Statistical conformity criteria for the compressive strength of concrete are a matter of debate. The criteria can have prejudicial effects on construction quality and reliability. Hence, the usefulness of statistical criteria for the small sample size n = 3 is questioned. These defects can cause a reduction in the quality of produced concrete and, consequently, too much risk for the recipient (investor). For this reason, the influence of conformity control on the value of the reliability index of concrete and reinforced concrete has been determined. The authors limited their consideration to the recommended standards PN-EN 206-1, PN-EN 1992 and ISO 2394 method of reliability index, which belongs to the analytical methods FORM (First Order Reliability Method). It assumes that the random variables are defined by two parameters of the normal distribution or an equivalent normal: the mean and the standard deviation. The impact of conformity control for n = 3 for concrete structures, designed according to the Eurocode 1992, for which the compressive strength of concrete is the capacity dominant parameter (sensitivity factor of dominating resistance parameter according to the FORM is 0.8), has been determined by evaluation of the reliability index.

The paper presents the experimental research and numerical simulations of reinforced concrete beams under torsional load. In the experimental tests Digital Image Correlation System (DIC System) Q-450 were used. DIC is a non-contact full-field image analysis method, based on grey value digital images that can determine displacements and strains of an object under load. Numerical simulations of the investigated beams were performed by using the ATENA 3D – Studio program. Creation of numerical models of reinforced concrete elements under torsion was complicated due to difficulties in modelling of real boundary conditions of these elements. The experimental research using DIC can be extremely useful in creating correct numerical models of investigated elements. High accuracy and a wide spectrum of results obtained from experimental tests allow for the modification of the boundary conditions assumed in the numerical model, so that these conditions correspond to the real fixing of the element during the tests.