For final storage of spent nuclear fuel it is suggested by the Swedish nuclear fuel and waste management company (SKB) to emplace the nuclear fuel into copper canisters which are surrounded by bentonite clay at approximately 500 meters’ depth into granitic rock. After emplacement of the canisters Bentonite swelling due to water saturation and hydrostatic pressure build up the canisters will be subjected to compressive loads. The canisters are constructed with a load carrying cast insert of ductile iron which is surrounded by a 50 mm thick corrosion resistant copper shell. The copper shell is not in itself load carrying but must retain its corrosion barrier ability (i.e thickness) when the compressive load is applied very slowly. For materials where the load is applied slowly, materials creep properties become of vital importance. It has been shown that creep ductility for oxygen free copper can be very low (< 1 %). SKB has for this reason proposed to use oxygen free copper alloyed with small amounts of phosph ous (30<P<100 ppm) in order to ensure that creep ductility is higher than 15 %. In order to verify the copper materials creep ductility, extensive creep testing has been performed by SKB. However, the challenge is to extrapolate test results lasting in a yearly scale into time scales lasting for several 10000 years, which is necessary to assess the repository long-term safety. Instead of performing creep tests, an alternative verification approach has been suggested to instead determine the minimum creep ductility needed to maintain sufficient safety margins for the canisters copper shell.
The objective of this work is to study an alternative approach to determine the minimum creep ductility needed to ensure the thickness of the canisters corrosion barrier.
In this report, two different kinds of finite element analysis have been conducted simulating copper canister deformation during buffer saturation and hydrostatic pressure build up. Instead of using a creep model for material behaviour, an alternative approach using an elastic-plastic material model has been applied in this work.
Results from the finite element analysis of strain levels in the area between the copper lid and copper tube show similar strain behaviour as earlier reported by SKB, using a time dependent material creep model. This result was expected since the strain level in this area primarily is deformation controlled in this load case. However, one benefit of using the alternative approach in this work is to study how different geometric parameters as well as how other material properties like friction, change the strain levels. The alternative approach can thus be one tool in optimizing the canister geometry in order to minimize copper strain. In the second part of the report the aim was to study the containment ability of canisters copper shell by means of creep brittleness. The safety function of the copper shell is to provide an approximately 50 mm thick corrosion barrier towards the oxygen free saline groundwater surrounding the canister after emplacement. Creep brittleness of the copper shell can potentially decrease the corrosion barrier by formation of creep cracks. The influence of creep brittleness was in this investigation studied by application of a damage mechanics approach based on defining a criterion for maximum allowable plastic strain. When the criterion is reached damage is initiated and the load bearing capacity is reduced to zero. The analysis involved a number of material parameters which were not known for the copper material in question. For this reason, some of the unknown material parameters were varied in order to investigate the method and get an idea of how the copper canister behaves under different assumptions. More knowledge about the actual material behaviour would be needed to be able to better evaluate the results. With these short-comings in mind, results from the damage mechanics analysis indicates that the minimum creep ductility for the copper material used for the canister should be in the order of 10% to withstand the pressure load. One attempt to use the damage mechanics approach for a load controlled case (internal pressure) was conducted, with limited success. It is suggested that a fracture mechanics approach might be more appropriate for analyzing such a load case. In summary, this study has shown that damage mechanics analyses are sensitive to several material parameters which are a necessary input in the analysis. More knowledge regarding the material properties are needed before accurate predictions using this method can be made. However, based on the damage evolution shown in this report for creep brittle copper, it can be concluded that creep brittleness of copper can potentially induce concentrated damage in certain directions meaning that the corrosion barrier of the copper shell can be reduced.