Elastic-Plastic Analysis of the PVRC Burst Disk Tests With Comparison to the ASME Code Primary Stress Limits

Abstract

This paper provides a comparison between finite element analysis results and test data from the Pressure Vessel Research Council (PVRC) burst disk program. Testing sponsored by the PVRC over 20 yr ago was done by pressurizing circular flat disks made from three different materials until failure by bursting. The purpose of this reanalysis is to investigate the use of finite element analysis (FEA) to assess the primary stress limits of the ASME Boiler and Pressure Vessel Code (hereafter the Code), and to qualify the use of elastic-plastic (EP-FEA) for limit-load calculations. The three materials tested represent the range of strength and ductility found in modern pressure vessel construction and include a low-strength, high-ductility material, a medium-strength, medium-ductility material, and a high-strength, low-ductility, low-alloy material. Results of elastic and EP-FEA are compared to test data. Stresses from the elastic analyses are linearized for comparison of Code primary stress limits to test results. Elastic-plastic analyses are done using both best-estimate and elastic-perfectly plastic (EPP) stress-strain curves. Both large strain-large displacement (LSLD) and small strain-small displacement (SSSD) assumptions are used with the EP-FEA. Analysis results are compared to test results to evaluate the various analysis methods, models, and assumptions as applied to the bursting of thin disks. The test results show that low-strength, high-ductility materials have a higher burst capacity than do high-strength, low-ductility materials. Linearized elastic FEA stresses and ASME Code primary stress limits provide excessive margins to failure for the burst disks for all three materials. The results of these studies show that LSLD EP-FEA can provide a best-estimate analysis of the disks, but the accuracy depends on the material stress-strain curve. This work concludes that SSSD EPP analysis methods provide a robust and viable alternative to the current elastic linearization method of satisfying the primary stress limits of the Code. [S0094-9930(00)01602-4]