Search

In this study, the meshless local radial point interpolation method (LRPIM) is used to model an electromagnetic device. This method is considered to be an interesting alternative to the finite element method and here it is used for evaluating the magnetic field and flux generated by electromagnetic devices—such as actuators. The results are validated by comparing them with those of the finite element method.

The purpose of this paper is to introduce periodic and antiperiodic boundary conditions in the natural element method (NEM). It is shown that as the NEM shape functions verify the Kronecker delta property, the imposition of such boundary conditions can be done in an easy way as in the finite-element method (FEM). These boundary conditions are important because they allow exploitation of the inherent symmetry of the electromagnetic devices. Therefore, with these techniques the NEM can now easily take advantage of the symmetry of electrical machines. The proposed approach is evaluated, and its results are compared with those obtained by the traditional FEM.

Energy planning at the neighborhood level is a major development axis for the energy transition. This scale allows the pooling of production and storage equipment, as well as new possibilities for demand-side management such as flexibility. To manage this growing complexity, one needs two tools. The first concerns modeling, allowing exhaustive simulation analyses of buildings and their energy systems. The second concerns optimization, making it possible to decide on the sizing or control of energy systems. In this article, we analyze, in the case of an existing residential neighborhood, the ability to study by modeling and optimization tools two scenarios of energy flexibility of indoor heating. We propose in particular a method allowing to rely on a varied set of data available to build the various models necessary for optimization tools or dynamic simulation. A study was conducted to identify the neighborhood’s flexibility potential in minimizing C O 2 emissions, through shared physical storage, or storage in the building envelope. The results of this optimization study were then compared to their application to the virtual neighborhood by simulation.

International audience; This paper addresses the topic of the electrical flexibility on the demand-side, by focusing on the residential buildings. Specifically, it aims to quickly formulate the optimal operation of building heat pumps, according to environmental or grid issues. Indeed the literature shows many examples of quantification of the flexibility impacts, but mostly relying on pre-defined strategies. To go from the evaluation of some strategies to the formulation of the optimal planning , we are introducing a methodology based on the automatic generation of models dedicated to mixed-integer linear programming (MILP) optimization. Finally , the method was applied to a new residential building during a month of winter.

This paper investigates how blocks of buildings could fit into load shedding strategies. It focuses in particular on what could be the effects on peak shaving, occupants’ thermal comfort or CO2 emissions reduction and how to quickly quantify them. To achieve this goal, we focused on a new residential district, thermally fed by heat pumps. Four modeling approaches were confronted in order to estimate buildings' responses to load shedding orders. On the one hand, a quick estimation of the peak shaving impact can rely on experimental results if the buildings' envelope and uses of the experimentation match those of the study case. On the other hand, thermal simulation models allow us to assess thermal comfort while considering the building physical response. Finally, a hybrid modeling approach can provide a good compromise between modeling rapidity and accuracy of the impacts estimation. At district scale, it may be necessary to mix modeling approaches, from experimental results to detailed thermal models. Accuracy is not guaranteed for all approaches so that the choice should be made carefully in regards to study needs. However, results are sufficient to compare the effects of load shedding strategies on peak shaving, thermal comfort, and CO2 emissions reductions.

International audience; Cet article présente un outil de modélisation et d’optimisation d’échanges d’énergie à l'échelle d’unquartier permettant de valoriser la flexibilité des acteurs appliqué à un cas d’étude. Cet outil permet de modéliserdes échanges à l’échelle macroscopique pour des projets multi-acteurs. Ecrit en langage Python, il utilise uneoptimisation MILP (Mixed-Integer Linear Programming). Le cas d’étude présenté dans cet article est situé sur laPresqu’île de Grenoble. Il s’agit du projet Valocal. Ce projet a pour but de valoriser la chaleur fatale du LaboratoireNational de Champs Magnétiques Intenses (LNCMI) en la réinjectant dans le réseau de chauffage urbain de laville. Cet article présente la modélisation des acteurs du projet – principalement l’opérateur du réseau de chaleur -dans le but de jouer sur les flexibilités des deux acteurs en fonction du taux de valorisation de la chaleur fatale, duprix de la calorie, ou encore des émissions CO2.

International audience; In this paper, the Mixed Finite Element and Natural Element (Mixed FEM-NEM) Method is applied to determine the flux distribution in magnetic devices using its reduced magnetic circuit models. In this solution, the NEM is applied exclusively in the gap of the devices where deformations of the mesh are expected in case of movement and the FEM is applied in the remainder magnetic circuit. In order to reduce the computational cost, the use of anti-periodic boundary conditions is proposed allowing the reduction of the entire domain model. To show the effectiveness of the proposed methodology, the problem is solved by using FEM and NEM in separate, and by the mixed FEM-NEM. The obtained results are compared using an error estimator and the computational cost.

International audience; In this paper, the interpolation functions behavior of the natural element method (NEM) and the finite element method (FEM) are compared. It is discussed how the unknown in both methods affects the stiffness matrix and contributes to NEM better accuracy. The visibility criterion and the constrained NEM are also addressed, and a pseudoalgorithm is proposed to implement the constrained Voronoï diagram, which is the base for the constrained NEM. A complex heterogeneous magnetic problem is solved by both FEM and NEM methods and their solutions are compared. It is shown that for the same discretization, the number of contributions for NEM is in general bigger than those related to the FEM and better accuracy results for NEM.

International audience; As energy transition is fundamental to have a chance to fight climate change, every stakeholder concerned by energy should be able to get a better knowledge of the consequences of these actions. However, it could be very complex to understand energy problematics without being an expert. This article focuses on giving the possibility to an energy intensive consumer of a district to make decisions about its energy planning while taking into account its specific operating constraints. A practical case has been studied in a heat recovery project to help the experiments planning of a research laboratory according to the thermal needs of the district. At first, the energy planning only aims to reduce its electricity consumption bill. In a second time, we consider re-using the thermal power from processes cooling. Then, two energy planning were realised: reducing district CO 2 emissions and reducing district supply cost. Finally, trade-offs between these two goals have been studied. The work is based on mixed-integer linear optimization models (MILP) gathered into a Python library to provide a modular decisions tool for energy stakeholders.

This paper compares the numerical performance of two numerical methods, the finite element method and the natural element method (NEM). NEM is relatively recent and is based on functions belonging to the Voronoi cell family. Although it has been proved that this method gives smoother and more accurate solutions than the finite elements, its computational cost is also known to be higher. In this paper, we compare computational efficiency, i.e., accuracy for a given cost, of finite elements and natural elements, for both Laplace and Sibson shape functions. We also bring into the comparison a Voronoi cell-based finite difference scheme which proves to be very efficient. The error is calculated using dual formulations or analytical solutions.

Purpose In FEM computations, the mesh quality improves the accuracy of the approximation solution and reduces the computation time. The dynamic bubble system meshing technique can provide high-quality meshes, but the packing process is time-consuming. This paper aims to improve the running time of the bubble meshing by using the advantages of parallel computing on graphics processing unit (GPU). Design/methodology/approach This paper is based on the analysis of the processing time on CPU. A massively parallel computing-based CUDA architecture is proposed to improve the bubble displacement and database updating. Constraints linked to hardware considerations are taken into account. Finally, speedup factors are provided on test cases and real scale examples. Findings The numerical experiences show the efficiency of parallel performance reaches a speedup of 35 compared to the serial implementation. Research limitations/implications This contribution is so far limited to two-dimensional (2D) geometries although the extension to three-dimension (3D) is straightforward regarding the meshing technique itself and the GPU implementation. The authors’ works are based on a CUDA environment which is widely used by developers. C\C++ and Java were the programming languages used. Other languages may of course lead to slightly different implementations. Practical implications This approach makes it possible to use bubble meshing technique for both initial design and optimization, as excellent meshes can be built in few seconds. Originality/value Compared to previous works, this contribution shows that the scalability of the bubble meshing technique needs to solve two key issues: reach a T(N) global cost of the implementation and reach a very fast size map interpolation strategy.

In this paper, a Mixed Method using the Finite Element (FEM) and the meshless Natural Element Method (NEM) is used to determine the magnetic flux distribution in magnetic circuit of an electrical machine. In this solution, the NEM is applied exclusively in the gap of the machine where the largest deformations of the mesh are expected when the movement occurs. Due the use of anti-periodic boundary conditions, only 1/4 of machine is modeled reducing the computational cost. The error estimator and computational time by using the FEM, the NEM and the mixed FEM-NEM are obtained and compared.

In a previous work a vector interpolation scheme based on the Natural Element Method (NEM) was introduced. For this, the standard NEM shape functions associated to the nodes of a mesh composed of arbitrarily shaped polygons were used to construct the vector functions linked to the edges of the polygonal elements. The scheme, which can be seen as an edge-based Finite Element Method extension to polygonal meshes, potentially simplifies the mesh generation and yields higher order approximations. Here, the sensitivity of this approach to the mesh regularity is examined.

In this paper, the Constrained Natural Element Method (C-NEM) is used to determine the magnetic flux distribution in an electrical machine. Due the use of anti-periodic boundary conditions, only % of machine is modeled. The obtained results are compared with the traditional Finite Element Method (FEM) with good agreement.

In this paper the coupling of natural element method with the boundary element method (Hybrid NEM — BEM method) is proposed to solve the electromagnetic scattering problem. A dielectric cylinder is solved by the proposed approach and the results are compared with the analytical solution showing the usefulness and accuracy of the method.

This paper presents a new methodology called Flex-CM for Flexible Compact Modeling to generate and reuse compact models, in order to build a full model of a complex electronic system at an early stage of the design flow. The system is split into independent parts. Each part is numerically modeled using a discrete approach then reduced to save memory and time at the simulation stage. The resulting pluggable and reduced model is called a micro-model. Then, a fast-to-simulate macro-model of a full system can be obtained by assembling the micro-models. The Flex-CM model is presented as a trade-off between reluctance networks and numerical models to perform a magnetic analysis at an early stage of the system design. The Flex-CM models are found to have number of advantages over both methods, resulting however on larger models than reluctance networks.

This paper deals with adaptive remeshing techniques. The traditional methods for remeshing are presented, raising the problem of quality. The bubble meshing technique that ensures high-quality meshes is exposed, and improvements brought to the technique are also presented. The procedure that consists in using this technique for generating high-quality meshes during the refinement iterations is showed. Some results for both the bubble meshing technique and its use for meshes refinement are presented. The whole study has been completed in 2-D and is extensible to 3-D.

A coupling of the partial element equivalent circuit method with the finite element method for solving magneto-harmonic electromagnetic problems is proposed. The new formulation, which combines the main advantages of the previous methods, is particularly well suited for modeling electromagnetic complex devices including a large number of conductors.

Despite its proven accuracy and computational efficiency, the Natural Element Method (NEM) suffers from the lack of a well-suited numerical integration technique. In this work different integration approaches are applied and tested on the NEM framework. Comparative analyses in terms of accuracy and computational cost are presented. The goal here is to elucidate the impact of these techniques on the NEM performance and thus assist the choice of the most adequate technique for a given scenario.

In this paper, we introduce some parallel techniques on NVIDIA compute unified device architecture GPU for the finite-element method applied in the magnetic field computation. To ensure the load balance, each parallel thread performs the integration of one element. In the assembly step, we introduced a fast procedure based on the sorting and rearrangement of non-zero entries on the GPU global memory. Then, a reducing process is executed to obtain the resulting coefficient matrix in a sparse format. About the solving step, we use the conjugate gradient iterative solver with a variety of preconditioning techniques. Our implementation does not require any preprocessing on mesh, but takes advantage of the parallel computing power of GPU. In our test, this parallel strategy improved the performance 30 times faster on the assembly process and four times faster on the solving process.

International audience; This paper presents a hybrid approach coupling natural element method (NEM) and boundary element method (BEM) for thetreatment of unbounded problems. The goal is to combine the accuracy of the NEM and the ability of the BEM in modeling linear anddeformable domains without a mesh. The performance of the proposed scheme is compared to a FEM-BEM coupling in terms ofaccuracy. Results show that for a given number of degrees of freedom the proposed approach (NEM-BEM) is able to provide moreaccurate solutions.

The problem of determining the electromagnetic scattering of arbitrary structures consisting of both dielectric and conducting materials are analyzed in this paper by the natural-element method coupled with the absorbing boundary condition (NEM-ABC). By this method, the target is first enclosed by a fictitious boundary. Maxwell’s equations are then solved by NEM inside and ABC is applied on the fictitious boundary. The results are compared with analytical and finite-element method-ABC ones which show the robustness, applicability and relevance of the method.

The set of infrared spectra recorded at different levels of hydrogen–deuterium (H/D) exchange as a function of time were processed using spectral decomposition. The most precise information about H/D exchange of the NH groups of the protein backbone was retrieved by observing the intensity change of the nearest C=O stretching vibration. The H/D exchange at the protein backbone begins with NH groups bonded to C=O with a characteristic frequency of 1683 cm−1. These amide groups were initially free and are the first to accept H-bonds from water during the hydration process. The NH groups, which are connected to C=O groups with a characteristic band at 1657 cm−1, exchange at a slower rate. For both populations the pairs of comparable exchange rates were calculated with rate constants of 10−3 min 1, 0.014 min 1, 0.0046 min 1, and 0.09 min 1. The appearance of two different exchange rates for each population is a consequence of the distinct exposure of particular molecular groups to the solvent. Two additional bands sensitive to exchange are attributed to NH bending modes in the side chains and are located at 1610 cm−1 and 1585 cm−1. These NH groups undergo H/D exchange at the beginning of the exchange with exchange rates of 0.019 min−1 and 0.17 min−1, respectively. The calculated exchange rate for hydrating water molecules is 0.037 min−1. The distribution of water's exchange rate is extremely broad and covers almost the entire interval of the time-dependent experiment. From the efficiency of the exchange (96%) it is evident that some parts of the protein are completely forbidden to water molecules.

The structure of the amide I band of bovine serum albumin (BSA) was determined using an H/D exchange experiment. The difference between the dry and hydrated exchange spectrum revealed the fine structure of the amide I band. The band at 1717 ± 2 cm−1 is due to the vibration of the COOH moieties from the protein side chains. Band components at 1682 ± 2 cm−1, 1655 ± 2 cm−1, and 1637 ± 2 cm−1 are assigned to the vibrations of the backbone C=O. These three bands belong to vibrations of three different populations of amide groups differing in the number of established H-bonds. The connectivity between the frequencies of various amide vibrations was determined by two-dimensional generalized correlation spectroscopy and spectral decomposition. About 7% of the whole exchangeable hydrogen atom population (NH, NH2, and OH groups from backbone and side chains) remains unexchanged, and these hydrogen atoms belong mainly to the NH groups, which are H-bonded to specific C=O groups. Moreover, this study concerns the approximately 10% of hydrogen atoms belonging to a particular HN … OC population with a characteristic amide A frequency at 3290 cm−1 and an amide I band at 1655 ± 2 cm−1, usually attributed to the α-helical structure that remains unexchanged. At higher temperature the exchange is more efficient. Upon heating, a further 4% of these NH groups are deuterated. The comparison of the exchange spectrum at higher temperature with the structural changes of the protein at the same temperature implies that the change in overall dynamics of the protein improve the level of exchange.

L'article presente les mecanismes qui permettent l'integration de differents solveurs dans la plate-forme SALOME. L'approche multiphysique est illustree par une application electrothermique.

Hydration of Na + and HN + (C 2 H 5 ) 3 sulfonated polyimide membranes is probed by infrared spectrometry using a recently described method. These membranes consist of identical sulfonated repeat unit (homopolymers) and form a good starting point to study more elaborated membranes, composed of these units plus similar ones with no sulfonic groups (block copolymers). These latter membranes may be used in fuel cells and will be described in a forthcoming article. We first identify the bands of hydrophilic groups, which will be anchor points for hydration. We then define three ‘elementary hydration spectra’ onto which all hydration spectra can be decomposed. Their analysis allows us to measure the water uptake of these membranes as a function of the hygrometry of the ambient atmosphere and to determine the structures of the dried membranes and their hydration mechanisms in terms of chemical reactions.

In this paper, we use an optimization technique based on the creation of response surfaces and the application of genetic algorithms to minimize the number of conductors of grounding grids, guaranteeing the safety levels defined by the maximum touch potential.

Presents a generic and extensible implementation of the physical property descriptor, embedded in the open source platform SALOME. To create a new physical model and use it in a complete problem description are the main facilities provided by the physical property descriptor. The description procedure used in this environment is shown on an electrostatic example.

A simple method for the decomposition of a set of spectra is proposed. The method is based on difference spectroscopy, which allows the decomposition of spectra down to their basic components. The basic components are easier to interpret and allow to monitor changes in the spectra induced by internal (chemical reaction) or external (temperature, pressure, concentration, pH, …) perturbations. Properly chosen components represent a linear basis for all spectra (or difference spectra) from the series. The coefficients of linear combination give us the amount of a given component in a particular region for the altered spectral parameter. A hydration of protein film is presented as a practical example.

The natural-element method (NEM) has proved to be able to provide more accurate and computationally efficient solutions than the finite-element method (FEM) for first-order consistency approximations. However, higher order approximations in the NEM framework are not obtained straightforwardly. This paper addresses a method of constructing higher order approximations out of the standard first-order NEM shape functions. This process is achieved through the de Boor algorithm. Accuracy of the scheme is compared with FEM. Results show that in second-order approximations context the NEM is still able to provide better accuracies for a given number of degrees of freedom.

Infrared studies of urea both in crystal and in water solution are presented. On the basis of temperature and concentration measurements of urea–water solutions, as well as from the spectra of crystalline urea films of hydrogenated and partially deuterated molecules, we have assigned the main bands in urea spectrum. It is shown that the series of spectra recorded at various concentrations in water can be decomposed into four basic spectra: the first one is the same as the spectrum of liquid water, the second one has features only due H 2 O but is slightly disturbed as compared with the first one and the other two display in addition features due to vibrations in urea. The relative intensities of these components show a strong concentration dependence, particularly for what concerns the ratio of the last two components. We ascribe one of these two components to a structure where urea–urea interaction is predominant in the solution while the other one is ascribed to a structure where spectral features are mainly due to urea–water interactions. The first component with urea–urea features can be observed only at urea concentrations higher than 1 M. Concentration dependence of the population of these two components shows that at higher urea concentrations the interaction between urea molecules can be observed presumably as a consequence of aggregation of urea molecules. Such aggregation of urea molecules in dimers and/or oligomers may explain some peculiarities connected with the unusual behaviour of urea solutions in water as predicted by the Shellman, Kreschek, Scheraga and Stokes model. Concentration measurements also show that urea molecules only slightly alter the structure of bulk water, and hence cannot be described as ‘structure breakers’.

The numerical tools required by the transient modeling of electrical machines are reviewed from an energetic viewpoint. The consistency with the finite element method is given. The impact of these modeling requirements is inspected on a high-sensitivity relay case-study and attention on motion-induced eddy currents is given. Due to the trend in energy savings expected in many markets, it is emphasized that future numerical software should take properly into account the loss processes within a 3-dimensional description.

A new algorithm to obtain the response surface of objective functions is presented. The diffuse-element method and quadtrees are used in the implementation of this algorithm. The approximation of a two-parameter analytical function and the optimization of a switched reluctance motor are performed to show the robustness of this new approach.

International audience

This paper deals with the three‐dimensional modeling of massive conductors carrying a total imposed current and involving multiply connected regions. Both stationary motion and time harmonic eddy current problems are considered. The aim of the modeling is to determine the current distribution and the eddy current’s effects inside the conductor, as well as to calculate the Lorentz force acting on a conducting region. The choice of a suitable formulation based on solving the current vector potential T and the magnetic scalar potential φ is argued, then various methods for the restoration of the Ampere law in the multiply connected regions are presented. Two methods for calculating an auxiliary vector potential T0 are applied in order to calculate the force acting on the armature of a railgun and the repulsion between the electrodes of a vacuum circuit breaker, then the results are compared.