Prediction of Long Term Stress Rupture Data for 2124 B. Wilshire a , H. Burt b and N.P. Lavery c Materials Research Centre, School of Engineering, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, United Kingdom a b.wilshire@swansea.ac.uk , b h.burt@swansea.ac.uk , c n.p.lavery@swansea.ac.uk Keywords: 2124, Creep Fracture, Creep Data Prediction. Abstract. The standard power law approaches widely used to describe creep and creep fracture behavior have not led to theories capable of predicting long-term data. Similarly, traditional parametric methods for property rationalization also have limited predictive capabilities. In contrast, quantifying the shapes of short-term creep curves using the θ methodology introduces several physically-meaningful procedures for creep data rationalization and prediction, which allow straightforward estimation of the 100,000 hour stress rupture values for the aluminum alloy, 2124. Introduction Because stress rupture tests are easier and cheaper to perform than creep tests, the long-term strengths of commercial alloys are usually defined by measuring the variations of the creep life (t f ) with stress (σ) at the relevant service temperatures (T). Yet, even when creep tests are carried out, in general, only the values of the minimum or secondary creep rate ( ε & m ) are also monitored, i.e. despite the distinctive shape of the normal creep strain/time curves displayed by most metals and alloys, the primary and tertiary stages are ignored. In both theoretical and practical studies, because t f often depends inversely on ε & m , the observed behavior patterns have been widely described using power-law equations of the form ) RT / Q exp( A t / M c n m f - σ = ε = & (1) where M, A and R are constants. However, using Eq. 1, large and variable values of the stress exponent (n) and the activation energy for creep (Q c ) are recorded for precipitation-hardened alloys. These observations are then commonly interpreted by assuming that different creep mechanisms become dominant in different stress-temperature regimes, but no agreement has been reached on the detailed mechanisms involved [1]. More importantly from an engineering perspective, the widespread adoption of power law approaches over almost half a century has not led to theories which can predict the properties expected during prolonged exposure at elevated temperatures. Similarly, the various parametric methods introduced in the 1950’s [2, 3] have limited predictive capabilities. Consequently, many tests lasting up to 100,000 hours have had to be completed to provide the required long-term design data for commercial aluminum alloys [4]. In contrast, using the Evans-Wilshire approach referred to as the θ Projection Concept [5, 6], long-term t f values independently determined for several aluminum alloys have been predicted accurately by quantifying the shapes of the normal creep curves obtained from tests lasting only a few thousand hours [7-10]. In addition, the θ equations relate directly to the deformation and damage processes controlling creep and creep fracture. As a result, the θ methodology allows rationalization of ε & m and t f data sets obtained over broad stress-temperature ranges [9, 10], with measurable properties replacing the empirical terms in conventional parametric methods. In turn, these straightforward rationalization procedures lead to physically-meaningful relationships [9], which are now shown to permit extended extrapolation of standard short-term t f measurements to predict the long-term behavior of 2124.