Stress-strain Simulation Research Articles

Development of a ReaxFF Reactive Force Field for Ettringite and Study of its Mechanical Failure Modes from Reactive Dynamics Simulations

Ettringite is a hexacalcium aluminate trisulfate hydrate mineral that forms during Portland cement hydration. Its presence plays an important role in controlling the setting rate of the highly reactive aluminate phases in cement paste and has also been associated with severe cracking in cured hardened cement. To understand how it forms and how its properties influence those of hardened cement and concrete, we have developed a first-principles-based ReaxFF reactive force field for Ca/Al/H/O/S. Here, we report on the development of this ReaxFF force field and on its validation and application using reactive molecular dynamics (RMD) simulations to characterize and understand the elastic, plastic, and failure response of ettringite at the atomic scale. The ReaxFF force field was validated by comparing the lattice parameters, pairwise distribution functions, and elastic constants of an ettringite crystal model obtained from RMD simulations with those from experiments. The predicted results are in close agreement with published experimental data. To characterize the atomistic failure modes of ettringite, we performed stress-strain simulations to find that Ca-O bonds are responsible for failure of the calcium sulfate and tricalcium aluminate (C3A) column in ettringite during uniaxial compression and tension and that hydrogen bond re-formation during compression induces an increase in plastic strain beyond the material's stress-strain proportionality limit. These results provide essential insight into understanding the mechanistic role of this mineral in cement and concrete degradation, and the ReaxFF potential developed in this work serves as a fundamental tool to further study the kinetics of hydration in cement and concrete.

Simulation of Semi-Solid Material Mechanical Behavior Using a Combined Discrete/Finite Element Method

As a necessary step toward the quantitative prediction of hot tearing defects, a three-dimensional stress–strain simulation based on a combined finite element (FE)/discrete element method (DEM) has been developed that is capable of predicting the mechanical behavior of semisolid metallic alloys during solidification. The solidification model used for generating the initial solid–liquid structure is based on a Voronoi tessellation of randomly distributed nucleation centers and a solute diffusion model for each element of this tessellation. At a given fraction of solid, the deformation is then simulated with the solid grains being modeled using an elastoviscoplastic constitutive law, whereas the remaining liquid layers at grain boundaries are approximated by flexible connectors, each consisting of a spring element and a damper element acting in parallel. The model predictions have been validated against Al-Cu alloy experimental data from the literature. The results show that a combined FE/DEM approach is able to express the overall mechanical behavior of semisolid alloys at the macroscale based on the morphology of the grain structure. For the first time, the localization of strain in the intergranular regions is taken into account. Thus, this approach constitutes an indispensible step towards the development of a comprehensive model of hot tearing.

An adaptive heat source model for finite-element analysis of keyhole plasma arc welding

Abstract An adaptive model of the heat source in numerical simulation of the keyhole plasma arc welding (PAW) requires that the model take into consideration the larger aspect ratio of welds and the volumetric distribution characteristic of heat intensity along the direction of the plate thickness. The available heat source models, either planar one like Gaussian or body ones like double-ellipsoidal and rotary Gaussian modes, are unable to describe the keyhole PAW process accurately. Based on the configuration feature of keyhole PAW welds, a combined heat source model is proposed for the numerical analysis of temperature fields in keyhole PAW process. It accounts for the keyhole effect indirectly and the volumetric distribution characteristics of plasma heat intensity along the direction of the workpiece thickness. With this adaptive heat source model, finite-element analysis of temperature profile in keyhole PAW is conducted and the weld geometry is determined. The results show that the predicted geometry and locus of the fusion line in the keyhole PAW weld cross-section are in good agreement with the experimental measurements. It lays solid foundation for the process optimization, metallurgical analysis and thermal stress–strain simulation of keyhole PAW process.

Development of Time Dependent Stress-Strain Simulation of Clay

AbstractThe objective of this study is to derive a time dependent effective based constitutive law on the basis of framework of the Modified Cam-Clay model. This model takes into account the anisotropic characteristics and creep behavior, based on the theory of viscoplasticity. The model sets the initial yield surface symmetric to theKoline for modeling the initialKocondition. A method is then developed to compute the gyration and expansion of the loading surface to simulate the anisotropic behavior due to the principal stress gyration after shear. The creep or time dependent behavior is considered in the model by adopting Kutter and Sathialingam's model, which was derived from Taylor's secondary consolidation theory and Bjerrum's delayed compression model. Compared with the Modified Cam-Clay model, the model requires five additional parameters to describe the soil behavior. All of the additional parameters can be obtained through conventional soil tests or parametric studies. The model is evaluated through a series of simulation of undrained shear tests and undrained creep tests.

The Mechanical Changes in the MEA of PEM Fuel Cells due to Load Cycling

The purpose of this work is to establish a theoretical framework for modeling of the interface changes between the MEA components of PEM fuel cells due to load cycles. A 2D finite element model with the cohesive zone model and the fractional contact model is proposed to simulate the interface changes due to hydrothermal cycles. The material properties as functions of temperature, humidity and strain rate are applied here. The stress-strain simulations show stress and strain distributions inside the MEA, how and where, the interface changes take place during the load cycle and the cause of membrane thinning.

Experimental evaluation of the electric pulse effect on residual stresses in a Ti-Al-Si-Ag coating

Experimental results and stress-strain state simulation by the finite element method show that a thin oxidation-resistant coating sputtered onto a titanium alloy substrate develops high-level residual stresses. The electric pulse passed through a coated specimen significantly reduces an average level of residual stresses in the coating.

A non-coaxial critical state soil model and its application to simple shear simulations

The yield vertex non-coaxial theory is implemented into a critical state soil model, CASM (Int. J. Numer. Anal. Meth. Geomech. 1998; 22:621–653) to investigate the non-coaxial influences on the stress–strain simulations of real soil behaviour in the presence of principal stress rotations. The CASM is a unified clay and sand model, developed based on the soil critical state concept and the state parameter concept. Without loss of simplicity, it is capable of simulating the behaviour of sands and clays within a wide range of densities. The non-coaxial CASM is employed to simulate the simple shear responses of Erksak sand and Weald clay under different densities and initial stress states. Dependence of the soil behaviour on the Lode angle and different plastic flow rules in the deviatoric plane are also considered in the study of non-coaxial influences. All the predictions indicate that the use of the non-coaxial model makes the orientations of the principal stress and the principal strain rate different during the early stage of shearing, and they approach the same ultimate values with an increase in loading. These ultimate orientations are dependent on the density of soils, and independent of their initial stress states. The use of the non-coaxial model also softens the shear stress evolutions, compared with the coaxial model. It is also found that the ultimate shear strengths by using the coaxial and non-coaxial models are dependent on the plastic flow rules in the deviatoric plane. Copyright © 2006 John Wiley & Sons, Ltd.

Simple Plasticity Sand Model Accounting for Fabric Change Effects

A simple stress-ratio controlled, critical state compatible, sand plasticity model is presented, first in the triaxial and then in generalized stress space. The model builds upon previous work of the writers, albeit the presentation here is made with extreme simplicity in mind, and three novel aspects are introduced. The first is a fabric-dilatancy related quantity, scalar valued in the triaxial and tensor valued in generalized stress space, which is instrumental in modeling macroscopically the effect of fabric changes during the dilatant phase of deformation on the subsequent contractant response upon load increment reversals, and the ensuing realistic simulation of the sand behavior under undrained cyclic loading. The second aspect is the dependence of the plastic strain rate direction on a modified Lode angle in the multiaxial generalization, a feature necessary to produce realistic stress-strain simulations in nontriaxial conditions. The third aspect is a very systematic connection between the simple triaxial and the general multiaxial formulation, in order to use correctly the model parameters of the former in the implementation of the latter. The simulative ability of the model is illustrated by comparison with data over a very wide range of pressures and densities.

Fatigue lifetime estimation of structures subjected to dynamic loading

Because fatigue is a typical dynamic phenomenon, a procedure is developed to estimate the lifetime of a structure, based on its dynamic characteristics. By means of modal analysis, a structure can be described by its modal parameters: mode shapes, natural frequencies, and damping values. Using a finite-element model, the modal stress-strain distribution of part of or an entire structure can be calculated. When the applied dynamic forces are known, combination of the modal strains yields the strain response of the structure. A stress-strain simulation controlled by Neuber's rule accounts for stress concentrations. A counting algorithm based on closed hysteresis loops divides the stress-strain history into single reversals. An appropriate damage rule is used to find the damage of one loading sequence. Finally, the lifetime is found from the number of loading sequences that can be applied to satisfy the failure criterion.

On the sonic fatigue life estimation of skin structures at room and elevated temperatures

Fighter/trainer empennages and STOL (Short Take-Off and Landing) aircraft flap systems are subjected to severe acoustic pressure levels as high as 150–170 dB. As a result, acoustic fatigue has become one of the major factors in design. Empennages and flap systems are also subject to high temperatures and thus the influence of thermal buckling on fatigue life must be taken into consideration. To estimate the sonic fatigue life of skin structures, combined use is made of the Monte Carlo method of non-linear panel response analysis and local stress-strain simulations with rheological models. Calculations were conducted and comparison of the results with experimental data shows that the method estimates the sonic fatigue life accurately when adequate values of fatigue notch factors are chosen. Example calculations with thermal as well as static pressure effects were also carried out and the effect of thermal buckling on sonic fatigue life is clarified.