Showing papers in "Nuclear Science and Engineering in 2001"
TL;DR: In this paper, a new burnup strategy CANDLE (Constant Axial shape of Neutron flux, nuclide densities and power shape During Life of Energy production) is proposed.
Abstract: The new burnup strategy CANDLE (Constant Axial shape of Neutron flux, nuclide densities and power shape During Life of Energy production) is proposed. With this burnup strategy, distributions of fu...
206 citations
TL;DR: Numerical results are presented that demonstrate the accuracy and efficiency of these methods for the SN equations on three-dimensional unstructured tetrahedral and hexahedral meshes.
Abstract: Discontinuous finite element methods for the SN equations on three-dimensional unstructured tetrahedral and hexahedral meshes are presented. Solution techniques including source iteration and diffu...
188 citations
TL;DR: In this paper, the performance of discontinuous finite element methods (DFEMs) on problems that contain optically thick diffusive regions is analyzed and tested, and the asymptotic analysis is quite general.
Abstract: The performance of discontinuous finite element methods (DFEMs) on problems that contain optically thick diffusive regions is analyzed and tested. The asymptotic analysis is quite general; it holds...
160 citations
TL;DR: A user-friendly and complete computer code package was developed comprising the multidimensional cross-section sensitivity and uncertainty code package SUSD3D, the secondary angular distribution (SAD) covariance data-processing module ERRORR34, and the cross- section covariance matrix library ZZ-VITAMIN-J/COVA.
Abstract: In the nuclear industry, an increased demand exists to provide modeling results with credible confidence bounds not only for simple but also for refined modeling. With the objective to facilitate a...
83 citations
TL;DR: In this paper, the 9Be(p,n) reaction was measured at several angles for proton-battering energies of 3.0, 3.4, 4.7, and 5.0 MeV.
Abstract: The thick-target neutron energy spectra of the 9Be(p,n) reaction were measured at several angles for proton-bombarding energies of 3.0, 3.4, 3.7, 4.0, and 5.0 MeV. Time-of-flight techniques were us...
77 citations
TL;DR: In this article, a new method for efficiently solving global Monte Carlo particle transport problems is presented, which is based on the use of a weight window that distributes Monte Carlo particles uniformly throughout the system.
Abstract: A new method for efficiently solving global Monte Carlo particle transport problems is presented. (In these problems, flux information is desired across the entire system, not just at a small number of detector locations.) The method is based on the use of a weight window that distributes Monte Carlo particles uniformly throughout the system. This (a) ensures that all subregions of the system are adequately sampled and (b) controls the particle weights, even in subregions far from sources. The weight window is constructed from an approximate deterministic solution of the forward transport problem. It is argued that a weight window based on the forward transport solution is more appropriate for global problems than the more familiar concept of basing a weight window on an adjoint solution for source-detector problems. It is also shown that by using Monte Carlo-generated Eddington factors in deterministic solutions of the quasi-diffusion equation, one can inexpensively compute updated forward-based weight windows and obtain a more efficient global Monte Carlo calculation.
76 citations
TL;DR: In this paper, a generalized Fokker-Planck (GFP) theory that robustly incorporates large-angle scattering is developed, and numerical experiments demonstrate that the resulting GFP equations are much more accurate than the standard GFP equation.
Abstract: The Fokker-Planck equation is often used to approximate the description of particle transport processes with highly forward-peaked scattering. Pomraning has shown that if the physical scattering kernel is sufficiently dominated by small-angle scattering, then the Fokker-Planck equation is an asymptotic approximation to the linear Boltzmann equation. However, most physically-meaningful scattering kernels contain a sufficient amount of large-angle scattering that the asymptotic criterion is not met. Thus, in many physical problems, solutions of the Fokker-Planck equation are substantially in error. In this paper, Pomraning's asymptotic results are generalized and a new generalized Fokker-Planck (GFP) theory that robustly incorporates large-angle scattering is developed. Numerical experiments demonstrate that the resulting GFP equations are much more accurate than the standard Fokker-Planck equation.
62 citations
TL;DR: In this paper, an improved quasi-static method is applied to the time-dependent, three-dimensional neutron transport equation with explicit representation of delayed neutrons, and the resulting new code, which uses a threedimensional, discrete ordinates code (TORT) to solve the static fixed-source equations, is tested using transient benchmark problems.
Abstract: An improved quasi-static method is applied to the time-dependent, three-dimensional neutron transport equation with explicit representation of delayed neutrons. The relevant equations are derived, and the corresponding implementation of the method is presented. The resulting new code, which uses a three-dimensional, discrete ordinates code (TORT) to solve the static fixed-source equations, is tested using transient benchmark problems that are available in the literature. Results obtained with the new time-dependent code, named TDTORT, are in satisfactory agreement with the benchmark problem results.
60 citations
TL;DR: In this paper, a method is presented to compute sensitivity coefficients for the eigenvalue of a critical assembly, including implicit effects associated with changes in resonance-shielded multigroup cross sectio...
Abstract: A method is presented to compute sensitivity coefficients for the eigenvalue of a critical assembly, including implicit effects associated with changes in resonance-shielded multigroup cross sectio...
59 citations
TL;DR: In this paper, a new method to obtain double differential cross sections of particles emitted in d+Li reactions is described, based on the diffraction approach, a modified intranuclear cascade model, and the usual evaluation techniques.
Abstract: For the generation of evaluated nuclear data sets required for the International Fusion Materials Irradiation Facility project, the basic features of the deuteron-lithium and neutron-lithium nuclear interactions are examined. Factors complicating the evaluation of deuteron-lithium reaction characteristics and weak points of previous calculations and evaluations are discussed. A new method to obtain double differential cross sections of particles emitted in d+Li reactions is described. The method is based on the diffraction approach, a modified intranuclear cascade model, and the usual evaluation techniques. The cross sections predicted by this method are in good agreement with existing experimental data for deuteron interactions in thick lithium targets. The study of neutron-lithium interactions is performed on the basis of different approaches: coupled channels, diffraction scattering, and direct breakup models.
49 citations
TL;DR: In the atmosphere, multiple scattering matters nowhere more than in clouds, and being a product of its turbulence, clouds are highly variable environments as discussed by the authors, which challenges three-dimensional (3D) modeling.
Abstract: In the atmosphere, multiple scattering matters nowhere more than in clouds, and being a product of its turbulence, clouds are highly variable environments. This challenges three-dimensional (3D) ra...
TL;DR: A transition metal hexacyanoferrate product CsTreat has been utilized at industrial scale for radioactive cesium separation at several nuclear power plants (NPPs) in several countries as mentioned in this paper.
Abstract: A transition metal hexacyanoferrate product CsTreat has been utilized at industrial scale for radioactive cesium separation at several nuclear power plants (NPPs) in several countries. A granular hexacyanoferrate ion exchanger has been used in packed-bed column mode operations for the removal of cesium from a variety of wastewater types. CsTreat beds have successfully purified both high-salt evaporator concentrates and dilute floor drain waters at NPPs in Finland and the United States. Furthermore, medium-active reprocessing solutions, containing high concentrations of sodium nitrate, have been decontaminated by a CsTreat bed at the Japan Atomic Energy Research Institute. These solutions are described as are other industrial applications of this ion exchange material, which, of all the commercial materials, has been shown to be the most selective exchanger for cesium. In addition, some prospective fields of hexacyanoferrate utilization, such as the use of CsTreat powder in a precoat filtration system, are discussed.
TL;DR: An adaptive grid method is presented for the solution of neutron diffusion problems in two dimensions and an a posteriori error estimator is developed to indicate the magnitude of local errors stemming from the low-order elemental interface approximations.
Abstract: An adaptive grid method is presented for the solution of neutron diffusion problems in two dimensions. The primal hybrid finite elements employed in the variational nodal method are used to reduce the diffusion equation to a coupled set of elemental response matrices. An a posteriori error estimator is developed to indicate the magnitude of local errors stemming from the low-order elemental interface approximations. An iterative procedure is implemented in which p refinement is applied locally by increasing the polynomial order of the interface approximations. The automated algorithm utilizes the a posteriori estimator to achieve local error reductions until an acceptable level of accuracy is reached throughout the problem domain. Application to a series of X-Y benchmark problems indicates the reduction of computational effort achievable by replacing uniform with adaptive refinement of the spatial approximations.
TL;DR: In this paper, the computational efficiencies of two nodal integral methods for the numerical solution of linear convection-diffusion equations are studied and compared with those of both a very traditional method and a recently developed state-of-the-art method.
Abstract: The computational efficiencies of two nodal integral methods for the numerical solution of linear convection-diffusion equations are studied. Although the first, which leads to a second-order spatial truncation error, has been reported earlier, it is reviewed in order to lead logically to the development here of the second, which has a third-order error. This third-order nodal integral method is developed by introducing an upwind approximation for the linear terms in the 'pseudo-sources' that appear in the transverse-averaged equations introduced in the formulation of nodal integral methods. This upwind approximation obviates the need to develop and solve additional equations for the transverse-averaged first moments of the unknown, as would have to be done in a more straightforwardly developed higher-order nodal integral method. The computational efficiencies of the second-order nodal method and the third-order nodal method - of which there are two versions: one, a full third-order method and the other, which uses simpler second-order equations near the boundaries - are compared with those of both a very traditional method and a recently developed state-of-the-art method. Based on the comparisons reported here for a challenging recirculating flow benchmark problem it appears that, among the five methods studied, the second-order nodal integral methodmore » has the highest computational efficiency (the lowest CPU computing times for the same accuracy requirements) in the practical 1% error regime.« less
TL;DR: An automated system for designing a loading pattern for boiling water reactors (BWRs) given a reference LP and control rod sequence has been developed and it was shown that the developed LPs using this system are equivalent or better than the manually optimized LPs.
Abstract: An automated system for designing a loading pattern (LP) for boiling water reactors (BWRs) given a reference LP and control rod (CR) sequence has been developed. This system employs the advanced nodal code SIMULATE-3 and a BWR LP optimization code FINELOAD-3, which uses a simple linear perturbation method and a modified Tabu search method to select potential optimized LP candidates. Both of these unique methods of FINELOAD-3 were developed to achieve an effective BWR LP optimization strategy and to have high computational efficiency. FINELOAD-3 also adjusts deep CR positions to compensate for the core reactivity deviation caused by fuel shuffling. The objective function is to maximize the end-of-cycle core reactivity while satisfying the specified thermal margins and cold shutdown margin constraints. This optimization system realized the practical application for real BWR LP design. Computer time needed to obtain an optimized LP for a typical BWR/5 octant core with 15 depletion steps is {approx}4 h using an engineering workstation. This system was extensively tested for real BWR reload core designs and showed that the developed LPs using this system are equivalent or better than the manually optimized LPs.
TL;DR: The refined AFEN method can describe the flux shape in the nodes more accurately, since the added flux expansion terms still satisfy the diffusion equation, and it can replace the corner-point fluxes with the interface flux moments without accuracy degradation, unless the problem consists of strongly dissimilar nodes.
Abstract: A refinement of the analytic function expansion nodal (AFEN) method is described. By increasing the number of flux expansion terms in the way that the original basis functions are combined with the transverse-direction linear functions, the refined AFEN method can describe the flux shape in the nodes more accurately, since the added flux expansion terms still satisfy the diffusion equation. The additional nodal unknowns introduced are the interface flux moments, and the additional constraints required are provided by the continuity conditions of the interface flux moments and the interface current moments. Also presented is an algebraically exact method for removing the numerical singularity that can occur in any analytic nodal method when the core contains nearly no-net-leakage nodes. The refined AFEN method was tested on the Organization for Economic Cooperation and Development (OECD)-L336 mixed-oxide benchmark problem in rectangular geometry, and the VVER-440 benchmark problem and a nearly no-net-leakage node embedded core problem, both in hexagonal geometry. The results show that the method improves not only the accuracy in predicting the flux distribution but also the computing time, and that it can replace the corner-point fluxes with the interface flux moments without accuracy degradation, unless the problem consists of strongly dissimilarmore » nodes. The possibility of excluding the corner-point fluxes increases the flexibility in implementing this method into the existing codes that do not have the corner-point flux scheme and may make it fit better for the nonlinear scheme based on two-node problems.« less
TL;DR: In this paper, a New Monte Carlo code (EBUF) was developed to calculate improved point isotropic photon exposure buildup factors in media, which accounts for coherent scattering and bound-electron Compton scattering.
Abstract: A New Monte Carlo code (EBUF) is developed to calculate improved point isotropic photon exposure buildup factors in media. Variance reduction techniques are used to perform calculations up to 60 mean free paths. EBUF accounts for coherent scattering and bound-electron Compton scattering. Bremsstrahlung photons and annihilation gamma rays as well as K and L X-rays are considered. The most recent cross-section data are used. The EBUF exposure buildup factors compare very well with those from the ANS-6.4.3 Working Group (ANS-6.4.3) when the same initial conditions are assumed: no coherent scattering, free-electron Compton scattering, and only K X-ray fluorescence. Next, a detailed physics treatment is used to calculate a representative set of exposure buildup factors in aluminum, iron, lead, water, air, and concrete over a large energy range (20 keV to 10 MeV). The effects of L X-rays are shown for lead at low energy. The EBUF factors are in good agreement with the SN1D code results for low-Z media. ...
TL;DR: In this article, an alternate formulation of the modified nodal integral method (MNIM) has been developed to further reduce computation time when solving nonlinear partial differential equations with a nonlinear convection term such as Burgers' equation and the Navier-Stokes equation.
Abstract: An alternate formulation of the recently proposed modified nodal integral method (MNIM) has been developed to further reduce computation time when solving nonlinear partial differential equations with a nonlinear convection term such as Burgers' equation and the Navier-Stokes equation. In this formulation, by adding and subtracting a linearized convection term, in which the node-averaged velocity at the previous time step multiplies the spatial derivative, the node-interior approximate analytical solution is developed in terms of this previous time-step node-averaged velocity. This leads to a set of discrete equations with coefficients that need to be evaluated only once each time step for each node, resulting in a significant reduction in computing time when compared with the original MNIM formulation. A numerical scheme using the node-averaged velocities at the previous time step - to be referred to as M{sup 2}NIM - for the two-dimensional, time-dependent Burgers' equation has been developed. The method is shown to be second order and to posses inherent upwinding. When compared with MNIM, numerical results show a significant reduction in the computation time without sacrificing accuracy.
TL;DR: In this article, the authors used the Finite Element Transient Criticality (FETCH) coupled neutronics/computational fluid dynamics code to simulate criticality transients in multiphase media and applied it to fissile solution transient criticality.
Abstract: Research on the incorporation of compressibility effects, for both the liquid and radiolytic gas phases, into the Finite Element Transient Criticality (FETCH) coupled neutronics/computational fluid dynamics code is described. The code has been developed to simulate criticality transients in multiphase media and is applied here to fissile solution transient criticality. The predicted fission and pressure transients obtained by the enhanced numerical model are benchmarked against the results from the SILENE series of experiments on criticality transients in uranium solutions. The amplitude and the form of the first pressure peak, following a step reactivity change, are well represented, and insight is gained into the form of subsequent pressure oscillations. An explanation is given on the absence of these oscillations in more energetic transients.
TL;DR: In this paper, the numerical stability, equilibrium diffusive limit, and accuracy of the variable Eddington factor (VEF) methods and flux-limited diffusion methods for radiation transport calculations are considered.
Abstract: The numerical stability, equilibrium diffusive limit, and accuracy of the variable Eddington factor (VEF) methods and flux-limited diffusion methods for radiation transport calculations are considered. The diffusive limit analysis proves that three VEF closures and their associated flux-limiters retain full first-order accuracy in the equilibrium diffusion limit while achieving the correct propagation speed in the optically thin streaming limit. The stability analysis reveals that the flux-limited diffusion methods are unconditionally stable, but the VEF equations with an arbitrary nonlinear closure can be numerically unstable for certain commonly used differencing schemes. However, regular solutions to the VEF equations are obtainable by Godunov-type schemes. Numerical comparisons among various solutions for a test problem show that flux-limited diffusion methods are only slightly less accurate than their corresponding VEF methods, and the Minerbo VEF method and the Minerbo flux-limited diffusion method are in general more accurate than other approximations.
TL;DR: In this paper, the application of burnable absorbers (BAs) to minimize power peaking, reactivity loss, and capture-to-fission probabilities in an accelerator-driven waste transmutation system has been investigated.
Abstract: The application of burnable absorbers (BAs) to minimize power peaking, reactivity loss, and capture-to-fission probabilities in an accelerator-driven waste transmutation system has been investigated. Boron-10-enriched B{sub 4}C absorber rods were introduced into a lead-bismuth-cooled core fueled with transuranic (TRU) discharges from light water reactors to achieve the smallest possible power peakings at beginning-of-life (BOL) subcriticality level of 0.97. Detailed Monte Carlo simulations show that a radial power peaking equal to 1.2 at BOL is attainable using a four-zone differentiation in BA content. Using a newly written Monte Carlo burnup code, reactivity losses were calculated to be 640 pcm per percent TRU burnup for unrecycled TRU discharges. Comparing to corresponding values in BA-free cores, BA introduction diminishes reactivity losses in TRU-fueled subcritical cores by {approx}20%. Radial power peaking after 300 days of operation at 1200-MW thermal power was <1.75 at a subcriticality level of {approx}0.92, which appears to be acceptable, with respect to limitations in cladding and fuel temperatures. In addition, the use of BAs yields significantly higher fission-to-capture probabilities in even-neutron-number nuclides. Fission-to-absorption probability ratio for {sup 241}Am equal to 0.33 was achieved in the configuration studied. Hence, production of the strong alpha-emitter {sup 242}Cm is reduced, leading to smallermore » fuel-swelling rates and pin pressurization. Disadvantages following BA introduction, such as increase of void worth and decrease of Doppler feedback in conjunction with small values of {beta}{sub eff}, need to be addressed by detailed studies of subcritical core dynamics.« less
TL;DR: In this article, an idealized lattice structure is considered of multilayer aerosol deposits, where every particle at the deposit surface is associated with a resuspension rate constant depending on a statistically distributed particle parameter and on flow conditions.
Abstract: An idealized lattice structure is considered of multilayer aerosol deposits, where every particle at the deposit surface is associated with a resuspension rate constant depending on a statistically distributed particle parameter and on flow conditions. The response of this generic model is represented by a set of integrodifferential equations. As a first application of the general formalism, the behavior of Fromentin's multilayer model is analyzed, and the model parameters are adapted to experimental data. In addition, improved relations between model parameters and physical input parameters are proposed. As a second application, a method is proposed for building multilayer models by using resuspension rate constants of existing monolayer models. The method is illustrated by a sample of monolayer data resulting from the model of Reeks, Reed, and Hall. Also discussed is the error to be expected if a monolayer resuspension model, which works well for thin aerosol deposits, is applied to thick deposits under the classical monolayer assumption that all deposited particles interact with the fluid at all times.
TL;DR: In this article, the reaction rate distributions measured with a special gamma-scanning machine employing twin germanium detectors were validated using the experimental results of the LWR-PROTEUS Phase I project corresponding to full-density water moderation conditions (core 1 B).
Abstract: HELIOS, CASMO-4, and MCNP4B calculations of reaction rate distributions in a modern, fresh 10 10 boiling water reactor fuel element have been validated using the experimental results of the LWR-PROTEUS Phase I project corresponding to full-density water moderation conditions (core 1 B). The reaction rate distributions measured with a special gamma-scanning machine employing twin germanium detectors consisted of total fission Ftot and 238U-capture C8. The average statistical errors for the gamma scans were better than 0.5% for Ftot and 0.9% for C8. The rod-by-rod measurements were performed on 60 different fuel rods selected from the central part of a test zone consisting of actual, fresh SVEA-96+ fuel elements, thus gaining in realism by departing from conventional fuel rod mockups. In the case of Ftot, the root-mean-square (rms) of the rod-by-rod distribution of differences between calculational and experimental (C-E) values has been found to be 1.1% for HELLOS and for CASMO-4, and 1.3% for MCNP4B. For C8, the rms values of the (C-E) distributions are 1.0, 1.3, and 1.4% as obtained with HELIOS, CASMO-4, and MCNP4B, respectively. The effects of using different data libraries (ENDF/B-V, ENDF/B-VI, and JEF-2.2) with MCNP4B were also studied and have been found to be small.
TL;DR: In this paper, the quality of existing data is not sufficient for mappings of neutron-induced reactions with oxygen to a wide range of neutron energies for many applied purposes, such as medical applications.
Abstract: Photon-production data from neutron-induced reactions with oxygen are important over a wide range of neutron energies for many applied purposes. The quality of existing data is not sufficient for m...
TL;DR: The neutron capture cross section of 232Th has been measured in the energy range from 5 to 225 keV at the Karlsruhe 3.7-MV Van de Graaff accelerator relative to the gold standard as mentioned in this paper.
Abstract: The neutron capture cross section of 232Th has been measured in the energy range from 5 to 225 keV at the Karlsruhe 3.7-MV Van de Graaff accelerator relative to the gold standard. Neutrons were pro...
CERN1
TL;DR: In this paper, the authors measured the neutron capture cross section relative to the JENDL-3 database values up to 800 keV and over 1.4 MeV for the reaction {sup 235}U(n,f).
Abstract: The neutron capture cross section of {sup 232}Th has been measured relative to {sigma}(n, {gamma}) for {sup 197}Au and {sigma}(n,f) for {sup 235}U in the energy range from 60 keV to 2 MeV. Neutrons were produced by the {sup 7}Li(p,n) and T(p,n) reactions at the 4-MV Van de Graaff Accelerator of CEN Bordeaux-Gradignan. The activation technique was used, and the cross section was measured relative to the {sup 197}Au(n,{gamma}) standard cross section up to 1 MeV. The characteristic gamma lines of the product nuclei {sup 233}Pa and {sup 198}Au were measured with a 40% high-purity germanium detector. Above this energy, the reaction {sup 235}U(n,f) was also used as a second standard, and the fission fragments were detected with a photovoltaic cell. The results, after applying the appropriate corrections, indicate that the cross sections are close to the JENDL-3 database values up to 800 keV and over 1.4 MeV. For energies in the intermediate range, our values are slightly lower than those from all the libraries.
TL;DR: For both PWR and Canada deuterium uranium (CANDU) tandem analysis, the Direct Use of spent PWR fuel In CANDU reactor (DUPIC) fuel cycle was studied in this paper.
Abstract: For both pressurized water reactor (PWR) and Canada deuterium uranium (CANDU) tandem analysis, the Direct Use of spent PWR fuel In CANDU reactor (DUPIC) fuel cycle in a CANDU 6 reactor is studied u...
TL;DR: In this article, a conceptual design of a lead-bismuth-eutectic (LBE)-cooled actinide burner core with innovative streaming fuel assemblies (FAs) is described.
Abstract: A conceptual design of a lead-bismuth-eutectic (LBE)-cooled actinide burner core with innovative streaming fuel assemblies (FAs) is described. The 1800-MW(thermal) core employs metallic, fertile-free fuel where the transuranics (plutonium plus minor actinides) are dispersed in a zirconium matrix. The core contains 157 streaming FAs that enhance neutron streaming by employing gas-filled, sealed streaming tubes at the FA periphery and center. The large reactivity excess at the beginning of life is compensated for by a system of double-entry control rods. The arrangement of top-entry and bottom-entry control rods in a staggered pattern allows the achievement of a very uniform axial power profile and a small reactivity change from control rod driveline expansion. The reactor can operate with an 18- to 24-month cycle length. Safety is provided through negative reactivity coefficients and tight neutronic coupling. The void coefficient is negative for a partially as well as a fully voided core. The effective delayed neutron fraction is 25% less than that of typical oxide-fueled fast reactors, making the requirements on reactor control performance more demanding. The Doppler coefficient is negative with a magnitude appreciably lower than the typical values of oxide fuels in sodium-cooled reactors, but comparable to the values observed in integral fast reactor (IFR) cores with metallic U-Pu-Zr fuels. The fuel thermal expansion coefficient is also negative, having a magnitude approximately equal to the Doppler coefficient. In terms of the transuranic destruction rate per MW(thermal) per effective full-power year, the design is comparable to accelerator-driven systems (ADSs). Long-lived fission products also can be transmuted, albeit at lower incineration efficiency than in ADSs.
TL;DR: In this article, measured neutron resonance cross sections are usually analyzed and parametrized by fitting theoretical curves to high-resolution point data, and the cross sections depend mainly on the “...
Abstract: Measured neutron resonance cross sections are usually analyzed and parametrized by fitting theoretical curves to high-resolution point data. Theoretically, the cross sections depend mainly on the “...
TL;DR: For nearly two decades, hydrogen water chemistry (HWC) has been used as a remedial measure to protect boiling water reactor (BWR) structural components against intergranular stress corrosion cracki...
Abstract: For nearly two decades, hydrogen water chemistry (HWC) has been used as a remedial measure to protect boiling water reactor (BWR) structural components against intergranular stress corrosion cracki...