Simplified Computation of Creep Damage Propagation

ABSTRACT:

The commonly accepted approach to dealing with material damage as the cause of structural failure is to treat the most highly distressed location in the structure as an equivalent simple test and to define failure of the structure as a whole as being failure at that point location.

The exception to this rule is plastic deformation. Yielding at a point was recognized several decades ago as being an excessively conservative definition of component failure and it is now standard design practice to accept failure as being the limit load, which is only reached, sometimes after extensive propagation of a plastic zone.

Other material failure mechanisms also occur after a finite period of damage propagation, but this additional strength, or life, is not usually taken into account, partly because the damage mechanisms themselves are not always well defined, and partly because of the computational difficulty involved in assessing the propagation of damage.

Creep rupture falls into the category of a mechanism which can enjoy an extensive period of damage propagation before structural failure occurs, but the difficulty of evaluating it quantitatively has meant that it continues to be dealt with as essentially a point failure phenomenon.

Relatively recently, many of the problems associated with assessing creep damage have been resolved, on the material side by increased use of so-called “continuum damage mechanics” based models such as Kachanov and Omega and, on the computational side, by the exponential growth in the capabilities of advanced Finite Element Analysis. It is now possible in principle to trace the entire life of a complex component, down to final disintegration. However, this capability still comes at a significant cost, and there is still room for simplification in order to bring this capability to a wider range of potential users.

This paper describes a process for evaluating the propagation of creep damage, down to the point of total disintegration, using approximations which exist within the standard capabilities of a typical FE design package. This innovation does not do anything that cannot be done today using the full repertoire of computational tools that exist, notably user subroutines, but provides a simpler platform which can be used to push damage evaluation further into the activities of day-to-day design with a significant reduction in the resource allocation currently required to do the job.

Results are compared with creep experiments on notched bars.

Keywords: creep, computation, damage propagation

Marriott, D., Stumph III, H., Sreeranganathan, A., Matice, C., “Simplified Computation of Creep Damage Propagation,” Paper No. PVP2016-6378, Proceedings of the ASME 2016 Pressure Vessels and Piping Conference PVP2016, July 17-21, 2016, Vancouver, British Columbia, Canada.