Potential for Standardisation of Techniques for Creep Testing of Internally Pressurised Tubular Components

Abstract

The design of components against creep is largely based on established methods which transform data obtained from conventional uniaxial creep tests into a form apparently acceptable for component design. It has long been established that there is not always a good relationship between uniaxial data and multiaxial data and in such cases tubular geometries should themselves be tested for reliable results. Indeed tests on tubular specimens reported by Clark (1928) pointed clearly to the advantages of tubular tests over conventional tests. Nevertheless, even with this early emphasis, the testing of tubular geometries has continued in the background with regard to usable data production since that time and it is only recently, with the aid of high precision design tools such as finite element analysis that very accurately designed components can be made which are of minimum investment level, maximum efficiency and of maximum potential life. Such design methods require data obtained from multiaxial tests and many facilities capable of such testing are being constructed or resurrected. However the development of testing standards for uniaxial creep testing which took place over the last 60 years has not been accompanied by the simultaneous and necessary development of standard approaches for internal pressure, multiaxial creep testing.