Assessment of experiment reproducibility by finite element analysis

Abstract eng:
Cylindrical metallic shell subjected to localized impact loading evoking wave response was considered. Experiments were carried out employing the Double pulse holographic interferometry (DPHI) for evaluation of displacements, the Doppler interferometry (DI) for velocities and the piezoelectric accelerometers (PA) for the registration of accelerations, while the finite element (FE) method was used for the numerical analysis. The paper compares experimental and numerical responses with the intention to discuss the accuracy limits of experimental and numerical approaches. Introduction The experimental investigation of high-speed phenomena accompanying the stress wave propagation in solids is a difficult task. Preparing and carrying out the experiment is expensive both in terms of required sophisticated hardware, material costs and manpower resources; furthermore the satisfactory reproducibility is hard to achieve in microsecond time scales. On the other hand employing the FE method for treatment of stress wave propagation in linear elastic continuum is a well reproducible routine, even if it is highly time and memory consuming. A lot of finite element results could be generated easily after the initial exertion is invested in proper design of the FE model. Still the detailed evaluation of finite element results requires additional both menial and mental effort. The former stems from the fact that the data sets needed for comparison with experimental data have to be mined out from extensive data files generated for every node at each time step. The latter emanates from an impossibility to distinguish precisely between the actual fast-frequency phenomena and spurious side effects due to time and space discretizations.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket