Part 1:
Eighteen pressure vessel heads having nine different geometries and made from two different steels have been subjected to monotonically increasing external pressure until collapse occurred. Deflection of the crown and strains in the most highly strained regions of the concave side of the heads were recorded as functions of pressure. All heads underwent permanent deformations. Four modes of failure were observed: (i) symmetrical buckling of the crown, (ii) asymmetrical buckling in the knuckle and transition region, (iii) asymmetrical buckling of the cylindrical wall, and (iv) progressive deformation without buckling.

These test results were compared with predictions by formulas proposed by Slember and Washington, the ASME Code, and by the "first yield" criteria. Four new formulas were developed and compared with the test results and selected data from Bart.

One formula was recommended as an improvement to the ASME Code predictions as was the use of the first yield criteria.

It was recommended that further effort be made to determine a more rigorous theoretical basis for the collapse of pressure vessel closure heads under external pressure.

Part 2:
The concept that is considered here proposes that the creep behavior of a structure under a given loading can be related to the results of a single uniaxial creep test at what is called the reference stress of the structure for that loading. In this review we discuss the evolution of the method and the determination of reference stresses for creeping structures under steady and variable loadings, stress relaxation, creep rupture and creep buckling. The accuracy of the method is shown by means of comparisons to experimental results for both simple and complicated structures. Recommendations for future effort are given and an assessment of the applicability of the method to elevated temperature design analysis is discussed.