The propagation of delamination in a layered anisotropic cylinder is studied under different loading conditions. Initially, a closed-form solution is obtained for an infinitely long cylinder by treating it as a thin shell, with an axisymmetric crack in the adhesive layer between two laminae, and an axisymmetric loading acting as the driving force. The axial and shear strains of the reference plane are considered negligible compared to its circumferential strain. The cylinder, after delamination, is modelled as three shell elements joined at the crack tip. A sharp, flat crack tip is assumed, and the slopes of the three shell elements at the crack tip are set to zero. The radial displacements of the three regions, obtained through the imposition of the proper boundary and matching conditions, are substituted into Clapeyron's theorem to obtain the total strain energy in the cylinder. The change in this strain energy for a unit change in the crack length is used in the Griffith energy balance criterion to evaluate the critical loading at which the delamination propagates. The critical loading is studied as a function of the half crack length. Due to the assumption of a flat crack tip, only the moment continuity across the crack tip is imposed. Normal shear continuity is not enforced and this leads to interesting results in a few cylinders. "Crack arrest" regions, or ranges of crack length within which no real value of the loading could make the delamination propagate, are observed in some cases. Two different loading conditions are treated: one with a pressure acting on the crack surface, and the other with a loading on the inner cylindrical surface. "Crack arrest" phenomenon occurs in the latter problem for certain layer thickness ratios. As the next step, a more general crack tip geometry is assumed and shear force continuity across the crack tip is enforced. This model establishes the fact that the abnormal delamination behaviour in the previous model, like the "crack arrest", is due to the absence of normal shear continuity across the crack tip. It also shows that the crack tip assumption made in the previous formulation - a flat, sharp crack - is valid. When the crack tip geometry is generalized the radial displacements in the cylinder are obtained numerically. Four different crack tip geometries are assumed. The axial and shear strains in the reference plane are included in a subsequent model to determine their effect on the delamination behaviour. A general observation is that the critical value of the loading is lowered for any initial crack length. The magnitude of the drop depends on the geometric and material properties of the plies that constitute the layered cylinder. When the loading is applied on the inner cylindrical surface, the effect is predominant for very small crack lengths. On the contrary, the decrease in the critical value of the pressure acting on the crack surface is considerable for large crack lengths. / Ph. D.

The Winland R35 method [Log R35 = 0.732 + 0.588 (Log Kair) 0.864 (Log O)] is based on the relationship between porosity, permeability, and pore throat radius at the point of 35% mercury saturation in capillary pressure measurements and is generally reliable in rocks with only intergranular porosity (such as sandstone) where pore and pore throat geometry are related closely to rock texture. Carbonate pores are not always so; consequently, the Winland method is not as reliable for assessing reservoir quality in carbonate reservoirs. To evaluate alternatives to the conventional Winland technique, based on rock facies characteristics, samples from the Jurassic Smackover Formation in Alabama and the Permian Clearfork Formation in Texas were tested for reservoir quality with use of the Winland R35 and Pittman methods to determine if either method is more reliable in carbonate reservoir studies. Pittman's modification of the Winland method was found to be more accurate graphically. A third method for evaluating reservoir rock character is provided by Lucia. This method is based on geological rather than petrophysical characteristics, and it revealed that pore throat sizes at 35% mercury saturation may include a variety of depositional and diagenetic rock fabrics. The Winland and Pittman petrophysical evaluation techniques, as well as the Lucia geological evaluation technique - when based on depositional facies alone - do not provide reliable measures of reservoir quality. An alternative method based on genetic pore type presented by Ahr in 2005 was tested for comparison. Using a porosity-permeability plot based on the pore type, the relationship between porosity, permeability, and pore type was found to be strong and reproducible. When the ratio of permeability to porosity was used in combination with Ahr genetic pore types, the results indicate that barriers, baffles, and flow units can be reliably defined. This study demonstrates that the use of pore types in conjunction with capillary pressure measurements is a more reliable method for evaluating carbonate reservoirs than any alternative method that is based on depositional facies or rock fabrics alone.

English The thesis explores the topic of cost planning of Public-Private-Partnership (PPP) projects in the Czech Republic, especially with respect to institutional settings and road infrastructure. First, the PPP concept is introduced from a theoretical perspective and compared to traditional public procurement. Then the financing of PPP projects is discussed in the context of project finance and the European PPP market. The main part of the thesis focuses on public procurement of road infrastructure and the advantages of the PPP organizational structure. Initially, flaws of the procurement institutions are identified and a number of solutions suggested. The solutions cover four main areas: improvement of procurement laws, better qualifications of public officials, strategic planning of needed roads and asset management perspective on the existing infrastructure. The question whether Czech institutions are hindering the potential of PPP projects is answered positively. Follows a thorough empirical analysis of a World Bank PPP model for highways through a Monte Carlo simulation. A particular case of R35 motorway is evaluated as a PPP project and key drivers of public and private NPV are identified and compared across three different scenarios. Heavyweight traffic intensity, its toll revenue and...