Technical Basis for Acceptance/Rejection Criteria for Flaws in High Pressure Gas Cylinder

The objective of this paper is to present the technical basis used for developing acceptance/rejection limits for seamless, high pressure gas cylinders that can be used at the time of retesting the cylinders. The development of acceptance/rejection limits for cylinders is done in three steps. First, the “critical flaw sizes” (e.g., depth and length or area) for selected types of flaws are established by an analysis procedure that has been verified by experimental tests. Next the “allowable flaw sizes” are calculated by modifying (reducing) the size of the critical flaw sizes for each cylinder by adjusting for fatigue crack growth that may occur during the use of the cylinder. Finally, the “acceptance/rejection criteria” is established to take into account other factors, such as all the expected operating conditions that the cylinders may see in service, and the reliability and detectability of the specific inspection equipment to be used and to adjust the allowable flaw sizes to provide an additional margin of safety. This acceptance/rejection limits have been incorporated in a recently published ISO Technical Report No. TR 22694:2008 (2007, “Gas Cylinders—Methods for Establishing Acceptance/Rejection Criteria for Flaws in Seamless Steel and Aluminum Alloy Cylinders at Time of Periodic Inspection and Requalification,” The International Standards Organization, Geneva, Switzerland, Technical Report No. 22694). In this work, the API 579 “Recommended Practice for Fitness-for-Service” (2000, API 579: Recommended Practice for Fitness-for-Service, 1st ed., American Petroleum Institute, Washington, DC) was used to calculate the critical flaw sizes for a range of cylinder sizes and strength levels. For this study, the critical flaw size is defined as the size of the flaw that will cause the cylinders to fail at the test pressure of the cylinder. The results of flawed-cylinder burst tests were used to experimentally verify the calculated critical flaw sizes. The allowable flaw sizes were then calculated by using well established fatigue crack growth rate data for steel and aluminum alloys to allow for the expected amount of fatigue crack growth that may occur during the specified retesting intervals. A limited number of tests was conducted to verify the allowable flaw size calculations. Further adjustments are made to the allowable flaw sizes to define the acceptance/rejection criteria to be used during cylinder retesting.