Comparison of existing performance prediction models for hard rock tunnel boring based on data collected at the Follo Line Project
MetadataShow full item record
In Norwegian hard rock tunnelling history, drill and blast has mainly been the excavation method. At the Follo Line Project, which will be the longest railway tunnel in Scandinavia when it opens in 2021, tunnel boring machines have been applied. Before and during construction of a tunnel project, a correct prediction of penetration rate is important, and the choice of the most accurate prediction model is crucial. The overall purpose of this thesis is to compare existing performance prediction models with the aim of finding the most accurate model for hard rock tunnel boring based on data collected at the Follo Line Project. To detect the most accurate model, the penetration rates have been calculated by using several models and further compared with the actual penetration rate achieved at the Follo Line Project. The prediction models used are listed below: NTNU model by Bruland (2000) NTNU model by Macias (2016) CSM model by Rostami (1997) MCSM model by Yagiz (2002) Gehring model by Gehring (1995) Alpine model by Wilfing (2016) Qtbm model by Barton (2000) Model by Yagiz (2008) Model by Hassanpour et al. (2011) Model by Farrokh et al. (2012) In general, most of the performance prediction models show promising results compared to the achieved penetration rate. The NTNU models and the Alpine model turned out to be the most accurate ones. If conservative results are sought for, the MCSM-model and the model by Yagiz show promising results. One of the secondary scopes was to determine the most influential parameters, both related to the achieved and the predicted penetration rate. A sensitivity analysis of the various parameters has been performed for this purpose. The outcome of this analysis shows that the applied cutter thrust, rock mass fracturing and uniaxial compressive strength are the most influential parameters on the penetration rate, both in reality and in the models. To strengthen the accuracy and reliability of the predictions, it is recommended to use more than one prediction model in the calculations. Thus, the sources of error connected to the models will be limited. The Follo Line Project is a well-suited project to compare hard rock prediction models due to the varying geology throughout the tunnel alignment.