Creep-fatigue assessments require inputs for both creep damage and fatigue damage. Then these damage terms are combined using some form of interaction diagram to enable estimation of the creep-fatigue life. Similarly, creep-fatigue crack growth assessments need separate calculations of creep crack growth, fatigue crack growth and again possibly some allowance for interaction in order to obtain the total crack growth. A consequence of these approaches is that uncertainties in creep calculations, uncertainties in fatigue calculations and uncertainties in the form of interaction all lead to significant uncertainties in the overall lifetime or crack growth assessment.
This paper first briefly describes how creep-fatigue and creep-fatigue crack growth assessments are performed using the UK R5 procedure and contrasts the methods with those in other codes. Then, the paper presents the guidance in R5 on addressing the many uncertainties in these assessments and discusses the use of probabilistic methods in order to avoid over-conservative lifetime and crack growth estimates.