Abstract :
A generalised Monkman—Grant relation, which can be derived from the 4- projection
technique, is proposed and then used to predict creep properties at non accelerated test
conditions. In this generalisation, creep rates at low strains are used to predict minimum creep
rates that are then used in the Monkman-Grant relation to predict times to failure. Predictions of
creep properties for 1CrMoV from this generalisation and from the 4- projection technique
were assessed using the mean absolute percentage error (MAPE) and mean square error
(MSE)—which was further decomposed into systematic and random components. When
considering the accuracy with which minimum creep rates were predicted, all but the
generalised Monkman-Grant relation using 0.1% strain had a lower MAPE compared to the
4- projection technique. The generalised Monkman-Grant relation using 0.5% and 1% strains
had larger random components of the MSE compared to the 4- projection technique. When
considering the accuracy with which times to failure were predicted, all of the generalised
Monkman-Grant relations produced lower MAPE compared to the 4- projection technique.
However, only when creep rates were measured at 0.2% strain, did the generalised
Monkman-Grant relation produce prediction errors that had a significantly higher random
component. C 2006 Springer Science + Business Media, Inc
