A fully implicit multi-axial solution strategy for direct ratchet boundary evaluation : theoretical development

Jappy, Alan and Mackenzie, Donald and Chen, Haofeng (2012) A fully implicit multi-axial solution strategy for direct ratchet boundary evaluation : theoretical development. In: 2012 ASME Pressure Vessel and Piping Conference, 2012-07-15 - 2012-07-20.

[thumbnail of Jappy_A_et_al_Pure_A_fully_implicit_lower_bound_multi_axial_solution_strategy_for_direct_ratchet_boundary_evaluation_implementation_and_comparison_Jul_2013.pdf] PDF. Filename: Jappy_A_et_al_Pure_A_fully_implicit_lower_bound_multi_axial_solution_strategy_for_direct_ratchet_boundary_evaluation_implementation_and_comparison_Jul_2013.pdf
Preprint

Download (821kB)

Abstract

Ensuring sufficient safety against ratchet is a fundamental requirement in pressure vessel design. Determining the ratchet boundary can prove difficult and computationally expensive when using a full elastic-plastic finite element analysis and a number of direct methods have been proposed that overcome the difficulties associated with ratchet boundary evaluation. Here, a new approach based on fully implicit Finite Element methods, similar to conventional elastic-plastic methods, is presented. The method utilizes a two-stage procedure. The first stage determines the cyclic stress state, which can include a varying residual stress component, by repeatedly converging on the solution for the different loads by superposition of elastic stress solutions using a modified elastic-plastic solution. The second stage calculates the constant loads which can be added to the steady cycle whilst ensuring the equivalent stresses remain below a modified yield strength. During stage 2 the modified yield strength is updated throughout the analysis, thus satisfying Melan’s Lower bound ratchet theorem. This is achieved utilizing the same elastic plastic model as the first stage, and a modified radial return method. The proposed methods are shown to provide better agreement with upper bound ratchet methods than other lower bound ratchet methods, however limitations in these are identified and discussed.