Reliability prediction of randomly loaded machine element is a complex issue. This thesis presents a method, which allows the reliability prediction of randomly loaded machine elements in the nodes of the finite element model. The method is based on a static reliability model, which requires knowledge of the probability densities of the load spectrum and fatigue strength at each stress level and the probability density of the equivalent stress amplitude. The probability density of the equivalent stress amplitude is obtained by proposed statistical transformation, with which the information about the distribution of the rainflow matrix is preserved.
D.09 Tutoring for postgraduate students
COBISS.SI-ID: 14419739The thesis provides a methodology to estimate parameters of constitutive material models that consider also the strain rate effects. The procedure for estimating the strain-rate governing material parameters is based on an impact experiment in which a ball indentates a flat metal sheet at high speed. A testing device was designed and mounted on a testing machine that was originally built for the burst tests of supercharger structures. In our aproach we linked a method for the design of experiments, a finite element code, a novel modelling of a response surface and a numerical optimisation algorithm in order to determine the strain rate dependent parameters of the Cowper-Symonds and Johnson-Cook material models, which are often used in complex numerical evaluations of impact structural loading. A comparison of the simulation and experimental results was carried out as well as a comparison between the estimated material-parameter values and the values from literature. The presented method that is a combination of a genetic algorithm optimisation method and global-localy modelled response surface significantly shortens the time for estimating the parameters of the material models and for this reason it is also aplicable on engineering field.
D.09 Tutoring for postgraduate students
COBISS.SI-ID: 14725403SIEMENS company has invited us to join a project the aim of which is to develop procedures for the prediction of the durability of lithium-ion batteries. It turns out that the damage mechanism that leads to the failure of the battery is fatigue. We have developed an algorithm that enables the modelling of the capacity of the battery as a function of voltage, current and temperature. We have presumed that voltage, current and temperature are random variables. Further on we have developed an algorithm that enables the conversion of voltage and the capacity of the battery into a damage parameter. Based on the calculated history of the damage parameter, the new procedure, based on Prandtl operators, can be used to predict the durability of the battery. We have tested the method preliminarily.
F.06 Development of a new product
COBISS.SI-ID: 15077403For different versions of trailing arms for our business partner a fatigue-life was numerically predicted. The basic material was AHSS sheet-metal. For the fatigue-life predictions its S-N curve with a scatter was experimentally determined. The fatigue life of welded joints was predicted according to EUROCODE 3-1.9. For the most critical weld (seam weld of two overlaping sheet metals) the standard S-N curves were checked by fatigue-life experiments. The selected stress concentration detail from EUROCODE 3-1.9 was validated by experiments for different slots between the welded sheet-metal plates.
F.17 Transfer of existing technologies, know-how, methods and procedures into practice
COBISS.SI-ID: 15084059In 2016 we started an extremely useful cooperation with Professor Francesca Cosmi, who works at Dipartimento di Ingegneria e Architettura Universita degli Studi di Trieste. Due to the vicinity of Trieste as well as to enrich the studies, we have decided to lecture in turn as invited lecturers on both faculties. Together with Professor Cosmi we also apply for European projects.
B.05 Guest lecturer at an institute/university
COBISS.SI-ID: 15114523