57949 research outputs found
Sort by
Modelling dissolved oxygen content in water for ultrasound power measurements
No full textPeer reviewed: YesNRC publication: Ye
Hybrid weld-bond joining technology of light metal alloys
No full textThe biggest challenge faced by transportation industries in structural assembly is the joints\u2019 integrity upon exposure to environmental conditions while maintaining excellent mechanical properties. Combination of mechanical and adhesive joining can provide high joint strengths, increased energy absorption and high fatigue lives. A weld-thru method has been previously employed where the parts to be assembled are friction stir welded (FSW) in the presence of a structural adhesive/sealant between the parts. However, this technology suffers adhesive damage due to weld-induced heat generated locally affecting the joints\u2019 durability and mechanical properties. Also, most adhesive/sealants have Tg below the temperature attained during welding (450 to 500 oC), unable to withstand the weld-induced heat. In this paper, an inverse technology, called flow-in or weld-bond was used to bond Al-Al and Al-steel using a low viscosity epoxy adhesive in combination with FSW and arc welding processes. In this method, the welding step in overlap mode is performed first. The adhesive is applied later through the welded gap via capillary forces. Since welding and adhesive application are independently performed, the adhesive safely fills the gap while maintaining the joints\u2019 integrity. The performance of the weld-bonded joints of pretreated Al and steel alloys as-assembled and after various environmental degradation (international standards) including combinations of heat, humidity, sub-zero temperatures, salt-spray and UV exposure, has been evaluated. The results show tremendous increase (1.7 times) in joint strengths on weld-bonded specimens in comparison to the weld-alone and adhesive-alone specimens with no change even after degradation. A simple surface treatment prior to adhesive application is found to help the joints sustain the salt-spray conditions. The fatigue life of the weld-bonded joints is found to be higher than those prepared using individual techniques.Peer reviewed: YesNRC publication: Ye
Calibration of AC shunts at frequencies up to 10 kHz using digital sampling
No full textThis paper describes a calibration system for AC shunts using digital sampling. It provides bot h ratio errors and phase displacements of a shunt under test with respect to a reference shunt. ln its basic configuration, the calibration system operates at currents of up to 10 A and frequencies of up to IO kHz. The expanded uncertainties (k = 2) of the described calibration system arc estimated to be better than 30 \ub5A/A for magnitude and 30 \ub5rad for phrase at 10 kHz, and lower than these at lower frequencies.Peer reviewed: YesNRC publication: Ye
A multi-objective control strategy for combined peak shaving and frequency regulation in V2B/V2G applications
No full text-The Canadian Government has set a target of reducing Greenhouse Gas (GHG) emissions by 30% in 2030 compared to emission levels in 2005. -The transportation sector is responsible for 23% of energy related GHG emissions. -EVs can significantly reduce transportation emissions in Canada. -The global EV market share is expected to be 54% of new car sales and 33% of the global car fleet by 2040. -V2X (X=home, buildings and grid) may create an additional stream of revenue for EV owners, enabling much faster adoption of EVs and earlier realization of large environmental benefits in Canada.Peer reviewed: NoNRC publication: Ye
The JCMT Gould Belt Survey: SCUBA-2 data reduction methods and Gaussian source recovery analysis
No full textThe James Clerk Maxwell Telescope (JCMT) Gould Belt Survey (GBS) was one of the first legacy surveys with the JCMT in Hawaii, mapping 47 deg2 of nearby (<500 pc) molecular clouds in dust continuum emission at 850 and 450 \u3bcm, as well as a more limited area in lines of various CO isotopologues. While molecular clouds and the material that forms stars have structures on many size scales, their larger-scale structures are difficult to observe reliably in the submillimeter regime using ground-based facilities. In this paper, we quantify the extent to which three subsequent data reduction methods employed by the JCMT GBS accurately recover emission structures of various size scales, in particular, dense cores, which are the focus of many GBS science goals. With our current best data reduction procedure, we expect to recover 100% of structures with Gaussian \u3c3 sizes of 6430'' and intensity peaks of at least five times the local noise for isolated peaks of emission. The measured sizes and peak fluxes of these compact structures are reliable (within 15% of the input values), but source recovery and reliability both decrease significantly for larger emission structures and fainter peaks. Additional factors such as source crowding have not been tested in our analysis. The most recent JCMT GBS data release includes pointing corrections, and we demonstrate that these tend to decrease the sizes and increase the peak intensities of compact sources in our data set, mostly at a low level (several percent), but occasionally with notable improvement.Peer reviewed: YesNRC publication: Ye
Editorial: Computational linguistics and literature
No full textComputational Linguistics\u2014or, more technically, Natural Language Processing\u2014has made great strides in the past several years. Machine learning (ML) is the technology of choice; deep learning, in particular, has pushed the envelope. These methods work best in narrow domains, given vast amounts of data and asked for information rather than for interpretation. Literary data are not limited by topic, and they are hardly ever plentiful enough. That is why work on such data has been, as yet, on the periphery of research and development in Computational Linguistics. This Research Topic aims to bring the processing of literary data to the attention of a broader audience.Peer reviewed: YesNRC publication: Ye
Data-driven modeling for energy consumption estimation
No full textEnergy consumption estimation for building energy management systems (BEMS) is one of the key factors in the success of energy saving measures in modern building operation, either residential buildings or commercial buildings. It provides a foundation for building owners to optimize not only the energy usage but also the operation to respond to the demand signals from smart grid. However, modeling energy consumption in traditional physical modeling techniques remains a challenge. To address this issue, we present a data mining-based methodology, as an alternative, for developing data-driven models to estimate energy consumption for BEMSs. Following the methodology, we developed data-driven models for estimating energy consumption for a chiller and a supply fan in an air handling unit (AHU) by using historic building operation data and weather forecast information. The models were evaluated with unseen data. The experimental results demonstrated that the data-driven models can estimate energy consumption for BEMS with promising accuracy.Peer reviewed: YesNRC publication: Ye
Abnormality detection in mammography using deep convolutional neural networks
No full textBreast cancer is the most common cancer in women worldwide. The most common screening technology is mammography. To reduce the cost and workload of radiologists, we propose a computer aided detection approach for classifying and localizing calcifications and masses in mammogram images. To improve on conventional approaches, we apply deep convolutional neural networks (CNN) for automatic feature learning and classifier building. In computer-aided mammography, deep CNN classifiers cannot be trained directly on full mammogram images because of the loss of image details from resizing at input layers. Instead, our classifiers are trained on labelled image patches and then adapted to work on full mammogram images for localizing the abnormalities. State-of-the-art deep convolutional neural networks are compared on their performance of classifying the abnormalities. Experimental results indicate that VGGNet receives the best overall accuracy at 92.53% in classifications. For localizing abnormalities, ResNet is selected for computing class activation maps because it is ready to be deployed without structural change or further training. Our approach demonstrates that deep convolutional neural network classifiers have remarkable localization capabilities despite no supervision on the location of abnormalities is provided.Peer reviewed: YesNRC publication: Ye
Determining strain, chemical composition, and thermal properties of Si/SiGe nanostructures via raman scattering spectroscopy
No full textStudies by Raman spectroscopy of two kinds of Si/SiGe nanostructures\u2014quantum dot multilayers and planar superlattices\u2014reveal a wide variety of spectral features including first- and second-order Raman scattering, polarized Raman scattering, and low-frequency inelastic light scattering associated with folded acoustic phonons. Here we overview how such features can be employed to semi-quantitatively analyze the strain, chemical composition, and thermal conductivity in these industrially important materials that are widely used for producing electronic and optoelectronic devices.Peer reviewed: NoNRC publication: Ye
Multi-scale impedance model for supercapacitor porous electrodes: theoretical prediction and experimental validation
No full textIn this paper, a multi-scale impedance model is developed for evaluating the charge storage capacity and ion-transfer rate of supercapacitor porous electrodes. This model can be used to theoretically understand the contributions of the intra-particle pore at nano-scale, the inter-particle pore at micron-scale, and the porous electrode at millimeter-scale to the performance of supercapacitor. Then, impedance spectrum characteristics of porous electrodes are screened via numerical simulation based on the developed multi-scale impedance model, especially for different electrode thicknesses and various faradaic processes with different time constants. Subsequently, seven supercapacitor samples, four of them with electrode thicknesses of 60, 100, 180, and 370\u202f\u3bcm respectively, and three of them with different types of current collectors (nickel, carbon-coated stainless-steel, and stainless-steel), are fabricated and characterized to validate the developed model. The fitting results and residuals of the measured impedance data validate the linear relationship of the scaled equivalent resistance of electrolyte v. s. electrode thickness and the scaled characteristic time constant v. s. the square of electrode thickness. The validated multi-scale impedance model might offer an effective method to rationally balance or optimize charge storage and charge transfer via morphology design for porous electrodes.Peer reviewed: YesNRC publication: Ye