Study of similarity with distorted configurations of geometry via numerical models

Abstract

Due to technical restrictions, the construction of a scaled structure (model) with all required dimensions can pose challenge in certain experimental impact tests. Consequently, one of the dimensions of the structure may be distorted, infringing the prerequisite conditions for the $\Pi$ theorem, and rendering similarity unattainable. In the present work, a method that changes the striker initial velocity in order to compensate for geometric distortion is analysed. The factors calculated in this technique allow estimating the real-size structure (prototype) by using scaled models. Initially, a structure comprising two plates clamped together and subjected to impact by a mass is employed to illustrate the method. Subsequently, two models with distorted thickness are tested through numerical simulations: a clamped tube struck by a rigid mass; a buffer bow colliding with a rigid wall. A comprehensive analysis of the results and the limitations of the method is done to ascertain the sources of errors and determine which structures types can benefit from this technique.