Fatigue and creep-fatigue crack growth in alloy 709 at elevated temperatures

Fatigue crack growth (FCG) and creep-fatigue crack growth (CFCG) in austenitic stainless-steel Fe-25Ni-20Cr (Alloy 709) were measured experimentally and simulated using the finite element method. Temperature conditions investigated were 550°C, 600°C and 700°C, with load hold times of 0 s, 60 s and 600 s. Fracture surface was investigated using scanning electron microscopy and optical imaging. Experiments indicated that crack growth exhibits minimal sensitivity to the various loading conditions evaluated. At 600°C, crack growth rates were independent of hold time or loading frequency. At 700°C, there was a small increase in crack growth rate as a function of hold time, with a 600s hold time causing a factor of 2 increase in crack growth rate over FCG. Finite element simulations were performed to compute plasticity-induced crack closure in the presence of creep deformations at the crack tip. The simulations produced FCG and CFCG rates similar to the experimental results.     

Creep crack growth behavior of type 316L steel

The creep crack growth behavior of a 316L type stainless steel was studied at 600°C and 650°C using widely different specimen geometries. A strong emphasis was laid on crack initiation, early crack growth, and slow crack growth rates (a). The applicability of four load parameters, the stress intensity factor, the nominal stress, the reference stress, and the contour integral C^*, was investigated. Finite element calculations were made to compare the experimental load line displacement rates with the calculated values.It is shown that it is possible to correlate time to initiation (T_i) with C^*. The obtained experimental relation is not in full agreement with a theoretical expression derived from models for creeping solids. The reasons for this discrepancy are discussed. Furthermore, it is observed that there is no unique correlation between a and any of the four investigated load parameters, especially at low crack growth rates (a0^?3mm/h). At large crack growth rates (a?10^?2 mm/h) the apparent correlation between a and C^* is essentially due to the fact that the displacement rate is controlled by crack growth and not by the overall creep behavior of the specimens. A simple model based upon the data on crack initiation and the creep ductility exhaustion concept is shown to give results in reasonable agreement with the experiments.     

Influence of oxidation on fatigue crack initiation and propagation in turbine disc alloy N18

Fatigue crack initiation and propagation behaviour in subsolvus heat treated turbine disc alloy N18 has been assessed in air and vacuum at 650 and 725 °C under three-point loading. Fatigue crack initiation processes have been evaluated using single edge U-notch specimens under a 1-1-1-1 trapezoidal loading waveform along with interrupted tests at 650 °C to allow intermittent observations of the notch surface. The results show apparent grain boundary (GB) oxidation can occur under an oxygen partial pressure of 10⁻²–10⁻³ Pa. Cracks mainly initiate from grain boundaries or γ/γ′ interfaces due to the formation and subsequent cracking of Cr-rich and/or Co-rich oxides, and occasionally initiate from surface pores. Fatigue life in these tests appears to be dominated by this crack initiation process and is significantly reduced by increasing temperature and/or application of an oxidizing environment. Crack growth tests conducted under 1-1-1-1 and 1-20-1-1 loading waveforms indicate that oxidation significantly degrades the crack growth resistance of N18 and is associated with more intergranular fracture surface features. Additional oxidation effects on propagation caused by higher temperature or prolonging dwell time appear limited, whereas a prolonged dwell period seems to instead promote additional creep process, which further enhance crack growth, especially at higher temperature.     

On the influence of rubbing fracture surfaces on fatigue crack propagation in mode III

Fatigue crack growth has been studied under fully reversed torsional loading (R = ?1) using AISI 4340 steel, quenched and tempered at 200°, 400° and 650°C. Only at high stress intensity ranges and short crack lengths are all specimens characterized by a microscopically flat Mode III (anti-plane shear) fracture surface. At lower stress intensities and larger crack lengths, fracture surfaces show a local hill-and-valley morphology with Mode I, 45° branch cracks. Since such surfaces are in sliding contact, friction, abrasion and mutual support of parts of the surface can occur readily during Mode III crack advance. Without significant axial loads superimposed on the torsional loading to minimize this interference, Mode III crack growth rates cannot be uniquely characterized by driving force parameters, such as ΔK_III and ΔCTD_III, computed from applied loads and crack length values. However, for short crack lengths (?0.4 mm), where such crack surface interference is minimal in this steel, it is found that the crack growth rate per cycle in Mode III is only a factor of four smaller than equivalent behaviour in Mode I, for the 650°C temper at $\text{ΔK}{}_{\mathrm{III}}\text{= 45 MPa m}\text{1}\text{2}$.     

Modelling of high temperature fatigue crack growth in Inconel 718 under hold time conditions

Inconel 718 is a frequently used material for gas turbine applications at temperatures up to 650 °C. The main load cycle for such components is typically defined by the start-up and shut-down of the engine. It generally includes hold times at high temperatures, which have been found to have a potential for greatly increasing the fatigue crack growth rate with respect to the number of load cycles. However, these effects may be totally or partly cancelled by other load features, such as overloads or blocks of continuous cyclic loading, and the actual crack propagation rate will therefore depend on the totality of features encompassed by the load cycle. It has previously been shown that the increased crack growth rate found in hold time experiments can be associated with a damage evolution, where the latter is not only responsible for the rapid intergranular crack propagation during the actual hold times, but also for the increased crack growth during the load reversals. In this paper, modelling of the hold time fatigue crack growth behaviour of Inconel 718 has been carried out, using the concept of a damaged zone as the basis for the treatment. With this conceptually simple and partly novel approach, it is shown that good agreement with experimental results can be found.     

OXIDATION-ASSISTED FATIGUE CRACK GROWTH BEHAVIOR IN ALLOY 718-PART II. APPLICATIONS

In previous work by the authors, a quantitative model based on a two-stage oxidation mechanism was developed to describe the oxidation-assisted crack growth behavior in alloy 718. This model is used here to predict the crack growth rate in this alloy at 650°C for two different loading conditions: one is a continuous cycle with a hold time duration of 300 s imposed at minimum load, the other is a continuous cycle without a hold time duration. The results obtained from applying the model to these loading profiles were then compared with those obtained experimentally. Good agreement was observed between the two data sets. Details of the model calculations are discussed, and suggestions to further extend the model capabilities are made in this paper.     

Creep crack growth behavior in 316 stainless steel at 594°C (1100°F)

The creep crack growth behavior of a type 316 stainless steel was characterized at 594°C (1100°F) using precracked single edge notch specimens loaded in displacement rate control. The steady-state crack growth rate, da/dt, correlated with J-integral and did not correlate with C ^*. The creep crack growth behavior in this material and temperature is compared with our previous creep crack growth rate data on a Cr-Mo-V steel at 538°C (1000°F) and on type 304 stainless steel at 594°C in which da/dt correlated with C ^*. A detailed discussion is included on why in some materials creep crack growth rate correlates with J integral and in others it correlates with C ^*.     

Accounting for crack tip cyclic plasticity and creep reversal in estimating (Ct)avg during creep‐fatigue crack growth

Creep‐fatigue crack growth (C‐FCG) rates in a P91 steel at 625°C were correlated as the average time rate of crack growth during hold time, (da/dt)_avg , with (C_t)_avg. At 60‐second hold time, the rates were lower than for 600‐second hold time. At 600‐second hold time, the crack growth rates converged on to the creep crack growth rate (CCGR) trend. Thus, the CCGR trend represents the upper bound for time‐dependent crack growth rates in P91 materials. The analytical expressions based on considering just the elastic and secondary creep deformation rates overestimated the magnitudes of (C_t)_avg by as much as a factor of 10 for the 600‐second hold time tests. After accounting for the effects of cyclic plasticity during unloading, and accounting for only partial reversal of creep strains accumulated during hold time, the estimates of (C_t)_avg compared well with the measured values. C_R represents the extent of crack tip creep strain reversal, and t_pl is the time required for the crack tip creep zone during the hold time to become equivalent in size to the cyclic plastic zone in terms of stress carried by that region. Together, these parameters accurately account for the effects of crack tip cyclic plasticity on the magnitude of (C_t)_avg. Both t_pl and C_R depend on material properties, and the latter also depends on the hold time. A parameter ? is introduced that is dependent only on material properties and from which C_R can be estimated for a given hold time. t_pl and ? can be reported as part of the test results from C‐FCG testing.     

Improving resistance to dynamic embrittlement and intergranular oxidation of nickel based superalloys by grain boundary engineering type processing

Abstract

Polycrystalline nickel based superalloys are prone to grain boundary attack by atmospheric oxygen either in the form of time dependent intergranular cracking during dwell time within a low cycle fatigue loading spectrum, known as hold time cracking, or in the form of intercrystalline oxidation at higher temperatures. In the case of hold time cracking of IN718 it has been shown that the crack propagation velocity is determined by local microstructure and environmental conditions, reaching values up to 10 μm s^?1 under four-point bending conditions at 650°C in air. The governing mechanism for this kind of time dependent quasi-brittle intergranular failure has been recognised to be 'dynamic embrittlement', i.e. diffusion of the embrittling element into the elastic stress field ahead of the crack tip, followed by stepwise decohesion. In a very similar way to intercrystalline oxidation, this damage mechanism seems to depend on the local microstructure. Assuming that oxygen grain boundary diffusivity is particularly slow for special coincident site lattice (CSL) grain boundaries, bending and oxidation experiments were carried out using specimens that underwent successive steps of deformation and annealling, i.e. grain boundary engineering. It has been shown that an increase in the fraction of special CSL grain boundaries yields a higher resistance to both intercrystalline oxidation and hold time cracking by dynamic embrittlement.     

Langzeitiges Kriechrißverhalten kennzeichnender Kraftwerksstähle

Long Term Creep Crack Behaviour of Typical Power Plant Steels The creep crack behaviour of the steels was investigated in a wide loading range up to a test duration of 40 000 h and down to a creep crack growth rate of 2 · 10^?5 mm/h with specimens of different shape and size. For steels of type l%Cr-l%Mo-0.6%Ni-0.3%V, 1%Cr-0.9%Mo-0.7%Ni-03.%V, 12%Cr-1%Mo-0.3%V-0.22%C and 12%Cr-l%Mo-0.3%V-0.20%C tested at 550°C, the creep crack growth rate could be described by the parameter C₂* with significantly smaller scatter bands than by the parameter C₁* or the stress intensity factor K_I. For steel 12%Cr-2%Ni-1%Mo tested at 450°C, parameter K_I leads to the smallest scatter band. The creep crack initiation can be described in a two-criteria diagram based on nominal stress and stress intensity factor. However the method is assumed to be over-conservative in case of increasing specimen size. As a result of several aperiodic creep fatigue crack tests, precracking under fatigue conditions gave a weak increase of the creep crack growth rate whereas by precracking under creep conditions the fatigue crack rate was strongly decreased.     

Prediction of Fatigue Lives of MAR‐M247 LC Based on the Crack Closure Concept

Low cycle fatigue (LCF), high cycle fatigue (HCF), and combined LCF and HCF tests are carried out on MAR‐M247 LC at 650 °C in air environment. Under combined LCF and HCF loading, block striations form on the fracture surface which are used to complete an effective crack growth curve by using the linear summation model. Crack growth lives starting from equivalent initial flaw sizes are calculated by the crack closure code FASTRAN and compared with experimental fatigue lives. Under HCF loading, predicted and experimental fatigue lives agree well for lifetimes above 10⁵ cycles. Lower lifetimes are overestimated indicating that the linear summation model is not valid for MAR‐M247 LC in this loading range. Interactions between the non‐crystallographic HCF crack growth and striated crack growth that is caused by the LCF loading are probably responsible for this behavior.     

Creep crack initiation and growth in AISI 316 (N) weld – FEM and experiments

Abstract

There are two aspects of the creep crack growth behaviour, namely, the crack initiation and the crack propagation. An incubation period is often observed prior to the onset of creep crack growth. In this study, creep crack initiation and propagation in pre-cracked compact tension (CT) specimens of a 316 (N) stainless steel weld at T = 550 and 625°C under static loading is investigated. Both the crack initiation time and the crack growth rate are measured as a function of fracture parameter C*. It is shown that it is possible to correlate the creep crack initiation time with the C* parameter. It is also shown that the creep crack growth rate can be correlated with the C* integral. Additionally, finite element analyses by using the ANSYS software have been performed at one test condition (T=625°C) in order to estimate numerically the crack mouth opening displacement rate history for a propagating crack using the node release technique. When the FEM results are compared with the experimental data, the results show a very satisfactory prediction capability.     

Impact of compressive hold on out‐of‐phase thermomechanical fatigue crack growth in IN 718

Accurate characterization and understanding of the fatigue crack growth behaviour of components in jet turbine engines is critical for successfully using a damage tolerant design method to maximise safety and efficiency. The hot section components experience changing loads and temperatures, and hence, fatigue crack growth rates are typically studied under thermomechanical loading. One question that remains unclear is the role of the compressive holds that are often part of an aircraft loading‐temperature spectrum. This experimental study was undertaken to investigate a turbine disk alloy, Inconel 718, subjected to different cycling and temperature profiles considering different lengths of hot compressive holds to determine its effect on the fatigue crack growth rate. It was found that the addition of a compressive hold at temperatures from 650 to 725 °C has no significant impact on the fatigue crack growth rate when compared with a cycle without a compressive hold. Fractographic analysis shows that crack growth is primarily transgranular in all cases studied suggesting that grain boundary oxidation, often observed during hot tensile holds, is insignificant.     

INFLUENCE OF TEMPERATURE ON FATIGUE CRACK GROWTH BEHAVIOUR OF STEELS AT ULTRASONIC FREQUENCY

Fatigue crack growth rates have been measured at 22 kHz for two types of carbon steel, between 20 and 500°C under symmetrical push-pull loading. The crack length was determined from the decrease in the resonant frequency of the specimen. For ?SN 412013 steel, an increased crack propagation resistance was observed between 250 and 300°C. The values of the constants C and n in the Paris equation, and K_a,th are dependent on temperature. The fatigue crack growth rate and C both increase with temperature, while n and K_a,th decrease with increasing temperature. Electron scanning and light microscopy have shown that intercrystalline fracture does not occur in ?SN 415313 steel at elevated temperature. Intercrystalline fracture was observed for specimens fatigued at 20°C and also in ?SN 412013 steel at temperatures of 200 and 500°C. The width of the plastic zone and the height of the surface relief around the fatigue crack increased with temperature.     

Fatigue and creep crack growth behaviour of Inconel MA 6000

Crack growth in MA 6000 under cyclic loading was studied at 24, 760, and 1000°C and under static loading at 1000°C in two matenal onentatwns. Correlatwns of fattgue crack growth rate with parameters ?K and ?J were examined. Also comparisons were made of experimental and predicted growth rates.

The rate of growth was influenced by temperature and onentatwn m addttwn to the loading mode. Fatigue crack growth rate generally increased with temperature. However in the L-T orientation at 1000°C secondary cracks developed perpendtcular to the primary crack and significantly altered its behaviour. Creep crack growth at 1000°C was strongly orientation dependent, mainly due to secondary crackmg m the L-T oriented specimen in the direction perpendicular to the main crack.

Fracture surfaces were examined by scanning electron microscopy. Also, comparisons were made between crack growth behaviour of MA 6000, MA 754 and MA 956.     

BEHAVIOUR OF A 1Cr-1Mo-0.25V STEEL AFTER LONG-TERM EXPOSURE-II. CREEP CRACK INITIATION AND CREEP CRACK GROWTH

Both the initiation and the propagation of creep cracks have been studied in a 1Cr-1Mo-0.25V steel at 550°C using CT type specimens. The material taken from a large turbine casing was investigated in two conditions: (i) unaged and (ii) after a long exposure in-service time of about 150,000 h at 540°C. In both cases the material was found to be creep ductile, which is justified in terms of fracture mechanics applied to creeping solids. It is shown that fracture mechanics is unable to provide unique correlations with global load-geometry parameters, either K or C* for all the stages of both crack initiation and crack growth. However there exists a unique correlation between C* and the time to initiation, t_i. This correlation does not depend on the initial conditions of the material. During crack growth two stages are defined. Stage I is a transient regime in which C* is almost constant, but the correspondence between the crack growth rate and C* is not unique since largely dependent on the initial loading applied to the specimens. It is shown that the apparent correlation between the crack propagation rate in stage II which corresponds to large crack growth rate is doubtful. A simplified method based on reference length and reference stress is used to calculate C* parameter and to simulate the load-line displacement rate. The results obtained from this method are compared to those derived from finite element calculations published in the literature.     

Crack growth in a new nickel-based superalloy at elevated temperature

Crack growth at elevated temperature has been examined in a new fine-grained nickel-based superalloy under triangular, fast-slow, slow-fast, dwell and sustained loading conditions at 650 and 725C. The effect of loading waveform seems to be minimal for base frequencies over 0.01 Hz with a mixture of time and cycle dependent crack growth observed for all but the fast-slow waveform, where the crack growth remained cycle-dependent and the crack growth rate mostly constant. For base frequencies less than 0.01 Hz, crack growth under dwell load clearly accelerated and the crack growth rates were comparable with those under sustained load. Creep contribution was found to be negligible while crack tip constraint may be relevant to the out-of-plane crack growth observed under predominantly sustained load conditions.     

Brittle intergranular fracture at elevated temperatures in low–alloy steel

Abstract

Recent studies of stress-relief cracking in low-alloy steels have focused attention on a novel mode of brittle intergranular fracture which occurs at elevated temperatures (300–650°C) in hard, coarse–grained heat–affected–zone microstructures. Fracture initiates at stress concentrators such as sharp cracks or inclusions, and can propagate under static loading at rates of 10^?11?10^?5 ms^?1 to produce intergranular facets with very little associated plastic deformation. The stress-intensity parameter K has been used to characterize crack growth, and three regimes of behaviour have been observed: (i) a threshold region at growth rates of 10^?11?10^?10 m S^?1, (ii)a plateau region, in which growth rates are independent of K between 10^?10 and 10^?8 m S^?1, and (iii) a region of highly K-sensitive crack growth between 10^?8 and ^?5 m S^?1. Independent Auger electron spectroscopy analyses have demonstrated that sulphur segregates locally to the high-temperature crack tip, giving rise to the embrittlement of a limited area of grain boundary. Together with other presegregated solutes, this enables brittle fracture to occur at high temperature, and the transfer of sulphur to the crack tip controls the rate of crack growth. Two models describing crack-tip sulphur segregation are currently proposed. In the first model, a quantitative analysis demonstrates that the crack-tip stress field will drive undersize solute atoms such as sulphur to the physical crack tip. In the second, the intergranular crack is modelled as a sharp cavity. Grain-boundary sulphides which are exposed by cavity formation become unstable and dissolve, saturating the cavity surface with sulphur, which is then drawn into the tip as part of the cavity growth process.

MSTj77     

Effect of dynamic strain aging on fatigue crack growth behaviour of reactor pressure vessel steels

Abstract

Fatigue tests under constant amplitude load were conducted on compact tension specimens of SA533B3 steels with four levels of sulphur content at different temperatures. A modified capacitance type crack opening displacement (COD) gauge was shown to be suitable for fatigue crack length measurement at high temperatures. Test results obtained with different measurement techniques show good consistency. The observation that the Young's moduli measured at a strain rate of 4 × 10^?3 s^?1 for the SA533B3 steels at 150 and 300°C do not decrease with increasing temperature seems to be related to the presence of dynamic strain aging. The fatigue crack growth rates at 150 and 300°C are about two and half times slower than those tested at 400°C because dynamic strain aging prevails at 150 and 300°C. Fractographic examination results suggest that inclusions embedded in secondary cracks enhanced the fatigue crack initiation rather than the fatigue crack growth.