Modeling Criteria for FE-Simulation of Shock Transfer Processes in Concrete Structures

Shocks and high frequency excitations on building structures due to impact loads (impact of wreckage and heavy objects from collisions, transport operations or explosions), but especially due to a postulated aircraft crash, lead to feasibility problems and large expenditures at safety-related installations in the designing of technological systems accommodated inside the building structures. A rational and cost-effective qualification and verification of the operability of such systems requires reliable information about the nature of the induced excitations (structural responses) to be anticipated at their particular areas of installation in the impacted structure. The analytical derivation of realistic and reliable structural responses requires the application of suitable, robust mathematical models as well as a critical evaluation of all influencing factors on entire shock transmission path, from the area of impact to the site of installation of the affected plant component or system. The present paper addresses especially to the required FE mesh refinement adequate to short duration and impact type of loading. In order not to mix various influences and for better understanding of the (mesh refinement-depended) wave propagation phenomena in real concrete structures physical uncertainties, for instance uncertainties in material properties (modulus of elasticity, yield strength, etc) as a consequence of the randomness in nature, were for the present not taken into consideration. Intense scientific studies (related to static loadings) has been devoted in the past to these research directions in several studies and the results were published in a number of papers (Schueller [1], Wiberg [2], Charmpis and Schueller [3] ). However, it is known that research in the directions of FE mesh refinement and quantitative uncertainties modeling has been progressing in parallel and practically independent research routes have been followed. Both of aforementioned scientific areas: the effect of quantitative uncertainties modeling as well as the influence of mesh density on the results obtained using FE models were especially examined in [3]. Despite extensive studies and computational analyses of impact-induced shocks performed using FE models of variable refinement limited and insufficient experimental results to date have precluded complete investigation and clarification of several "peculiarities" of shock transmission phenomena in FE models. This refers mainly to the divergence of numerical results obtained using FE models (versus analytically derived