Author Description

This report is based on the results and experience gained in assessing both public domain and MATTER data, some previously reported in the MATTER deliverable D4.5: "Creep-fatigue interaction rules for P91" and some assessed here. A number of methods, including interaction diagram based methods and simplified methods, have been compared for predicting the creepfatigue life of P91 steel. The effect of cyclic softening on creep properties have been considered in the evaluations presented in this report. The purpose of this report is to give recommendations for creep and creep-fatigue assessment for components made of X10CrMoVNb9-1 (P91) steel. Based on the conclusions of the assessments the following recommendations are given: 1.^The creep properties of P91 steel suffer from cyclic loading and it is recommended that the current RCC-MRx creep strain equations should be modified or replaced by a model that can implement time factors or stress reduction factors to correct for softening. Suitable models have been identified. The softening of P91 potentially causes unconservativeness in significant creep conditions where accumulated strain is a limiting design factor. 2. The interaction diagram models currently applied in design codes are complicated to apply and include several complex modelling challenges. It has been recommended that alternative modelling concepts, such as using simplified models should be considered for use in design codes. The simplified models optimized for P91 have been shown to give good and robust predictions on cycles to failure. 3. Currently available P91 data mainly consists of large strain range and short hold period data, where creep is mainly causing additional strain.^Data with verified creep cavitation damage is limited or totally lacking. It is recommended that more data is generated at low strain ranges and long hold times to improve long term extrapolation robustness. Tensile property data for softened material is also needed to determin