Scipedia: Documents published in 2015

Pseudo forces and stress evaluation in hyper beam models

Scipedia content — Thu, 13 Oct 2016 16:34:09 +0200

This paper deals with the unified formulation of linear models in rod elastostatics and, specifically, with the class of models including generalized displacements that are additional to those defining rigid body motions of the cross-sections. The main feature of these models, named hyper beam models by the authors, is the coupling between static and kinematic variables in the equilibrium equations. New static variables called pseudo forces have been introduced to represent the action of the internal constraints which control the deformation of the cross-section. Using the pseudo forces, a systematic procedure to evaluate the model-consistent stress distributions on cross-sections –which is an approximation of the 3D elastic solution– has been developed. The application of the unified formulation to a problem requiring the introduction of non rigid-body-motional degrees of freedom –warped bending– shows its ability to systematically build the equations, and to assess the influence of the intervening parameters. The response-defining variables, parameters and equations, as well as the expressions of consistent stress distributions on the sections, are obtained by means of the proposed unified procedure, parting exclusively from the cross-sectional kinematics expressed in terms of the respective generalized displacements.

Computational wind engineering and its application in civil engineering. Aerodynamic and aeroelastic analysis

Scipedia content — Thu, 13 Oct 2016 16:33:58 +0200

Experimental tests in wind tunnels have been traditionally employed as a fundamental tool to evaluate aerodynamic and aeroelastic effects due to wind action on civil engineering structures, such as bridges and slender buildings. In the last decades, due to the versatility presented by numerical methods to change physical as well as geometrical parameters, numerical simulation has become a very attractive tool. Computational Fluid Dynamics (CFD), Computational Structural Dynamics (CSD) together with Fluid‐Structure Interaction (FSI) techniques are employed in aerodynamic and aeroelastic analysis in several engineering fields. Aerodynamic and aeroelastic behavior due to wind action on the Guama River Bridge, located at Pará State, Brazil, is first studied, taking into account experimental tests performed in the Wind Tunnel Joaquim Blessman of the Building Aerodynamic Laboratory, UFRGS. Numerical procedures are used to simulate experimental tests in order to determine aerodynamic and aeroelastic characteristics of the bridge, which is idealized by sectional models. Finally, an aeroelastic analysis of a flexible slender building is presented. Good results are obtained using numerical simulation, when compared with experimental tests.

A calculation methodology for the determination of the dynamic alongwind response of tall structures under wind action

Scipedia content — Thu, 13 Oct 2016 16:33:40 +0200

The optimization of computing methods along with the advance of construction technology of civil structures, allow nowadays to design and build lighter and low stiffness and damping structures, turning them into elements of high sensitivity against wind-induced dynamic effects. The purpose of this work is to describe and to apply a simple calculation methodology that enables the determination of the longitudinal dynamic response of tall buildings in time domain. The wind velocity field is reproduced by velocity series that incorporate the coherence function which then are transformed into fluctuating nodal forces using the quasi-static loading model. Furthermore, to include the effect of imperfect averaging associated with the discretized nodes, the aerodynamic admittance function is added. Through the proposed methodology two structures widely studied by other authors are analyzed: a steel tower and the CAARC Standard Tall Building. The results thus obtained show agreement with those reported in the literature.

Simplified model for the study of the asymmetric rocking motion of slender rigid bodies

Scipedia content — Thu, 13 Oct 2016 16:33:06 +0200

This paper deals with the numerical modelling of rigid blocks by means of a rigid body spring model (RBSM). The hypothesis of small rotations has been considered in order to simplify the mathematical formulation. The asymmetric rocking motion is described. Extensive numerical simulations have been carried out in order to validate the numerical model, as well as to study the dynamical behaviour of asymmetric rigid bodies. The maximum response depends on the type of asymmetry and the direction of the load, which should become new variables in the study of asymmetric rocking motion.

Hemodynamic on abdominal aortic aneurysm: Parametric study

Scipedia content — Thu, 13 Oct 2016 16:33:00 +0200

The aim of this study is to assess how the shape of the abdominal aortic aneurysms (AAA) affects the hemodynamic wall stresses. With this purpose, different AAAs are studied through simplified models based on geometrical parameters of the aneurism such as its maximum and minimum diameter, length and asymmetry. Then, a computational fluid dynamics analysis is performed on the simplified models in order to compute pressure and wall shear stresses on the aneurysm sac. The results obtained show that blood pressure is the main dynamic load acting on the artery wall, and that the morphology of the aneurysm could be a good indicator of risk of failure. Furthermore, the computational results are compared with patient-specific real models with the objective to assess the reliability of the proposed simplified approach.

Development of a decision support system for optimization of the performance of sailing yachts

Scipedia content — Thu, 13 Oct 2016 16:32:53 +0200

In this paper, the conception and design of a new monitoring system for a racing yachts rig is presented. The sensors developed are able to process the measured strain data, by applying artificial neural networks (ANN) algorithms, and then evaluate the load acting on an element and identify the direction of the action of that force. This way, it is possible to identify the actual operating conditions of the yacht rig. The required data for ANN training is generated from the results obtained from different finite element method (FEM) computational models of the device. Furthermore, during the design phase of the system, different experimental campaigns were carried out. The experimental tests were designed to serve as proof of concept, as well as to validate the different procedures used in the system development and application. The developed monitoring system is wireless, low-intrusive and easily adaptable to any yacht configuration. This work also presents the integration of the monitoring system into a coupled fluid-structure computation model for the sails and rig of a boat. The resulting system is an efficient tool for evaluating performance and decision support in the adjustment of a sailboat rig.

Design of a movil manipulator with optimum mechanical energy using differential evolution

Scipedia content — Thu, 13 Oct 2016 16:32:16 +0200

In this work, the design of kinematic and dynamic parameters of links of a mobile manipulator is stated as a numerical optimization problem (NOP). The NOP aims to minimize the mechanical energy of the system in critical positions given by the vertices of a workspace and consequently reduce the energy consumption of the control system during its operation in the workspace. An exhaustive exploitation mechanism is included in the traditional differential evolution algorithm to improve the search into a neighborhood in the design space. Simulation results show the performance of the algorithm. The resulting design shows less energy consumption than other three designs.

Numerical analysis of the vibrational behavior of the moving assembly of a dynamic loudspeaker

Scipedia content — Thu, 13 Oct 2016 16:32:07 +0200

This paper describes the process undertaken to analyze numerically the vibrational behavior of the moving assembly of a dynamic loudspeaker with double bottom suspension and in the absence of the above one using the finite element method (FEM). The study focuses on the low frequency range. Model calibration was performed based on experimental measurements of the resonance frequency and diaphragm displacement. The importance of the participation factors associated to the excitation force and the changes produced in these when unbalanced forces are introduced are emphasized. The analysis also provides data to decide on the distance between suspensions, their optimal number of folds and the placement of the lead wires, parameters of great interest on the design of this type of loudspeakers.

Modeling of the free border singularity in 3D cracks using XFEM and spherical harmonics

Scipedia content — Thu, 13 Oct 2016 16:32:00 +0200

One type of singularity that may appear in a three-dimensional fracture problem under elastic behavior is the free corner singularity, which occurs at the intersection of the crack with a free boundary and whose effect is usually ignored. The extended finite element method (XFEM) is a technique that allows the efficient numerical modeling of fracture problems, by using enrichment functions within a finite element model that incorporate the geometry and effect of the crack. However, when applied to the study of a problem with corner singularity, the singular behavior is not properly captured. The reason is that the usual enrichment in the XFEM only describes the typical crack front singularity. Hence, in order to include the effect of the free border singularity, the enrichment has to be modified. In this work, we present a new set of functions which relies on spherical harmonics and is able to capture the behavior of the free border singularity.

Changes in wave propagation in a guitar top plate due to the fan bracing and the bridge

Scipedia content — Thu, 13 Oct 2016 16:31:56 +0200

Wave propagation on a guitar top plate was explored using transient analysis. The goal of using time domain, instead of frequency domain as it is commonly reported, was to achieve a more simple approach about the influence of the fan bracing and the bridge. For this purpose, the elastic properties of a real top plate were measured through simple experiments, in order to build the respective finite element model. The model exhibited good agreement with experimental behaviors of the literature. In the same top plate, the propagation of two kinds of waves was simulated; one analysis was done using one wave of high harmonic content, and another applying a wave with low harmonic content. Results showed, to naked eye, that the velocity of propagation over the top plate clearly changed due to their interactions with the attached structures. Such change was particularly noticeable for waves of short wavelength, in other words, for high frequencies. Then, it is evident that for guitars, designs of fan bracing and bridge are crucial, mainly, in treble sounds radiated by the instruments.

Study of the optical behavior of skins using Monte Carlo techniques

Scipedia content — Thu, 13 Oct 2016 16:31:46 +0200

In both biomedical optics and biophotonics, one of the main challengers is the simulation of light spread in biological tissues. The approximation which is most used is Monte Carlos method; it is a standard because of its flexibility and its reliability modeling heterogeneous tissues. This paper shows how to analyse the optical behaviour of the skin, and how some of the deductions drawn from this analysis can add information for the diagnosis and treatment of skins. The optic qualities of healthy skins from different races have been applied; and as a possible alteration in the skin, basalioma presence has been studied. The results provide us with a method to distinguish between “healthy” and “ill” skin, which could make the procedure to identify cancer skins easier. From the study of the different healthy skins an example of immediate application also appears which benefits from the knowledge acquired from the values obtained (photodepilation). In this work the most valued results of the simulation programs based in Monte Carlos technics are presented, which allow studying the skins behaviour facing an optical radiation; the approximation to the light trajectory interacting with the tissues is obtained from the optic qualities known before. First of all, a general approach to the theme is carried out, identifying the main problems that appear in this kind of studies and a commercial programme is used.

An efficient simulation of structures under impulsive loadings and an expedited application

Scipedia content — Thu, 13 Oct 2016 16:31:28 +0200

The class of physical problems of metal structures subject to rapid, large amplitude loads, such as occurs when they are submitted to an explosion blast, involves large deformations and inelastic action. Current state of art in explicit dynamic Finite Element Analysis codes require time steps limited by stability requirements rather than by accuracy. For very high rate processes, accuracy would require a very small time step so that the stability limit does not penalize. Acceptable accuracy may not require such small steps if implicit methods are considered. The implicit methods used in quasi‐static analysis, however, require the solution of large systems of equations, and several such systems must be solved at each time step to satisfy the dynamic equilibrium when using Newton iteration. The direct matrix solvers that are used incur in computational costs that grow as n m2, where n is number of degrees‐of‐freedom and m is the band of the front width, thus resulting in prohibitive costs for very large problems. In this work, the use of higher order, implicit time integration and matrix‐free solution of the system of equations is explored. Computational efficiency of explicit codes is combined with the inherent stability of implicit methods. The model developed is applied in the simulation of a complex structure, with elevated computational cost, showing its validity.

Nonlinear dynamical model for the assessment of the seismic behaviour of ferrocement houses

Scipedia content — Thu, 13 Oct 2016 16:31:21 +0200

This article evaluates the cyclic behaviour of a prototype of a precast ferrocement house, which is assembled using precast thin walls, in order to identify its structural parameters and propose a nonlinear dynamical model which is able to simulate the behaviour of those houses when subjected to seismic loading. For this purpose, a full-scale precast ferrocement house was tested under cyclic loading conditions. From the experimental results, the hysteretic behaviour, the elastic stiffness, the shear resistance, the maximum strength, the ductility, the energy dissipation, the equivalent damping and its coefficient of energy dissipation were assessed; those parameters are needed in order to evaluate the dynamic behaviour of ferrocement dwellings when subjected to earthquakes. At the end, the Bouc-Wen-Baber-Noori (BWBN) model of hysteresis was implemented. The model allowed performing the seismic evaluation of the precast ferrocement house under seismic activity.

Derivation of nodular cast iron elastic properties via computational micromechanics

Scipedia content — Thu, 13 Oct 2016 16:31:08 +0200

Nodular cast iron is a Fe-C-Si metallic alloy whose microstructure consists of a matrix, composed in general of ferrite and pearlite, with graphite nodules embedded in it. Owing to its microscopical heterogeneity, the material response is governed by the properties, morphology and typology of the phases involved. This work reports on the evaluation of the elastic properties, i.e., Youngs modulus and Poissons ratio, of an equivalent homogeneous material that characterizes the macroscopic response of a nodular cast iron. Asymptotic homogenization is used to this end. This approach is applied to both 3D and 2D multiparticle cells simulated via the finite element method. Two different physically and geometrically-based criteria are considered to estimate the representative volume element (RVE), where the size of the RVE is found to be sensitive to the chosen criterion. The main microstructural features are obtained from a computational simulation of the solidification process of the material. The numerical predictions computed for the 3D and 2D cases are compared and discussed in terms of the resulting elastic properties. It is observed that the models employing 3D multiparticle cells require lower RVE sizes than the corresponding 2D models, where the latter present a stiffer elastic response.

Obtaining foot bone structure applying global and adaptive thresholding

Scipedia content — Thu, 13 Oct 2016 16:30:55 +0200

The description of the mechanical behavior of hard tissues by means of discrete models goes through various stages of analysis, which range from digital image processing to the specification of tissues physical properties to the discrete model. To achieve good results it is essential to decompose these models into their constituent parts. In this paper we discuss a method for geometrical description of foot bones from a sequence of computed tomography (CT) images. This research proposes a combination between global and adaptive thresholdings to determine the geometric domain of bones in each slice and the analysis of the spatial relationships between contours in consecutive planes in order to obtain bones’ isosurfaces. The algorithm proposed is based on 4 stages: the reading of computed tomography (CT) images; the determination of the contours that define the bone tissue present on each slice; the grouping of contours whose relationship meet a given criteria; the elimination of non-valid volumes. As a result, it is possible to obtain the geometrical domain of a great number of foot bones whose range in the Hounsfield is [–1000; 1383].

Design of slender steel structures sensitive to buckling. Geometric imperfection definition

Scipedia content — Thu, 13 Oct 2016 16:30:43 +0200

The purpose of this paper is to present a proposal for the design of steel structures sensitive to buckling to fill the gaps in the current Standard EN 1993-1-1, providing guidelines to obtain the magnitude of the imperfection generalizing the procedure given in clause 5.3.2(11) of EN 1993-1-1 for steel structures sensitive to flexural buckling under compression. According to the standard there are still uncertainties about how to obtain the imperfection when the structures are sensitive to torsional or flexural-torsional buckling due to either compression or bending. In this paper a general procedure to obtain the imperfection is provided and it can be proved that the recommendation given in EN 1993-1-1 is a particular case.

A proper generalized decomposition approach for high-order problems

Scipedia content — Thu, 13 Oct 2016 16:29:55 +0200

In this paper two different approximations for the solution of high-order problems by proper generalized decompositions (PGD) are developed. The first one is based upon the use of collocation techniques, along with Chebyshev polynomials, while the second employs Hermite polynomials in a Galerkin framework. Both approaches having pros and cons, they are studied with the help of some classical benchmark tests.

New thermal compensation algorithm for piezoresistive pressure sensors based on the linear approximation by segments

Scipedia content — Thu, 13 Oct 2016 16:29:45 +0200

The piezoresistive pressure sensors are widely used in monitoring and control applications within the industrial sector. These sensors have problems of temperature dependence, and non-linearities that affect the measurement accuracy. For this reason, it is necessary to use signal conditioning circuits for to obtain high accuracy outputs. One of the most common techniques currently used to conditioning the signals from the sensors is based on the use of digital signal processors and compensation algorithms. This paper proposes a new calibration and compensation algorithm for piezoresistive pressure sensors based on the linear segment approximation with thermal compensation. The algorithm was simulated using the mathematical software MathCAD and using data from a piezoresistive pressure dummy sensor. The algorithm has been validated experimentally implementing it into two digital signal processors using vented piezoresistives pressure sensors of 10 bar in the range of temperatures from 0 °C to 80 °C. The result was that the maximum error in the sensor output due to the temperature dependence of the offset was 0,081% of full scale output. The maximum error obtained at the output caused by the temperature dependence of the sensitivity was 0,106% of full scale output. The effectiveness of the algorithm is demonstrated in the thermal compensation of piezoresistive pressure sensors. This algorithm can be applied to other types of sensors having thermal dependencies and incorporating a microprocessor or digital signal processor.

Control parameter for polygonal simplification applied to medical images

Scipedia content — Thu, 13 Oct 2016 16:29:39 +0200

The description of mechanical behavior of hard tissues by means of discrete models goes through various stages of analysis, which range from digital image processing and the specification of physical properties of tissue to the discrete model. Nowadays, technologies for acquiring medical images allow obtaining high resolution images. Obtaining the geometrical domain of these images requires high levels of hardware to process and store them, and polygonal simplification, with a defined minimum error, is one of the ways to compress the geometrical domain. Polygonal simplification is important in order to have the necessary and sufficient information to reconstruct surfaces and volumes that geometrically describe tissues and organs of the human body for their analysis by means of finite elements. In this paper an experimental control parameter is defined for polygon simplification. Experimental results proved that is possible to obtain a tridimensional model with an average volume variation of 87.49%, using only 36.3% of initial data, with a maximum error of 1.0% between initial and final polygon areas. All this process takes less processing time.

An implicit implementation of the Arruda-Boyce viscoplastic model

Scipedia content — Thu, 13 Oct 2016 16:29:22 +0200

The Arruda-Boyce viscoplastic model has been the cornerstone for the development of sophisticated constitutive models for polymers and soft tissues. A simple implicit finite element implementation for the Arruda-Boyce viscoplastic model, which is coded as user material routine in ABAQUS-Standard, is presented in this work. The implementation uses the Backard-Euler method and standard stress-update procedures in the relaxed configuration; it has no iterative procedures; it completely avoids the use of eigen-decompositions, polar decompositions and rotation matrices while square roots, exponentials and natural logarithms of second order tensors are computed using high-order Padé approximants. The tangent-stiffness matrix is computed using a finite difference scheme. As a result, the stress-update algorithm turns out to be very simple to code and to implement, but still very accurate and quite efficient in computational terms. The precision, stability and efficiency of the proposed formulation are assessed and demonstrated by means of a number of examples. As a result, a number of recommendations are given to best set the size of the inelastic strain increment and the residuals for the Newton-Raphson procedures; and to select the most effective algorithms for the evaluation of the tangent stiffness matrix and to solve the system of equations.

Foam injection as a mechanism of mobility control in reservoir rocks

Scipedia content — Thu, 13 Oct 2016 16:29:08 +0200

One of the most important tasks in Petroleum Engineering is a reservoir characterization. A good knowledge of the properties of the reservoirs is essential for reliable production forecasts and the application of special methods of secondary recovery. A very useful tool for this purpose is the numerical simulation of oil reservoirs, which through the application of numerical methods allows for the solution of complex differential equations by nature. This paper deals with the computational modeling of foam injection in oil reservoirs using the Eclipse® software. The main objective of this research is to understand what the parameters that maximize the increase of the oil recovery factor retained in the pore space after using conventional methods are. Foam is injected from the surface into a reservoir of 8.400 feet deep containing oil viscosity which is dependent on the pressure, this oil is conducted to the surface through a production well. Both wells are completed over the entire thickness of the formation of interest. Equations that express the detailed model and the foam are discretized together with the governing equations of the fluid flow. Comparisons at different foam concentration are shown for light, intermediate and heavy oils. Results show that there is a direct correlation between the concentration of the displacing fluid foam and the oil recovery factor. Note also that a critical concentration c* foam is evidenced with direct influence on the efficiency of the process.

Implementation of a moving finite element mesh within the catenary-pantograph dynamic interaction

Scipedia content — Thu, 13 Oct 2016 16:29:03 +0200

This paper presents a technological application of a general methodology to analyze cable structures under moving loads, particularly on the catenary-pantograph dynamic interaction. This work firstly describes the modeling of a finite element moving mesh which integrally moves over the contact wire following the pantograph along the whole catenary, prior focusing on its comparison against the classical finite element mesh. The case study corresponds to the standard EN-50318 one, whose results fulfill the ranges proposed by this validation rule. The main advantage of the moving mesh method is presented in the falling of computational costs about 4 times lower than the classic model with similar precision.

Non-linear sloshing effect on storage tanks subjected to high earthquake ground motion

Scipedia content — Thu, 13 Oct 2016 16:28:50 +0200

A numerical method is proposed for the seismic response analysis of rectangular and cylindrical tanks accounting for non-linear hydrodynamic effects. The equations of motion are mapped from the physical to the computational domain by means of a change of variable. The numerical solution is then obtained by means of finite differences. The system of second order equations obtained from the finite difference discretization are linearized using the Newton method and solved numerically by means of the biconjugate gradient method. The numerical formulation was applied to study the response of tanks subjected to ground motion records from the Mexico earthquake of 19 September 1985, obtained in Mexico City. Response results are given in terms of wave height on the liquid surface, and shear force and overturning moment at the base wall, and are compared with those obtained for the linear solution.