Prediction of rock mass rating using TSP method and statistical analysis in Semnan Rooziyeh spring conveyance tunnel. (September 2018)

• Record Type: Journal Article
• Authors: Esmailzadeh, Akbar; Mikaeil, Reza; Shafei, Erfan; Sadegheslam, Golsa
• Is Part Of: Tunnelling and underground space technology. Volume 79(2018)
• Page Start: 224
• Page End: 230
• Publication Date: 2018-09
• Subjects: TSP method -- Statistical modelling -- Water conveyance tunnel -- Rock mass rate; Tunneling -- Periodicals; Underground construction -- Periodicals; Tunnels -- Periodicals; Underground areas -- Periodicals; 624.193
• DOI: 10.1016/j.tust.2018.05.001
• Languages: English
• ISSNs: 0886-7798

Abstract

Highlights: Geological hazard assessment of the multiple tunneling paths are introduced. Progress paths of the tunnel and project quality are investigated using combined TSP and statistical modeling methods. Nonlinear correlation models with case-sensitive parameters are extracted. The prediction reliability of presented study in evaluation of rock mass quality for tunnel advancing path in approved. Abstract: In this study, the rock mass rating (RMR) value is assessed for a case-defined water transfer project, namely Rooziyeh spring. The progress paths of the tunnel and project quality are investigated using combined TSP and statistical modeling methods. The accuracy and validity of results are verified with respect to experimental RMR values obtained in the site plan. The developed models are investigated using the t- and F-statistical tests (with the confidence level of 99%), the determination coefficient and the estimation error. The investigation results showed that the developed models can reliably predict the RMR values based on the TSP test results of the Rooziyeh spring tunnel path. In addition, in order to validate the developed statistical models and to generate the prediction method to the other tunnel cases, the datasets of Koohrang water conveyance tunnel are provided as inputs. Therefore, the RMR and TSP results are studied in 11 zones of this tunnel with RMR values of 40–73. The average prediction error for linear and nonlinear models is 43% and 22%, respectively. Based on results, the nonlinear statistical models provide the highest accuracy than other linear models and thus are approved as accurate statistical models.