GUIDE FOR NUMERICAL MODELLING OF DROP TESTING OF PACKAGES FOR THE TRANSPORT OF RADIOACTIVE MATERIALS

To demonstrate compliance of package designs to applicable regulatory requirements [1], the good behaviour of the package under drop test conditions has to be justified. In order to do this, more and more analyses based on numerical calculations are performed to complete or replace drop tests with package specimens. These analyses are commonly used to: identify the worst drop conditions: angle, impact area, material properties taking into account temperature effects in the complete range of design temperatures, order of the sequence of drops (puncture drop before or after free drop test), justify the similarity between the specimens used for the drop tests and the worst package configuration, define stress fields for brittle fracture analysis, study the effects of the potential added features to the package at the time of transport (frame for example). But such numerical models should be established with care, to ensure they do not lead to incorrect results. In this context, the French Institute for Radiation protection and Nuclear Safety (IRSN) has prepared a guide on transient dynamic calculation providing relevant modelling options and associated validation criteria. This guide has been built, using both the feedback experiments from the results of a large number of numerical calculations using LS-DYNA® calculation code [2] and purpose built benchmarks to determine about the most suitable options in case of doubt. The main parts of this guide concern: the general principles for modelling and the consequences of numerical choices on the results (explicit or implicit calculation, simple or double precision, numbers of processors, symmetric or parallel processing, scale factor, speed), the specific options relative to meshing and elements formulation, material library, behaviour laws, contacts algorithms and bolt calculations, the quality results evaluation (Hourglass and penalty energies, sampling and filtering results). The results show the high sensitivity of calculation options, in particular for the shock absorber (wood, foam) elements formulation, the method used for applying bolt initial stress, the size and the homogeneity of the meshing, the contacts and gaps management. The use of appropriate options has been proven to provide more accurate results with a finer depiction of reality.