EHTC 2009

Drop Test Behavior of a Low Radioactive Waste Package in Concrete

Sylvain Tholance, ATR Ingénierie

Background:
In the context of achieving a deep storage of long life radioactive waste ANDRA is studying different concepts of packages to store LLMA (Long Life Medium Activity) waste and C ones (High Activity) and for spent fuel. Studies on the handling of LLMA storage packages showed that the risk of falling exists, especially at the time of placing the package into the storage cell. The overall objective of the studies is to assess the ability of the LLMA storage package to keep the containment of nuclear waste in case of fall. All dynamic simulations are performed with RADIOSS solver, product of Hyperworks Tools.

The LLMA package is made of fiber-reinforced concrete. Their size and weight vary according to the dangerousness of the inner substance. The B2.1 package has a generally cubic shape (1.5 m square) for a total weight of approximately 6-7 tons. The body is composed of 4 cells where wastes are placed. A precast concrete cover is anchored to the body by 5 bolts. Two concepts of the body are assessed: a standard model using reinforcement and low density steel fiber and an alternative one unarmed but with a greater fiber density.

Simulation and Testing:
The first stage of simulations is to characterize the behaviour of fiber reinforced concrete and to identify the parameters of concrete law n°24 of RADIOSS from a confrontation testing - calculations on specimens: tests of simple compression, simple tension, bending 4 points, etc.. The concrete laws are then injected into the complete FE models of B2.1 package. Drop tests at scale 1 were carried out to validate the concepts and qualifying simulations.

Results:
The 2 concepts B2.1 packages have successfully completed the drop test on the corner of 6 meters high. The wastes placed inside are weakly damaged and the behaviour of the concrete storage package is satisfactory: the closing of the package is maintained and cracking generally does not affect the handling ability after fall. For the 2 concepts, the lid is kept on the body through the anchoring system. The extent of cracks and openings are particularly more pronounced on the alternative concept unarmed. The safety margin seems less important.
The simulations provide results close to reality in terms of overall behaviour in terms of crush at the corner, cracking and general solicitation on the waste. It remains to better control the dispersion caused by the manufacturing of the package body and of fiber concrete itself on fall resistance of packages to reach more robust design.

Perpectives:
Subsequent studies will be done with stochastic analysis (Design of experiments, Optimization, Robustness) implementing the abilities of HyperStudy (DSS).

 

<< Back to Agenda