COMPUTATIONAL CONTINUUM DAMAGED MECHANICS: ITS USE IN THE PREDICTION OF CREEP IN STRUCTURES - PAST, PRESENT AND FUTURE

摘要

The development is reviewed of the field of computational Creep Continuum Damage Mechanics. Emphasis is placed on four principal drivers: firstly, high-quality laboratory data; secondly, mechanisms-based constitutive equations and their calibration against laboratory data; thirdly, recent applications of computational CDM in structural analysis; and, fourthly, the rapidly increasing power of supercomputers and associated techniques. The need for high-quality laboratory data is expressed through an examination of the sources of errors in uni-axial creep testing; in particular, testpiece design and variations in test temperature. The importance is stressed of using mechanisms-based constitutive equations which reflect the underlying physics of the processes of deformation, damage, and failure. Strategies are also discussed for the identification of the dominant mechanisms for inclusion in the constitutive equation set and calibration against laboratory data. A recent Continuum Damage Mechanics solution is presented to the problem of failure in a welded pressure vessel obtained using two and three-dimensional finite element techniques. Finally, advances are addressed which are likely to take place over the next decade in both computer processor speed and in large-scale fast access computer memory. The impact of these likely developments is addressed on the ability to numerically solve problems in the creep of structures. Barriers to progress are identified and their solution discussed.