Multiscale experimental characterization and modelling validation of microstructure and mechanical properties of engineering alloys

Abstract

Abstract Although Al-Cu alloys were discovered more than one century ago and have been widely used since then in structural applications, there are still a lot of uncertainties in the quantitative understanding of the processing-microstructure-mechanical properties link for this system as well as for many other engineering alloys. Current developments in simulation tools (first principles calculations, molecular mechanics, phase-field modeling, dislocation dynamics, crystal plasticity) as well as in multiscale modeling strategies (transition state theory, homogenization) are trying to bridge this gap and provide bottom-up strategies that are able to predict the mechanical performance of alloys based on the composition and processing steps. Nevertheless, one important limitation of the current efforts for the development of multiscale models to predict the microstructure induced by processing and the mechanical properties associated with the microstructure is the lack of controlled experimental data sets. They are necessary to establish rigorously the processing-microstructure-mechanical properties link and to validate or disprove the results of multiscale models. A multiscale experimental characterization approach and subsequent model validation is used in this thesis to study the microstructural development during solidification as well as a result of thermal treatments after casting in Al-Cu alloys. The Al-Cu system was selected for the thesis because its importance for the viewpoint and also because of the richness of its microstructure, which includes a variety of metastable precipitates. The processing conditions have been controlled as accurately as possible to provide quantitative results of the microstructure (primary dendrite arm spacing, precipitate size, shape, volume fraction and spatial distribution) for each condition. Moreover, the information of this data set was used to rationalize the predictions of the microstructural development during directional solidification and of the nucleation and growth of precipitates provided by state-of-the art phase field models. Directional solidification experiments in two Al-Cu alloys have been carried out in an in house built mold in order to perform a more industrial like solidification process. The temperature was monitored with 7 thermocouples placed along the mold and the heat extraction was performed on one end of the mold, while the rest was thermally isolated. The solidification conditions were extracted from the thermocouple readings and the primary dendrite arm spacing was measured from cross sections along the ingot. The results obtained with this set-up were in agreement with those in literature, validating this approach to study directional solidification. Moreover, the primary dendrite stable range was calculated through a phase field model. These results were used to validate a dendrite needle network model that was used to predict the dendritic stable range for the alloy with higher Cu content. The effect of the thermal treatments have been addressed by studying the microstructure-properties link regarding the effect of precipitates on the mechanical properties of Al-Cu alloys. To that end, microstructural analysis of samples with different heat treatments have been performed. Vickers hardness, transmission electron microscopy and differential scanning calorimetry have been used to extract the precipitate sequence as well as the size, shape, volume fraction and spatial distribution of the precipitates. The microstructural information was also used to validate a phase field model of nucleation of growth of precipitates. With this microstructural information, samples with only one type of precipitates were selected to perform study the mechanical properties in order to assess the effect of individual precipitates on the strengthening. Mechanical properties of single crystals were determined by means of micropillar compression test in crystals with different orientations. The results of the mechanical tests, together with detailed microscopic analysis of the deformed samples, were used to determine the individual contribution of the different precipitates to the strength of the alloy. Moreover, this information was used to validate the current models of precipitation strengthening based in either atomistic or dislocation dynamics simulations. Finally, the experimental stress-strain curves from micropillar compression tests have been used to calibrate a crystal plasticity model to predict the mechanical performance of polycrystals, showing the potential of nanomechanical testing techniques in combination with crystal plasticity model to design new alloys.