KIa Crack Arrest Toughness Assessment Using Thermal Shock on Notched Disks

The aim of the study is to validate the KIa-T curve on a thermal shock experiment performed on a notched disk (DTSE) taken from a A533-B type steel. Several experiments have been performed. Non linear thermal analyses were carried out using the finite element method in order to obtain the full thermal field within the specimen during crack propagation. The results obtained are in excellent agreement with the experimental results. The DTSE is also interpretated in static terms to compare the obtained KIa (T) values with the limit curve. Finally, dynamic F.E. simulations allow to estimate the influence of dynamic effects in the DTSE and thus validate the methodology. According to the computations, the crack arrested when dK/da>0 and dKd/da = 0. The comparison between stress intensity factor computed from elastic-static analysis (or dynamic) and RCC-M code demonstrated the conservatism of the approach. Then static analysis is sufficient to analyse the result, since no wave interference with the crack propagation was identified.

Assessment of Large-Scale Pressurized Thermal Shock Experiments Using the FAVOR Fracture Mechanics Computer Code

Large-scale experiments of pressure vessels performed at the Oak National Laboratory (ORNL)1 in the mid 1980s validated the applicability of the linear-elastic fracture mechanics (LEFM) computational methodology for application to fracture analysis of reactor pressure vessels (RPVs) in nuclear power plants. The current federal regulations to insure that nuclear RPVs maintain their structural integrity, when subjected to transients such as pressurized thermal shock (PTS) events, were derived in the early-mid 1980s from a comprehensive computational methodology of which LEFM is a major element. Recently, the United States Nuclear Regulatory Commission (USNRC) has conducted the PTS re-evaluation project that has the objective to establish a technical basis for a potential relaxation to the current PTS regulations which could have profound implications for plant license-extension considerations. The PTS re-evaluation project has primarily consisted of the development and application of an updated risk-based computational methodology that has been implemented into the Fracture Analysis of Vessels: Oak Ridge (FAVOR) computer code. LEFM continues to be a major element of the updated computational methodology. As part of the PTS re-evaluation program, there has been an extensive effort to validate that FAVOR has an accurate implementation of the LEFM methodology. This effort has consisted of the successful benchmarking of thermal analysis, stress analysis, and LEFM fracture analysis results between FAVOR and ABAQUS, a commercial general-purpose finite element computer code that has fracture mechanics capabilities, for a range of transient descriptions. The NRC has also participated in international round-robin benchmarking exercises in which FAVOR-generated solutions to well-specified PTS problems have been compared to solutions generated by other research institutions. A more fundamental aspect of the ongoing validation of FAVOR is demonstration that FAVOR can be used to successfully predict the results of large-scale fracture experiments. The objective of this paper is to document the FAVOR analysis of the first large-scale pressurized thermal shock experiment (PTSE) performed at ORNL. Results of these analyses provide validation that FAVOR accurately predicts the cleavage fracture initiation of a long surface breaking flaw in a large-scale thick-walled pressure vessel.

Temperature-Dependent Micro-Crack Propagation

The crack-lattice trapping phenomenon introduce by R. Thomson et al[1] is studied for the conditions of the Frenkel-Kontrova-type experiment. By using a new method, which allows further model extension for a finite temperature case we are able to describe an equilibrium crack energetics for arbitrary externa conditions and ascertain the crack propagation conditions. Specifically, the system free energy F as a function of nonlinear bond displacement ul for an external forces P and for a finite temperature T is found. The equilibrium values for the displacement ul = ul* and for G* = G(ul*), are obtained. The free-energy barrier height G = Gmax − G* dependence upon P and T is determined. With the help of the exact solution of the equilibrium equations we obtained the free energy as function of crack length G(l,T,P). We found that local free energy barriers take place for every crack length l, which is in contrast to the Thomson model. From the microscopic viewpoint it means that crack advance is controlled by local free energy barriers. We found that near the equilibrium length the crack energy barrier is relatively high, while far from equilibrium crack position, energy barrier height decreases to a finite value. It is worth to note that the barrier height monotonically decrease with the increase of the environment temperature. On the basis of our model the temperature dependence of the crack surface energy will be found, the global energetics of the crack will be described.

Shear-Dominated Fracture of Ductile Metal Under Mixed Loading and Preliminary Assessment

At room temperature, fracture of ductile metal often appears ductile, which is thought as the result of nucleation, dilatation and coalescence of microvoids in material. The remnant of microvoids in fracture surface is dimple. Dimple-mode fracture (void-dominated) is often accompanied with dilatation of volume of loaded material [1–4]. Fracture not accompanied with dilatation of volume was often studied at low temperature. Researchers think that the modes and mechanisms of fracture are affected by temperature and when temperature decreases ductile fracture becomes brittle fracture. No-volume-dilatation fracture is often related with low temperature fracture or brittle material’s fracture. But at room temperature, there is also no-volume-dilatation fracture, i.e., shear fracture of ductile material. There are very large differences between brittle fracture and shear fracture of ductile material, in which the deformation may be considerably larger before collapse [5]. Yamamoto [6] thought that heterogeneity of material resulted in local shear band in which there were lots of microvoids, and this led to fracture. Yang and Lee [7] described analogous process of initiation of fraction in their work. They indeed attributed shear fracture of ductile material to evolution of microvoids. In this presentation, instead of explaining shear-dominated fracture of ductile material as the result of evolution of microvoids, shear-dominated fracture of ductile metal is thought to be due to another mesomechanism, micro-shear-band. Some experiment results and simulation of finite element method of shear-dominated fracture of ductile metal are given. A kind of preliminary assessment about shear-dominated fracture of ductile material is proposed.

Embedded Crack Treatments and Fracture Toughness Evaluation Methods in Probabilistic Fracture Mechanics Analysis Code for the PTS Analysis of RPV

Using the probabilistic fracture mechanics analysis code PASCAL, we studied the treatment method of an embedded crack and the fracture toughness evaluation methods on the probability of crack initiation and fracture of a reactor pressure vessel (RPV). For calculating the stress intensity factor (SIF) of an embedded crack, the ASME and CRIEPI procedures were introduced into the PASCAL code. The CRIEPI method enables us to calculate the SIF values at three points on the crack tip. Under a severe pressurized thermal shock (PTS) condition, the crack growth analysis methods with different SIF calculation points and crack growth directions are compared. To evaluate precisely the fracture toughness after neutron irradiation, the new fracture toughness curves based on the Weibull distribution were incorporated into the PASCAL code. The calculated results with these new curves showed little difference in the conditional probabilities of RPV fracture as compared to the curve currently used in the U.S.

Fracture Toughness of Ductile Cast Iron and Applicability of Fracture Mechanics to DCI Casks

The JSME Cask Code, Rules for Construction of Metallic Casks, has a scope of using three types of casks fabricated by stainless steels, forged low alloy steels and ductile cast irons (DCI). On the other hand, the use of DCI for Cask material is not within the scope of the ASME Code, Section III, Division 3. In this paper, the fracture toughness of DCI for JSME cask code was compared with those of quenched and tempered low alloy steels. Furthermore, applicability of fracture mechanics to DCI casks is demonstrated by the results of fracture tests of reduced scale model casks. The results deny criticism arising from the fact that the DCI is not homogeneous material due to precipitation of spheroidal graphite particles for the application of the fracture mechanics.

Mechanical and Hydraulic Loads on the Supports of the Rotary Machine Due to Shaft Seizure in a Nuclear Power Plant

In a nuclear power plant there are two major equipment with high mechanical inertia, which have a rotating shaft. These are Pumps in the Primary Heat Transport System and Turbine in the secondary system. In both cases, the shaft seizure leads to transfer of very large loads to the supports. These supports, if not designed for seizure loads may fail. If the supports fail, there is a good possibility of a missile generated and hit the safety equipment. Seizure loads in these machines have three components namely mechanical inertial load, electrical load and hydraulic load. While the electrical and hydraulic loads have a limited peak value, the inertial load depends on the seizure time. For the normal observed seizures the three have a similar order of magnitude during seizure. As the casing is overdesigned the combined load is experienced by the supports. The pump of the Primary Heat Transport System (PHTS) of a nuclear power plant is centrifugal type run by an induction motor. If the pump shaft seizes, the seizure load will be experienced by the support structure. Due to the presence of the flywheel, the total moment of inertia of the pump motor assembly is quite high. Hence the resisting torque may be higher than the support’s design torque. Besides, the electric torque will continue to be applied as the motor trip on the overload current is delayed by several seconds as the corresponding relay is a thermal relay. Seizure torque will depend on pump seizure time. Lesser the seizure time, higher would be the load on the pump supports. The turbine in the secondary system has a large inertia due to blades. In case of a seizure the generator is tripped in hundreds of milliseconds. The load experienced by turbine supports due to seizure is significantly enhanced in the first few seconds due to sustained steam supply before it is cut off. This paper discusses the estimation of the three types of loads during seizure of the shafts in the pumps and turbine. It also discusses the possible safety consequences of these loads.

A Two-Parameter Description of the Behaviour of a Process Zone at a Crack Tip: Part II — Determination of the Parameters

An earlier paper (Part I) has shown how key parameters associated with the uniform stress process zone model of a crack: crack tip opening displacement, process zone size, crack opening area and the effective opening area of the process zone, depend upon parameters that are associated with the relevant terms in the expansion of the expression, for the purely elastic situation, for the relative displacement of the crack faces or the stress ahead of an elastic crack. The earlier paper focussed upon the case where the non-linear (with regards to applied stress) contributions to the crack-process zone parameters were determined to the first two terms in increasing powers of the applied loading stress parameter. These terms depend upon the first two terms in the expressions for the crack face relative displacement on the stress ahead of the crack in the elastic situation. The first of these terms is related to the stress intensity factor. In this paper we show how the parameter g0, which defines the second term, can be determined for some idealised situations.

Comparison of J-Integral Values and J-R Curves Between Carbon Steel Compact Tension Specimen and Pipe Specimen

In the present study, the transferability of elastic-plastic fracture toughness from a small-scale to a large-scale specimen was experimentally confirmed for carbon steel pipe with mild toughness. Fracture toughness tests were carried out on a pipe specimen 318.5 mm in outer diameter, 10.3 mm in thickness and having a through-wall crack, and also on a compact tension specimen 9.7mm in thickness, 25.4 mm in width, that had been cut out from the pipe specimen. Test results indicated the J-integral value of the pipe specimen at the crack initiation to be nearly twice that of the CT specimen. Finite element analysis conducted on the two specimens indicated this difference to arise primarily from the constraint near the crack front. Discussion was also made of the effects of crack orientation on elastic-plastic fracture toughness of CT specimens. The J-integral value at crack initiation in the specimen whose crack direction coincided with the pipe axial was found to be almost 54 % more than for specimens whose crack direction was circumferential.

Evaluation of Stress Intensity Factors by Finite Element Alternating Method

The finite element alternating method (FEAM), in conjunction with the finite element analysis (FEA) and the analytical solution for an elliptical crack in an infinite solid subject to arbitrary crack-face traction, is used for evaluating the stress intensity factor (SIF) of surface cracks. The major advantage of this method is that the SIF can be calculated by using the FEA results for an uncracked body. A newly developed system allows the FEAM to be performed by a simple method, which consists of the conventional FEA for an uncracked body and a subroutine for the FEAM alternating procedure. The SIFs are evaluated for semi-elliptical surface cracks on a plate and in a cylinder as well as interacting cracks on a plate. It is also shown that, by using fine mesh, the maximum error of the evaluation by the FEAM can be suppressed less than 2 percent.