A Lennard-Jones Embedded-Atom Potential and Its Application to the Study of Melting

The nearest-neighbour Lennard-Jones potential from the embedded-atom method is extended to a form that includes more than nearest neighbours. The model has been applied to study melting with molecular dynamics. The calculated melting point, fractional volume change on melting, heat of fusion and linear coefficients of thermal expansion are in good agreement with experimental data. We have found that the second and third neighbours influence the melting point distinctly.

Embedded cell model of three-dimensional two-phase composites

An embedded cell model is presented to obtain the effective elastic moduli and the elastic-plastic stress-strain relations of three-dimensional two-phase particulate composites. Each cell consists of an ellipsoidal inclusion surrounded by a finite ellipsoidal matrix that embedded in an infinite matrix. When both matrix and particle are elastic, the effective elastic moduli are derived which is an exact analytic formula without any simplified approximation that can be expressed in an explicit form. Further, the elastic-plastic stress-strain relations are obtained for spherical cells and oblate spheroid cells, in which the matrix is elastic and the particle is elastic-plastic. In addition, the macroscopic elastic-plastic constitutive relation of particle reinforced composites (PRC) is investigated by a systematic approach [1] in which the matrix is elastic-plastic and the particle is elastic.

Direct parameter extraction in feedthrough-embedded capacitive MEMS resonators

This paper presents a method for fast and accurate determination of parameters relevant to the characterization of capacitive MEMS resonators like quality factor (Q), resonant frequency (fn), and equivalent circuit parameters such as the motional capacitance (Cm). In the presence of a parasitic feedthrough capacitor (CF) appearing across the input and output ports, the transmission characteristic is marked by two resonances: series (S) and parallel (P). Close approximations of these circuit parameters are obtained without having to first de-embed the resonator motional current typically buried in feedthrough by using the series and parallel resonances. While previous methods with the same objective are well known, we show that these are limited to the condition where CF ≪ CmQ. In contrast, this work focuses on moderate capacitive feedthrough levels where CF > CmQ, which are more common in MEMS resonators. The method is applied to data obtained from the measured electrical transmission of fabricated SOI MEMS resonators. Parameter values deduced via direct extraction are then compared against those obtained by a full extraction procedure where de-embedding is first performed and followed by a Lorentzian fit to the data based on the classical transfer function associated with a generic LRC series resonant circuit. © 2011 Elsevier B.V. All rights reserved.

Scaling the iHMM: Parallelization versus Hadoop

This paper compares parallel and distributed implementations of an iterative, Gibbs sampling, machine learning algorithm. Distributed implementations run under Hadoop on facility computing clouds. The probabilistic model under study is the infinite HMM [1], in which parameters are learnt using an instance blocked Gibbs sampling, with a step consisting of a dynamic program. We apply this model to learn part-of-speech tags from newswire text in an unsupervised fashion. However our focus here is on runtime performance, as opposed to NLP-relevant scores, embodied by iteration duration, ease of development, deployment and debugging. © 2010 IEEE.

Effective elastic moduli of inhomogeneous solids by embedded cell model

An embedded cell model is presented to obtain the effective elastic moduli for three-dimensional two-phase composites which is an exact analytic formula without any simplified approximation and can be expressed in an explicit form. For the different cells such as spherical inclusions and cracks surrounded by sphere and oblate ellipsoidal matrix, the effective elastic moduli are evaluated and the results are compared with those from various micromechanics models. These results show that the present model is direct, simple and efficient to deal with three-dimensional tyro-phase composites.

Plasma brain-derived neurotrophic factor levels, learning capacity and cognition in patients with first episode psychosis

Background: Cognitive impairments are seen in first psychotic episode (FEP) patients. The neurobiological underpinnings that might underlie these changes remain unknown. The aim of this study is to investigate whether Brain Derived Neurotrophic Factor (BDNF) levels are associated with cognitive impairment in FEP patients compared with healthy controls. Methods: 45 FEP patients and 45 healthy controls matched by age, gender and educational level were selected from the Basque Country area of Spain. Plasma BDNF levels were assessed in healthy controls and in patients. A battery of cognitive tests was applied to both groups, with the patients being assessed at 6 months after the acute episode and only in those with a clinical response to treatment. Results: Plasma BDNF levels were altered in patients compared with the control group. In FEP patients, we observed a positive association between BDNF levels at six months and five cognitive domains (learning ability,immediate and delayed memory, abstract thinking and processing speed) which persisted after controlling for medications prescribed, drug use, intelligence quotient (IQ) and negative symptoms. In the healthy control group, BDNF levels were not associated with cognitive test scores. Conclusion: Our results suggest that BDNF is associated with the cognitive impairment seen after a FEP. Further investigations of the role of this neurotrophin in the symptoms associated with psychosis onset are warranted.