The mechanical test procedures that address fuel cladding failure during a RIA are reviewed with an emphasis on the development of test procedures that determine the deformation and fracture behavior of cladding under conditions similar to those reached in a RIA. An analysis of cladding strain data from experimental research reactor test programs that have simulated the RIA is presented. These data show that the cladding undergoes deformation characterized by hoop extension subject to a range of multiaxial stress states and strain paths comprised between plane-strain (no axial extension of the cladding tube) and equal-biaxial tension (equal strain in both the hoop and the axial orientations). Current mechanical test procedures of cladding material are then reviewed with a focus on their ability to generate the appropriate deformation response and to induce the prototypical multiaxial stress states and failure modes activated during a RIA. Two main groups of tests currently exist. In the first group, the deformation behavior of the cladding is examined by several variations of hoop tensile tests in which an axial contraction of the specimen gage section occurs such that a near-uniaxial tension stress state results; finite element analyses are then usually employed to deduce the deformation response, often under conditions of an assumed coefficient of friction between the specimen and test fixtures. The second group includes test procedures which attempt to reproduce the deformation and failure conditions close to those seen during a RIA such that any stress-state corrections of the failure conditions are comparatively small. The advantages and disadvantages of all of these deformation/fracture tests are discussed with special reference to testing high burnup fuel cladding.

Specimen geometries have been developed to determine the mechanical properties of irradiated Zircaloy cladding subjected to the mechanical conditions and temperatures associated with reactivity-initiated accidents (RIA) and loss-of-coolant accidents (LOCA). Miniature ring-stretch specimens were designed to induce both uniaxial and plane-strain states of stress in the transverse (hoop) direction of the cladding. Also, longitudinal tube specimens were also designed to determine the constitutive properties in the axial direction. Finite-element analysis (FEA) and experimental parameters and results were closely coupled to optimize an accurate determination of the stress-strain response and to induce fracture behavior representative of accident conditions. To determine the constitutive properties, a procedure was utilized to transform measured values of load and displacement to a stress-strain response under complex loading states. Additionally, methods have been developed to measure true plastic strains in the gauge section and the initiation of failure using real-time data analysis software. Strain rates and heating conditions have been selected based on their relevance to the mechanical response and temperatures of the cladding during the accidents.

Based on the behavior of a three-void cluster embedded within a representative volume element, this study utilizes three-dimensional finite element analyses to examine the sensitivity of void growth and coalescence to strain hardening, multiaxial stress state and inter-void spacing. The strain-induced growth of voids within the cluster is accelerated when the voids are closely spaced in a low strain-hardening material subject to high levels of stress triaxiality. Far-field deformation causes strain to concentrate within the inter-void ligament, and the resulting behavior induces a load–loss response of the inter-void region. Based on the load–loss criterion for the onset of void coalescence, the results show that coalescence is accelerated by increasing stress triaxiality and decreasing strain hardening and inter-void spacing. A straightforward analysis is then presented that relates void coalescence to (a) the strain-hardening exponent and (b) the dependence of the plastic constraint factor within the inter-void ligament on strain, the latter being sensitive to far-field stress triaxiality and void geometry.

Ductile fracture of HY-100 steel at high stress triaxialities occurs by a void-sheet mode of failure in which large elongated voids, formed at MnS inclusions, coalesce as a result of a localized deformation instability that develops between neighboring voids. In this study, micro-mechanical modeling using finite element analysis has been employed to examine the deformation localization behavior within void arrays based on experimentally observed inclusion microstructures of HY-100 steel. Treating the elongated voids as through-thickness holes, we utilize image-based multi-hole models, each depicting roughly 125 voids, to identify the significance of the critical features (size, spacing, clustering) of the void microstructure on the deformation localization process and ultimately void-sheet coalescence and failure. The deformation localization is especially sensitive to the presence of a few large voids spaced within roughly 30 hole diameters of each other and oriented on planes 45°±15° to the maximum applied principal stress. The results also show that deformation localization develops more readily at high stress triaxialities. Smaller, “secondary” voids can promote the onset of strain localization between large voids, even if they nucleate after a rather large void nucleation strain. Within microstructures consisting of solely of small voids, high density clusters can cause intense strain localization, but it is confined within the scale of the cluster.

Damage accumulation in the form of the volume fractions and number densities of strain-induced voids has been experimentally characterized for HSLA-100 steel subjected to tensile failure over a range of stress-states (stress triaxiality ratios of 0.8 to 1.4). The dependence of void volume fraction on strain indicates the presence of a void growth stage that is sensitive to stress-state in a manner that can be described by a relationship with the form that can be predicted by Rice and Tracey but with an increased dependence on stress triaxiality. The damage results also suggest a transition of stable void growth to rapid void growth and imminent material fracture at a critical void volume fraction that decreases slowly with increasing stress triaxiality ratio. A straight-forward analysis, based on the experimental observations, relates the observed experimental dependence of failure strains on stress triaxiality for this steel. The damage accumulation behavior of this steel, which fails due to the growth and coalescence of equiaxed voids, is also contrasted to that of HY-100 steel which forms elongated voids.

This study has examined the effects of nickel alloying additions on the microstructural characteristics and mechanical properties of Fe–xNi–0.85Mo–0.4C-base steels that were powder processed using double-press double-sinter processing to maximize density. The steels were examined in the as-processed condition as well as in a quench-and-temper heat treated condition. Tensile behavior indicates that while nickel content (at levels of 2,4, and 6%) increased tensile strength in the as-sintered condition, it did not significantly affect tensile strength in the quenched and tempered condition. In both conditions increasing Ni content decreased elongation to fracture. The 4% Ni steel, which tended to have the smallest maximum pore size, also exhibited the greatest fatigue strength.

The influence of tensile specimen geometry on the deformation behavior of flat Zircaloy-4 tensile specimens has been examined for gauge length-to-width ratios that range from 1:1 to 4:1. Specimen geometry has only minor effects on the values of the yield stress, tensile strength, apparent uniform strain at maximum load, and strain-hardening exponent. However, in all geometries but the 4:1 configuration, diffuse necking occurs before maximum load. As a result, strain distributions at maximum load are uniform only in the 4:1 geometry. The elongation to failure is also affected by specimen geometry with the shorter gauge sections exhibiting much higher total elongation values, due in large part to the concomitant specimen necking behavior.

Dry sliding block-on-ring wear tests were performed on a squeeze cast A390 Al alloy, a high pressure die cast 20%SiC–Al MMC, and a newly developed as-cast 50%SiC–Al MMC. The testing conditions spanned the transition that control the mild to severe wear for all materials. The results show that the sliding wear resistance increases as SiC particle volume fraction increases. The critical transition temperature, at which wear rates transit from mild to severe, also increases with increasing SiC content. Examination of the wear surfaces, the subsurface characteristics, and the wear debris indicate that a hard ‘mechanically alloyed’ layer, high in SiC content, forms on the sliding surface of the 50%SiC composite. This layer prevents the surface adhesion wear mechanisms active in the A390 alloy, and it inhibits delamination wear mechanisms that control the mild wear of the 20%SiC composite. As a result, mild wear of the 50%SiC composite occurs by an oxidation process. In the 20%SiC material, severe wear occurs as a consequence of material removal by a flow-related extrusion-like process. In contrast, the high SiC content prevents plasticity in the 50%SiC composite, which eventually is susceptible to severe wear at very high temperatures (≈450°C) due to a near-brittle cracking processes.

During operation of nuclear power reactors, irradiated Zircaloy-4 cladding tubes contain circumferentially oriented hydrides concentrated in a layer near the outer surface of the cladding. This study has investigated the effect of a hydride layer or “rim” located near the outer surface of the cladding tube on the failure of unirradiated Zircaloy-4 cladding tubes. Utilizing plane-strain ring-stretch tests with the maximum principal stress along the circumferential or hoop direction, we examined the influence of a hydride rim on the failure of unirradiated Zircaloy-4 cladding at both room temperature and 300°C. Fracture is found to be sensitive to hydride-rim thickness such that cladding tubes with a hydride-rim thickness >140 μm (≈700 wppm total hydrogen) exhibit brittle behavior, while cladding tubes with a rim thickness 95 μm and >110 μm under uniaxial and biaxial hoop tension, respectively. These results suggest that a ductile-to-brittle transition occurs in the deformation behavior of cladding tubes containing a a finite hydride rim thickness. Additionally, previous studies [18,19] have investigated the effect of a hydride rim on the failure path of Zircaloy-4 cladding tubes irradiated to high fuel burnups and subjected to RIAsimulation tests. Although not quantitative, according to post-test examinations, Garde and coworkers [18] and Fuketa and co-workers [19] suggest that fracture initiation occurs in the hydride rim by brittle crack growth and, depending on temperature and loading path, followed by either ductile or brittle fracture of the remaining cladding ligament. The purpose of this study is to explore the response of unirradiated Zircaloy-4 cladding tubes that contain hydrides concentrated in the form of a thin layer near the outer surface, as is typical of high-burnup cladding [19]. Using “ring-stretch” specimen geometries in order to impose multi-axial stresses and near plane-strain tension in the hoop direction of the cladding tube, we examine the influence of thin layers of hydrides on the ductility of Zircaloy-4 cladding subject to stress states relevant to potential in-service accidents such as the RIA. Experimental Procedures Material As in previous studies [12,15,17], Zircaloy-4 cladding tubes were obtained from Sandvik Metals and Westinghouse Electric Corporation in a CWSR condition with outer diameter of

Interfacial morphology and reaction products in thermal barrier coating systems were investigated by scanning and transmission electron microscopy (SEM and TEM). The samples consist of yttria-stabilized zirconia (YSZ; 6–8 wt.% yttria) deposited by air plasma spraying onto either of two types of bond coats: a layer consisting of Ni–15.9Cr–5.3Al–0.6Y with 5 wt.% of alumina particulate added, or one that was only the base Ni–Cr–Al–Y composition. In samples thermally cycled to failure in a burner rig, numerous interfacial protrusions of several microns or more in size are observed. These have a complex microstructure and contain elemental Ni intermixed with Ni(Al,Cr)2O4 spinel, (Al,Cr)2O3, and other oxides. Unlike some prior studies, nickel oxide (NiO) was not detected. Protrusion microstructures were similar for the two bond coat systems, but interfacial protrusions for the case of the base composition (i.e. no added alumina particulate) did not contain any spinel phase. Comparison of cross-sectional samples before and after oxidation indicates that the protrusions arise from the encapsulation of isolated segments of the bond coat. The intermixing of metallic Ni grains with oxides in the reaction zone may contribute to failure by affecting local stresses during thermal cycling.

The extended use of Zircaloy cladding in light water reactors degrades its mechanical properties by a combination of irradiation embrittlement, coolant-side oxidation, hydrogen pickup, and hydride formation. The hydrides are usually concentrated in the form of a dense layer or rim near the cooler outer surface of the cladding. Utilizing plane-strain ring-stretch tests to approximate the loading path in a reactivity-initiated accident (RIA) transient, we examined the influence of a hydride rim on the fracture behavior of unirradiated Zircaloy-4 cladding at room temperature and 300 C. Failure is sensitive to hydride-rim thickness such that cladding tubes with a hydride-rim thickness >100 {micro}m ({approx}700 wppm total hydrogen) exhibit brittle behavior, while those with a thickness <90 {micro}m ({approx}600 wppm) remain ductile. The mechanism of failure is identified as strain-induced crack initiation within the hydride rim and failure within the uncracked ligament due to either a shear instability or damage-induced fracture. We also report some preliminary results of the uniaxial tensile behavior of low-Sn Zircaloy-4 cladding tubes in a cold-worked, stress-relieved condition in the transverse (hoop) direction at strain rates of 0.001/s and 0.2/s and temperatures of 26-400 C.

The influence of temperature and strain rate on the deformation and failure behavior of HY-100 steel has been examined as a function of stress state using notched and un-notched axisymmetric tensile specimens. Behavior over the range of temperatures/strain rates from −85°C and 1 s−1 to 27°C and 10−3 s−1 shows an equivalence of decreasing test temperature or increasing strain rate on deformation behavior in a manner that can be predicted by the thermally activated flow theory. Over the entire range of temperatures/strain rates, the influence of stress state on failure is such that two void coalescence mechanisms control failure; at low stress triaxialities, relatively equiaxed voids grow to impingement, while at high triaxialities, a void-sheet process intervenes linking elongated MnS-initiated voids by a shear instability. The failure strains decrease rapidly with increasing stress triaxiality ratio in a similar manner for all temperatures and strain rates except for an intermediate stress triaxiality condition where the void-sheet mode of failure extends to lower stress triaxialities under cryogenic test conditions.

Ductile fracture of engineering alloys frequently occurs by a mechanism of void coalescence in which void-sheets form between the primary voids. Based on the microstructural features that control failure of HY-100 steel, computational modeling has been performed to examine the deformation localization behavior between primary voids and to predict ductile fracture by the void-sheet coalescence mechanism. Elongated inclusion-initiated voids are simulated as two distinct, hole-like voids on a plane inclined to the stress axis based on the inclined nature of the fracture surface. Consistent with experimental behavior, the micro-mechanical model identifies a strong tendency for strain localization between the voids (and therefore void-sheet failure) but only at a high degree of stress triaxiality. Furthermore, based on the formation of a secondary void population, the analysis also predicts with reasonable accuracy both the magnitude and stress-state dependence of the experimentally determined failure strains.

The high cycle fatigue behavior of squeeze cast SiC whisker-reinforced aluminum composites based on either A356 Al or A390 Al matrices has been studied. Squeeze cast Al/SiCw specimens, which contain roughly 17 vol.% whiskers from two different sources, have been examined for their high cycle fatigue strength under fully reversed test conditions using a staircase method to determine the mean fatigue strength at 107 cycles. The results show 30–40% increases in the fatigue strength of the A356 Al-based composites when compared to the unreinforced matrix alloy, but much less fatigue strengthening in the A390 Al-based composites. A fractographic analysis indicates that about 80% of the composite specimens fail as a result of crack initiation within regions which are characterized by low volume fractions of the SiC whiskers. These reinforcement-free regions assume two forms: continuous ‘veins,’ which are the more deleterious to fatigue, and discontinuous ‘unreinforced areas,’ which are deleterious only in certain shapes and sizes. Both finite element analysis and an Eshelby-based analysis indicate that the localized stresses within the unreinforced regions appear to be high enough to initiate fatigue cracks, especially if unreinforced areas are elongated and their major axis is aligned to the stress axis. The fractographic analysis also identifies the importance of primary Si particles in limiting the fatigue strength of the A390 Al–based composites.

Based on experiments and computational modeling, this study examines the failure behavior of structural steel, HY-100, which has been pre-strained at a high stress triaxiality and subsequently failed at a lower stress triaxiality. Both tensile tests of circumferentially notched specimens and the associated fractography show that even a small pre-strain at high stress triaxiality promotes an extension of the low ductility, ‘void-sheet’ mode of failure to lower stress triaxialities. Thus, there is a decrease in the failure strain compared to that if the material is deformed only at the lower stress triaxiality. These results imply that the pre-strain damage nucleates elongated voids whose growth is critical to the development of void-sheet failure. Micro-mechanical modeling using finite element analysis confirms that localization of plastic flow should occur between elongated ‘hole shaped voids’, despite their rather small initial cross-section size (2.5 μm) and comparatively large spacing (70 μm). Furthermore, employing a local failure criterion, the computational analysis predicts failure strains which are in good agreement with those observed after the pre-strain and strain-path change.

The high cycle fatigue behavior of a cast SiC particle-reinforced aluminum composite, Duralcan F3D20S, has been studied. The material which contains about 20 vol. % of SiC particles with an average size of 15 Gun, has been examined in the following two conditions: (a) in the as-cast condition after high-pressure die casting and (b) after hot isostatic pressing (HIP). Somewhat surprisingly, there were no significant differences between the high cycle fatigue lives of as-cast and HIP'd specimens. While porosity limited the fatigue lives in all cases. surface porosity was found to be more deleterious to fatigue life than internal pores of the same size. Primarily as a result of this factor, the HIP'd specimens, which retained small porosity on the subsequently machined surface, did not exhibit improved fatigue lives. Particle-free regions rich in Cu and Ni contents were also present and. when present in the vicinity of porosity, contributed to crack initiation, reducing fatigue life.

The contrasting characteristics of damage evolution have been examined in a multidirectional carbon/epoxy composite laminate (IM7/8551-7) subjected to both quasi-static and dynamic loading. Our experiments were performed on bend-test bars that were loaded either in 'supported' four-point bending or under 'unsupported' conditions with a Hopkinson pressure bar to induce dynamic loading. We found differences in the damage that occurred in specimens loaded by the two techniques, in terms of the number of cracks and the length of the cracks. In the case of quasi-static loading, there were many matrix cracks within individual plies and only a few delamination cracks between plies; the maximum ratio of numbers of matrix to delamination cracks observed was 6:1. Despite their small number, the delamination cracks had a greater total length than the matrix cracks, and specimen failure occurred as a result of delamination crack propagation. During dynamic loading, the ratio between numbers of matrix and delamination cracks was 3:1, and in this case the ratio between the total crack lengths was unity. A quantitative assessment of damage induced during quasi-static bending was made from specimen stiffness results. Using simple beam theory and knowing the location of the damage, we correlated beam stiffness to the materials effective elastic modulus. We found that the composite's effective modulus decreased rapidly with small amounts of initial damage, but that subsequent increases in damage decreased the effective modulus at a much lower rate.

Slow strain-rate tensile testing of Ni-based Alloys X-750 and 625 was performed in high-purity, deaerated water in order to determine whether hydrogen embrittlement occurs in these alloys at room temperature and 288 degreesC. These tests were conducted at an initial strain-rate of 4.6x10(-7)/sec under both 30 psig nitrogen (0 cc H-2/kg H2O STP) and 40 psig hydrogen (60 cc H-2/kg H2O STP), on Alloy X-750 in two heat-treatment conditions and on Alloy 625 in a solution annealed and aged condition. At room temperature in the hydrogenated environment, both alloys showed a pronounced susceptibility to hydrogen embrittlement. The presence of hydrogen reduced both the elongation to failure and the reduction in area at fracture by about 50%. Fractography indicated a transition from ductile, transgranular failure in the nitrogen environment to predominantly intergranular fracture under hydrogenated conditions. Fractography of the specimens tested at room temperature showed a transition in crack growth behavior from mixed mode slip band decohesion in nitrogen environments to Mode I intergranular fracture path in hydrogenated environments. In contrast, hydrogen had no discernible effect on the measured properties at 288 degreesC under the slow strain-rate conditions imposed.

Failure of two high strength Navy steels, HY-100 and HSLA-100, have been examined as a function of stress state using notched and unnotched axisymmetric tensile specimens. The failure strains of both steels decrease in an exponential manner with increasing stress triaxiality. The magnitudes of the failure strains are also sensitive to the material chemistry, particularly with regard to the addition of Ca as an inclusion shape modifier. Two failure mechanisms are identified: a global damage accumulation process involving coalescence of relatively equiaxed voids and a void-sheet process which links large, elongated inclusion-initiated voids by a shear instability. The void-sheet mechanism intervenes and limits ductility at high stress triaxialities in the HY-100 steel when it is oriented such that the elongated inclusions are normal to the tensile direction.

Experiments have been performed to examine the ductility of Zircaloy 4 cladding tubes under conditions of near plane-strain deformation in the hoop direction (transverse to the tube axis) at temperatures of 25 and 300°C and at strain rates of 10−3 and 102 s−1. To conduct these experiments, a specimen configuration was designed in which near plane-strain deformation is achieved, and a test methodology was established to determine two failure conditions: the limit strain at the onset of localized necking and the fracture strain. Experiments performed on cold-worked stress relieved material using the transverse plane-strain specimen geometry indicate major differences in failure behavior from that observed in uniaxial tension, although both test conditions result in failure by a localized necking process. The experimental results also indicate that while plane-strain fracture strains increase with temperature between 25 and 300°C, at a given temperature they are insensitive to strain rate. The limit strains at localized necking also increase with temperature but only at the high 102 s−1 strain rate. Finally, the failure data indicate a strong sensitivity to surface flaws, as predicted by localized necking theory.