS) ranges from 162-226 MPa, that contains high strength rocks. In this group AR decreases linearly with increase of thrust. Reason for this trend is very clear from data analysis that is due to very high strength, presence of less joints and very low fracture class. For high strength rocks a prediction model for AR may be used with slight variations from case to case. On other hand group two comprises Queen water tunnel, Vereina and Varzo tunnels. Rock mass strata mainly consists of Micaschists and Gneiss.

Rock strength (UCS) varies between 55 - 162 MPa, which is low to medium strength rocks. In group two AR linearly increases with increase of TBM thrust. The reason is low rock strength and presence of medium to high frequency of joints and a high rock fracture class. For the low strength rocks, another AR prediction model is suggested.