Creep-Fatigue Fracture: Analysis of Internal Damage

This chapter first presents the engineering and scientific background on high-temperature creep-fatigue. Secondly, research and problems regarding conventional high-temperature creep-fatigue are pointed out. Finally, the three objectives and significance of this book are described.

This chapter presents analytical and experimental results on representative creep damage parameters for grain boundary cavities that initiate under creep-fatigue conditions. First, equations relating creep damage parameters to cavity and grain boundary distributions were derived using a probabilistic model. Then, the size and shape distribution of the grain boundary cavities that initiated under creep-fatigue conditions in SUS304, which is a typical polycrystalline heat-resistant steel, and their change from the initial to the middle of the life of the steel were shown. Based on the observations of cavities, four kinds of creep damage parameters (area fraction of cavities on cross section, fraction of cavities on grain boundary lines, areal cavity density, and A-parameter) were evaluated. For all damage parameters, the damage value increases with increasing life fraction. However, under creep-dominated conditions, where only the grain boundary is preferentially damaged, only the fraction of cavities on grain boundary lines has a clear physical meaning as creep damage.

This Chapter presents analytical results based on the experimental results and observations of cracks in creep-fatigue tests of a stainless steel SUS304. Two fracture morphologies were observed in high-temperature creep-fatigue fracture: inner cracking type fracture due to coalescence of small internal intergranular cracks and surface cracking type due to growth of small surface intergranular cracks. The fracture morphologies are strongly dependent on loading conditions. The inner cracking type fracture occurs at low tensile strain rate and high compressive strain rate, as well as at high temperatures, on the other hand, the surface cracking type fracture occurs under the opposite conditions. Three-dimensional fracture mechanism maps were created for tensile strain rate, compressive strain rate, and temperature to visualize the relationship between loading conditions and fracture morphologies. For the same strain range at high temperatures, the fatigue life at fatigue cracking type fracture is the longest, followed by the fatigue life at surface intergranular cracking type fracture and the fatigue life at internal intergranular cracking type fracture.

In this chapter, using austenitic stainless steel SUS304 and steam turbine rotor material 1Cr-1Mo-1/4 V forged steel, creep-fatigue interruption tests were conducted at high temperatures and low tensile strain rates. The small cracks that initiated and grew inside the specimens were observed on the longitudinal split sections along the stress axis, and their changes with time were clarified. In addition, the size of intergranular cavities, which are considered to be closely related to the inner cracking type creep-fatigue fracture, was measured. The main results obtained can be summarized as follows. Small inner cracks are initiated and grow with a temporal and spatial distribution on grain boundaries nearly perpendicular to the stress axis. However, there are few cases in which neighboring cracks coalesce more than one grain size (about 50 µm) distance apart in the direction of the stress axis. Small inner cracks initiate in the early stage of the lifetime. The density of the cracks increases rapidly after the mid-life. Three-dimensional observation of a 0.75N_f specimen of SUS304 shows that the areal crack density of the longitudinally split section is about 35 cracks/mm² at this point, but the volumetric crack density is approximately 1400 cracks/mm³, which is very high. Multiple cavities are initiated and grow in a series on the same grain boundary, and they merge to form a crack with the length of a grain boundary as a unit. The difficulty of crack initiation is almost independent of the grain boundary length, but cracks tend to be initiated at grain boundaries that are perpendicular to the stress axis. Crack initiation is dominant until late in the fatigue life. Small inner crack growth also progresses discretely with one grain boundary length as a unit due to the growth and coalescence of cavities on almost one grain boundary adjacent to the crack tip and is similar to crack initiation in mechanism. However, there is no significant difference in the crack length distribution up to the late fatigue life, and no significant crack growth is observed. At the late stage of fatigue life, the overall crack length distribution rapidly shifts toward the longer side, suggesting that a large number of small cracks distributed on almost the same cross section grow slightly to induce mutual coalescence and rapidly form a main crack.

In this chapter elucidates modeling and simulation methods of one-dimensional and three-dimensional models of random fracture resistance of grain boundaries for an initiation and propagation of multiple small cracks in creep-fatigue. In the one-dimensional model of random fracture resistance of grain boundaries, the concept of the model is explained, and it is shown that the one-dimensional model can be used to simulate the behavior of small cracks from the early to mid-life. The three-dimensional model incorporates both the microstructural sensitivity of small cracks and the interaction of initiation and propagation of multiple cracks, and thus, enables the simulation up to the later stage of the life, which cannot be conducted by the one-dimensional model. The crack initiation and propagation under a slow-tensile fast-compressive creep-fatigue at 923 K (650 °C) in a vacuum were simulated using the heat-resistant steel with the grain boundaries which were created based on a nucleation-growth model and subsequently whose fracture resistance was set by the model of random fracture resistance. Consequently, it became evident that the experimental results on the crack density, angular distribution of the initiated crack, crack length distribution, and crack growth rate were well simulated.

In this chapter, regarding small inner cracks in creep-fatigue of austenitic stainless steel SUS304, we proposed a numerical simulation method to estimate the spatial distribution of small inner cracks and its changes that cannot be directly and continuously observed, based on the observations of intergranular cracks on the stress axis longitudinal section of specimens obtained from interrupted tests. The small inner crack model, which is the basis of this simulation, is created by a computer model of a stack of planes (multiplanes) consisting of grain boundary facets projected on a plane perpendicular to the stress axis, and cracks are assumed to initiate and grow on grain boundary facets with a temporal and spatial distribution in units of grain boundary facets. By selecting an appropriate crack initiation driving force F and crack propagation driving force K, the crack density, mean crack length, and crack length distribution actually measured on a longitudinal section can be well reproduced from early to late life.

In this chapter, time-dependent and cycle-dependent crack propagation tests were conducted using notched specimens of SUS304 stainless steel in 4 major cases to investigate the effect of small inner cracks introduced by high-temperature cp-type creep-fatigue on macrocrack propagation. For the case of a macrocrack propagation in the zone containing no small crack (Case 1), the macrocrack propagation under cc-type and cp-type conditions is time-dependent, while the macrocrack propagation under pp-type and pc-type conditions is cycle-dependent. For the case of a macrocrack propagation in the zone containing uniformly distributed pre-small cracks (Case 2), the existing small inner cracks accelerate both time-dependent and cycle-dependent macrocrack propagation. For the case of a macrocrack propagation in the process zone with small crack nucleation (Case 3), many small intergranular cracks were observed near the main crack inside the specimen, and the macrocrack propagation rate was approximately 4 times higher than that of the time-dependent crack propagation of case 1. For the case of a macrocrack propagation combined with Case 2 and Case 3 (Case 4), the macrocrack propagation rate is almost equal to that in Case 3 under same conditions.

In this chapter, a damage healing due to compressive creep was examined by applying pc-type fatigue loading to Cr–Mo–V forged steel specimens in which cavities and small cracks were introduced to the inside by cp-type creep-fatigue tests. In the slow-tensile fast-compression (cp-type) creep-fatigue tests at high temperature, many small cracks and cavities initiate inside the specimen, and at the end of the life, they coalesce and grow, leading to rapid fracture. In the pc-type fatigue tests, small cracks and cavities do not initiate inside the specimen, and the cracks on the specimen surface grow stably and lead to fracture. After introducing damage to the specimen by loading it to 3/4 of the life of cp-type creep-fatigue, applying pc-type fatigue loading causes the small inner cracks and cavities to shrink and disappear. In other words, pc-type fatigue loading has the effect of healing the damage caused by cp-type fatigue. This healing effect appears even with a small number of cycles (5 cycles) of pc-type fatigue loading. Materials in which small inner cracks and cavities are annihilated and healed by pc-type fatigue loading exhibit a creep-fatigue life equivalent to that of virgin materials. The above-mentioned healing effect is thought to be due to the coalescence of the crack surfaces during compressive creep and the shrinkage of the cavity due to grain boundary diffusion under compressive stress.

This book mainly describes the fracture behavior of heat-resistant austenitic stainless steel SUS304 under high-temperature creep-fatigue, especially under the conditions that exhibit inner cracking type fracture, from the onset to the final failure. Typical damage of austenitic stainless steel SUS304 caused by creep-fatigue is grain boundary cavities, small intergranular cracks, and large cracks.