In this paper, flysch is presented as a representative material of a wide section of the Carpathian Mountains, with some areas in Poland highlighted. The geological structure of this area is complex due to the alternating layers of blocky rock masses and soil (Vessia et al., 2017). Such a complex pattern is seen in some Alpine flysch slopes, such as the Ingelsberg landslide area (Romeo et al., 2015). Many authors are monitored, predicted landslides (Allasia et al., 2013; Bertacchini et al., 2009; Casagli et al., 2010) by sophisticated sensors. The rock-soil flysch successions have become intensively fissured as a result of their geological history, weathering (precipitation and snowmelt), and long-term water retention, especially on the surface layers. These complex materials are characterised by heterogeneous lithologies, whose mechanical properties are largely uncertain. These geological structures have also been confirmed by monitoring and control studies performed on a large number of landslides (Bednarczyk, 2014). One of the most striking phenomena is the sudden decrease in the strength parameters in the studied rocks in the direction parallel to the layers due to watering. The process is made possible by heterogeneous fractured strong rock layers with high permeability coefficients for water. This study precisely describes the phenomena occurring at the contact area between the component layers of flysch under the wet conditions of a weak plane. An elastic-plastic analysis method that considers the developed strength model at the surfaces of the contact areas (Biernatowski & Pula, 1988; Pula, 1997) has been used to estimate the load capacity for piles working under a horizontal load. The piles are part of a reliability chain (Pula, 1997) in a given construction and are the first element of concern for monitoring (Muszynski & Rybak, 2017). A particular device intended to study the dependence of the shear stress on a fixed failure surface in a controlled consolidation condition was utilized. The study was conducted for a wide range of displacements and for different values of stabilized vertical stresses of consolidation. The complexity of the processes occurring in the shear zone, presented as a detailed study of the material crack mechanics, is highlighted. The laboratory results were used to construct the mechanical model of the slip surface between the soil and rock with the description supported by a neural network (NN) approximation. The artificial NN was created as a multi-layered, easy to use approach for interpreting results and for quick reconstruction of approximated values useful for the calculations presented in laterally loaded piles. For the calculations, long, sheared strips of material were considered in a semi-analytical procedure to solve a differential equation of stability. The calculations are intended to reveal the safety indexes for a wide range of boundary tasks as the most significant indicator for design decisions.
The road tunnel in Laliki was excavated in highly heterogeneous, severely tectonically damaged and mainly very weak rocks of the Western Carpathians flysch. In particular, the conditions were characterized by a high percentage of very weak laminated shale and weathered rock mass, an unfavorable and very steep slope of the rock layers and unstable hydrological conditions with outflows of water in loosened tectonic zones. That structure and properties of the rock mass highly uncertain. This paper describes the influence of geological engineering and geotechnical conditions on the primary lining of a main road tunnel. The deformation of the primary lining was analyzed in terms of the percentage share of sandstones and shale, geomechanical classifications RMR (Bieniawski 1989) and QTS (Tesar 1979), types of the primary lining and the use of rock bolts and micropiles. The analysis was preceded by characterization of geological engineering conditions and technological characterization of applied primary linings. Displacements of the primary lining, greater than acceptable, occurred several times in a top heading during tunneling. The primary lining was reinforced by additional rock bolts and wire mesh, a thicker layer of shotcrete and micropiles if deformation reached the emergency state for some types of linings and they didn't indicate any tendency for stabilization. The reinforcement was used until the deformation stabilization was achieved. In the most difficult conditions, the lining was reinforced by a longer micropile umbrella. Parameters for the primary lining were selected on the basis of ongoing geological engineering and geotechnical measurements, in accordance with NATM's principles. The rock mass around the tunnel in Laliki is an example of weak carrying capacity. The observed displacements in the rock mass indicate that the disturbed zone around the tunnel was heavily developed. The primary lining used in such conditions must bear a relatively high load capacity from overlying loosened material and therefore the lack of interaction with the surrounding rock mass should be assumed. The data obtained indicate that the use of the primary lining in the highly variable conditions in the Carpathian flysch requires accurate geological engineering and geotechnical analysis during the day-to-day process of tunneling in order to verify the projected assumptions. The primary linings should be reinforced as needed based on the results of geotechnical measurements, monitoring the interaction between the rock mass and the system of lining.