International Journal of Optimization in Civil Engineering

TOPOLOGICAL OPTIMIZATION OF VIBRATING CONTINUUM STRUCTURES FOR OPTIMAL NATURAL EIGENFREQUENCY

B. Hassani

Keeping the eigenfrequencies of a structure away from the external excitation frequencies is one of the main goals in the design of vibrating structures in order to avoid risk of resonance. This paper is devoted to the topological design of freely vibrating continuum structures with the aim of maximizing the fundamental eigenfrequency. Since in the process of topology optimization some areas of domain can potentially be removed, it is quite possible to encounter the problem of localized modes. Hence, the modified Solid Isotropic Material with Penalization (SIMP) model is here used to avoid artificial modes in low density areas. As during the optimization process, the first natural frequency increases, it may become close to the second natural frequency. Due to lack of the usual differentiability of the multiple eigenfrequencies, their sensitivity are calculated by the mathematical perturbation analysis. The optimization problem is formulated by a variable bound formulation and it is solved by the Method of Moving Asymptotes (MMA). Two dimensional plane elasticity problems with different sets of boundary conditions and attachment of a concentrated nonstructural mass are considered. Numerical results show the validity and supremacy of this approach.

A CELLULAR AUTOMATA BASED FIREFLY ALGORITHM FOR LAYOUT OPTIMIZAION OF TRUSS STRUCTURES

S. Gholizadeh

In this study an efficient meta-heuristic is proposed for layout optimization of truss structures by combining cellular automata (CA) and firefly algorithm (FA). In the proposed meta-heuristic, called here as cellular automata firefly algorithm (CAFA), a new equation is presented for position updating of fireflies based on the concept of CA. Two benchmark examples of truss structures are presented to illustrate the efficiency of the proposed algorithm. Numerical results reveal that the proposed algorithm is a powerful optimization technique with improved convergence rate in comparison with other existing algorithms.

DEVELOPMENT OF RELATION FOR THE SEISMICALLY BASE ISOLATED STRUCTURES USING MODIFIED ABC ALGORITHM

M. Daei

In this paper, the effective parameters on the ductility demand of the seismically base isolated structure are investigated, and then a relation between the strength reduction factor and the target ductility is presented. The investigation has been conducted by modelling the base isolated structure as a two degree of freedom model in the OpenSees software, and the possibility of yielding in the superstructure has been considered in the model. Results show that the period of isolator and superstructure have the most effect on the ductility demand, therefore these two parameters beside the strength reduction factor and the target ductility have been used as variables of relation. A nonlinear regression model has been developed for forecasting the relation and the constant parameters of the proposed scheme has been obtained based on an optimization model solved by modified artificial bee colony (ABC) algorithm. A database including 224 models under 20 earthquake records with 2% probability of exceedance in 50 years have been generated for this purpose. Since there is not any explicit closed form formula to calculate the strength reduction factor for a specific target ductility; another optimization model has been developed to calculate the data used as input of the nonlinear regression model. The proposed relation includes two nonlinear functions and it is able to quantify the inelastic performance of base isolated structures for a wide range of earthquake records accurately.

A COMPARISON OF PERFORMANCE OF SEVERAL ARTIFICIAL INTELLIGENCE METHODS FOR ESTIMATION OF REQUIRED ROTATIONAL TORQUE TO OPERATE HORIZONTAL DIRECTIONAL DRILLING

H. Fattahi

Horizontal Directional Drilling (HDD) is extensively used in geothechnical engineering. In a variety of conditions it is essential to predict the torque required for performing the reaming operation. Nevertheless, there is presently not a convenient method to accomplish this task. To overcome this problem, in this research, the application of computational intelligence methods for data analysis named Support Vector Regression (SVR) optimized by differential evolution algorithm (DE) and Adaptive Neuro-Fuzzy Inference System (ANFIS) to estimate of the required rotational torque to operate horizontal directional drilling is demonstrated. Three ANFIS models were implemented, ANFIS–subtractive clustering method (ANFIS-SCM), ANFIS–grid partitioning (ANFIS-GP) and ANFIS–fuzzy c–means clustering method (ANFIS-FCM). The estimation abilities offered using SVR-DE, ANFIS-FCM, ANFIS-SCM, ANFIS-GP were presented by using field data given in open source literatures. In these models, the rotational torque (M) is used as the output parameter, while the length of drill string in the borehole (L), axial force on the cutter/bit (P), rotational speed (revolutions per minute) of the bit (N), the radius for the i^th reaming operation (D_i), the mud flow rate (W), the total angular change of the borehole (K_L), and the mud viscosity (V) are the input parameters. To compare the performance of models for rotational torque to operate horizontal directional drilling prediction, the coefficient of correlation (R²) and mean square error (MSE) of the models were calculated, indicating the good performance of the ANFIS-SCM model.

A COMPREHENSIVE STUDY ON THE CONCRETE COMPRESSIVE STRENGTH ESTIMATION USING ARTIFICIAL NEURAL NETWORK AND ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM

K. Behfarnia

This research deals with the development and comparison of two data-driven models, i.e., Artificial Neural Network (ANN) and Adaptive Neuro-based Fuzzy Inference System (ANFIS) models for estimation of 28-day compressive strength of concrete for 160 different mix designs. These various mix designs are constructed based on seven different parameters, i.e., 3/4 mm sand, 3/8 mm sand, cement content, maximum size of aggregate, gravel content, water-cement ratio, and fineness modulus. In this study, it is found that the ANN model is an efficient model for prediction of compressive strength of concrete. In addition, ANFIS model is a suitable model for the same estimation purposes, however, the ANN model is recognized to be more fitting than ANFIS model in predicting the 28-day compressive strength of concrete.

INFLUENCE OF FIBER ASPECT RATIO ON SHEAR CAPACITY OF DEEP BEAMS USING ARTIFICIAL NEURAL NETWORK TECHNIQUE

U. Naik

This paper deals with the effect of fiber aspect ratio of steel fibers on shear strength of steel fiber reinforced concrete deep beams loaded with shear span to depth ratio less than two using the artificial neural network technique. The network model predicts reasonably good results when compared with the equation proposed by previous researchers. The parametric study involves deep beams of M55 grade concrete with fiber volume fraction 0.5% to 2% of fiber aspect ratio ranging from 50 to 100 and longitudinal steel percentage varying from 0% to 2.5%. The analysis reveals that the fiber aspect ratio also affects the shear strength and needs to be combined with fiber volume fraction.

AN ADAPTIVE IMPORTANCE SAMPLING-BASED ALGORITHM USING THE FIRST-ORDER METHOD FOR STRUCTURAL RELIABILITY

M. A. Shayanfar

Monte Carlo simulation (MCS) is a useful tool for computation of probability of failure in reliability analysis. However, the large number of samples, often required for acceptable accuracy, makes it time-consuming. Importance sampling is a method on the basis of MCS which has been proposed to reduce the computational time of MCS. In this paper, a new adaptive importance sampling-based algorithm applying the concepts of first-order reliability method (FORM) and using (1) a new simple technique to select an appropriate initial point as the location of design point, (2) a new criterion to update this design point in each iteration and (3) a new sampling density function, is proposed to reduce the number of deterministic analyses. Besides, although this algorithm works with the position of design point, it does not need any extra knowledge and updates this position based on previous generated results. Through illustrative examples, commonly used in the literature to test the performance of new algorithms, it will be shown that the proposed method needs fewer number of limit state function (LSF) evaluations.

OPTIMIZATION FORMULATION FOR NONLINEAR STRUCTURAL ANALYSIS

M. Rezaiee-Pajand

In this paper, the effect of angle between predictor and corrector surfaces on the structural analysis is investigated. Two objective functions are formulated based on this angle and also the load factor. Optimizing these functions, and using the structural equilibrium path’s geometry, lead to two new constraints for the nonlinear solver. Besides, one more formula is achieved, which was previously found by other researchers, via a different mathematical process. Several benchmark structures, which have geometric nonlinear behavior, are analyzed with the proposed methods. The finite element method is utilized to analyze these problems. The abilities of suggested schemes are evaluated in tracing the complex equilibrium paths. Moreover, comparison study for the required number of increments and iterations is performed. Results reflect the robustness of the authors’ formulations.

FORM FINDING FOR RECTILINEAR ORTHOGONAL BUILDINGS THROUGH CHARGED SYSTEM SEARCH ALGORITHM

P. Sharafi

Preliminary layout design of buildings has a substantial effect on the ultimate design of structural components and accordingly influences the construction cost. Exploring structurally efficient forms and shapes during the conceptual design stage of a project can also facilitate the optimum integrated design of buildings. This paper presents an automated method of determining column layout design of rectilinear orthogonal building frames using Charged System Search (CSS) algorithm. The layout design problem is presented as a combinatorial optimization problem named multi-dimensional knapsack problem by setting some constraints to the problem, where the minimum cost and maximum plan regularity are the objectives. The efficiency and robustness of CSS to solve the combinatorial optimization problem are demonstrated through a numerical design problem. The results of the algorithm are compared to those of an ant colony algorithm in order to validate the solution.

THE PERIODIC GREEN VEHICLE ROUTING PROBLEM WITH CONSIDERING OF TIME-DEPENDENT URBAN TRAFFIC AND TIME WINDOWS

S. H. Mirmohammadi

The travel times among demand points are strongly influenced by traffic in a supply chain. Due to this fact, the service times for customers are variable. For this reason, service time is often changes over a time interval in a real environment. In this paper, a time-dependent periodic green vehicle routing problem (VRP) considering the time windows for serving the customers and multiple trip is developed with this assumption that urban traffic would disrupt timely services. The objective function of proposed problem is to minimize the total amount of carbon dioxide emissions produced by the vehicle, earliness and lateness penalties costs and costs of used vehicles. At first, a novel linear integer mathematical model is formulated and then the model is validated via solving some test problems by CPLEX solver. Finally, the sensitivity analysis is carried out to study the role of two critical parameters in the optimal solution.