A Combined Experimental and First-Principles Based Assessment of Finite-Temperature Thermodynamic Properties of Intermetallic Al3Sc

We present a first-principles assessment of the finite-temperature thermodynamic properties of the intermetallic Al3Sc phase including the complete spectrum of excitations and compare the theoretical findings with our dilatometric and calorimetric measurements. While significant electronic contributions to the heat capacity and thermal expansion are observed near the melting temperature, anharmonic contributions, and electron–phonon coupling effects are found to be relatively small. On the one hand, these accurate methods are used to demonstrate shortcomings of empirical predictions of phase stabilities such as the Neumann–Kopp rule. On the other hand, their combination with elasticity theory was found to provide an upper limit for the size of Al3Sc nanoprecipitates needed to maintain coherency with the host matrix. The chemo-mechanical coupling being responsible for the coherency loss of strengthening precipitates is revealed by a combination of state-of-the-art simulations and dedicated experiments. These findings can be exploited to fine-tune the microstructure of Al-Sc-based alloys to approach optimum mechanical properties.

Phase-field model of precipitation processes with coherency loss

A phase-field model is proposed to simulate coherency loss coupled with microstructure evolution. A special field variable is employed to describe the degree of coherency loss of each particle and its evolution is governed by a Ginzburg-Landau type kinetic equation. For the sake of computational efficiency, a flood-fill algorithm is introduced that can drastically reduce the required number of field variables, which allows the model to efficiently simulate a large number of particles sufficient for characterizing their statistical features during Ostwald ripening. The model can incorporate size dependence of coherency loss, metastability of coherent particles, and effectively incorporate the underlying mechanisms of coherency loss by introducing a so-called differential energy criterion. The model is applied to simulate coarsening of Al3Sc precipitates in aluminum alloy and comprehensively compared with experiments. Our results clearly show how the particle size distribution is changed during coherency loss and affects the coarsening rate.

Impact of Local Site Conditions on Simulation of Non-stationary Spatial Variable Seismic Motions

It is commonly accepted that multi-scale structures are subject to spatially variable seismic motions. This spatial variability of seismic motions is described by different intensities at different locations due to the coherency loss effect, wave passage effect and local site conditions. For multi-scale structures, the estimation of seismic excitations must consider these factors. Often, the influence of the spatial variability of seismic motion on the dynamic response of structures is performed by neglecting the site effect. In several cases, it has been observed that the high intensities of seismic motion are caused by the site amplification besides coherency loss and wave passage effects. This study aims to analyze the impact of local site conditions on seismic motions. For this purpose, a method of simulation of spatially variable seismic motions is performed. The seismic signals on the bedrock are defined by considering a target power spectral density and a coherency loss model. According to the seismic wave propagation theory, the projection of these seismic motions on the surface is realized. The results of this study show that neglecting the local site conditions induces an undervaluation of spatially variable seismic excitations.

ProPaCoL-Net: A Novel Recursive Stereo Image SR Network with Progressive Parallax Coherency Learning

Recently, stereo cameras have been widely packed in smart phones and autonomous vehicles thanks to low cost and smallsized packages. Nevertheless, acquiring high resolution (HR) stereo images is still a challenging problem. While the traditional stereo image processing tasks have mainly focused on stereo matching, stereo super-resolution (SR) has drawn less attention which is necessitated for HR images. Some deep learning based stereo image SR works have recently shown promising results. However, they have not fully exploited binocular parallax in SR, which may lead to unrealistic visual perception. In this paper, we present a novel and computationally efficient convolutional neural network (CNN) based deep SR network for stereo images by learning parallax coherency between the left and right SR images, which is called ProPaCoL-Net. The proposed ProPaCoL-Net progressively learns parallax coherency via a novel recursive parallax coherency (RPC) module with shared parameters. The RPC module is effectively designed to extract parallax information in prior for the left image SR from its right view input images and vice versa. Furthermore, we propose a parallax coherency loss to reliably train the ProPaCoL-Net. From extensive experiments, the ProPaCoL-Net shows to outperform the very recent state-of-the-art method with average 1.15 dB higher in PSNR.