Filtern
Erscheinungsjahr
Dokumenttyp
- Beitrag zu einem Tagungsband (33)
- Vortrag (15)
- Zeitschriftenartikel (13)
- Beitrag zu einem Sammelband (7)
- Posterpräsentation (1)
- Forschungsbericht (1)
Schlagworte
- Simulation (8)
- Drop test (7)
- Fracture mechanics (7)
- Finite element method (6)
- Safety assessment (6)
- Container design (5)
- Damping concrete (5)
- Ductile iron (5)
- Duktiles Gusseisen (5)
- Dynamic loading conditions (5)
Organisationseinheit der BAM
Eingeladener Vortrag
- nein (15)
Cylindrical casks made of ductile cast iron are used for transport, interim storage and final disposal of radioactive waste in Germany. A basic design criterion is the ability to withstand a horizontal drop without impact Binders onto a foundation representative for the real ground of a storage facility. The increasing use of more cost-effective material compositions requires optimized cask geometries to reduce stresses and strains in the cask structure. For example, a non-uniform wall thickness was introduced to reduce maximum wall bending stresses. As a result, the load characteristics changed from line load to point load at bottom and lid side under horizontal drop fest conditions. Hereby, the position of highest stress inside the structure has shifted significantly. This was the reason for a systematic investigation of effects caused by small design changes or small variations of fest conditions. It led to a better understanding of the positions of high local stresses, their time history and maximum value for cylindrical casks under horizontal drop test conditions. Hence, the paper presents the lessons learnt from modeling and simulating such scenarios considering an impact without limiters onto a realistic target.
The safety of transport packages may be demonstrated by numerical calculation of load scenarios defined in the IAEA regulations. Possible handling accidents of casks at interim storage sites or in a final repository are typically analyzed by dynamic finite element computations. In each case the investigated load scenario must be transferred into a mathematical model. Secondly the mathematical model must be transferred into a numerical model. Reliable finite element models should be developed by assembling verified sub-models of components. The finite element mesh, material modeling, initial and boundary conditions, contact definitions, and time integration as well as the benefit of pre- and post-calculations are discussed. The paper presents lessons learnt from modeling dynamic test scenarios for finite element analyses over the years.
Finite element analysis (FEA) has been carried out for investigation of damping concrete under different impact loading conditions with a built-in material model and damage criteria available in FEA code ABAQUS.
At first, all parameters for the selected material model had been derived from compression Tests of cubic specimens. After that, a validation was carried out with different static and dynamic penetration tests. Finally, a 5 meter real drop test with a 23 Mg cylindrical cask could successfully be simulated.
Validation of Numerical Simulation Models for Transport and Storage Casks using Drop Test Results
(2012)
Validation of numerical simulation models for transport and storage casks using drop test results
(2012)
The safety assessment of new designs for transport and storage casks for radioactive materials is a challenging task accomplished using different methods such as prototype tests, model tests, calculations and analogy reflections. At BAM (Federal Institute for Materials Research and Testing), the test procedures for the mechanical IAEA (International Atomic Energy Agency) test conditions often start with preliminary finite element (FE) calculations mostly with a small-scale cask model for verification of the proposed test cask instrumentation and test plan. On that basis the extensive test cask instrumentation is applied and checked. After that, a series of drop tests consisting of different test sequences is performed.
Following the drop tests, numerical post-analyses are carried out. These analyses offer the possibility of a detailed calculation and assessment of stresses and strains in the entire test cask construction. The calculation results have to be carefully compared with the measurement data over the impact history to find out all relevant parameters for a realistic simulation of the impact scenario. The desired ideal boundary test conditions often cannot be met exactly during the drop tests. Therefore, the numerical post-analyses are carried out by using the real boundary conditions of the drop tests. The objective is to find a validated model, where the results of the numerical simulations satisfactorily meet the experimental results.
Under test conditions according to the IAEA transport regulations, casks are usually equipped with impact limiters and dropped onto a so-called unyielding target. In general, it is difficult to verify a complex FE model by using results from only one drop test because of the complex impact process and the complex structure of such packages. After each drop test, numerical post-analyses should be carried out. Only if all drop tests were simulated successfully by using the same FE model under different test conditions, it is possible to obtain a validated numerical model for further investigations. In this case the results of the numerical simulations meet satisfactorily the experimental results. In this paper a study is presented, where the influence of different components on the cask loading is investigated systematically.
This paper demonstrates exemplarily how numerical and experimental approaches can be combined reasonably in mechanical assessment of package integrity according to the IAEA regulations. The paper also concentrates on the question about how static mechanical approaches can be applied, and what their problems are in relation to dynamic calculation approaches. Under defined impact tests, which represent accident transport conditions, the package has to withstand impact loading, e.g. resulting from a 9 m free drop onto an unyielding target in sequence with a 1 m puncture drop test. Owing to the local character of the interaction between the puncture bar and the cask body, it is possible to develop a dynamic numerical model for the 1 m puncture drop which allows an appropriate simulation of the interaction area. Results from existing experimental drop tests with prototype or small scale cask models can be used for verification and validation of applied analysis codes and models. The link between analysis and experimental drop testing is described exemplarily by considering a regulatory 1 m puncture bar drop test onto the cask body of a recently approved German high level waste transport package. For the 9 m drop test of the package, it is difficult to develop a dynamic numerical model of the package due to the complexity of the interaction between cask body, impact limiters and unyielding target. Dynamic calculations require an extensive verification with experimental results. The simulation of a 9 m drop of a package with impact limiters is thereby often more complex than the simulation of a 1 m puncture drop onto the cask body. A different approximation method can be applied for the consideration of dynamic effects on the impact loading of the package. In a first step, maximum impact force and rigid body deceleration of the cask body during the impact process can be calculated with simplified numerical tools. This rigid body deceleration can subsequently be applied on a verified static numerical model. Dynamic effects, which cannot be covered by the static numerical analysis, have therefore to be considered by using an additional dynamic factor. The paper describes this approach exemplarily for a 9 m horizontal drop of a typical spent fuel cask design.
The 9 meter drop onto an unyielding target is one of the important mechanical tests within the safety assessment of transport casks for radioactive material. In general, the cask is equipped with impact limiters to reduce the dynamic load on the cask body by absorbing a major part of the kinetic energy. The impact limiters are often made of wood or aluminium. In this study an elastic-plastic material model with volume change was used to describe the stress-strain behaviour of wood found in crush tests. For aluminium, an elastic-incremental plastic material model with Cowper-Symonds parameters for strain rate depending material hardening was used to model the adiabatic stress-strain relations measured at specimens at constant ambient temperature. Hereafter simulations with a sophisticated finite element model were carried out and compared with different drop tests. Four drop tests of a half-scale cask model equipped with wood and aluminium impact limiters with different drop positions were selected to investigate the impact limiter behaviour during a 9 meter drop test. All drop tests were simulated with the same FE mesh but under different boundary and initial conditions.