Lehigh University

About: Lehigh University is a education organization based out in Bethlehem, Pennsylvania, United States. It is known for research contribution in the topics: Catalysis & Fracture mechanics. The organization has 12684 authors who have published 26550 publications receiving 770061 citations.

On the dynamics of near-wall turbulence

Abstract: A model of the dynamic physical processes that occur in the near-wall region of a turbulent flow at high Reynolds numbers is described The hairpin vortex is postulated to be the basic flow structure of the turbulent boundary layer It is argued that the central features of the near-wall flow can be explained in terms of how asymmetric hairpin vortices interact with the background shear flow, with each other, and with the surface layer near the wall The physical process that leads to the regeneration of new hairpin vortices near the surface is described, as well as the processes of evolution of such vortices to larger-scale motions farther from the surface The model is supported by recent important developments in the theory of unsteady surface-layer separation and a number of `kernel' experiments which serve to elucidate the basic fluid mechanics phenomena believed to be relevant to the turbulent boundary layer Explanations for the kinematical behaviour observed in direct numerical simulations of low Reynolds number boundary-layer and channel flows are given An important aspect of the model is that it has been formulated to be consistent with accepted rational mechanics concepts that are known to provide a proper mathematical description of high Reynolds number flow

The Supervisory Working Alliance, Trainee Self‐Efficacy, and Satisfaction

Abstract: Theoretically, when the supervisory working alliance is strong, the trainee and supervisor share a strong emotional bond and agree on the goals and tasks of supervision. Tested was Bordin's (1983) proposition that changes in counselor trainees' perceptions of the supervisory alliance over the course of supervision would predict supervisory outcomes. A national sample of beginning practicum- to intern-level trainees were assessed at the beginning and end of an academic semester. Contrary to predictions, changes in the alliance were not predictive of changes in trainees' self-efficacy. However, improvements in the emotional bond between the trainees and supervisors were associated with greater satisfaction.

Modeling of transconductance degradation and extraction of threshold voltage in thin oxide MOSFET's

Abstract: In this paper, we discuss the low-drain voltage transconductance behavior of the MOSFET due to surface mobility variation, interface states and small geometry, and its application in threshold voltage determination. We modify the Pao-Sah drain current model to incorporate a mobility model and obtain 3% accuracy from subthreshold to very strong inversion for a wide range of substrate biases. The effects of non-ideal scaling, finite inversion layer thickness, surface roughness mobility degradation under high normal electric fields and interface states on the transconductance behavior are discussed. We observe the peak transconductance increases with substrate bias in short-channel devices and decreases with substrate bias in long-channel devices. Finally, we show the threshold voltage can be determined from the gate voltage at which the rate of transconductance change ( ∂g m ∂V GS ) is a maximum. This threshold voltage is identifiable with a known band-bending (surface potential) of the substrate (φ s ⋍ 2φ F + V SB ) , from which the band-bending at all gate biases can be calculated. The transconductance change (TC) method is insensitive to device degradations (e.g. mobility, series resistance, hot-carrier) in contrast to the conventional method of linear extrapolation to zero drain current.

SPDA-CNN: Unifying Semantic Part Detection and Abstraction for Fine-Grained Recognition

Abstract: Most convolutional neural networks (CNNs) lack midlevel layers that model semantic parts of objects. This limits CNN-based methods from reaching their full potential in detecting and utilizing small semantic parts in recognition. Introducing such mid-level layers can facilitate the extraction of part-specific features which can be utilized for better recognition performance. This is particularly important in the domain of fine-grained recognition. In this paper, we propose a new CNN architecture that integrates semantic part detection and abstraction (SPDACNN) for fine-grained classification. The proposed network has two sub-networks: one for detection and one for recognition. The detection sub-network has a novel top-down proposal method to generate small semantic part candidates for detection. The classification sub-network introduces novel part layers that extract features from parts detected by the detection sub-network, and combine them for recognition. As a result, the proposed architecture provides an end-to-end network that performs detection, localization of multiple semantic parts, and whole object recognition within one framework that shares the computation of convolutional filters. Our method outperforms state-of-theart methods with a large margin for small parts detection (e.g. our precision of 93.40% vs the best previous precision of 74.00% for detecting the head on CUB-2011). It also compares favorably to the existing state-of-the-art on finegrained classification, e.g. it achieves 85.14% accuracy on CUB-2011.