The prediction of crack propagation is an important engineering problem. In this work, combined with triangular plane stress finite elements, a new remeshing algorithm for crack opening problems was developed. The proposed algorithm extends the crack iteratively until a threshold maximum crack length is achieved. The crack propagation direction is calculated using the maximum tangential stress criterion. In this calculation, in order to smoothen the stress field in the vicinity of the crack tip, a weighted average of the stresses of the integration points around the crack tip is considered. The algorithm also ensures that there are always at least eight elements and nine nodes surrounding the crack tip, unless the crack tip is close to a domain boundary, in which case there can be fewer elements and nodes around the crack tip.

Four benchmark tests were performed showing that this algorithm leads to accurate crack paths when compared to findings from previous research works, as long as the initial mesh is not too coarse. This algorithm also leads to regular meshes during the propagation process, with very few distorted elements, which is generally one of the main problems when calculating crack propagation with the finite element method.

This paper presents a GPU-accelerated implementation of the Locally Optimal Block Preconditioned Conjugate Gradient (LOBPCG) method with an inexact nullspace filtering approach to find eigenvalues in electromagnetics analysis with higher-order FEM. The performance of the proposed approach is verified using the Kepler (Tesla K40c) graphics accelerator, and is compared to the performance of the implementation based on functions from the Intel MKL on the Intel Xeon (E5-2680 v3, 12 threads) central processing unit (CPU) executed in parallel mode. Compared to the CPU reference implementation based on the Intel MKL functions, the proposed GPU-based LOBPCG method with inexact nullspace filtering allowed us to achieve up to 2.9-fold acceleration.

The purpose of this paper is to propose a critical assessment of Young’s modulus determination of coated materials using Impulse Excitation Technique (IET). In this technique, the coated substrate is excited by an impulse and the acoustic vibrations are recorded. The frequency of the first bending mode is then used in a mechanical model to obtain the Young’s modulus of the coating. The existing models are based on two different theories: the flexural rigidity of a composite beam and the Classical Laminated Beam Theory (CLBT). The aim of the present paper is to assess the accuracy (trueness and precision) of the technique. For this, different models proposed in the literature are compared with a finite element model of the specimen for various conditions. The trueness and precision of models were evaluated and the best model was identified. Then a detailed uncertainty budget is performed to identify and quantify the most influent factors on the global uncertainty.

The aerodynamic and structural design of a pitching blade tip with a double-swept planform is presented. The authors demonstrate how high-fidelity finite element (FE) and computational fluid dynamic (CFD) simulations are successfully used in the design phase. Eigenfrequencies, deformation, and stress distributions are evaluated by means of a three-dimensional (3D) FE model. Unsteady Reynolds-averaged Navier-Stokes (RANS) simulations are compared to experimental data for a light dynamic stall case at Ma = 0.5, Re = 1.2 × 106. The results show a very good agreement as long as the flow stays attached. Tendencies for the span-wise location of separation are captured. As soon as separation sets in, discrepancies between experimental and numerical data are observed. The experimental data show that for light dynamic stall cases at Ma = 0.5, a factor of safety of FoS = 2.0 is sufficient if the presented simulation methods are used.

L’objectif principal de cet article est la mise au point d’une méthodologie numérique deprévision du comportement local d’un procédé de renforcement des poteaux en bois par lafibre de carbone. En particulier, on s’intéresse à l’étude du comportementélasto-plastique du matériau bois avec effet du renforcement. Une modélisation basée surla thermodynamique des processus irréversibles avec variables d’état est utilisée pourtraduire le couplage entre le comportement plastique à écrouissage isotrope et l’effet durenforcement. Les aspects théoriques et numériques de cette formulation sont décrits endétail. La résolution du problème d’équilibre global est assurée par un schéma DynamiqueExplicite. La validation de la procédure de calcul implémentée dans ABAQUS/Explicit estfaite sur la simulation d’un essai de compression des éprouvettes circulaires en boisrenforcées par la fibre de carbone. Les résultats des simulations sont confrontés à ceuxde l’expérience.

This paper presents a ductile damage-gradient based nonlocal and fully coupledelastoplastic constitutive equations by adding a Helmholtz equation to regularize theinitial and boundary value problem (IBVP) exhibiting some damage induced softening. First,a thermodynamically-consistent formulation of gradient-regularized plasticity fullycoupled with isotropic ductile damage and accounting for mixed non linear isotropic andkinematic hardening is presented. For the sake of simplicity, only a simplified version ofthis model based on von Mises isotropic yield function and accounting for the singlenonlinear isotropic hardening is studied and implemented numerically using an in house FEcode. An additional partial differential equation governing the evolution of the nonlocalisotropic damage is added to the equilibrium equations and the associated weak formsderived to define the IBVP (initial and boundary value problem). After the time and spacediscretization, two algebraic equations: one highly nonlinear associated with theequilibrium equation and the second purely linear associated with the damage non localityequation are obtained. Over a typical load increment, the first equation is solvediteratively thanks to the Newton-Raphson scheme and the second equation is solved directlyto compute the nonlocal damage \hbox{$\Bar{{D}}$}D̅ at each node. All the constitutive equations are “strongly” affected bythis nonlocal damage variable transferred to each integration point. Some applicationsshow the ability of the proposed approach to obtain a mesh independent solution for afixed value of the length scale parameter. Comparisons between fully local and nonlocalsolutions are given.

We are concerned with a model of ionic polymer-metal composite (IPMC) materials that consists of a coupled system of the Poisson and Nernst-Planck equations, discretized by means of the finite element method (FEM). We show that due to the transient character of the problem it is efficient to use adaptive algorithms that are capable of changing the mesh dynamically in time. We also show that due to large qualitative and quantitative differences between the two solution components, it is efficient to approximate them on different meshes using a novel adaptive multimesh hp-FEM. The study is accompanied with numerous computations and comparisons of the adaptive multimesh hp-FEM with several other adaptive FEM algorithms.

Transient temperature field of laminated plate which considering radiation is analyzed by laying temperature lines in every layer in this paper. Finite element equations are established by using high-order displacement field which considering shear effect based on the temperature analysis. Then thermally induced vibration of laminated plate is investigated by solving the finite element equation with numerical method. Compared with numerical results of ANSYS by using 3-D coupled element (SOLID5), the methods presented in this paper have a satisfied precision and better efficiency. It can give reference to the thermally induced vibration analysis of thin-large flexible laminated appendages of spacecraft.

As 3D scanners and numerical simulation are now mature, we have made a benchmark by comparing simulation results with real forged parts. We did consider the complete forging process of a cylindrical part using hammer from the billet to the final part. This process has been simulated using a thermo-elasto-viscoplastic constitutive equation including the ductile damage with the implemented with the help of VUMAT subroutine provided by ABAQUS software. Before and after each step, the forged part is scanned and compared with the result of the FEM simulation in order to tune some of the process numerical simulation parameters.

Over the past few years several technical efforts have been focused on continuous improvement of quench and tempered product at ArcelorMittal Burns Harbor plate mill facility with particular attention being given to optimization of the flatness of quenched product. Incremental improvement in quench line equipment and practices is being aided by computer-based modelling tools and specialized instrumentation. The paper describes the development of different tools used to control flatness after quench and tempering units. The detailed description of a 3D model designed to predict distortion such as edge waves, center buckles, longbows and crossbows during quenching is presented. The model developed with Abaqus is able to predict the evolution of plate temperature, thermal stresses, strains, and distortions during cooling. The water cooling system is represented by top and bottom movable surfaces where a heat flux depending on width, length and temperature is applied. The effect of phase transformation is approximated by a variation of the plate thermal expansion coefficient.

We determine the steady-state structures that result from liquid-liquid demixing in afree surface film of binary liquid on a solid substrate. The considered model correspondsto the static limit of the diffuse interface theory describing the phase separationprocess for a binary liquid (model-H), when supplemented by boundary conditions at thefree surface and taking the influence of the solid substrate into account. The resultingvariational problem is numerically solved employing a Finite Element Method on an adaptivegrid. The developed numerical scheme allows us to obtain the coupled steady-state filmthickness profile and the concentration profile inside the film. As an example wedetermine steady state profiles for a reflection-symmetric two-dimensional droplet forvarious surface tensions of the film and various preferential attraction strength of onecomponent to the substrate. We discuss the relation of the results of the present diffuseinterface theory to the sharp interface limit and determine the effective interfacetension of the diffuse interface by several means.

Dans cette étude, on présente une contribution à la modélisation expérimentale etnumérique du comportement élastoplastique endommageable anisotrope de tôles destinées à lamise en forme. L’accent est mis sur la localisation des instabilités plastiques induitespar l’anisotropie plastique et l’endommagement. La procédure d’identification desparamètres matériau du modèle de comportement est basée sur une approche inversed’optimisation s’appuyant sur des essais de traction orientés, associée à des simulationsnumériques (Abaqus/Standard et Matlab). La validation de la procédure proposée a pu êtremenée grâce à une confrontation entre simulations numériques et résultats expérimentauxobtenus sur des éprouvettes caractéristiques des tôles minces texturées en acier doux.L’application est ensuite faite à la simulation de procédés de gonflement hydraulique deces tôles avec prévision de l’endommagement ductile. Plusieurs cas en expansion libre(matrices circulaires et elliptiques) et en expansion dans des cavités de matrice serontprésentés et des comparaisons avec les résultats expérimentaux seront réalisées.

Thirty six kinds of chamfered and unchamfered main cutting edge carbide tips were used in turning of high-strength glass-fiber-reinforced plastics (GFRP) materials to study the cutting temperature of tip's surface. The frictional forces and heat generated on elementary cutting tools are calculated by using the measured cutting forces and the theoretical cutting analysis. The heat partition factors between the tip and chip are solved by using the inverse heat transfer analysis, which utilizes temperature on the K type carbide tip's surface measured by infrared as the input. The tip's surface temperature of the carbide is solved by finite element analysis (FEA) and compared with those obtained from experimental measurements. A good agreement demonstrates the accuracy of the proposed model.

The incompressible MHD equations couple Navier-Stokes equations with Maxwell's equationsto describe the flow of a viscous, incompressible, and electrically conducting fluid ina Lipschitz domain $\Omega \subset \mathbb{R}^3$ .We verify convergence of iterates of different coupling anddecoupling fully discrete schemes towards weak solutions forvanishing discretization parameters. Optimal first order of convergence is shown in the presence of strong solutions for a splitting scheme which decouplesthe computation of velocity field, pressure, and magnetic fields atevery iteration step.

Ce travail est dédié à l'approche avancée pour modéliser et simuler un procédé de mise en forme de tôles minces par emboutissage dans l'objectif d'optimiser l'évolution de l'endommagement au cours du formage. La méthode proposée permet de prédire l'instant et l'endroit d'apparition de l'endommagement dans une pièce mécanique avant sa réalisation physique. Une approche locale basée sur le couplage fort entre le comportement élasto-plastique anisotrope à écrouissage mixte (isotrope et cinématique) non-linéaire et un endommagement ductile supposé isotrope est proposée. Les aspects numériques et théoriques des équations constitutives du modèle couplé seront présentés ainsi que la réduction des équations. La résolution globale du système d'équations est réalisée par un schéma dynamique explicite. La procédure est appliquée à la prévision de défauts dans un procédé d'emboutissage de type Swift. En dernier les résultats des simulations sont comparés à ceux de l'expérience.

Lattice strains around a platelet oxygen precipitate in Si wafer is studied by energy filtering convergent-beam electron diffraction (CBED) and calculations based on the finite element method (FEM). Local lattice strains are measured from CBED patterns obtained with a probe size less than 2 nm in a specimen thicker than 450 nm. Strains measured are compressive along a direction normal to a plate of the precipitate and tensile along a direction parallel to the plate. Two-dimensional stress fields near the precipitate are obtained with FEM computer analyses by fitting the measured strains. It appears that shear stresses are concentrated at the end of the precipitate edge and the maximum shear stress at an interface between the precipitate and the Si-matrix is 1.9 GPa. It is demonstrated that a combination of the energy filtering CBED and FEM is very useful for the study of local strains near interfaces in semiconductor devices, in particular for the study of stress fields that are too steep for application of the conventional CBED technique.