On the use of models for high-performance scientific computing applications: an experience report

This paper reports on a four-year project that aims to raise the abstraction level through the use of model-driven engineering (MDE) techniques in the development of scientific applications relying on high-performance computing. The development and maintenance of high-performance scientific computing software is reputedly a complex task. This complexity results from the frequent evolutions of supercomputers and the tight coupling between software and hardware aspects. Moreover, current parallel programming approaches result in a mixing of concerns within the source code. Our approach relies on the use of MDE and consists in defining domain-specific modeling languages targeting various domain experts involved in the development of HPC applications, allowing each of them to handle their dedicated model in a both user-friendly and hardware-independent way. The different concerns are separated thanks to the use of several models as well as several modeling viewpoints on these models. Depending on the targeted execution platforms, these abstract models are translated into executable implementations by means of model transformations. To make all of these effective, we have developed a tool chain that is also presented in this paper. The approach is assessed through a multi-dimensional validation that focuses on its applicability, its expressiveness and its efficiency. To capitalize on the gained experience, we analyze some lessons learned during this project.     

The evolution of the exhaust diesel soot adsorptive properties after NO2 exposure and according to their SOF content

Several samples of exhaust diesel soot are investigated by inverse gas chromatography and linear solvation energy relationship (LSER) modelling according to their soluble organic fraction content and their time of exposure in oxidative conditions. The results demonstrate the evolution of the adsorptive properties of the studied materials towards volatile compounds during the oxidation under NO₂.     

PEFC Stack Operating in Anodic Dead End Mode

The polymer electrolyte fuel cell (PEFC) system is considered as an alternative power source and is particularly promising for mobile applications. This paper reports some experimental results performed on a PEFC stack, operating in anodic dead end mode. The flush frequency of the anode is investigated. In order to reduce hydrogen consumption in an embedded application, the flush frequency could be calculated from system modelling, according to the operating conditions. The first step of such a model is presented. It simulates the water exchanges between the anode and cathode channels in the anode dead end case. It has been implemented in a MATLAB® environment.     

Study of the Role of the Tyrosine Kinase Receptor MerTK in the Development of Kidney Ischemia-Reperfusion Injury in RCS Rats

Renal ischaemia reperfusion (I/R) triggers a cascade of events including oxidative stress, apoptotic body and microparticle (MP) formation as well as an acute inflammatory process that may contribute to organ failure. Macrophages are recruited to phagocytose cell debris and MPs. The tyrosine kinase receptor MerTK is a major player in the phagocytosis process. Experimental models of renal I/R events are of major importance for identifying I/R key players and for elaborating novel therapeutical approaches. A major aim of our study was to investigate possible involvement of MerTK in renal I/R. We performed our study on both natural mutant rats for MerTK (referred to as RCS) and on wild type rats referred to as WT. I/R was established by of bilateral clamping of the renal pedicles for 30′ followed by three days of reperfusion. Plasma samples were analysed for creatinine, aspartate aminotransferase (ASAT), lactate dehydrogenase (LDH), kidney injury molecule -1 (KIM-1), and neutrophil gelatinase-associated lipocalin (NGAL) levels and for MPs. Kidney tissue damage and CD68-positive cell requirement were analysed by histochemistry. monocyte chemoattractant protein-1 (MCP-1), myeloperoxidase (MPO), inducible nitric oxide synthase (iNOS), and histone 3A (H3A) levels in kidney tissue lysates were analysed by western blotting. The phagocytic activity of blood-isolated monocytes collected from RCS or WT towards annexin-V positive bodies derived from cultured renal cell was assessed by fluorescence-activated single cell sorting (FACS) and confocal microscopy analyses. The renal I/R model for RCS rat described for the first time here paves the way for further investigations of MerTK-dependent events in renal tissue injury and repair mechanisms.     

Influence of a metallic filler on polyurethane formation

The addition of metallic fillers is recognized to impart improvement of acoustic properties to polyurethane-based materials used for marine applications. The system under consideration was obtained by the reaction of a triisocyanateterminated prepolymer with 1,4-butanediol, in the presence of various concentrations of lead powder. With increasing lead content, gelation time is reduced and an Einstein-type law does not apply. The kinetics of the reaction were determined by Fourier transform infrared spectroscopy. Deviation from the classic second order is observed for the filled systems, and the higher the lead content, the more important the deviation is, showing undoubtedly the catalytic effect of the metallic powder.     

Time-resolved vibration measurement with temporal speckle pattern interferometry

Temporal speckle pattern interferometry (TSPI) is an optical measurement procedurefor measuring the displacement of rough technical surfaces. The time-dependent speckle modulation due to optical path difference changes is tracked during the whole displacement of the surface and then evaluated pointwise without referring to neighboring pixels. This feature allows for its use as independent point sensors. This aspect of incremental phase tracking enables TSPI to be used to measure time-resolved mechanical vibrations. It also reduces the deteriorating effect of the decorrelation. Therefore large displacements can be measured. A concept for an inexpensive fiber-optical point sensor was developed and the theoretical accuracy for vibration measurement was investigated. The TSPI measurement of a loudspeaker membrane is compared with a high-precision vibrometer measurement. The first results show good agreement.     

Integration of smart windows into building design for reduction of yearly overall energy consumption and peak loads

The purpose of this work is to investigate the potential of diminishing the energy consumed by typical low thermal mass office buildings for heating, cooling and lighting by using smart windows. The windows considered consisted of a double pane glazing unit in which a controllable absorbing layer is added on the interior surface of the exterior glass pane. This absorbing layer allows to change the optical properties of the window, resulting in a direct potential of control of the incident solar heat flux entering the building through the windows. A corresponding numerical model is developed showing that optimizing the solar heat flux absorption rate of the absorbing layer in regard of the necessary heating, cooling and lighting needs helps reducing significantly the total yearly energy consumption, and cooling peak loads. The simulations were done considering a building located in Quebec City, Canada.     

Thermogravimetric analysis as a new method to determine the lignocellulosic composition of biomass

Biomass energy uses organic matter such as wood or plants - lignocellulosic biomass - for creating heat, generating electricity and producing green oil for cars. Modern biomass energy recycles organic leftovers from forestry and agriculture, like corn stovers, rice husks, wood waste and pressed sugar cane, or uses special, fast-growing “energy crops” like willow and switchgrass, as fuel. Biomass is composed of three major components: cellulose, hemicelluloses, and lignin. Their differences in chemical structures lead to different chemical reactivities, making the relative composition in cellulose, hemicelluloses and lignin in the biomass a crucial factor for process design. In this paper thermogravimetric analysis is investigated as a new method to obtain lignin, hemicellulose and ??-cellulose contents in biomass. It is shown that this alternative method lead to comparable results than common methods used for the determination of the ??-cellulose content, with an enhancement of the accuracy in the determination of the hemicellulose content. Unfortunately, this method cannot be adopted for the determination of the lignin amount.     

Comparison of oxidation model predictions with gasification data of IG-110, IG-430 and NBG-25 nuclear graphite

A phenomenological oxidation kinetics model of graphite is presented and its results are compared with the reported experimental gasification data for nuclear graphite of IG-110, IG-430 and NBG-25. The model uses four elementary chemical kinetics reactions, employs Gaussian-like distributions of the specific activation energies for adsorption of oxygen and desorption of CO gas, and accounts for the changes in the effective surface areas of free active sites and stable oxide complexes with weight loss. The distributions of the specific activation energies for adsorption and desorption, the values of the pre-exponential rate coefficients for the four elementary chemical reactions and the surface area of free active sites are obtained from the reported measurements using a multi-parameter optimization algorithm. At high temperatures, when gasification is diffusion limited, the model calculates the diffusion velocity of oxygen in the boundary layer using a semi-empirical correlation developed for air flows at Reynolds numbers ranging from 0.001 to 100. The model also accounts for the changes in the external surface area, the oxygen pressure in the bulk gas mixture and the effective diffusion coefficient in the boundary layer with weight loss. The model results of the total gasification rate and weight loss with time in the experiments agree well with the reported measurements for the three types of nuclear graphite investigated, at temperatures from 723 to 1226 K and weight loss fractions up to ～0.86.     

High-Temperature Oxidation Behavior of Refractory High-Entropy Alloys: Effect of Alloy Composition

The high-temperature oxidation behavior of a new family of refractory high-entropy alloys (HEAs) with compositions of W–Mo–Cr–Ti–Al, Nb–Mo–Cr–Ti–Al and Ta–Mo–Cr–Ti–Al was studied at 1000 and 1100 °C. Based on these equimolar starting compositions, the main incentive of this study was to select the most promising alloy system whose properties may then be successively improved. Despite the high amount of refractory elements, Ta–Mo–Cr–Ti–Al showed good oxidation resistance at 1000 and 1100 °C. Moderate values of mass gain and complex oxidation kinetics were observed for the W- and Nb-containing HEAs. These alloys formed inhomogeneous oxide scales possessing regions with thick and porous layers as well as areas revealing quite thin oxide scales due to the formation of discontinuous Cr- and Al-rich scales. The most promising behavior was shown by the alloy Ta–Mo–Cr–Ti–Al which followed the parabolic rate law for oxide growth due to the formation of a thin and compact Al-rich layer.