Keiji Nakacho

Improvement of Residual Stresses of Circumferential Joint of Pipe by Heat-Sink Welding

Intergranular stress corrosion cracking may occur in some specific conditions on the inner surface of the welded joints of stainless steel pipes which are furnished in nuclear plants. One of the remedies for this cracking is to convert welding residual stress on this surface into compression. In this research, in order to improve welding residual stress, the authors investigated the effectiveness of the heat-sink welding (water cooling) by conducting theoretical analyses and experiments on SUS 304 pipes of different sizes in comparison with the conventional welding. The mechanisms of production of residual stresses by both methods are clarified and conditions for effective application of the heat-sink welding such as limitation of heat input, procedure of welding are indicated.

Measurement of Welding Residual Stresses of Reactor Vessel by Inherent Strain Method : Measurement of Residual Stresses of Pipe-Plate Penetration Joint

This study aims to ensure the safety of nuclear power plants. The accidents involving leaks from the welded zones at the pipe penetration part of a reactor vessel or at a coolant pipe are reported at home and abroad. One of the main causes is the welding residual stress. So, it is important to know the welding residual stress for maintaining high safety of the plants, the estimation of plants life cycle and the plan of maintenance. The welded joints of the nuclear power plants have complex shapes, and the welding residual stresses also have complex distributions three-dimensionally. In this study, the inherent strain method combined with finite element method is used to measure the welding residual stresses accurately. The mock-up is idealized for the welded joint at the pipe penetration part of the actual reactor vessel. The inherent strain method is applied to measure the residual stresses. In this method, the inherent strains are unknowns. When the residual stresses are distributed complexly in a three-dimensional stress-state, the number of unknowns becomes very large. So, the inherent strains are expressed with some functions to decrease the number largely. The theory, the experiment process and the analysed results are explained. The characteristics of the distributions of residual stresses and their production mechanisms are discussed. The inherent strain method gives the most probable values and the deviations of the residual stresses. The deviations are small enough for the most probable values. It assures the high reliability of the estimated results.

A Simple Estimating Method for Reduction of Welding Residual Stresses in Thick Welded Joint From Stress-Relief Annealing—Part III: Development of Estimating Equations for Multiaxial Stress State in Thick Welded Joint

Stress-relief annealing (SR treatment) is often applied to relieve welding residual stresses in the fabrication process of pressure vessels, etc. This study aims at development of an efficient method as simple as hand calculation to estimate reduction of residual stresses of a very thick welded joint by SR treatment. In the first report, an estimating method was developed for relaxation tests, in uniaxial stress state, at changing and constant temperatures because the stress relaxation phenomenon is considerably similar to that observed in the SR treatment of a joint. In the second report, the stresses relaxed by SR treatment in a very thick welded joint were analyzed accurately by the finite element method based on thermal elastic-plastic creep theory. The characteristics of the changes of the welding residual stresses in multiaxial stress state were studied in detail for further development of the estimating method to SR treatment of a very thick welded joint, of which the stress state and boundary condition are very complex. In this report, the estimating equations in multiaxial stress states are developed for the stress relaxation phenomenon in the thick welded joints, based on the foregoing characteristics.

Application of the plastic node method to thermal elastic-plastic and dynamic problems

Abstract In 1968, the author and his colleagues developed a new mechanism of the plastic hinge, based on the incremental theory of plasticity, and derived the elastic-plastic stiffness matrices for one-dimensional members. In 1979, expanding the basic idea of the plastic hinge, the authors developed a new method of plastic analysis of plates and solid bodies using the ordinary finite element method, and derived the elastic-plastic stiffness matrices. Later, they named the new method of plastic analysis ‘the plastic node method’. In this paper, the basic theory of the plastic node method has been further developed for analyses of thermal elastic-plastic and dynamic behaviors of structures, and stiffness equations thus derived. Using this new theory, several examples have been analyzed and the applicability of this method has been demonstrated.

Measurement of welding residual stresses of reactor vessels by inherent strain method – Diagnosis of inherent strain distribution function

The fundamental objective of this study is to ensure the safety of nuclear reactors. A few accidents involving leaks from welded zones at the pipe penetration part of reactor vessels or at coolant pipes have been reported at home and abroad. One of the main causes is welding residual stress. Therefore, it is very important to know the welding residual stress in order to maintain the high safety of the plant, estimate the plant life cycle and design an effective maintenance plan. Welded joints of nuclear reactor vessels have complex shapes, and the welding residual stresses also have three-dimensional (3D) complex distributions. In this study, inherent strain-based theory and method are applied to measure the welding residual stresses. The inherent strain method is an analytical method as an inverse problem, using the least squares method, based on the finite element method. So the method gives the most probable value and deviation of residual stress. The reliability of the estimated result is discussed. In this method, inherent strains are unknowns. When residual stresses are distributed complexly in a 3D stress-state, the number of unknowns becomes very large. So, the inherent strain distribution is expressed with an appropriate function to decrease largely the number. A mock-up is idealized for a welded joint at the pipe penetration part of an actual reactor vessel. The inherent strain method is applied to the measure the residual stress of the joint. In this paper, the applicability of the inherent strain distribution function is diagnosed. Ten kinds of functions are applied to estimate the residual stress, and the accuracy and reliability of the analysed results are judged from three points of view, i.e. residuals, unbiased estimate of variance of errors and welding mechanics. The most suitable function is selected, which brings the most reliable result.

Utility of Step Iterative Total Strain Method with Large Increment of Temparature

The thermal elastic-plastic analysis for the welding transient and residual stresses and deformations needs the huge computation time to obtain the accurate results. In this study, a new method, a step iterative total strain method, was developed to shorten the computation time largely, keeping high accuracy. The theory can take account of not only the temperature-dependencies of the mechanical properties, as usual thermal elastic-plastic theory, but also the change from elastic state to elastic-plastic state in one large temperature increment. Based on the theory (the thermal elastic-plastic constitutive equation), the equilibrium equation (the stiffness equation) was developed in the finite element method (FEM). New FEM program for thermal elastic-plastic analysis was developed, based on the above theory. A simple important problem of thermal elastic-plastic behavior was analyzed to verify the capability (accuracy) of the new method. In this report, some complex typical welding problems were analyzed. The residual stresses and deformations were compared with the accurate results obtained by the usual incremental method with very small temperature increments. As a result, it was confirmed that the computation time is very largely shortened, compared with the usual incremental method, keeping the high accuracy.

A Simple Estimating Method for Reduction of Welding Residual Stresses in Thick Welded Joint From Stress-Relief Annealing—Part IV: Applicability of the Simple Estimating Method for Stress-Relief Annealing of Thick Welded Joint

Stress-relief annealing (SR treatment) is often applied to relieve welding residual stresses in the fabrication process of pressure vessels, etc. This study aims at development of an efficient method as simple as hand calculation to estimate reduction of residual stresses of a very thick welded joint by SR treatment. In the first report, an estimating method was developed for relaxation tests, in uniaxial stress state, at changing and constant temperatures because the stress relaxation phenomenon may be considerably similar to that observed in the SR treatment of a joint. In the second report, the stresses relaxed by SR treatment in a very thick welded joint were analyzed accurately by the finite element method based on thermal elastic-plastic creep theory. The characteristics of the changes of the welding residual stresses in multiaxial stress state were studied in detail for further development of the estimating method to SR treatment of a very thick welded joint, of which the stress state and boundary condition are very complex. In the third report, the estimating equations in multiaxial stress states were developed for the stress relaxation phenomenon in the thick welded joints, based on the foregoing characteristics. In this report, the applicability of the simple estimating method is investigated for SR treatment of the thick welded joint, by comparing the estimated results with the accurate ones obtained by FEM.

THEORIES OF ANALYSIS AND MEASUREMENT OF WELDING RESIDUAL STRESSES AND PWHT.

2.2 Theory of Analysis 2.2.1 General forms of constitutive equation and stiffness equation 2.2.2 Constitutive equation (Incremental relation between stress and total strain) for thermal elastic-plastic state 2.2.3 Constitutive equation (Incremental relation between stress and total strain) for thermal elastic-plastic-creep state 2.2.4 Summary of constitutive equations 2.2.5 Basic equations for thermal elastic-plastic-creep analysis by finite element method (Stiffness equation)

A New Accurate Incremental Theory of Thermal Elastic-Plastic Analysis for Large Shortening of Computation Time

The thermal elastic-plastic analysis for the welding transient and residual stresses and deformations needs the huge computation time to obtain the accurate results. The main cause is that the increment of temperature has to be small enough to calculate accurately. In this study, the thermal elastic constitutive equation and the thermal elastic-plastic constitutive equation are refined so that the very large increment of temperature can be used. Moreover, in the latter, the definition of differential coefficient of the yield stress for change of temperature is modified to be ingenious. The new constitutive equations are introduced to the FEM program.Simple typical problem of thermal elastic-plastic analysis was calculated to confirm the reliability and the accuracy of the above theory. Then, some complex welding problems were analyzed. The welding residual stresses and deformations were compared with the accurate results obtained using small temperature increment. As a result, it was confirmed that the computation time is very largely shortened, compared with the usual incremental method using small temperature increment, keeping the high accuracy.

Prediction of creep life of thick welded joints by simple analytical model. Creep properties of thick welded joints and their improvement (2nd report)

To guarantee the safety of high-temperature vessels, such as high-speed breeder reactors, throughout their service life, it is necessary to ensure accurate creep deformation and life prediction. The welds produced in heavy plates have a particularly complex thermal history. This results in materials being highly inhomogeneous and implies their creep properties and behaviour are extremely sophisticated. This study focuses on prediction of the creep life of thick welded joints in nuclear reactors. A theoretical analysis and experiments are used to clarify the creep properties of welded joints until rupture as one of their critical performance features. The results obtained are intended to improve the creep properties of welded joints (reduce the creep strain rate and prolong the creep life) as well as to identify suitable material properties and welding methods from the perspective of constitutional control. The creep strength of welded joints has so far been extensively researched (see eg papers 2 and 3 for experimental research and paper 4 for a theoretical finite element analysis). The purpose of the present research study, however, is to provide a simple analytical model (and finally an analytical solution) for prediction of the creep behaviour of multi-pass thick welded joints. The first part of this study describes the development of a simple one-dimensional finite element model for numerical consideration of the effect of a multiaxial stress state. The creep curves and weld metal local strain distributions determined in the model are compared with the corresponding experimental results and the model shown to have a good prediction accuracy. This paper focuses on creep curves together with the creep rupture as an important aspect of the creep properties of welded joints, using the foregoing simple analytical model to predict the time to rupture (creep life) of welded joints. Aspects such as the method for introduction of rupture phenomena (conditions) into the simple analytical model, model accuracy, and local creep properties predicted by the model are also reported. Outline of simple analytical model