Application of an incomplete similarity approach in the analysis of impact problems

Resumo

The assessment of safety to the impact of motor vehicles requires costly experimental tests, such as
crash-tests and advanced methods of structural analysis. The design of bus structures in Brazil, for example, does
not have adequate calculations methods and regulatory standards, and safety is difficult to assess. Therefore, a
methodology is proposed for conducting impact tests on small-scale models that have incomplete similarity, in
order to facilitate access to experimental data about vehicle crashworthiness. The methodology is applied to the
simplified problem of a drop test on a fully clamped tubular steel beam. The test was numerically simulated using
the Abaqus® software, from which it was possible to reproduce results from prototypes and models. Through a
methodology presented in the literature, two models were used to correct the incomplete similarity characterized
by the parameter regarding the beam thickness. This approach demonstrated a good correlation with the numerical
results in the range of the dimensionless group in which the models were made, and started to diverge from the
numerical results as it moved away from that range. A second approach, with a second non-similar parameter, was
used for comparison with experimental data, which demonstrated a good correlation.