Warwick M. Payten

Weld performance under creep using finite element modelling

The service life of a high temperature plant is frequently limited by creep damage in welded joints. The factors affecting weld performance under creep conditions are poorly understood. This paper presents the results of modelling of creep in some 240 different girth welds. The effects of weld angle, weld metal and heat affected zone material properties, heat affected zone width, and added axial loading were noted on the stress response of the models. A significant increase in stresses was seen with added axial loadings. The results include a number of results which allow us to specify construction and operation parameters, to assess which welds are at risk in service, and to define inspection or replacement strategies.

Optimal topology design using a global self-organisational approach

Abstract A method based on a self-organisational approach has been developed, where a local state operator defines the state of each finite element at each iteration. The algorithm is based on the principle of local adaptation with global feedback in the form of a derivative equation based on von Mises stresses that allows the local remodelling function to vary at each time step. The state operator has the form of a nonlinear differential equation solved iteratively based on the material density and strain energy density within each element. A constraint equation is formulated based on a maximum deviatoric strain energy criteria, with the objective to minimise the mass of the design domain subject to the above constraint. A number of examples are presented to demonstrate the use of this approach.

Advanced Ductility Exhaustion Methods for the Calculation of Creep Damage during Creep-Fatigue Cycling

The thermal mechanical cycling of high temperature components can result in creep-fatigue cycles in which the creep dwell has a wide variety of positions within the cycle. During in-phase cycling the creep dwell is placed at the maximum strain in the cycle. Out of phase cycling gives compressive dwells and phase cycling between 0° and 180° gives a cycle in which the dwell is placed before the maximum strain is reached, i.e., at an intermediate position within the cycle. Furthermore, components often experience cycles with a variety of different strain amplitudes, which can result in cycle sequencing effects. The effects of these different types of creep dwells have been investigated as part of the development programme for the R5 High Temperature Life Assessment Procedure. The materials included in the R5 programme were two low alloy ferritic steels; a cast 1CrMoV and a cast 1/2 CrMoV and three austenitic stainless steels; a type 316H steel, a cast type 304L, and a type 347 weld metal. The analysis of these tests has resulted in the proposal of a new method to calculate creep damage. This has been shown to give better predictions for the creep damage at failure in laboratory tests compared with both the current R5 ductility exhaustion approach and the time fraction approach, which is used in typical design codes such as ASME III and RCC-MR. In particular, the new method gives significantly improved predictions of creep damage at failure for creep-fatigue cycles with intermediate dwells and for cycles with low strain ranges, which are of particular relevance to the service cycles in real plant. This paper reviews the findings of the work on the new method, the effects of multiaxial states of stress and the effects of compressive dwells.

Modelling creep of pressure vessels with thermal gradients using Theta projection data

Pressure vessels are often exposed to through-wall temperature gradients. Thermal stresses occur in addition to pressure stresses. The resulting creep response is calculated using the Theta projection creep algorithm within a finite element code. It was found that the stress and temperature dependence of the creep response may lead to complex stress evolution.

Finite element analysis of creep using Theta projection data

The Theta projection creep algorithm has been implemented within a finite element code. This extends the predictive capability of Theta projection data to complex geometries, multiple material problems such as welded joints, and non-steady temperature conditions. Validation of the finite element methodology has been undertaken by re-modelling creep data based on the original Theta coefficients. The stress redistribution in a cylindrical pressure vessel was examined and compared to that predicted by the Norton equation. The effects of temperature variation were also modelled.

Estimating the plastic collapse of pressure vessels using plasticity contours

Abstract This investigation concerns the use of non-linear finite element analysis (FEA) to predict the plastic collapse of pressure vessels with and without defects. Computational non-linear plastic analysis of pressure vessels to estimate the collapse pressure is an iterative process. As a consequence, if the models are large and complex, considerable computer resources and time are required to generate solutions. As normal materials tend not to be ideally elastic–plastic (no sudden plastic collapse), a number of techniques are used to estimate the collapse pressure based on the FEA solution. These techniques often rely on plotting the pressure–strain curve; however, other techniques are also possible and are considered here. The aim of this paper is to assess the different methodologies in conjunction with FEA. Firstly, to consider their accuracy and secondly, to ascertain whether a particular technique would result in a reduction in the analysis time necessary to recover the plastic collapse pressure. Of the methods analysed, one based on the measurement (as a percentage) of the leading edge of the effective plastic strain (EPS) contour band showed promise. Under increasing pressures, regardless of geometry or material yield curve, the %EPS contour area coverage displays linear behaviour before through-wall plasticity is achieved. Based on this it is possible to extrapolate from low levels of plastic deformation up to plastic collapse pressures without requiring additional non-linear analyses, resulting in significant cost reductions in computational analysis times.

Creep-Fatigue Interaction Models for Grade 91 Steel

Different approaches for modelling creep–fatigue (CF) interaction are used on strain controlled creep fatigue data of 9Cr–1Mo-VNb (P91) steel and assessed with the target of finding suitable candidates for use in design rules. The assessed models include time, ductility, and strain energy-based creep-fatigue interaction methods and two simplified models. For the interaction diagram-based models, the challenge of acquiring representative creep damage fractions from the dynamic material response, i.e., cyclic softening with changing relaxation behaviour is addressed. In addition, the interaction diagram approaches are discussed in the light of known (fatigue) material scatter and defining representative cycles for CF data. The performance of the model are presented and also compared against the RCC–MR design code methodology. It is shown that the fitting accuracy of the complex interaction models vary significantly and that modified ductility based models seem to be less susceptible to changes in supporting creep and relaxation models. Successful and also superior prediction of the CF number of cycles to failure for Grade 91 steel was accomplished by simplified methods with much less fitting parameters. The practicality in using interaction diagram methods for design purposes, where simplicity is a key issue, is questioned.

Development of a Modular Ceramic Knee Prosthesis

Degenerative joint disease, recognized as an increasing problem for society, is a direct result of an aging population (1). When patients present with joint pain, their primary concern is the relief of pain and return to a mobile life style. This often requires replacement of skeletal parts, such as hips, knees, elbows, finger joints, shoulder, and teeth, or fusion of vertebrae, and repair or augmentation of the jaw and bones of the skull. The result is a current worldwide orthopedic market valued at over

Creep Modeling of Welded Joints Using the Theta Projection Concept and Finite Element Analysis

5 billion; joint replacement represents 68% of this market. The demand for knee replacements is increasing at approx 17%/yr, with some 300,000 knee joints replaced each year in the United States alone (2). This increase results in part from increased confidence in using such prostheses. Unfortunately, results do not reinforce this confidence: Long-term clinical results are scattered (3), and, although the overall rate of failure is reasonably low, it remains unacceptable. A further complication arises because the increase in younger patients undergoing total knee arthroplasty (TKA) may well lead to a higher incidence of eventual failure.

Modelling the creep behaviour of a reheat header longitudinal weld

Modeling of welded joints under creep conditions with finite element analysis was undertaken using the theta projection method. The results were compared to modeling based on a simple Norton law. Theta projection data extends the accuracy and predictive capability of finite element modeling of critical structures operating at high temperature and pressure. In some cases analyzed, it was found that the results diverged from those gained using a Norton law creep model.