Hamide Tekeli

Comparison of critical column buckling load in regression, fuzzy logic and ANN based estimations

In structural stability analyses, determining the critical buckling load is a crucial issue. Regression, fuzzy logic and Artificial Neural Network ANN algorithms can be used to determine critical buckling loads. This study compares the results of different approaches for column buckling load prediction. Regression, Fuzzy logic and ANN algorithms were employed in the analyses, representing material properties to take uncertainties into account and the results were compared. The results show that uncertainties play an important role in stability analyses and should be considered in the design. The elastic modulus results predicted by regression, fuzzy logic and ANN are also compared to those obtained using empirical results of the buildings codes and various models. These comparisons show that obtained results in the present study give closer results than the different design codes. Therefore, proposed models can be used for critical buckling loads and regression, fuzzy logic and ANN have strong potential as a feasible tool for estimating column buckling loads within the range of input parameters considered.

Determination of elastic moduli effects on storey-drifts by fuzzy logic algorithm

Concrete compression strength is a standard input for obtaining the elastic modulus of concrete. Calculated values of elastic modulus for the same concrete strength may vary according to different codes throughout the world since the formulas of the elastic modulus are different. Elastic modulus is one of the important parameters for structural design. Storey-drifts are also among the important structural design criteria directly related to the elastic modulus. Different elastic moduli values may lead lateral storey-drift calculations to differ from the code requirements. More dependable elastic moduli definition results in more dependable storey-drift calculations. Fuzzy logic algorithm, which generates more dependable results in engineering problems, is applicable to solve this problem. This study investigates the effects of storey-drifts on structural design; upper and lower boundaries of elastic moduli are predicted using a fuzzy logic algorithm in order to determine the effects of the elastic moduli on the storey-drift limitations found in different design codes. The study considers various design codes that are obtained worldwide. Study results also show that using a range of values for the elastic moduli is a more dependable approach for structural design safety, and that dependable storey-drifts are crucially important for structural design.

An approximation method for design applications related to sway in RC framed buildings

In this paper, an approximate method is proposed for determining sway of multistory RC buildings subjected to various types of lateral loads. The calculation of both fundamental period and stability index in RC building requires the sway term at each story level. Using approximate method design engineers can estimate sway terms at each story level. The developed analytical expressions are inserted into fundamental period and stability index equations to replace the sway terms, which yields modified equations for fundamental period and stability index without any sway terms. It is fairly easier to employ these equations developed by eliminating all sway terms. Results obtained from the equations are remarkably close to those generated by the related computer program. Consequently, design engineers can reliably use the simple equations to calculate stability index and fundamental period, which enables the determination of these parameters without referring to the complex sway terms. The capability and accuracy of the proposed equations are demonstrated by a numerical example in which computer program results are compared with the proposed methodology.