Wissenschaftliche Artikel der BAM
Filtern
Dokumenttyp
Sprache
- Englisch (3)
Referierte Publikation
- ja (3)
Schlagworte
- Stress relaxation (3) (entfernen)
Organisationseinheit der BAM
Paper des Monats
- ja (1)
This article refers to the research article entitled “Creep-Fatigue of P92 in Service-Like Tests with Combined Stress- and Strain-Controlled Dwell Times” [1]. It presents experimental mechanical data from complex service-like creep-fatigue experiments performed isothermally at 620 °C and a low strain amplitude of 0.2 % on tempered martensite-ferritic grade P92 steel. The data sets in text file format provide cyclic deformation (min. and max. stresses) and the total (hysteresis) data of all recorded fatigue cycles for three different creep-fatigue experiments: 1) a standard relaxation fatigue (RF) test with symmetrical dwell times of three minutes introduced at minimum and maximum strain, 2) a fully strain-controlled service-like relaxation (SLR) test combining these three-minute peak strain dwells with a 30-minute dwell in between at zero strain, and 3) a partly stress-controlled service-like creep (SLC) test combining the three-minute peak strain dwells with 30-minute dwells at constant stress. Such service-like (SL) tests with additional long-term stress- and strain-controlled dwell times are non-standard, rare, and expensive, making these data very valuable. They may be used to approximate cyclic softening in the technically relevant range, for the design of complex SL experiments, or for detailed analyses of stress-strain hystereses (e.g., for stress or strain partitioning methods, for the determination of hysteresis energies (work), inelastic strain components, etc.). In addition, the latter analyses may supply important input for advanced parametric lifetime modeling of components under creep-fatigue loading or model calibration parameters.
Creep-fatigue of P92 in service-like tests with combined stress- and strain-controlled dwell times
(2023)
Complex service-like relaxation- and creep-fatigue tests with strain- and stress-controlled dwells and fatigue cycle durations of approx. 2200 s were performed exemplarily on a grade P92 steel at 620 ◦C in this study. The results indicate deviations in the prevailing creep mechanisms of long-term relaxation and creep dwells, affecting subsequent dwells, load shifts, and the macroscopic softening behavior quite differently. In addition, fracture surfaces and longitudinal metallographic sections reveal intergranular crack growth for complex loading with stress-controlled dwells, whereas complex strain-controlled tests enhance oxidation and transgranular crack propagation. These findings substantiate the limited transferability of relaxation-fatigue to creep-fatigue conditions.
To improve the predictive capability of long-term stress relaxation of elastomers during thermo-oxidative ageing, a method to separate reversible and irreversible processes was adopted.
The separation is performed through the analysis of compression set after tempering. On the Basis of this separation, a numerical model for long-term stress relaxation during homogeneous ageing is proposed. The model consists of an additive contribution of physical and chemical relaxation.
Computer simulations of compression stress relaxation were performed for long ageing times and the results were validated with the Arrhenius treatment, the kinetic study and the time-temperature superposition technique based on experimental data. For chemical relaxation, two decay functions are introduced each with an activation energy and a degradative process. The first process with the lower activation energy dominates at lower ageing times, while the second one with the higher activation energy at longer ageing times. A degradation-rate based model for the evolution of each process and ist contribution to the total system during homogeneous ageing is proposed. The main advantage of the model is the possibility to quickly validate the interpolation at lower temperatures within the range of slower chemical processes without forcing a straight-line extrapolation.