Dynamic Fracture Mechanics Assessments for Cubic Ductile Cast Iron Containers

Abstract

An improved BAM safety assessment concept for the cask material ductile cast iron (DCI) to cover higher stresses in the cask body, highly dynamic load scenarios, and a broader range of material qualities will require more extensive fracture mechanics analyses based on a combination of material testing, calculation of applied stresses, and inspection standards. As an example, the brittle fracture mechanics assessment of a surface crack in a plate due to the dynamic load from the 5 m drop of a cubic container (not equipped with impact limiters) onto a reinforced concrete target is investigated. The numerically calculated time-dependent stress intensity factor is compared with a previous static solution with the same loading history inserted. For the scenario studied the differences between the curves are negligible because a dynamic load of the cask within a time scale of millisec9nds can be considered as a quasi static load for the crack.     

A simple sizing optimization technique for an impact limiter based on dynamic material properties

According to IAEA regulations, a transportation package for radioactive material should perform its intended function of containing the radioactive contents after a drop test, which is one of the hypothetical accident conditions. Impact limiters attached to a transport cask absorb most of the impact energy. So, it is important to determine the shape, size and material of impact limiters properly. The material data needed in this determination is a dynamic one. In this study, several materials considered as those of impact limiters were tested by drop weight equipment to acquire the dynamic material characteristics data. The impact absorbing volume of the impact limiter was derived mathematically for each drop condition. A size optimization of the impact limiter was conducted. The derived impact absorbing volumes were applied as constraints. These volumes should be less than the critical volumes generated based on the dynamic material characteristics. The derived procedure to decide the shape of the impact limiter can be useful at the preliminary design stage when the transportation package's outline is roughly determined and applied as an input value.     

Topical BAM cask design evaluation using drop tests and numerical calculations: accidental cask drop without impact limiters onto storage building foundation

Abstract

Cask impacts without impact limiters onto unyielding targets result in totally different mechanical reactions from those of relatively smooth impacts using impact limiters. During the licensing procedure of the new GNS CASTOR HAW 28M design for vitrified high activity waste, BAM therefore decided to perform an additional drop test with a 1 : 2 scale test cask (CASTOR HAW/TB2). In spite of a small drop height of only 0˙3 m onto the unyielding target of the BAM drop test facility, which conservatively covers any storage building foundation, the impact caused considerable stresses to the cask structure with high stress and strain rates. This paper presents the evaluation strategy of BAM including the drop test results and the development and qualification of appropriate finite element modelling to achieve sufficient agreement between test and calculation results. Further steps include mechanical analyses of reduced and full scale cask designs to determine the most critically stressed areas of the structure, verify scaling factors and demonstrate safety with respect to cask integrity and tightness.     

Dynamic impact characteristics of KN-18 SNF transport cask – Part 1: An advanced numerical simulation and validation technique

Domestic and international regulations for the transportation of radioactive materials strictly prescribe the design requirements for spent nuclear fuel (SNF) transport casks. According to the applicable codes, a transport cask must withstand a free-drop impact of 9 m onto an unyielding surface and a free-drop impact of 1 m onto a mild steel bar. However, the structural performance of a transport cask is not easy to evaluate precisely because the dynamic impact characteristics of the cask, which includes impact limiters to absorb the impact energy, are so complex.     

Numerical Simulation of Dynamic Deformation Of Spent Fuel Transport Packages for Nuclear Power Plants in Accidental Mechanical Impacts

Abstract

Numerical simulation results and their analysis are presented for dynamic deformations of the TUK-117 package, intended for the transport of spent nuclear fuel from nuclear power plants, subjected to accidental 9 m drops onto an unyielding surface at different angles. This paper focuses on the analysis of the deformation behaviour of container shock limiters. It is demonstrated that maximum loading affects the package during a side drop. For a side drop, the maximum strain levels are determined for the different construction elements, including the cask's body and the bolts securing the sealing lid. Dynamic simulation of the behaviour of the construction elements was carried out using the LS-DYNA code, version 970.     

Experimental testing of impact limiters for RAM packages under drop test conditions

In context with new cask designs and their approval procedure, the experimental testing of impact limiters under drop test conditions becomes more and more important in order to assess the damage mechanics behaviour and safety margins for validation reasons. In recent years, various designs of impact limiters have been tested by the Federal Institute for Materials Research and Testing within specific component testing and particularly with regard to type B package design approval procedures. The paper focuses on the experimental realisation of impact limiter tests and presents implemented measurement techniques to determine the amount of deformation and to explain the impact behaviour by means of photogrammetric metrology and three-dimensional fringe projection method, high speed motion analysis and adjusted deceleration measurements.     

Response of a concrete container at different drop orientations

Impact analyses of a nuclear waste dry storage container to simulate an accidental drop onto an unyielding surface have been carried out for three different drop orientations; namely (1) C.G. directly above the initial impact point, (2) 5° tilted vertical drop and (3) 5° tilted horizontal corner edge drop. The finite element technique is employed using MSC/PATRAN for pre- and post-processing while the analyses are carried out using the hydrodynamic code DYNA3D. Responses in terms of energy, momentum, contact force, stresses, strains and deformation during and after impact are closely examined and compared. It is found that, with impact limiters for protection, the container can survive the drop impact; however, local responses are different at different drop orientations.     

Safety Analysis of Casks Under Extreme Impact Conditions

Abstract

The determination of the inherent safety of casks under extreme impact conditions has been of increasing interest since the terrorist attacks of 11 September 2001. For nearly three decades BAM has been investigating cask safety under severe accident conditionslike drop tests from more than 9 m onto different targets and without impact limiters as well as artificially damaged prototype casks. One of the most critical scenarios for a cask is the centric impact of a dynamic load onto the lid-seal system. This can be caused, for example, by a direct aircraft crash (or just its engine) as well as by an impact due to thecollapse of a building, e.g. a nuclear facility storage hall. In this context BAM is developing methods to calculate the deformation of cask components and — with respect to leak-tightness — relative displacements between the metallic seals and their counterparts. This paper presents reflections on modelling of cask structures for finite-element analyses and discusses calculated results of stresses and deformations. Another important aspect is the behaviour of a cask under a lateral impact by aircraft or fragments of a building. Examples of the kinetic reaction (cask acceleration due to the fragments, subsequent contact with neighbouring structures like the ground, buildings or casks) are shown and discussed in correlation to cask stresses which are to be expected.     

Characterising polyurethane foam as impact absorber in transport packages

Abstract

Transport and storage packages used for the safe transport of radioactive materials are required to satisfy IAEA regulations. One key design requirement for a radioactive material transport package is that under a 9 m regulatory drop test, containment functions are maintained. For certain payload types, such as fuel assemblies, impact loads on the payloads may need to be controlled in order to maintain spacing and confinement. To achieve all of this, detailed and accurate characterisation of the impact absorbing material is important in order to design an effective shock absorber. Polyurethane foam is an excellent energy absorbing material because it has a relatively high specific strength, a large compressive deformation, much of this at constant force, and a predictable compressive strength characteristic. Traditionally various types of wood have been used for this purpose, however foams are a more cost effective alternative, which are readily available, and can be formed and shaped easily. Some grades may have the added advantage of providing an almost isotropic crush response, combined with significant thermal protection. The general compressive strength properties of foams and their temperature dependencies are well documented by manufacturers; however, strain rate sensitivity and stiffness variation with orientation are not readily available. Hence impact compression tests for polyurethane foams for a range of densities from 56 to 320 kg m^–3 were specified by Rolls-Royce and performed by the Health and Safety Laboratory. These tests included dynamic conditions for a range of strain rates and temperatures and a selection of orientations of the foam. Following collation of the test results, property curves were derived for the range of temperatures at which the package was expected to operate in service between –10 and +75°C. The properties for a given specification of foam will vary within a defined tolerance range, mainly due to the variables inherent during manufacture. Hence nominal static curves were derived for each foam and a number of factors were taken into account to derive the full range of foam properties: density, compressive strength, temperature and manufacturer supplied tolerance. The net result of this work was a series of force displacement plots, depicting upper and lower bounds to account for the cumulative effects of many variables. Accounting for these upper and lower performance bounds is an essential approach in justification of any modern package design. This paper describes the characterisation and mathematical modelling of polyurethane foam for use as the main impact energy absorber in a new design of package for transporting fresh fuel. The non-linear finite element (FE) code LS-DYNA was used to carry out simulation of the tests. The HONEYCOMB material model available in LS-DYNA was used to accurately predict the test measurements of the foam material. The properties derived for the foam were then used as input to the full FE model used for the licensing of the new package design. Full scale drop testing of the package demonstrated good correlation of deformations between test and FE model analysis, providing good validation evidence of the foam characterisation in the transport package.     

Plastic impact testing of shipping cask fin specimens

An impact test program was conducted to determine the energy absorption characteristics of typical heat transfer or structural fins on shipping casks as these fins undergo dynamic, gross plastic deformation. It was assumed that the energy absorption characteristics of fins welded parallel to the longitudinal axis of a cylindrical cask are linear with respect to their length, and the test program was designed to determine the energy absorption capability per inch of fin length along the cask body. Specimens of typical fin geometries were impacted with free-falling weight dropped from known heights, and the force-time relationships occurring during each impact were recorded. A computer program was used to convert the test data into deformation-time and energy-time relationships. These relationships were correlated in a set of curves that enable the cask designer or analyst to compute the energy absorption capacity of a longitudinal fin on a cask as a function of its geometry and the percentage of deformation it is expected to experience as well as the peak force the fin would experience during the impact resulting from a 30-ft free drop into an essentially unyielding horizontal surface.     

Vulnerabilities of legacy fuel casks when evaluated using modern structural analysis tools

Abstract

Cylindrical fuel casks often have impact limiters surrounding the ends of the cask shaft in a typical 'dumbbell' arrangement. The primary purpose of these impact limiters is to absorb energy to reduce loads on the cask structure during impacts associated with a severe accident. Impact limiters are also credited in many packages with protecting closure seals and reducing peak temperatures during fire events. For this credit to be taken in safety analyses, the impact limiter attachment system must be shown to retain the impact limiter following normal conditions of transport (NCT) and hypothetical accident conditions (HAC) impacts. Large casks are often certified by analysis only because of the cost associated with testing. Therefore, some cask impact limiter attachment systems have not been tested in real impacts. A recent structural analysis of the T-3 spent fuel containment cask found problems with the design of the impact limiter attachment system. Assumptions in the original safety analysis for packaging (SARP) concerning the loading in the attachment bolts were found to be inaccurate in certain drop orientations. This paper documents the lessons learned and their applicability to impact limiter attachment system designs.     

Numerical Assessment of Spent Fuel Casks Impacting on Real Targets

Abstract

A reference container of high capacity was analysed for loads beyond those it has to withstand during a 9 m IAEA drop test onto an unyielding target. In doing this a lid-end drop with shock absorber onto a real target was simulated. This is a possible accident for the rail transport of such casks. In this case the most critical components of the containment system are the primary lid bolts. The behaviour of the lid system and its sealing function were investigated with finite element (FE) analysis. To correlate the findings with a corresponding impact velocity onto real targets an analytical method was used. Despite the conservative assumptions made in this study a two-fold safety factor compared to the 9 m drop tests onto the unyielding target could be shown. The quantification of the additional safety the cask might provide requires further basic investigations on the behaviour of the real targets considered as well as the reduction of the conservatism included in the assumptions made up to now.     

Japanese perspectives and research on packaging,transport and storage of spent fuel

Abstract

The present paper gives an overview of Japanese experimental studies of dual-purpose metal casks. The studies included: cask drop without impact limiters, drop of a heavy weight onto a cask due to building collapse, burial of a cask in debris from building collapse, tipping over of a cask during an earthquake, long-term containment of metal gaskets and transportability of casks after long-term storage. Most of the studies employed full-scale casks for the experiments.     

Computer modelling of wood-filled impact limiters

The performance of impact limiters, composed of balsa wood encased in steel shells, has been studied by experimental tests and computer modelling. The aim of the work was to develop a validated capability for simulating the impact behaviour with the code DYNA3D. The code is thus available to support new package designs employing similar shock absorbing features—in principle, any crushable material encased in a metallic shell.

A brief review is given of the mechanical properties of balsa wood, including methods for modelling orthotropic crushable materials. With DYNA3D, the material model for metallic honeycombs was adapted for application to wood. Both static crush tests and impact tests have been simulated, with good results for deformation and energy absorption. Dynamic strength enhancement and friction effects are examined.     

Spent fuel transportation risk assessment: cask impact analyses

Abstract

The Nuclear Regulatory Commission (NRC) has recently completed an updated Spent Fuel Transportation Risk Assessment, NUREG-2125. This assessment considered the response of three certified casks to a range of impact accidents in order to determine whether or not they would lose their ability to contain the spent fuel or maintain effective shielding. The casks consisted of a lead shielded rail cask that can be transported either with or without an inner welded canister, an all-steel rail cask that is transported with an inner welded canister, and a DU shielded truck cask that is transported with directly loaded fuel. Finite element analyses were performed for impacts at speeds of 48, 97, 145 and 193 kilometres per hour into a rigid target. Impacts in end-on, side-on, and CG-over-corner orientations were analysed for each cask and impact speed. Calculations were performed to equate these impacts onto rigid targets with higher speed impacts onto the yielding targets that exist in the real world. These analyses indicated that a cask with an inner welded canister or a truck cask would not release radioactive material in any impact accident and that only very high-speed impacts onto hard rock targets could result in either release of material or significant degradation of shielding for rail casks without an inner canister. Impacts other than those onto flat unyielding targets were also considered. Analyses show that an impact that bypasses the impact limiters on the ends of the casks does not result in seal failure and neither does an impact by a locomotive also between the impact limiters.     

Protecting against corner impacts: sensitivities discovered during a rail cask impact limiter design

Type B packages for the transportation of radioactive materials must remain ‘essentially leak tight’ under severe regulatory accident conditions, defined in the US Nuclear Regulatory Commission’s 10 CFR 71·73 and the International Atomic Energy Agency’s TS-R-1. The 9-m free drop test requirement onto an unyielding surface is performed in an orientation ‘for which maximum damage is expected’. Analytical techniques are used to evaluate various possible impact orientations before testing, and historically these maximal damage orientations have been side, slap-down, end, and centre-of-gravity over corner. Other orientations are rarely considered. Sandia National Laboratories (SNL) was asked by Equipos Nucleares SA (ENSA) to design, analyse, and test an impact limiter system for a newly designed rail cask. During the conceptual design process, SNL performed due diligence and evaluated a wide spectrum of possible impact orientations, in order to assure that peak cask body acceleration design goals were not exceeded. However, design of the impact limiter, including not only crush strength of constituent materials (which can be orientation and temperature dependent), but also the shape of the impact limiter, greatly affects peak acceleration response during 9-m drops in various orientations. Although many impact limiter design shapes resemble truncated right circular cylinders attached to each end of the cask, some tend to round the outer corners or truncate those corners with conical sections. SNL’s original conceptual design followed a similar theme, intending to use polyurethane foam or aluminium honeycomb within a bevelled corner shaped cylindrical shell. Detailed finite element analyses indicated excellent impact resistance at regulatory cold temperatures in the stereotypically tested side, slap-down, end, and CGOC impact orientations. Shortly before proceeding to engineering design, a rarely-considered impact orientation of 45° from horizontal indicated that cask body acceleration levels jumped unexpectedly, exceeding the design goal due to insufficient crushable material protecting the sharp corner of the cask. A complete re-design of the impact limiter was necessary, and the lessons learned from this experience could have implications for future impact limiter designs, and possibly existing designs that may not have considered this atypical impact orientation during the design process.     

Behaviour of a Package Dropped Onto Yielding and Unyielding Targets

Abstract

IAEA transport regulations require the 9 m drop test for type B packages. The target used in this drop test must be unyielding. However, the real target that a transport package might encounter at an accident during transport is a yielding target, such as concrete, asphalt, or soil. To compare the impact acceleration between a real target and an unyielding target, analyses of the drop test onto a real target and an unyielding target were performed by using a LS-DYNA-3D computer code. A road surface is usually concrete or asphalt, and such surfaces are very hard; but there are soil, sand, etc. under these materials, so the impact caused by a drop accident is relatively small. It becomes clear that the drop height onto a road, corresponding to the 9 m drop height required by IAEA regulations is about 50 m, and the impact velocity is about 110 km.s^?1     

Benchmark calculations for mechanical stresses upon a transport cask

In this paper, benchmark calculations are carried out using the FEM code ADINA. Three drop tests with a DCI cask carried out in Japan are investigated. The drop orientations comprise the 9 m drop with impact flat onto the cylindrical shell of the shock absorbers at the top and bottom part as well as the 1 m drop with impact of the bar in the center of the cylinder wall. Calculated results are compared to the experimental ones. It is found that the calculations meet the measurements conservatively.     

Moisture content of wood: influence on mechanical behaviour of wood filled impact limiters and importance for quality surveillance during manufacturing

Abstract

The moisture content of wood is known to have a significant influence on the wood’s mechanical properties. Using wood as an energy absorber in impact limiters of packages for the transport of radioactive material, it is of particular importance to ensure the moisture content and thus relevant mechanical properties to be in specified limits. The paper surveys the influence of wood moisture content on the mechanical properties of wood. Different measuring methods are discussed with respect to in situ applicability, accuracy and effort. The results of an experimental analysis of the accuracy of hand held moisture metres using the electrical resistance method are discussed. Conclusions are drawn regarding the measurement of moisture content of wood upon delivery as well as of complete impact limiter assemblies. Requirements for quality surveillance during manufacturing of wood filled impact limiters are derived and it is exemplified how to meet them. Construction, manufacturing and inspection of impact limiter encapsulation with regard to leak tightness are addressed.