Creep rupture of structures

Abstract

A study is made of the failure times of structural components which operate at temperatures sufficiently high to cause material deterioration due to creep rupture. Expressions are derived which give lower bounds on failure times and which take into consideration the different stress criteria known to affect rupture mechanisms. The formulae are used to predict failure times of a variety of components, and it is found convenient, from a practical point of view, to express the times in terms of an equivalent representative rupture stress. By using this stress, failure times are obtained directly from uniaxial stress rupture data. It is found in the examples studied that the values for the representative rupture stress are almost independent of the constants used to define the deformation and rupture processes. Experimental evidence supports the prediction of the theory; for example, copper bars in torsion show better rupture characteristics than bars of aluminium alloy. The position is reversed in notched tensile specimens, with the aluminium specimens showing better characteristics than those of copper. It can be deduced that it is the form of rupture mechanism which affects behaviour rather than ductility as might be expected, since the creep ductility of the aluminium alloy is much less than that for copper.