Effects of temperature and grain size on deformation and fracture in recrystallized Ni3Al doped with boron

The effects of temperature and grain size on the deformation and fracture behaviour of recrystallized Ni₃Al doped with boron were investigated by tensile tests at temperatures up to 973 K as a function of grain sizes from 1.6 to 105m. The yield stress showed a positive temperature dependence to a peak temperature in somewhat different manners depending on the grain size. For coarse-grained specimens, a rapid drop in elongation was observed with increasing temperature. The predominant fracture mode changed with temperature from the transgranular fracture of {1 1 1} cracking to brittle intergranular fracture. This embrittlement at elevated temperatures was considered to occur by a high stress concentration at grain boundaries arising from increased flow stress level and the occurrence of grain boundary sliding (GBS). In contrast, the elongation was not so markedly decreased with temperature for intermediate- and fine-grained specimens which exhibited ductile intergranular fracture and cavitation fracture, respectively, at elevated temperatures, and a slant-type fracture and cup-cone fracture, respectively, at low temperatures. The suppression of serious high-temperature embrittlement for intermediate-grained specimens was explained in terms of the slow propagation of a crack formed by GBS, owing to stress relaxation by dynamic recrystallization (DR) and plastic deformation. In the case of ultra-fine-grained specimens a large elongation was developed at elevated temperatures, which was interpreted as that the further occurrence of DR with increasing volume fraction of grain boundaries reduces the cavitation promoted by GBS, and that the limited sliding length due to extremely small grain diameter raises the stress for cavity formation.     

Visual tracking using convolutional features with sparse coding

Artificial Intelligence Review - Visual object tracking has become one of the most active research topics in computer vision, and it has been applied in several commercial...     

Support NNEF execution model for NNAPI

The Journal of Supercomputing - With growing applications such as image recognition, speech recognition, ADAS, and AIoT, artificial intelligence (AI) frameworks are becoming popular in various...     

Deep neural network-based fusion model for emotion recognition using visual data

The Journal of Supercomputing - In this study, we present a fusion model for emotion recognition based on visual data. The proposed model uses video information as its input and generates emotion...     

High-performance simulations of turbulent boundary layer flow using Intel Xeon Phi many-core processors

The Journal of Supercomputing - Direct numerical simulations (DNS) of turbulent flows have increasing importance because they not only provide fundamental understanding of turbulent flows but also...     

State estimation and experimental verification of a robotic vehicle with six in-wheel drives using Kalman filter

Microsystem Technologies - In this research we present an algorithm for a six-wheeled robotic vehicle with articulated suspension (RVAS) to estimate the vehicle velocity and acceleration states,...     

Modeling and control of a finger-like mechanism using bending shape memory alloys

Microsystem Technologies - In this research a biologically inspired finger-like mechanism similar to human musculoskeletal system is developed based on Shape Memory Alloys (SMAs). SMA actuators are...     

Contact printing pressure uniformization in roll-to-roll process using individual drive cross-coupled torque control

Microsystem Technologies - Printed electronics such as solar cells, RFIDs, and display panels can be made using printed electronic technology by printing viscous liquids having various properties...     

Multitask learning for health condition identification and remaining useful life prediction: deep convolutional neural network approach

Predicting remaining useful life (RUL) is crucial for system maintenance. Condition monitoring makes not only degradation data available for RUL estimation but also categorized health status data for health state identification. However, RUL prediction has been treated as an independent process in most cases even though potential relevance exists with health status detection process. In this paper, we propose a convolution neural network based multi-task learning method to reflect the relatedness of RUL estimation with health status detection process. The proposed method applied to the C-MAPSS dataset for aero-engine unit prognostics supported superior performances to existing baseline models.

     

A run-to-run controller for a chemical mechanical planarization process using least squares generative adversarial networks

Achieving high processing quality for chemical mechanical planarization (CMP) in semiconductor manufacturing is difficult due to the distinct process variations associated with this method, such as drift and shift. Run-to-run control aims to maintain the targeted process quality by reducing the effect of process variations. The goal of controller learning is to infer an underlying output–input reverse mapping based on input–output samples considering the process variations. Existing controllers learn reverse mapping by minimizing the total mapping error for sample data. However, this approach often fails to generate inputs for unseen target outputs because conditional input distributions on target outputs are not captured in the learning. In this study, we propose a controller based on a least squares generative adversarial network (LSGAN) that can capture the input distributions. GANs are deep-learning architectures composed of two neural nets: a generator and a discriminator. In the proposed model, the generator attempts to produce fake input distributions that are similar to the real input distributions considering the process variation features extracted using convolutional layers, while the discriminator attempts to detect the fake distributions. Competition in this game drives both networks to improve their performance until the generated input distributions are indistinguishable from the real distributions. An experiment using the data obtained from a work-site CMP tool verified that the proposed model outperformed the comparison models in terms of control accuracy and computation time.