Search

In 1965, the Atomic Energy Commission (AEC), at the advice of the Advisory Committee on Reactor Safeguards (ACRS), initiated the process that resulted in the establishment of the Heavy Section Steel Technology (HSST) Program at Oak Ridge National Laboratory (ORNL). Dr. Spencer H. Bush of Battelle Northwest Laboratory, the man being honored by this symposium, representing the ACRS, was one of the Staff Advisors for the program and helped to guide its technical direction. In 1989, the Heavy-Section Steel Irradiation (HSSI) Program, formerly the HSST task on irradiation effects, was formed as a separate program, and this year the HSST/HSSI Programs, sponsored by the U.S. Nuclear Regulatory Commission (USNRC), celebrate 40 years of continuous research oriented toward the safety of light-water nuclear reactor pressure vessels. This paper presents a summary of results from those programs with a view to future activities. The HSST Program was established in 1967 and initially included extensive investigations of heavy-section low-alloy steel plates, forgings, and welds, including metallurgical studies, mechanical properties, fracture toughness (quasi-static and dynamic), fatigue crack-growth, and crack arrest toughness. Also included were irradiation effects studies, thermal shock analyses, testing of thick-section tensile and fracture specimens, and non-destructive testing. In the subsequent decades, the HSST Program conducted extensive large-scale experiments with intermediate-size vessels (with varying size flaws) pressurized to failure, similar experiments under conditions of thermal shock and even pressurized thermal shock (PTS), wide-plate crack arrest tests, and biaxial tests with cruciform-shaped specimens. Extensive analytical and numerical studies accompanied these experiments, including the development of computer codes such as the recent Fracture Analysis of Vessels Oak Ridge (FAVOR) code currently being used for PTS evaluations. In the absence of radiation damage to the RPV, fracture of the vessel is improbable. However, exposure to high energy neutrons can result in embrittlement of radiation-sensitive RPV materials. The HSSI Program has conducted a series of experiments to assess the effects of neutron irradiation on RPV material behavior, especially fracture toughness. These studies have included RPV plates and welds, varying chemical compositions, and fracture toughness specimens up to 4 in. thickness. The results of these investigations, in conjunction with results from commercial reactor surveillance programs, are used to develop a methodology for the prediction of radiation effects on RPV materials. Results from the HSST and HSSI Program are used by the USNRC in the evaluation of RPV integrity and regulation of overall nuclear plant safety.

Two intermediate test vessels with inside nozzle corner cracks have been pressurized to failure at Oak Ridge National Laboratory (ORNL) by the Heavy Section Steel Technology (HSST) Program. Vessel V-5 leaked without fracturing at 88°C (190°F), and Vessel V-9 failed by fast fracture at 24°C (75°F) as expected. The nozzle corner failure strains were 6.5 and 8.4 percent, both considerably greater than pretest plane-strain estimates. The inside nozzle corner tangential strains were negative, implying transverse contraction along the crack front. Therefore, both vessels were reanalyzed, considering the effects of partial transverse restraint by means of the Irwin β I c formula. In addition, it was found possible to accurately estimate the nozzle corner pressure-strain curve by either of two semi-empirical equations, both of which agree with the elastic and fully plastic behavior of the vessels. Calculations of failure strain and fracture toughness corresponding to the measured final strain and flaw size are made for both vessels, and the results agree well with the measured values.

Proper identification of the transition temperature and the shape of the lower-bound (K{sub Ic}) fracture toughness curve in the transition range has been a long-term objective of work at Oak Ridge National Laboratory. A past practice has been to test a large number of specimens of varying sizes, from 1/2 to 8T compacts, in expectation that size effects and statistical variability of K{sub Jc} could be resolved empirically. Recently, statistical and constraint-based models have been developed that purport to explain much of what has been seen. Weakest-link theory has been successfully used to predict specimen size effects for the lower part of the transition curve. Constraint-based models of {beta}{sub c}-{beta}{sub Ic} and J{sub ssy} (small-scale yield) also can model size effects, but these tend to conflict among themselves with regard to the prediction of full constraint K{sub Jc}. All lack potential for defining the absolute lower bound of fracture toughness. Statistically based models have the benefit of quantifying data scatter characteristics and provide a basis for making lower-bound toughness estimates with assigned error estimates. The appropriate characterization of transition temperature is of value in industrial problems and is of particular importance to the nuclear industry where safety issues are involved.more » Here the K{sub Jc} data are obtained from small specimens, the size of which is dictated by volume limitations of surveillance capsule size. A basis has been explored for establishing a lower-envelope curve from such data.« less

A method for estimating the decrease in crack-initiation toughness, from a reference plane strain value, due to positive straining along the crack front of a circumferential flaw in a reactor pressure vessel is presented in this study. This method relates crack initiation under generalized plane strain conditions with material failure at points within a distance of a few crack-tip-opening displacements ahead of a crack front, and involves the formulation of a micromechanical crack-initiation model. While this study is intended to address concerns regarding the effects of positive out-of- plane straining on ductile crack initiation, the approach adopted in this work can be extended in a straightforward fashion to examine conditions of macroscopic cleavage crack initiation. Provided single- parameter dominance of near-tip fields exists in the flawed structure, results from this study could be used to examine the appropriateness of applying plane strain fracture toughness to the evaluation of circumferential flaws, in particular to those in ring-forged vessels which have no longitudinal welds. In addition, results from this study could also be applied toward the analysis of the effects of thermal streaming on the fracture resistance of circumferentially oriented flaws in a pressure vessel. 37 refs., 8 figs., 1 tab.

This lecture begins with a overview of applied fracture mechanics pertinent to safety of pressure vessels. It then progresses to a chronological panorama of experimental and analytical results. To be useful and dependable in safety analysis of real structures, new analysis developments must be physically realistic, which means that they must accurately describe physical cause and effect. Consequently, before mathematical modeling can begin, cause and effect must be established from experimental data. This can be difficult and time consuming, but worth the effort. Accordingly, the theme of this paper is that the search for patterns is constant and vital. This theme is illustrated by the development of small, single-specimen, fracture toughness testing techniques. It is also illustrated by the development, based on two different published large-strain, elastic-plastic, three-dimensional finite-element analyses, of a hypothesis concerning three-dimensional loss of constraint. When a generalization of Irwin`s thickness-normalized plastic-zone parameter, reaches a value close to 2{pi}, the through-thickness contraction strain at the apex of the near-tip logarithmic-spiral slip-line region becomes the dominant negative strain accommodating crack opening. Because slip lines passing from the midplane to the stress-free side surfaces do not have to curve, once these slip lines are established, stresses near the crack tip are only elevated by strain hardening and constraint becomes significantly relaxed. This hypothesis, based on published three-dimensional elastic-plastic analyses, provides a potentially valuable means for gaining additional insight into constraint effects on fracture toughness by considering the roles played by the plastic strains as well as the stresses that develop near a crack tip.

A series of six wide-plate crack-arrest tests was recently completed by the Heavy-Section Steel Technology program at the National Bureau of Standards, Gaithersburg, MD, using tensile-loaded specimens of A533 Grade B Class 1 steel. Crack-arrest data were obtained at temperatures in the transition range and above the onset of the Charpy upper shelf, thereby providing a basis for the development and evaluation of improved fracture-analysis methods. The 1 by 1 by 0.102-m single-edge-notched (SEN) specimens were welded to long straight pull tabs and subjected to a transverse linear temperature gradient before loading. The crack tips were sharpened by hydrogen-charging an electron-beam weld. The tests were designed to obtain crack arrest near the middle of the specimen where the temperature would produce a high-toughness level in the upper transition region of the material. The specimens were instrumented with strain gages and thermocouples. Initial static design calculations were made using textbook formulas. Additional calculations, using an assumed set of K I D versus a and T relations and an effective stress wave concept, confirmed the reasonableness of tentative design parameters. Pretest and posttest dynamic finite-element calculations were performed for each test. Computed results are compared with transient data for crack-line strains, crack speed, crack-opening displacement, arrest location, and postarrest tearing. Results from both application-mode and generation-mode dynamic analyses are presented. The arrest toughness values calculated from the test data are summarized for temperatures ranging from the transition into the Charpy upper-shelf range.

Information is provided on the series of pressurized-thermal-shock experiments at the Oak Ridge National Laboratory, motivated by a concern for the behavior of flaws in reactor pressure vessels having welds or shells exhibiting low upper-shelf Charpy impact energies, approx. 68J or less. (JDB)

An American Society for Testing and Materials (ASTM) standard method (E 1921-97) has been developed that exclusively uses fracture mechanics test practices and advanced statistical methods to establish the ductile-to-brittle transition range of fracture toughness for structural steels. The development of suitably accurate analyses had been slowed in the past due to an incomplete understanding of the operational mechanisms that control the fracture toughness behavior of structural steels. New perspectives taken are (1) that dominant linear-elastic conditions need not be rigidly enforced in test specimens and (2) that the effect of weakest-link behavior is defined from local cleavage crack initiators such as precipitates, inclusions, and grain boundary embrittlement; namely, all microstructural features in steel. Statistical models can be built upon such mechanisms that result in defined fracture probability levels and, when coupled to a master curve concept, can more accurately define the true loctition of the ductile-to-brittle transition temperature. An integral part of the ASTM test standard development work has been the production of a supporting technical basis document. This document presents substantial background data and supporting theoretical aspects that have been used to justify the method development. The paper will include some of the salient features presented.

The Master Curve method for determination of fracture toughness in the transition range in ASTM standard E1921 [1] brought an opportunity for the ASME Code to adopt a much better fracture toughness curve based on directly measured fracture toughness data. This also enables obtaining statistically based fracture toughness data. The industry, through PVRC Task Group (subsequently Section XI Task Group on Master Curve Fracture Toughness), took a two-phase approach to implement the adoption of the Master Curve method in the ASME Code. First, Phase I was completed with the issuance of ASME Code Cases N-629/N-631 [9, 10], published in 1999 which allowed the existing Code KIc curve to be used by means of an alternate indexing reference temperature RTT0 . This provided an important new approach to allow material specific, measured fracture toughness curves for ferritic steels in the code applications. However, this only rectified part of the shortcomings of the present Code KIc curve. In Phase II, it is intended to develop a direct means to utilize a tolerance bound of the Master Curve itself in place of the ASME KIc curve. This paper summarizes a proposal for such a procedure whereby a Master Curve fracture toughness tolerance bound is made usable in the ASME flaw evaluation processes, i.e. in Appendix A and Appendix G to Section XI of the ASME Boiler and Pressure Vessel Code. A draft code case is presented in Appendix in this paper.Copyright © 2008 by ASME

There is a rule in Acceptance Standards of Class 1 ferritic vessel in the ASME B&PV Code Section XI describing that allowable subsurface flaws near a component surface are transformed to allowable surface flaws. This is the flaw-to-surface proximity rule for the subsurface flaws. If the subsurface flaw is located closer to the component surface, the subsurface flaw not allowed close to the component surface but may be acceptable after transforming it to a surface flaw. This is a discontinuity of a subsurface flaw in the current Acceptance Standard for Class 1 ferritic vessels. Authors had developed a solution for the discontinuity and proposed the proximity rule at the last ASME PVP 2006 conference. After the discussion at the ASME Code Working Group on Flaw Evaluation, the proposed proximity rule was revised. This paper describes the revision of new flaw-to-surface proximity rule to minimizing the changes to the current allowable subsurface flaws of the Acceptance Standards.Copyright © 2007 by ASME

To enable determination of the fracture toughness reference temperature, T0, with reactor pressure vessel surveillance specimens, the precracked Charpy (PCVN) three-point bend, SE(B), specimen is of interest. Compared with the 25-mm (1 in.) thick compact, 1TC(T), specimen, tests with the PCVN specimen (10×10×55 mm) have resulted in T0 temperatures as much as 40°C lower (a so-called specimen bias effect). The Heavy-Section Steel Irradiation (HSSI) Program at Oak Ridge National Laboratory developed a two-part project to evaluate the C(T) versus PCVN differences, (1) calibration experiments concentrating on test practices, and (2) a matrix of transition range tests with various specimen geometries and sizes, including 1T SE(B) and 1TC(T). The test material selected was a plate of A533 grade B class 1 steel. The calibration experiments included assessment of the computational validity of J-integral determinations, while the constraint characteristics of various specimen types and sizes were evaluated using key curves and notch strength determinations. The results indicate that J-integral solutions for the small PCVN specimen are comparable in terms of J-integral validity with 1T bend specimens. Regarding constraint evaluations, Phase I deformation is defined where plastic deformation is confined to crack tip plastic zone development, whereas Phase II deformation is defined where plastic hinging deformation develops. In Phase II deformation, the 0.5T SE(B) BxB specimen (slightly larger than the PCVN specimen) consistently showed the highest constraint of all SE(B) specimens evaluated for constraint comparisons. The PCVN specimen begins the Phase II type of deformation at relatively low KR levels, with the result that KJc values above about 70 MPa√m from precracked Charpy specimens are under extensive plastic hinging deformation. For the second part, about twenty specimens of each type and size (65 PCVN) have been tested to enable statistically reliable comparisons of T0 for the various cases, with tests completed at temperatures from −28 to −37°C. The KJc test data obtained for HSST Plates 13B and 13A show PCVN bias values from −30 to −40°C relative to 1TC(T), but the standard 1TSE(B) specimen shows a bias of about −10°C. However, the SE(B) specimens exhibit a tendency for decreasing T0 with decreasing specimen size (B and/or b). The results are compared with those from other materials and observations are noted regarding potential effects of test temperature and metallurgical factors.

Intermediate test vessel V-8, a 152-mm-thick vessel fabricated of SA533, grade B, class 1 steel, was pressurized to failure at -23/sup 0/C. The vessel contained a fatigue-sharpened notch adjacent to a half-bead weld repair that had not been stress relieved. Residual stresses and fracture toughnesses were determined before the pressure test by measurements on a prototypical weld, and fracture predictions were made by linear elastic fracture analysis. Predictions agreed well with test results, demonstrating the important influence of high residual stresses on fracture behavior.

An important fracture mechanics problem is the determination of fracture toughness values with small specimens that fail after yielding. The J integral has an explicit meaning in terms of notch tip conditions in the plastic range and can be calculated from a single specimen test data. In this paper, an improved J integral analysis is developed for the compact specimen by considering the combined loading that exists on the net section. The analysis agrees with linear elastic fracture mechanics in the elastic range and gives a result close to that given by the equivalent energy method in the plastic range.

A pressurized-thermal-shock (PTS) facility was developed in the Heavy-Section Steel Technology Program at Oak Ridge National Laboratory for performing experiments that challenge predictions of analytical methods applicable to full-scale reactor pressure vessels under combined loading. The first experiment (PTSE-1) was designed to address three principal issues: (1) warm-prestressing phenomena; (2) crack propagation from brittle to ductile regions; and (3) transient crack stabilization in ductile regions. The paper presents a description of the PTS facility at ORNL and a review of the objectives and results of the first test. Also included are elastodynamic finite-element analyses of the two crack run-arrest events that occurred in the second and third phases of the test. Finally, some conclusions and recommendations are presented based on the outcome of the first experiment.

When small specimens are used to measure the cleavage fracture toughness of pressure vessel steels in the transition range of temperature, specimen thickness size effects and large amounts of data scatter are often observed. The size effects are manifested by an increase in the average value of fracture toughness with decreasing specimen thickness, eventually resulting in a change in fracture mode from cleavage to ductile tearing. It has been shown that a semiempirical adjustment for the interacting effects of specimen thickness, yield stress and toughness originally proposed by Irwin is capable of reducing the calculated values of toughness and data scatter to levels consistent with large specimen test data. This is true for dynamic as well as for static initiation toughness values. The nature of the size effect described by the Irwin ..beta../sub Ic/ equation is illustrated and specific cases are shown in which ..beta../sub Ic/ adjustment has eliminated size effects, for both static and dynamic fracture toughness data.

The equation of the normalised COD design curve recently proposed in the UK as a basis for determining allowable crack sizes is derived from the Equivalent Energy approximation for the J Integral. It is also shown that another approximation for the J Integral recently proposed by Begley et al . 8 is mathematically equivalent to the normalised COD approach.

This paper summarizes the capabilities and applications of the general-purpose and special-purpose computer programs that have been developed at the Oak Ridge National Laboratory (ORNL) for use in fracture mechanics analyses of HSST pressure vessel experiments. Emphasis is placed on the OCA/USA code, which is designed for analysis of pressurized-thermal-shock (PTS) conditions, and on the ORMGEN/ADINA/ORVIRT system which is used for more general analysis. Fundamental features of these programs are discussed, along with applications to pressure vessel experiments.

The first pressurized-thermal-shock test of a 148 mm thick steel pressure vessel with a 1 m long flaw was performed to investigate fracture behavior of a vessel under conditions relevant to a flawed nuclear reactor pressure vessel during an overcooling accident. The objectives were to observe crack arrest and stability on the ductile upper shelf and the effects of warm prestressing on crack initiation. Three coordinated pressure and thermal transients were imposed on the vessel, which was preheated to ∼290°C. Two episodes of crack propagation and arrest occurred. The thermal transients were induced by coolant at −29 to 15°C. Pressure transients were as high as 94.4 MPa. The experimental objectives were attained. The inhibiting effects of warm prestressing were definitely demonstrated. Crack propagation was nearly pure cleavage, and arrest at ∼30 K above the onset of the Charpy upper-shelf was experienced in a positive K1 gradient and with K1 = 300 MPa√m. Fracture-mechanics analysis of brittle fracture based on small-specimen toughness measurements was reasonably accurate. Flaw evaluation by procedures of the ASME Boiler and Pressure Vessel Code conservatively predicted vessel failure, which did not occur. No ductile tearing occurred after each crack arrest, although some stable tearing had been predicted on the basis of tearing resistance data.

Eight flawed 990-mm-OD, 152-mm-thick vessels of ASTM A508 class 2 forging steel or A533, grade B, class 1 steel plate have been pressurized hydraulically to burst or rupture. The rupture test of one vessel (V-7) was repeated pneumatically to study the effects of the sustained load thus attainable. Test temperatures ranged from 0 to 91°C so that transitional to upper shelf fracture toughness behavior was observed. Pretest predictions of failure pressure were based on tensile and toughness properties determined from Charpy-size to 102-mm-thick specimens representative of the test vessels. Linear elastic fracture mechanics based on strain and plastic instability analyses were generally adequate for determining the failure pressure, which ranged from 2.15 to 3.28 times design pressure.

In nonlinear applications of computational fracture mechanics, energy release rate techniques are used increasingly for computing stress intensity parameters of crack configurations. Recently, deLorenzi used the virtual-crack-extension method to derive an analytical expression for the energy release rate that is better suited for three-dimensional calculations than the well-known J-integral. Certain studies of fracture phenomena, such as pressurized-thermal-shock of cracked structures, require that crack tip parameters be determined for combined thermal and mechanical loads. A method is proposed here that modifies the isothermal formulation of deLorenzi to account for thermal strains in cracked bodies. This combined thermo-mechanical formulation of the energy release rate is valid for general fracture, including nonplanar fracture, and applies to thermo-elastic as well as deformation plasticity material models. Two applications of the technique are described here. In the first, semi-elliptical surface cracks in an experimental test vessel are analyzed under elastic-plastic conditions using the finite element method. The second application is a thick-walled test vessel subjected to combined pressure and thermal shock loading.

The second pressurized-thermal-shock experiment (PTSE-2) of the Heavy-Section Steel Technology Program was conceived to investigate fracture behavior of steel with low ductile-tearing resistance. The experiment was performed in the pressurized-thermal-shock test facility at the Oak Ridge National Laboratory. PTSE-2 was designed primarily to reveal the interaction of ductile and brittle modes of fracture and secondarily to investigates the effects of warm prestressing, A test vessel was prepared by inserting a cracklike flaw of well-defined geometry on the outside surface of the vessel. The flaw was 1 m long by ≈ 15 mm deep. The instrumented vessel was placed in the test facility in which it was initially heated to a uniform temperature and was then concurrently cooled on the outside and pressurized on the inside. These actions produced an evolution of temperature, toughness, and stress gradients relative to the prepared flaw that was appropriate to the planned objectives. The experiment was conducted in twoseparate transients, each one starting with the vessel nearly isothermal. The first transient induced a warm-prestressed state, during which K I , first exceeded K Ic . This was followed by repressurization until a cleavage fracture propagated and arrested. The final transient was designed to produce and investigate a cleavage crack propagation followed by unstable tearing. During this transient, the fracture events occurred as had been planned.

This paper contains example calculations of ductile flaw instability stresses for hypothetical flaws in nuclear pressure vessels. For comparison, three different methods of estimating upper shelf toughness as a function of Charpy impact energy were used, namely a power law R-curve correlation, the Rolfe-Novak correlation, and the Paris J50 correlation. All three methods were used in LEFM calculations including a plastic zone size correction, and gave similar results, with the Paris J50 method being the most conservative at low Charpy upper shelf energy levels. Safety factors based on the tearing modulus ratio T mat T appl can exceed those based on load by considerable amounts and use of them at this time is not recommended. The use of resistance curve data obtained from actual vessel material test specimens is recommended over the use of correlations. Furthermore, evaluation of a recently proposed modified crack extension adjustment procedure for R-curve data, which is not over-conservative, is recommended.

The inelastic deformation of most porous materials produces plastic volume changes, a phenomenon that does not occur in the case of most metals. These plastic volume changes may be of either sign, depending on the state of stress. The inelastic deformation of some porous materials can be approximately described by means of an ellipsoidal yield function. The equations for an ellipsoidal yield function are derived, initially for the case of transverse isotropy, isotropic strain hardening, and principal axes coinciding with the axes of anisotropy. These equations are applied to a nuclear grade graphite, EGCR-type AGOT graphite. An example problem is solved concerning the thermal loading of a graphite fuel sleeve. Possible applications of the ellipsoidal yield function to creep analysis and to concrete, are discussed. Finally, the shear stress terms in the ellipsoidal yield function are derived to make possible a more general application of the analysis.

It is the intent of most structural designs that general yielding should always precede fracture, even in the presence of flaws. The objective of the gross strain approach is to make direct measurements of flaw tolerance, in terms of gross strain to maximum load, as a function of flaw size and temperature. Gross strain is defined as the strain at the crack location, normal to the crack plane, that would exist if the crack were not there. Gross strain is measured as the axial strain in a tensile specimen containing a fatigue sharpened part-through surface flaw. Measurements of critical gross strain and critical crack opening displacement have been made for A533-B pressure vessel steel. These values undergo an abrupt increase with temperature, at the temperature at which gross section yielding and fracture coincide. The transition temperature increases with flaw size, as expected. Specimen thickness and width criteria have been developed for ensuring adequate constraint at the crack tip. The gross strain approach shows definite promise of becoming a measure of toughness that can be used in a quantitative procedure for design against brittle fracture when the object is to insure yielding before fracture in the presence of the largest expected flaw.

The second series of intermediate vessel tests were crack initiation fracture tests of 6-in.-thick 39-in.-OD steel vessels with sharp surface flaws approximately 2$sup 1$/$sub 2$ in. deep by 8 in. long in the longitudinal weld seams of the test cylinders. Fracture was initiated by means of hydraulic pressurization. One vessel was tested at each of three temperatures: 75, 130, and 190$sup 0$F. Pretest analyses were made to predict the failure pressures and strains. Fracture toughness data obtained by equivalent-energy analysis of precracked Charpy-V tests and compact-tension specimen tests were used in the fracture analyses. The vessels behaved generally as had been expected. Posttest fracture analyses were also performed for each vessel. Detailed discussions of the fracture analysis methods developed in support of the vessel tests described are included. 34 references. (auth)

HSST intermediate test vessel V-7 was repaired after being tested hydrostatically to leakage and was retested pneumatically as vessel V-7A. Except for the method of applying the load, the conditions in both tests were nearly identical. In each case, a sharp outside surface flaw 547 mm long (18 in.) by about 135 mm deep (5.3 in.) was prepared in the 152-mm-thick (6-in.) test cylinder of A533, grade B, class 1 steel. The inside surface of vessel V-7A was sealed in the region of the flaw by a thin metal patch so that pressure could be sustained after rupture. Vessel V-7A failed by rupture of the flaw ligament without burst, as expected. Rupture occurred at 144.3 MPa (20.92 ksi), after which pressure was sustained for 30 min without any indication of instability. The rupture pressure of vessel V-7A was about 2 percent less than that of vessel V-7.

Papers are presented on such topics as the experimental and numerical validation of a ductile fracture local criterion based on cavity growth simulation, plasticity near a blunt flaw under remote tension, the evaluation of steel toughness using various elastic-plastic fracture toughness parameters, and fracture mechanics tests on welded joints. Consideration is also given to a methodology for ductile fracture analysis based on damage mechanics, the prediction of critical crack size in plastically strained welded panels, and a combined statistical and constraint model for the ductile-brittle transition region.

Eight 6-in.-thick 39-in.-OD steel pressure vessels containing carefully prepared and sharpened surface cracks have been tested to provide an improved quantitative basis for evaluating the safety margins against fracture of nuclear reactor pressure vessels. The cylindrical regions of the test vessels were fabricated from either A508 class 2 forging steel or A533, grade B, class 1 steel plate. All the eight vessels tested were loaded to failure by internal hydraulic pressure. Flaws in the test vessels ranged from 1.20 to 5.30 in. in depth, and test temperatures ranged from 32 to 196$sup 0$F. Extensive fast fracture was observed, as expected, at 32$sup 0$F at a pressure near the gross yield pressure of the test vessel, and two vessels leaked without fracturing at or slightly above 190$sup 0$F. One of these vessels contained a flawed nozzle, and the other contained a very deep external surface flaw. Another vessel containing a smaller external surface flaw achieved 2 percent strain in the plastic range at 190$sup 0$F before developing a shear fracture that arrested before reaching the ends of the vessel. Extensive pre- and posttest fracture analyses have been performed for all the intermediate vessels tested. 11 references. (auth)

Failure testing is described for two 99-cm-diam (39-in.), 15.2-cm-thick (6-in.) steel pressure vessels, each containing one flawed nozzle. Vessel V-5 was tested at 88/sup 0/C (190/sup 0/F) and failed by leaking without fracturing after extensive stable crack growth. Vessel V-9 was tested at 25/sup 0/C (75/sup 0/F) and failed by fracturing. Material properties measured before the tests were used for pretest and posttest fracture analyses. Test results supported by analysis indicate that inside nozzle corner cracks are not subject to plane strain under pressure loading. The preparation of inside nozzle corner cracks is described in detail. Extensive experimental data are tabulated and plotted.

The High Flux Isotope Reactor (HFIR) pressure vessel has been in service for 20 years, and during this time, radiation damage was monitored with a vessel-material surveillance program. In mid-November 1986, data from this program indicated that the radiation-induced reduction in fracture toughness was greater than expected. As a result, a reevaluation of vessel integrity was undertaken. Updated methods of fracture-mechanics analysis were applied, and an accelerated irradiations program was conducted using the Oak Ridge Research Reactor. Results of these efforts indicate that (1) the vessel life can be extended 10 years if the reactor power level is reduced 15% and if the vessel is subjected to a hydrostatic proof test each year; (2) during the 10-year life extension, significant radiation damage will be limited to a rather small area around the beam tubes; and (3) the greater-than-expected damage rate is the result of the very low neutron flux in the HFIR vessel relative to that in samples of material irradiated in materials-testing reactors (a factor of approx.10/sup 4/ less), that is, a rate effect.

It is shown that the J-integral can be directly evaluated from single load-displacement records for a series of crack toughness specimens having the common feature that their only significant length dimension is that of the uncracked ligament. For the special case of bending loads on the ligament of a deeply cracked bar, J is shown to be twice the work of deformation divided by the ligament area. This and like results are employed to discuss Charpy and 'equivalent energy' toughness measures and also to evolve yet simpler estimating procedures for the J-integral.

It has recently been shown by experiment that a quantity known as the J-Integral may be a useful fracture criterion in the inelastic range. The mathematical definition of the J-Integral is first used here, in combination with some simplifying assumptions, to derive the known relationship between the conventional elastic stress concentration factor of a sharp notch and the elastic stress intensity factor for a crack of the same shape. The same assumptions, together with a generalization of Neuber's equation for the inelastic stress and strain concentration factors for a sharp notch, are then used to derive a relationship between the J-Integral and the parameter K I c d of the Equivalent Energy Method. The purpose of this derivation is to identify a set of assumptions which, in combination, lead to an explicit relationship between the two parameters. A graphical procedure for estimating upper and lower bounds on the full restraint value of the fracture toughness is also discussed.

To satisfy the need for a measure of fracture toughness at stress levels above yield, a quantity called gross strain crack tolerance has been proposed. It is the total elastic-plus-plastic strain measured at maximum load in a laboratory test of a specimen containing a surface crack and is, of course, a function of crack size and temperature. The object has been to develop a test to produce results that can be used to estimate the tolerance of material in full-scale hardware for strain in the presence of a crack. Previous work utilized precracked tension specimens of A533-B steel to study the effects of specimen geometry and crack size as a function of temperature. Now, the effects of strain gradients have been studied by loading the specimen in bending or by placing the crack in the bottom of a groove. Analysis of the displacement measurements showed that proper characterization of all the strain gradients tested required consideration of the strain values at two locations. The first, defined as the gross strain crack tolerance (GSCT), was located in the net section normal to the crack plane at mid-crack depth, the approximate location along the crack border where fracture originated. The second strain value was the gross strain on the back face in the gross section above and below the crack. At temperatures within the transition region, GSCT increased in direct proportion to the degree of tensile strain on the back face of the gross section. Tests in bending, which put the neutral axis between the crack tip and the back face, produced the lowest value of GSCT, while those in which the crack propagated into regions of increasing strain yielded the highest GSCT. These effects were attributed to the influence of back face strain on the development of the plastic zone and consequent relief of constraint.

To satisfy the need for a measure of fracture toughness at stress levels above yield, a quantity called gross strain crack tolerance has been proposed. It is the total elastic-plus-plastic strain measured at maximum load in a laboratory test of a specimen containing a surface crack and is, of course, a function of crack size and temperature. The object has been to develop a test to produce results that can be used to estimate the tolerance of material in full-scale hardware for strain in the presence of a crack. Previous work utilized precracked tension specimens of A533-B steel to study the effects of specimen geometry and crack size as a function of temperature. Now, the effects of strain gradients have been studied by loading the specimen in bending or by placing the crack in the bottom of a groove. Analysis of the displacement measurements showed that proper characterization of all the strain gradients tested required consideration of the strain values at two locations. The first, defined as the gross strain crack tolerance (GSCT), was located in the net section normal to the crack plane at mid-crack depth, the approximate location along the crack border where fracture originated. The second strain value was the gross strain on the back face in the gross section above and below the crack. At temperatures within the transition region, GSCT increased in direct proportion to the degree of tensile strain on the back face of the gross section. Tests in bending, which put the neutral axis between the crack tip and the back face, produced the lowest value of GSCT, while those in which the crack propagated into regions of increasing strain yielded the highest GSCT. These effects were attributed to the influence of back face strain on the development of the plastic zone and consequent relief of constraint.