Hirpa Lemu

This paper presents a method of determining the friction coefficient in metal forming using multilayer artificial neural networks based on experimental data obtained from strip drawing test. The number of input variables of the artificial neural network has been optimized using genetic algorithm. This process is based on surface parameters of the sheet and dies, sheet material parameters and clamping force as input parameters to train the neural network. In addition to demonstrating the fact that regression statistics model using genetic selection and intelligent problem solver are better than models without preprocessing of input data, the sensitivity analysis of the input variables has been conducted. This avoids the time-consuming testing of neurons in finding the best network architecture. The obtained results from this study have also pointed out that genetic algorithm can successfully be applied to optimize the training set and the outputs agree with experimental results. This...

ABSTRACT In part I of this study an optimum NURBS curve fitting by two evolutionary optimization techniques was successfully designed. These methods were implemented to optimize the location of a set of NURBS control points for the measured point cloud of four segments of a gas turbine compressor airfoil shape. The purpose of the optimization was to demonstrate the good ability of evolutionary techniques, in particular Genetic Algorithms, in optimizing such curve fitting problems. The objective of part II is to examine two alternative solutions for NURBS curve fitting of the same airfoil point cloud with swarm intelligence optimization technique. Indeed, the same work has been done by applying two basically different optimization approaches that is Particle Swarm Optimization and Invasive Weed Optimization. Results allow seeing a number of advantages as well as some disadvantages in this optimum curve fitting approach in comparison to the previous techniques applied by authors.

ABSTRACT This paper presents the study on deformability analysis of sheet metal forming by using finite element methods (FEM). The paper particularly focuses on application of dynamic explicit and static implicit approaches to simulate metal forming of rectangular cups where different material models and contact conditions with friction are involved. Comparative studies of the results on formability of the forming process using the two approaches are presented. The simulated results are verified using results from experimental study of the deformation on the same material. Further, a comparison between a quadratic Hill anisotropic yield criterion and von Mises yield criterion with isotropic hardening has been presented. The results confirm that the dynamic explicit procedure is more efficient in simulating sheet metal forming processes. The study shows also that the finite element analysis undoubtedly gives good approximate numerical results to real processes when the material and friction anisotropy are considered.

ABSTRACT In the classical product development process, design of a product is verified by testing physical prototypes. This approach falls short of fulfilling the demands for providing products at lower cost and short delivery time. This article illustrates the use of the digital data from CAD model in analysis and simulation of complex structures. Two industrial examples are shortly discussed to demonstrate the advantages of integrating analysis in the early phase of the product development process. Full Text at Springer, may require registration or fee