Print ISSN: 1681-6900

Online ISSN: 2412-0758

Author : A. Mohsen, Ali


An Experimental Approach and Constructing a New Non-Linear Regression Model for Prediction the Anisotropy Parameters of Annealing Treated Commercially Pure Aluminum Sheets

Jabbar H. Mohmmed; Ali A. Mohsen; Bassam A. Ahmed; Najmuldeen Yousif Mahmood

Engineering and Technology Journal, 2016, Volume 34, Issue 14, Pages 2775-2783

Earing is a common phenomenon in deep drawing process that increases the waste of metal. This phenomena is affected by material anisotropy, thus, it is important to study the effects of material parameters on this material behavior. This paper focuses on identify the optimal condition of annealing treatment which result in higher value of normal anisotropy and lower value of planar anisotropy which lead to reduce the waste material in subsequent forming processes. Therefore, in this study, anisotropic behavior and formability of commercially pure aluminum thin sheets was investigated after annealing the samples at different temperatures (350, 400, and 450) °C. Uniaxial tensile tests were carried out at room temperature (25°C) to evaluate formability parameters. For this purpose different tensile test samples in the directions of 0°, 45° and 90° in respect to the rolling direction were prepared. In addition to, metallographic test was carried out to as-received and annealed samples to observe the changes in microstructure.
Plastic strain ratio and planar anisotropy of samples were calculated from the tensile test data. Based on the tensile test results of samples, the earing phenomenon due to planar anisotropy in commercially pure aluminum sheet was analyzed. The results indicate that the annealing at 400°C brought the optimum conditions.
Moreover, new regressions model for prediction the anisotropy parameters of sheet metal using statistical techniques (SPSS software) were constructed in this work. The experimental data were compared to those predicting values. A comparison clearly indicates that there are good identification between measured and predicted values with multiple correlation coefficients R of 0.932 and accuracy of about 87 %. The results reveal that the proposed model is effective and reliable tool to obtain accurate prediction of the anisotropy behavior of metal sheets.