Dottorato in INGEGNERIA INDUSTRIALE E GESTIONALE

Titolo della tesi: Smart engineering and implementation of ultimate strength criteria for ductile materials in structural applications

In the last decades, huge efforts have been made to predict the conditions of onset of plastic failure for ductile metal alloys. Mathematical models, called ductile damage models, have been developed in the literature, which estimate the amount of plastic deformation the material can accumulate before reaching fracture under any loading conditions, which results in an estimation of ductility. A demanding calibration strategy is required to use ductile damage models. Several experimental tests including complex multiaxial ones, a post-processing phase via Finite Element to derive the local information that cannot be directly measured from the experimental tests, and considerable theoretical know-how are usually required. This limits the use of damage models on an industrial scale. Indeed, nowadays they are only used in academic research, providing a limited contribution to the field of industrial design which cannot exploit the advantages of employing such highly innovative analysis techniques. To enhance the usability of damage models on an industrial scale, it was proposed to engineer and simplify the tests and calibration procedure, turning the damage criteria into a ready-to-use analysis tool without having to rely on complex experimental setups, non-trivial and time-consuming numerical simulations and specific know-how in the field of damage criteria. Peculiar multiaxial test specimens were designed that allow damage models to be calibrated using only a universal testing machine, simplifying the experimental phase. An analytical-numerical method was developed to directly identify the calibration parameters from the experimental tests in order to circumvent the challenging numerical simulations required to postprocess the executed tests. In addition, novel methodology was proposed to reduce the number of experimental tests required to characterise ductility, which also provides information on the minimum guaranteed strength of materials. Furthermore, a database was compiled containing information on the elasto-plastic behaviour up to fracture of the most widely used metal alloys. Finally, the different solutions developed in the thesis will then be merged into a numerical software tool tailored to the industrial user which will prove to be a valid design aid in the structural assessment of parts and mechanical systems. Concerning the results achieved, the performance of the investigated multiaxial specimens is remarkable, indeed they are able to generate multiaxial tensile-shear stress states under static and dynamic conditions by applying a uniaxial load. Therefore they can replace the mechanical tests that require a complex experimental setup. Reliable information on the stress and strain state are obtained from the developed analytical-numerical expressions, whose resulting accuracy is fairly good and leads to equivalent ductility information compared to those derived from finite element analyses. In addition, it was proven that an initial ductility estimation of material can be achieved only conducting a tensile test thanks to the new simplified expressions of damage criteria and developed material database, which it is suitable in all those scenarios where high accuracy is not strictly required. Furthermore, the designed numerical framework hides to the end user the complexity of the damage models and, due to its scalability features to control the required level of accuracy. To conclude, the results obtained in this research are promising, and prove the possibility of engineering the calibration procedure and the use of damage models, enhancing the technology transfer from the academic research field to the industrial world.