Predicting mechanical properties of metals from big data is of great importance to materials engineering. The present work aims at applying artificial neural network (ANN) models to predict the tensile properties including yield strength (YS) and ultimate tensile strength (UTS) on austenitic stainless steel as a function of chemical composition, heat treatment and test temperature. The developed models have good prediction performance for YS and UTS, with R values over 0.93. The models were also tested to verify the reliability and accuracy in the context of metallurgical principles and other data published in the literature. In addition, the mean impact value analysis was conducted to quantitatively examine the relative significance of each input variable for the improvement of prediction performance. The trained models can be used as a guideline for the preparation and development of new austenitic stainless steels with the required tensile properties.

To study the influence of poly(ethylene oxide) (PEO) on the relaxation process of poly(vinyl acetate) (PVAc) in the glass to rubber softening dispersion, the mechanical loss behaviour of PEO/PVAc blends (with PEO contents up to 20 wt%) were measured as a function of temperature and frequency. It is shown that the PEO component has a plasticizing effect on the α and α′ relaxation modes of PVAc, where the α mode is ascribed to the local segmental mode and the α′ mode is related to the motion of longer chain segments in the softening dispersion, composed of the sub-Rouse modes and the Rouse modes. Both the α and α′ modes shifted to lower temperatures with increase of PEO concentration. Time–temperature superposition breaks down for PEO/PVAc blends in the entire temperature range due to the different friction coefficients of the α and α′ modes. A single master curve of the α′ mode for pure PVAc and the PEO/PVAc blends is obtained, suggesting that the size of chain segments containing on the order of 10 to 50 or more backbone bonds does not change upon blending. Furthermore, the dynamics of the α′ mode of the PVAc component in the blends is found to show a characteristic crossover through the temperature dependence of relaxation time and relaxation strength. The crossover temperature T_B decreases with increasing PEO content, while the crossover relaxation time for the blends is constant with a value about 0.08 s, much longer than 10^{−7 ± 1} s for the α mode. According to the coupling model, the crossover is suggested to be due to the variation of intermolecular coupling at T_B.

Bubbles seriously degrade the mechanical properties of tungsten and thus threaten the safety of nuclear fusion devices, however, the underlying atomic mechanism of bubble growth from clustered hydrogen and helium atoms is still mysterious. In this work, first-principles calculations are therefore carried out to assess the stability of tungsten atoms around both hydrogen and helium clusters. We find that the closest vacancy-formation energies of interstitial hydrogen and helium clusters are substantially decreased. The first-nearest and second-nearest vacancy-formation energies close to vacancy–hydrogen clusters decrease in a step-like way to ∼0, while those close to vacancy–helium clusters are reduced almost linearly to ∼−5.46 eV when atom number reaches 10. The vacancy-formation energies closest to helium clusters are more significantly reduced than those nearest to hydrogen clusters, whatever the clusters are embedded at interstitial sites or vacancies. The reduction of vacancy-formation energies results in instability and thus emission of tungsten atoms close to interstitial helium and vacancy–helium clusters, which illustrates the experimental results, that the tungsten atoms can be emitted from the vicinity of vacancy–helium clusters. In addition, the emission of unstable tungsten atoms close to hydrogen clusters may become possible once they are disturbed by the environment. The emission of tungsten atoms facilitates the growth and evolution of hydrogen and helium clusters and ultimately the bubble formation. The results also explain the bubble formation even if no displacement damage is produced in tungsten exposed to low-energy hydrogen and helium plasma.