Elastic Plastic Pressure Vessel Analysis per ASME Section VIII Div 2
- Sadat Mohammed
- Nov 22, 2015
- 2 min read
Traditionally, when the computational resources were limited, the focus of pressure vessel FEA analysis was primarily linear static analysis with simplified material definitions using young’s modulus as the primary material behavior property. This approach is listed in detailed in ASME section VIII Div 2 and Div 1. As long as the stresses observed in FEA post process are within yield, the material behavior is well understood. Stress Linearization is accurate when classification lines are drawn through any thickness of the pressure vessel. But, as soon as some stresses exceed yield, the material behavior modeled linearly becomes erroneous.
Stress characterization through classification lines would provide membrane, membrane plus bending and peak stresses. Often analysts are required to classify stresses as primary or secondary stresses. There will be usually debate if the stress is self limiting or not. This question remains unanswered in many reports and assumptions are made.
With the advent of High performance computing, multi processor technologies and higher capacity of RAM in engineering computers, non linear analysis becomes feasible and may unanswered questions from the linear approach can be addressed.
In the Non linear analysis, the material behavior is modeled beyond the yeild point, thereby estimating the load carrying capacity of the structure with more confidence.
There are two load factor based approaches for modeling the nonlinear behavior listed in the ASME code. A elastic-fully plastic approach and elastic plastic approach with the plastic regions taking some load before collapse is predicted.
These approaches enables a more confident approach to pressure vessel analysis while identifying real issues and steering clear of over design.
Many of the Pressure vessel analysis gurus prefer the elastic plastic approach over linear approach due to the advantages of using proper material behavior model and lower post process time.
Comentários