نشریه روش های عددی در مهندسی
سال سی و پنجم شماره 2 (زمستان 1395)

Weighted Simulation-based Design Method (WSDM) is recently developed as an efficient method for Reliability- Based Design Optimization (RBDO). Despite simplicity, this method degrades effectiveness to obtain accurate optimum design for high dimension RBDO. Besides, its application range is restricted to RBDOs including only random design variables. In the present study, local search strategy is employed to enhance the accuracy of conventional WSDOM, and to reduce the computational cost. Besides, a shifting strategy is proposed to increase the application range of WSDM for handling general RBDO problems. The efficiency of the proposed methods is investigated by solving some structural reliability problems. Comparisonof the obtained results with exact solutions confirms accuracy and superiority of the proposed method for solving various engineering problems.

In this article, the vibration analysis of a set of parallel Timoshenko beams connected by intermediate flexible connections, with arbitrary numbers, is studied. The moving load is a vehicle, which is modeled by a two-axle six degrees of freedom system, as a mass-spring-damper system, in a plane motion. For the solution, a new method is proposed which uses a change of variables strategy to decouple the system of differential equations. For this purpose, the stiffness matrix obtained from each column of intermediate connections should have the same normalized eigenvectors. The displacements and the bending moments of the beams and the vehicle due to changes in the stiffness of connections and changes in speeds will be examined. Finally, the validity of the results are measured.

In this study, for the first time natural convection heat transfer of Al2O3-water nanofluid with constant and variable properties is investigated within square shape porous media using the lattice Boltzmann method. The horizontal walls of the cavity are insulated, and left and right vertical walls are hot and cold, respectively. The Study have been carried out for Rayleigh numbers of 103, 104, 105, 106, Darcy numbers of 10-2, 10-4, porosity coefficients of 0.4, 0.6, 0.9 and solid volume fraction of 0, 0.01, 0.02 and 0.03. In order to consider the effect of porous media, Darcy-Forchheimer model is used. The results show that the presence of the porous media decreases the velocity of nanofluid and consequently decreases the strength of the flow. With decreasing Darcy number and porosity coefficient, natural convection heat transfer weakens and the mechanism of natural convection of nano-fluids tends to that of thermal conduction. With increasing Rayleigh number, the strength of flow in cavity and average Nusselt number increases. In all cases studied, increase in volume fraction improves heat transfer. In constant properties model, by increasing solid volume fraction, average Nusselt number increases more than that of variable properties model. The results show that Lattice Boltzmann method has the ability to simulate flow in porous media.

In this paper, the solution of a non-linear model of Cell Manufacturing (CM) in certain and dynamic conditions is studied, considering intracellular and extracellular costs, cell constructing costs, the cost of restoration and the cost of equipment transportation per distance travelled. Since the number of cells in each stage of production is important, by optimizing the number of cells, additional costs can be minimized. Therefore, the main objective of this study is to investigate the optimal number of cells located. Bio-geographical Based Optimization (BBO) algorithm is applied in the CM for the first time in the literature and the obtained results from this algorithm are compared with the results of well-known genetic algorithm. The results shows the good performance of genetic algorithm. Finally, the conclusion and future research are provided.

The purpose of the present study is to investigate and analyze numerically, the effective mechanisms on the flow field and pollutant dispersion around a simple and long street canyon by means of Large Eddy Simulation (LES) approach using various inflow turbulence generation methods. For this purpose, four methods i.e. vortex, mapping, synthetic and no-inlet perturbation methods are used as inflow turbulence generators in LES. Results suggest that all methods are capable of capturing the two important structures of canyon vortex and corner eddy, which have great influences on air ventilation inside the street canyon. The magnitudes of concentration on the leeward wall of the first building are approximately four times as much as those of windward wall of the second building. Among the various inflow turbulence generation methods, the vortex method is the most precise method and no-inlet perturbation method is the least precise method.

Many efforts have been done in recent years to decrease the required time for analysis of FKP (Forward Kinematics Problem) of parallel robots.This paper starts with developing kinematics of a parallel robot and finishes with a suggested algorithm to solve forward kinematics of robots. In this paper, by combining the artificial neural networks and a 3rd-order numerical algorithm, an improved hybrid strategy is proposed in order to increase the accuracy and speed of forward kinematics analysis of parallel manipulators. First, an approximate solution of the forward kinematics problem is produced by the neural network. This approximate solution is then considered as the initial guess for the 3rd order Newton-Raphson numerical technique. By applying Stewart-Gough parallel manipulator, the efficiency of the proposed method is evaluated. It is shown that replacing the Newton-Raphson algorithm by the 3rd-order one leads to a reduction of the iterations required to reach the desired accuracy level and thus a reduction of the FKP analysis time. Finally, Stewart robot is used to simulate the movement of jaw. This novel algorithm can be applied to any forward kinematics of serial or parallel robots.

In this paper, free vibration of carbon nanotube-reinforced functionally graded circular plates with hole has been investigated. Distribution of carbon nanotubes are continuous and the gradual and graded changes of materials through the plate thickness are considered as volume fraction. Considering the linear and non-linear variation of circular plates through the radial direction and also considering the proposed function for the thickness, the plate thickness can be convex or concave. Moreover, the motion equations of plate were obtained based on the third-order shear deformation theory. These equations are coupled differential equations which can convert Ordinary Differential Equations (ODE) using the Trigonometric series expansion of displacement fields such that they satisfy the axial symmetry condition. Solving the converted ODE equations is too difficult. For this reason, the differential quadrature method is employed to solve these equations. The obtained results are compared with the results reported by other researchers and an excellent agreement is observed between them. Finally, the effects of different geometric parameters as well as different volume fracture of nanotubes on natural frequency have been studied.

For solving the dynamic equilibrium equation of structures, several second-order numerical methods have so far been proposed. In these algorithms, conditional stability, period elongation, amplitude error, appearance of spurious frequencies and dependency of the algorithms to the time steps are the crucial problems. Among the numerical methods, Newmark average acceleration algorithm, regardless of existence of spurious frequencies, is very popular in the structural dynamics due to its unconditionally stability status of the method. Recently, several first-order methods have been introduced for resolving the accuracy and stability issues. However, in these methods stability, accuracy and error in inversion of the state matrix are known as major issues. When the state matrix became singular or ill conditioned, numerical errors will occure in the computational process. Many of the available first-order methods were to improve the stability and accuracy and also to remove the error of inversion. Even though the introduced methods are conditionally stable, no investigation on errors, occuring during dynamic loading, has been reported for them. The main purpose of this paper is to utilize a specific decomposition method based on Singular Value Decomposition (SVD) for modifying PIM algorithm. Using the SVD inversion technique, the singularity problem of the state matrix has been resolved. In this paper, the modified method is called PIMS. As well, by applying the developed method for dynamic loading, the error of responses has been investigated. The results show that PIMS algorithm is stable and, comparing with secoend order Newmark and other available first order methods, has more accuracy.

In this study, Circular Disk Model (CDM) has been developed to determine the residual stresses in twophase and three- phase unit cell. The two-phase unit cell is consisting of carbon fiber and matrix. The three-phase unit cell is consisting of carbon fiber, carbon nanotubes and matrix in which the carbon fiber is reinforced with the carbon nanotube using electrophoresis method. For different volume fractions of carbon nanotubes, thermal properties of the carbon fiber and carbon nanotube in different linear and lateral directions and also different placement conditions of carbon nanotubes have been considered. Also, residual stresses distribution in two and three phases has been studied, separately. Results of micromechanical analysis of residual stresses obtained from Finite Element Method and CDM, confirms the evaluation and development of three dimensional CDM.

In this paper, wave propagation method was applied to detect damage of structures. Spectral Finite Element Method (SFEM) was used to analyze wave propagation in structures. Two types of structures i.e. rod and Euler-Bernoulli beam were modelled using spectral elements. The advantage of spectral finite element over conventional Finite Element Method (FEM), in wave propagation problems, is its accuracy and lower computational time. Two examples of rod and Euler-Bernoulli beam with embeded concentrated mass were presented to illustrate the superiority of SFEM to FEM. Finally, a cracked beam was modeled and analyzed using spectral finite elements and the location of the crack was determined using time history response of beam structure.