Publication details
Structure orientation analysis of Nové spojení Tunnel (Vítkov hill, Barrandian, Czech Republic)
| Authors | |
|---|---|
| Year of publication | 2014 |
| Type | Conference abstract |
| MU Faculty or unit | |
| Citation | |
| Description | Discontinuity data of the Southern railway tunnel of Nové spojení were analysed in order to characterize structure orientation of Vítkov hill. Tunnels of Nové spojení encounter Ordovician Šárka and Dobrotivá Formations. This locality is close by Prague Fault and sedimentary rocks in tunnels are deformed by it. Over 3000 major bedding planes and joint set traces were selected as representative data for statistical analysis of discontinuity strike. The tunnel was divided into total of 25 sections, 50 m each. However Western tunnel terminal section turned out to be less than 20 m and Eastern ending section 51 m. Twenty-five individual data sets were created from strike measurement of each section. In order to determine data sets’ properties, strike measurements were divided into 12 intervals, 30° each. Strike frequency and cumulative relative frequency was calculated. All of twenty-five data sets shown non-Gaussian distribution. Five of twenty-five data set didn’t met Chi-Square Test conditions for expected intervals counts. On the basis of data sets properties and analysis purpose, Kolmogorov-Smirnov Test was selected as alternative to more common Chi-Square Test. Kolmogorov-Smirnov Test is a non-parametric test for one dimensional probability distribution. Test permits the comparison of two data sets without making assumptions about their distributions. Assuming discontinuities have similar strike distributions for statistical homogeneity of domains, adjacent sections were compared. In addition to Komogorov-Smirnov Test, every result for two adjacent sections was verified by visual comparison of corresponding pair of contoured diagrams. As a result of statistic analysis of discontinuity strike, Southern tunnel of Nové spojení was divided into five domains. In every domain, strikes measurements form three to four dominant discontinuity clusters. These dominant clusters correspond with assumption that statistically homogenous domain have similar strike distribution. |