No abstract

SUMMARYA canonical form for the representation of elastic predictor radial return plasticity algorithms is presented which is deceptive in its simplicity. Iterative application of the corrector is in a form used universally in the solution of ordinary di erential equations and substitution of di erent yield functions and state variable evolution equations is trivial. The consistency condition is maintained by an additional constraint equation via what is, in e ect, a Lagrange multiplier. A consistent material Jacobian may be obtained automatically by applying partial elimination to the same set of equations. The process is numerical and requires no additional algebraic manipulation. To demonstrate the validity of such a simple technique, existing, apparently more complex, formulations are derived through the simple expedient of static condensation. As a practical example of the application of the method, a standard von Mises plasticity model is implemented and results are presented for two standard benchmarks that test quadratic convergence for large increments.

SUMMARYA canonical form for the representation of material constitutive equations within standard finite element codes has been developed which is deceptive in its simplicity. Substitution of different equation systems is trivial and a consistent material Jacobian may be obtained automatically. The process is essentially numerical and does not involve the difficult algebraic manipulation associated with more traditional approaches. It is nevertheless exact because partial derivatives are derived analytically and simple because calculation of these derivatives is the only algebraic manipulation required. The remainder of the process is generic. In this paper, the algorithm is simplified further. A much more detailed and transparent explanation of the key to the method is given, namely calculation of the consistent Jacobian. A trivial modification also extends the method to plane stress. The original algorithm was validated for a simple material model. The new form is used to implement the significantly more difficult modified Gurson model for porous metal plasticity with hydrostatic yield dependence. It is tested for single element cases involving total collapse and also used to simulate necking of a notched cylindrical bar.

This paper provides an overview over the work of the European Network on Neutron Techniques Standardization for Structural Integrity (NeT). The network involves some 35 organisations from industry and academia and these partners undertake the application of modern experimental and numerical techniques to problems related to the structural integrity of components, mainly relevant to nuclear applications. While being built around neutron scattering techniques, which are predominantly applied for analyses of welding residual stresses, one of the major strengths of the consortium is the diversity in available experimental and numerical techniques. In the residual stress area, for example, many types of materials characterizations testing, several methods for residual stress analysis, including neutron and X-ray diffraction, deep hole drilling, the contour method and others, and many different ways of numerical analysis employing several commercially available FEM codes can be covered by the partners. Currently the network has embarked on five different Task Groups. Four of these are dealing with welding residual stress assessment, and one applies Small Angle Neutron Scattering for studying thermal ageing processes in duplex stainless steels used for reactor core internals. The work already performed in the context of NeT and the envisaged investigations for the ongoing Task Groups are briefly outlined in this paper. The aim is to give the reader a comprehensive overview of the work of NeT and to shed some light on the potential present in this kind of collaborative effort.

SUMMARYIn order to solve problems of re-radiation and reflection in finite element thermal analysis, a discretization consisting of flat isothermal surfaces is usually assumed. Such an approximation is clearly inconsistent if an isoparametric model is used for any surrounding solid. In addition, geometric view-factors are often required explicitly, but their calculation is generally considered to be tedious and error prone. The representation of radiative exchange by a boundary integral equation is a standard analytical technique. The use of such a representation in conjunction with finite element analysis, however, appears to be quite recent. The present work describes a scheme whereby standard finite elementiboundary element coupling theory may be applied to yield a method which is both simple and consistent. The use of isoparametric elements allows varying temperatures and curved surfaces to be modelled. For an empty enclosure the kernel of the integral equation is simply related to the differential geometric view-factor. Re-radiation and reflection are implicit.