Complex light-matter interactions in azobenzene polymers have limited our understanding of how photoisomerization induces deformation as a function of the underlying polymer network and form of the light excitation. A unified modeling framework is formulated to advance the understanding of surface deformation and bulk deformation of polymer films that are controlled by linear or circularly polarized light or vortex beams. It is shown that dipole forces strongly respond to polarized light in contrast to higher order quadrupole forces that are often used to describe surface relief grating deformation through a field gradient constitutive law. The modeling results and comparisons with a broad range of photomechanical data in the literature suggest that the molecular structure of the azobenzene monomers dramatically influences the photostrictive behavior. The results provide important insight for designing azobenzene monomers within a polymer network to achieve enhanced photo-responsive deformation.
Restricted Boltzmann Machines (RBMs) have been proposed for developing neural networks for a variety of unsupervised machine learning applications such as image recognition, drug discovery, and materials design. The Boltzmann probability distribution is used as a model to identify network parameters by optimizing the likelihood of predicting an output given hidden states trained on available data. Training such networks often requires sampling over a large probability space that must be approximated during gradient based optimization. Quantum annealing has been proposed as a means to search this space more efficiently which has been experimentally investigated on D-Wave hardware. D-Wave implementation requires selection of an effective inverse temperature or hyperparameter ( $$\beta $$ ) within the Boltzmann distribution which can strongly influence optimization. Here, we show how this parameter can be estimated as a hyperparameter applied to D-Wave hardware during neural network training by maximizing the likelihood or minimizing the Shannon entropy. We find both methods improve training RBMs based upon D-Wave hardware experimental validation on an image recognition problem. Neural network image reconstruction errors are evaluated using Bayesian uncertainty analysis which illustrate more than an order magnitude lower image reconstruction error using the maximum likelihood over manually optimizing the hyperparameter. The maximum likelihood method is also shown to out-perform minimizing the Shannon entropy for image reconstruction.
11 July 2016 in Superconductor Science and Technology
PIT and RRP® Nb3Sn strands are being developed for high field accelerator magnet upgrades for the high luminosity LHC. Here we report a quantitative study of the shape and position of PIT filaments and RRP® sub-elements after rolling lengths of unreacted PIT and RRP® round wires to simulate cabling deformation. In the as-drawn condition, filament shape distortion occurs preferentially in the outer ring filaments. By contrast, rolling induces non-uniform shear bands that generate greater distortion of inner ring filaments. By making a full digitization of the shapes of all filaments, we find that a critical distortion occurs for thickness reductions between 10% and 20% when filament shapes in inner filament rings heavily degrade, especially in the vicinity of the strong 45° shear bands imposed by the rolling. It is well known that maintaining diffusion barrier integrity is vital to retaining adequate RRR in the stabilizing copper needed for magnet stability. Diffusion barrier breaks occur preferentially in these distorted inner filaments and drive local Sn leakage during reaction, increasing RRR degradation.
25 May 2017 in Journal of Intelligent Material Systems and Structures
The effect of flexoelectricity on the formation and evolution of domain structures in ferroelectric materials is developed by integrating strain gradient theory into a finite element phase field model. Length scales associated with elastic strain gradients and the corresponding polarization gradient across a domain wall are integrated into the time-dependent Ginzburg–Landau theory and numerical simulated. Theoretical relations of a shear strain gradient along an electrode/dielectric interface are first solved and verified numerically using the finite element model. A singularity in the strain gradient–induced polarization is shown to occur as the elastic strain gradient length scale approaches a flexoelectric length scale. The theory and finite element modeling is then extended to quantify strain gradient electromechanics near 180° and 90° tetragonal phase domain structures. It is shown that the strain gradient length scale strongly influences changes in strain across the domain walls but has a negligible effect on the polarization. Strain gradient effects become negligible when , where is the polarization domain wall length scale.
8 October 2021 in Journal of Intelligent Material Systems and Structures
The quantification of uncertainty in intelligent material systems and structures requires methods to objectively compare complex models to measurements, where the majority of cases include multiple model outputs and quantities of interests given multiphysics coupling. This creates questions about constructing appropriate measures of uncertainty during fusion of data and comparisons between data and models. Novel materials with complex or poorly understood coupling can benefit from advanced statistical analysis to judge models in light of multiphysics data. Here, we apply the Maximum Entropy (ME) method to more complicated ferroelectric single crystals containing domain structures and soft electrostrictive membranes under both mechanical and electrical loading. Multiple quantities of interest are considered, which requires fusing heterogeneous information together when quantifying the uncertainty of lower fidelity models. We find that parameters, which were initially unidentifiable using a single quantity of interest, become identifiable using multiple quantities of interest. We also show that posterior densities may broaden or narrow when multiple data sets are fused together. This is likely due to conflict or agreement, respectively, between the different quantities of interest and the multiple model outputs. Such information is important to advance our predictions of intelligent materials and structures from multi-model inputs and heterogeneous data.