Should we divide Creep Rupture up into (1) Time Dependent Plastic Collapse, trans-granular fracture and (2) Creep Failures, inter-granular fracture?
Trans-granular failures caused by time dependent plastic collapse are all too common in short to medium term uniaxial creep rupture tests. However, this failure does not tell us anything about the inter-granular failures that occur in real plant components. Could we improve the accuracy of design codes and life assessments procedures by analysing the test data and predicting the plant life using different models for:
(1) Time Dependent Plastic Collapse, trans-granular fracture and
(2) Creep Failures, inter-granular fracture?
Yes. I believe so. I have addressed this issue in my previous publications:
Wu, X. J, S. Williams, and D. Gong. 2012. A true-stress creep model based on deformation mechanisms for polycrystalline materials. J. Mater. Eng. Perform. 21(11): 2255‑2262.
And more so in my recent book:
Xijia Wu, Deformation and Life Evolution in Crystalline Materials--An Integrated Creep-Fatigue Theory, CRC Press, 2019.
The answer is yes, but not simply
Because of the shorter test life, we are probably not seeing the ingress of oxidation along grain boundaries we would expect from longer term tests that results in weakening and thus inter-granular failure.
This sub surface 'creep-corrosion' would impact creep rate and rupture time.
I think looking to reduction of area for existing data is going to be a credible approach most likely combined with end of life creep rate.
It now begs the questions:
Should we look at short terms tests and apply an oxidation-inter granular adjustment?
Are these modified values going to be better for extrapolating long term data?
Are we currently region splitting because of rupture mechanism instead of creep mechanism?
Over the past 20 years, I have been concentrating my FFS evaluations at the stresses applicable to the components operating in the creep range. The lower stress and longer creep rupture test times are much more relevant to predicting remaining lives in power piping systems (e.g., girth weldments subject to redistributed multiaxial stresses (axial, hoop, and through-wall redistribution of stresses).) Empirical calibration to actual creep rupture failures provides confidence in more accurate FFS methodologies. Appropriate extrapolations to lower stresses and longer times can reduce uncertainties. The application of appropriate boundary condition loads should consider significant unintented field anomalies.