Determination of material/geometry of the section most adequate for a static loaded beam subjected to a combination of bending and torsion

Abstract

The acceleration of industrial machines mobile parts has been increasing over the last few years, due to the need of higher production in a short period of time. The machines were dimensioned for a lower value of acceleration, which means there is not enough rigidity for the correct operation at much higher accelerations. Nowadays, the accelerations can be near 12 times the acceleration of gravity. There is the need of improving rigidity to make possible the correct machine operation without undesired vibrations that can ultimately lead to failure. The main applications of this work are plotters and laser cutting machines. To improve rigidity, one must improve the relevant material properties, and the relevant geometric variables of the model.[1] A novel Finite Element Model Updating methodology is presented in this paper. The considered models were : a ribbed plate and a tubular beam. The models were built by means of the Finite Element Method (FEM), and MATLAB was used to control the optimization process, using a programming code. Both material properties and geometric parameters were optimized. The main aim of the materials modeling is to know how the value of the objective function changes with the value of the material properties. Materials selection was performed, using material selection charts, to select the best material for the application. The value of these properties was not in the catalogue, and the properties used to perform the material selection were related to a material sub-class, Eg. Steel. The final material selection determined the best specific material for the application, and that material was mechanically tested. The mechanical tests performed were: Tensile Test and Extensometry Test, to obtain the relevant material properties, mainly Young Modulus, Poisson Coefficient and Yield Stress. The deflection of the optimized models reduced strongly in comparison to the initial models.