Blade Alloy Research Articles

Thresholds for high-cycle fatigue in a turbine engine Ti–6Al–4V alloy

The characterization of critical levels of microstructural damage that can lead to fatigue-crack propagation under high-cycle fatigue loading conditions is a major concern for the aircraft industry with respect to the structural integrity of turbine engine components. The extremely high cyclic frequencies characteristic of in-flight loading spectra necessitate that a damage-tolerant design approach be based on a crack-propagation threshold, Δ K TH. The present study identifies a practical lower-bound large-crack threshold under high-cycle fatigue conditions in a Ti–6Al–4V blade alloy (with ∼60% primary α in a matrix of lamellar α+β). Lower-bound thresholds are measured by modifying standard large-crack propagation tests to simulate small-crack behavior. These techniques include high load-ratio testing under both constant- R and constant- K max conditions, performed at cyclic loading frequencies up to 1 kHz and R-ratios up to 0.92. The results of these tests are compared to the near-threshold behavior of naturally-initiated small cracks, and to the crack initiation and early growth behavior of small cracks emanating from sites of simulated foreign object damage.

High‐cycle fatigue of Ti–6Al–4V

The definition of the critical levels of microstructural damage that can lead to the propagation of fatigue cracks under high‐cycle fatigue loading conditions is a major concern with respect to the structural integrity of turbine‐engine components in military aircraft. The extremely high cyclic frequencies characteristic of in‐flight loading spectra, coupled with the presence of small cracks resulting from fretting or foreign object damage (FOD), necessitate that a defect‐tolerant design approach be based on a crack‐propagation threshold. The present study is focused on characterizing such near‐threshold fatigue‐crack propagation behaviour in a Ti–6Al–4V blade alloy (with ~60% primary α in a matrix of lamellar α + β), at high frequencies (20–1500 Hz) and load ratios (0.1–0.95) in both ambient temperature air and vacuum environments. Results indicate that ‘worst‐case’ thresholds, measured on large cracks, may be used as a practical lower bound to describe the onset of naturally initiated small‐crack growth and the initiation and early growth of small cracks emanating from sites of simulated FOD.

Age-hardening behavior of Inconel X-750 superalloy

In order to develop highly efficient aircraft engines, the use of alloys for turbine blades that combine high creep resistance and lower density brings substantial benefits. That may be achieved by the introduction of a reinforcing phase, such as a discontinuous reinforcement. So far, most researched particle-reinforced Ni-based superalloys made use of sub-micrometer particles, such as oxides and borides, often produced by in-situ reactions. These are proposed solely as dislocation immobilizers, in a role comparable to that carried out by γ’ precipitates, while metal matrix composites (MMCs) containing micro-scale particles offer a load-transfer potential. To combine both approaches, a recently designed polycrystalline Ni-based γ’-strengthened superalloy reinforced with TiC particles was produced by low energy mixing of powders, followed by pressure sintering. The composite, along with a non-reinforced reference, was submitted to compression creep tests at 973 K (700 °C), with stresses varying from 280 to 500 MPa. Selected specimens had their microstructure evaluated with techniques such as EBSD and XRD. When simulating a working turbine, the material density influences directly the centripetal forces acting on a section of a turbine blade, and consequently the creep rates. In the proposed composite, not only an increase in the threshold stress by a load-transfer component was obtained, but also the density reduction affected remarkably the creep rates. Moreover, the presented MMC not only presented a superior creep resistance in relation to the non-reinforced alloy, but also exhibited higher stability of primary and secondary γ’ volumes after 500 h in creep tests.

GTD111 Alloy Material Study

Very few property data on this common turbine blade alloy have been published. As longer hours of service are accumulated, maintenance considerations such as developing optimum component life strategies and repair processes become important. The lack of specific material data hampers the effort of users and repair facilities to achieve optimum service from this alloy. This study measured some of the basic mechanical and metallurgical characteristics of this polycrystalline nickel base superalloy. Tensile and short-term creep rupture properties as well as microstructural and fracture characteristics are presented. Both the as-heat-treated and thermally exposed characteristics at two different temperatures are examined.

The fatigue and fracture of turbine blade alloys in high-pressure hydrogen : DeLuca, D.P. and Cowles, B.A. JOM (1996) 48 (11), 61–64

The roles of microstructure and environment in influencing mechanisms of damage during high-temperature, low cycle fatigue were investigated in a lithium-containing aluminium alloy in the peak-aged condition, and tested over a range of plastic strains. Reduction in fatigue life and resistance were observed at an elevated test temperature. Localized oxidation and embrittlement at grain boundaries are promoted by the applied stress, and play an important role in accelerating crack initiation and subsequent crack propagation. Evidence is presented to rule out contributions from microstructural changes through the coarsening of precipitate particles and the formation of new precipitate phases, and damage due to creep. The apparent acceleration of crack initiation and enhanced crack propagation are ascribed to a complex interaction of four mechanisms: intrinsic microstructural features, the embrittling effect of the environment, the slip mode arising from a combination of mechanical and microstructural contributions, and applied stress. The abnormal plastic strain-fatigue life by behaviour is attributed to differences in the distribution of deformation as a function of plastic strain amplitude.

Development of alloyed and dispersion-strengthened MCrAlY coatings

An on-going research program has investigated the effects of the separate and combined additions of Ta, Pt and an oxide dispersion to basic MCrAlY overlay coatings. The goals were improved oxidation and hot corrosion resistance, greater diffusional stability, and coating thermal expansion reduction in order to be closer to nickel base turbine blade alloys. It was found in cyclic burner rig testing at 1050°C that each individual addition improved resistance of the baseline compositions. Oxide and Pt additions reduced thermal expansion. Creep resistance was greatly increased by oxide additions, which also strongly increased yield strength. Yield strength was modestly increased by Pt at low temperature and significantly increased by Ta at low to moderate temperatures.

The fatigue and fracture of turbine blade alloys in high-pressure hydrogen

In June 1995, the NASA space shuttle Discovery lifted off for the first time with an improved high-pressure oxidizer turbopump fitted to one of its three main engines. In 1997, an advanced high-pressure fuel turbopump will enter service as well. Both systems will make extensive use of nickel-based superalloy microstructures specially tailored for improved resistance to hydrogen embrittlement. This article reviews a study on cast single-crystal and equiaxed nickel superalloys and the subsequent approach used to improve their hydrogen resistance.

Coatings for the Protection of Turbine Blades From Erosion

Many types of turbines, including aircraft gas turbines, steam turbines, and power recovery turbines, suffer from solid particle erosion caused by a variety of materials ingested into the machines. Utilization of various laboratory erosion tests tailored to the specific application by using various erodents, temperatures, velocities, and angles of impact, have been shown to be effective in the development and selection of coatings for the erosion protection of turbine blades and other components. Detonation gun coatings have demonstrated their efficacy in providing substantial protection in many situations. It has now been shown that several tungsten carbide and chromium carbide Super D-Gun™ coatings not only have better erosion resistance than their D-Gun analogs, but cause little or no degradation of the fatigue properties of the blade alloys. Nonetheless, caution should be employed in the application of any laboratory data to a specific situation and additional testing done as warranted by the turbine designer.

The application of the Graham and Walles creep equation to aeroengine superalloys

The paper briefly discusses the charcteristics of creep deformation and the requirements of any creep equation developed to quantify this behavior. It is shown how the Graham and Walles methodology (1)† fulfils the identified objectives and how the equations have been applied to simple and complex variable loading situations. By reference to an appropriate blade alloy it will be shown that the approach can be extended to produce deformation maps and the results will be compared with theoretical creep models such as dislocation shearing of precipitate particles. Finally it is shown how the Graham and Walles model has been applied to single crystal materials to predict creep in randomly oriented crystals.

Dislocation behavior and surface morphology of high purity copper single crystals fatigued at low-strain amplitude at elevated temperatures: Hidaka, K. Diss. Abstr. Int. (Dec. 1992) 53 (6), 135 pp

The characterization of critical levels of microstructural damage that can lead to fatigue-crack propagation under high-cycle fatigue loading conditions is a major concern for the aircraft industry with respect to the structural integrity of turbine engine components. The extremely high cyclic frequencies characteristic of in-flight loading spectra necessitate that a damage-tolerant design approach be based on a crack-propagation threshold, ΔKTH. The present study identifies a practical lower-bound large-crack threshold under high-cycle fatigue conditions in a Ti–6Al–4V blade alloy (with ∼60% primary α in a matrix of lamellar α+β). Lower-bound thresholds are measured by modifying standard large-crack propagation tests to simulate small-crack behavior. These techniques include high load-ratio testing under both constant-R and constant-Kmax conditions, performed at cyclic loading frequencies up to 1 kHz and R-ratios up to 0.92. The results of these tests are compared to the near-threshold behavior of naturally-initiated small cracks, and to the crack initiation and early growth behavior of small cracks emanating from sites of simulated foreign object damage.

A contribution to the problem of the endurance limit of nitrided service parts: Kern, T.-U. and Spies, H.-J. Proc. Conf. Surface Engineering, Bremen, Germany, 1993, pp. 187–189

Mixed-mode, high-cycle fatigue-crack growth thresholds are reported for through-thickness cracks (large compared to microstructural dimensions) in a Ti–6Al–4V turbine blade alloy with a bimodal microstructure. Specifically, the effect of combined mode I and mode II loading, over a range of phase angles β=tan−1(ΔKII/ΔKI) from 0° to 82° (ΔKII/ΔKI∼0–7), is examined for load ratios (ratio of minimum to maximum loads) ranging from R=0.1 to 0.8 at a cyclic loading frequency of 1000 Hz in ambient temperature air. Although the general trend for the mode I stress-intensity range at the threshold, ΔKI,TH, is to decrease with increasing mode mixity, ΔKII/ΔKI, and load ratio, R, if the crack-driving force is alternatively characterized in terms of the strain-energy release rate, ΔG, incorporating contributions from both the applied tensile and shear loading, the threshold fatigue-crack growth resistance increases significantly with the applied ratio of ΔKII/ΔKI. The pure mode I threshold, in terms of ΔGTH, is observed to be a lower bound (worst case) with respect to mixed-mode (I+II) behavior. These results are compared with mixed-mode fatigue thresholds for short cracks, where the precrack wake has been machined to within ∼200 μm of the precrack tip. For such short cracks, wherein the magnitude of crack-tip shielding which can develop is greatly reduced, the measured mixed-mode fatigue-crack growth thresholds are observed to be markedly lower. Moreover, the dependence of the mixed-mode fatigue-crack growth resistance on the applied phase angle is significantly reduced. Comparison of the large- and short-crack data suggests that the increase in the large-crack fatigue threshold, ΔGTH, with an increasing mode mixity (ΔKII/ΔKI) is largely due to shielding from shear-induced crack-surface contact, which reduces the local crack-driving force actually experienced at the crack tip. Quantification of such shielding is described in Part II of this paper.

Fatigue Crack Initiation In 403 Stainless Steel In Simulated Steam Cycle Environments: Hydroxide And Silicate Solutions

AbstractThe fatigue crack initiation behaviour of a 12%Cr turbine blade alloy (403 stainless steel) was studied at a single alternating stress level of 448 MN m−2 in deaerated solutions of NaOH at pH 10,12, and 14 and in 1M Na4SiO4 solution (pH 10) at 100°C. Fatigue life in the hydroxide solution increased with increasing pH and was also good in the silicate solution. Long cracks of similar appearance were found to be initiated on the surface of the steel in all the solutions except NaO H at pH 14, in which short incipient microcracks of a transgranular nature initiated on the specimen surface, regardless of the selective attack that occurred at grain and phase boundaries. It was found that increasing the pH of the hydroxide solution and the presence of the silicate anion both cause the initiated cracks to propagate into the bulk of the material in a transgranular mode, thereby resulting in a longer fatigue life.

The root area parameter model for evaluation of effects of various artifical small defects and mutual interaction of small defects on fatigue limit: Toriyama, T. and Murakami, Y. J. Soc. Mater. Sci. Jpn (Oct 1993) 42(481), 1160–1166 (in Japanese)

There have been few studies to date describing fatigue-crack propagation thresholds under mixed-mode loading conditions in the presence of cracks, that are small as compared to the characteristic microstructural dimensions. To address this need, the variation in mixed-mode, high-cycle fatigue-crack growth thresholds with crack size is reported for a Ti-6Al-4V turbine blade alloy with a fine-grained bimodal microstructure. Specifically, threshold behavior is examined for large through-thickness cracks >4 mm in length), short through-thickness cracks (~200 μm in length), and microstructurally-small surface cracks (10–50 μm in diameter) under combined mode I and mode II loading at load ratios (ratio of minimum to maximum load) ranging from 0.1 to 0.8. For mode-mixities ranging from pure mode I to predominantly mode II, large crack, mode I ΔKI,TH thresholds were found to decrease substantially with increasing phase angle. However, by characterizing in terms of the range in strain energy release rate, ΔGTH, incorporating both mode I and mode II contributions, it was observed that the pure mode I threshold could be regarded as a ‘worst case’ under mixed-mode loading in this alloy. By estimating the effective crack-driving force actually experienced at the crack tip, the observed increase in the mixed-mode ΔGTH threshold with mode-mixity was attributed to an increasing influence of crack-tip shielding due to crack closure and crack-surface interference. Equivalent thresholds for through-thickness short cracks, where the crack wake and hence the effect of such shielding is minimized, were consequently far less sensitive to mode-mixity and corresponded in magnitude to the shielding-corrected large-crack thresholds. This effect was accentuated for the measured thresholds of microstructurally-small surface cracks; such small-crack, mixed-mode ΔGTH thresholds not only displayed a minimal influence of mode-mixity but were up to two orders of magnitude smaller than those for corresponding large cracks at the same load ratio and mode-mixity conditions.

Test methods for accelerated development, design and life assessment of high-temperature materials

A new conceptual framework involving an integrated approach to materials development, component design and life prediction for high-temperature applications is outlined. Termed ‘Design for Performance’, this provides an alternative to the traditional approach which requires the creep to rupture testing of many specimens over long periods. Accelerated development of new alloys for higher-temperature turbine blades and discs, and assessment of remaining life of operating components, is now possible. The tests described are a refined stress relaxation test — from which extensive stress versus strain-rate data may be generated over a range of temperatures — and a constant displacement rate crack growth test. These tests provide a separation of the creep strength and fracture resistance criteria required in design. Extensive studies on the cast alloy IN738 have shown that the creep strength is quite insensitive to prior exposure and microstructure, but fracture resistance is very sensitive to thermal history. Relative to the standard heat treatment, fracture resistance can be seriously impaired or substantially improved. It is also shown that calibration with traditional design methods for creep strength is possible. Once minimum performance standards have been established for the original design, these may be used as a basis for life management. For example, decisions regarding component replacement, repair or refurbishment may be made as one or more of these minimum performance standards are approached during service. Thus, the methodology provides a consistent basis for life management for both new and serviced components. It is argued that not only is this approach much faster, more efficient and less costly, but that the framework on which it is based is technically superior to the traditional approach.

Contemporary heat-resistant and cast alloys for gas turbine blades

Contemporary heat-resistant and cast alloys for gas turbine blades

Case histories on surgical implants and their causes: Zitter, H. Werkst. Korros. Sept. 1991 42, (9), 455–466 (in German)

Mixed-mode, high-cycle fatigue-crack growth thresholds are reported for through-thickness cracks (large compared to microstructural dimensions) in a Ti–6Al–4V turbine blade alloy with a bimodal microstructure. Specifically, the effect of combined mode I and mode II loading, over a range of phase angles β=tan−1(ΔKII/ΔKI) from 0° to 82° (ΔKII/ΔKI∼0–7), is examined for load ratios (ratio of minimum to maximum loads) ranging from R=0.1 to 0.8 at a cyclic loading frequency of 1000 Hz in ambient temperature air. Although the general trend for the mode I stress-intensity range at the threshold, ΔKI,TH, is to decrease with increasing mode mixity, ΔKII/ΔKI, and load ratio, R, if the crack-driving force is alternatively characterized in terms of the strain-energy release rate, ΔG, incorporating contributions from both the applied tensile and shear loading, the threshold fatigue-crack growth resistance increases significantly with the applied ratio of ΔKII/ΔKI. The pure mode I threshold, in terms of ΔGTH, is observed to be a lower bound (worst case) with respect to mixed-mode (I+II) behavior. These results are compared with mixed-mode fatigue thresholds for short cracks, where the precrack wake has been machined to within ∼200 μm of the precrack tip. For such short cracks, wherein the magnitude of crack-tip shielding which can develop is greatly reduced, the measured mixed-mode fatigue-crack growth thresholds are observed to be markedly lower. Moreover, the dependence of the mixed-mode fatigue-crack growth resistance on the applied phase angle is significantly reduced. Comparison of the large- and short-crack data suggests that the increase in the large-crack fatigue threshold, ΔGTH, with an increasing mode mixity (ΔKII/ΔKI) is largely due to shielding from shear-induced crack-surface contact, which reduces the local crack-driving force actually experienced at the crack tip. Quantification of such shielding is described in Part II of this paper.

Possibility of increasing the service life of damaged blades

Results are given of an investigation of the effectiveness of a method of strengthening the working titanium alloy blades of a gas turbine engine compressor. It is established that the proposed strengthening method can lengthen to more than double the service life of damaged blades. In addition it is shown that in these blades, in comparison with nonstrengthened blades, cracks develop in a narrower zone and predominantly on the side of the inlet edge.

Statistical design of fatigue crack growth test programmes: Salivar, G.C. and Hoeppner, D.W. ASTM J. Test. Eval. Nov. 1988 16, (6), 508–515

Mixed-mode, high-cycle fatigue-crack growth thresholds are reported for through-thickness cracks (large compared to microstructural dimensions) in a Ti–6Al–4V turbine blade alloy with a bimodal microstructure. Specifically, the effect of combined mode I and mode II loading, over a range of phase angles β=tan−1(ΔKII/ΔKI) from 0° to 82° (ΔKII/ΔKI∼0–7), is examined for load ratios (ratio of minimum to maximum loads) ranging from R=0.1 to 0.8 at a cyclic loading frequency of 1000 Hz in ambient temperature air. Although the general trend for the mode I stress-intensity range at the threshold, ΔKI,TH, is to decrease with increasing mode mixity, ΔKII/ΔKI, and load ratio, R, if the crack-driving force is alternatively characterized in terms of the strain-energy release rate, ΔG, incorporating contributions from both the applied tensile and shear loading, the threshold fatigue-crack growth resistance increases significantly with the applied ratio of ΔKII/ΔKI. The pure mode I threshold, in terms of ΔGTH, is observed to be a lower bound (worst case) with respect to mixed-mode (I+II) behavior. These results are compared with mixed-mode fatigue thresholds for short cracks, where the precrack wake has been machined to within ∼200 μm of the precrack tip. For such short cracks, wherein the magnitude of crack-tip shielding which can develop is greatly reduced, the measured mixed-mode fatigue-crack growth thresholds are observed to be markedly lower. Moreover, the dependence of the mixed-mode fatigue-crack growth resistance on the applied phase angle is significantly reduced. Comparison of the large- and short-crack data suggests that the increase in the large-crack fatigue threshold, ΔGTH, with an increasing mode mixity (ΔKII/ΔKI) is largely due to shielding from shear-induced crack-surface contact, which reduces the local crack-driving force actually experienced at the crack tip. Quantification of such shielding is described in Part II of this paper.

Cyclic crack growth and closure effects in low alloy ferritic steels during creep-fatigue at 550 °C : Skelton, R.P. High Temp. Technol. Aug. 1989 7, (3), 115–128

There have been few studies to date describing fatigue-crack propagation thresholds under mixed-mode loading conditions in the presence of cracks, that are small as compared to the characteristic microstructural dimensions. To address this need, the variation in mixed-mode, high-cycle fatigue-crack growth thresholds with crack size is reported for a Ti-6Al-4V turbine blade alloy with a fine-grained bimodal microstructure. Specifically, threshold behavior is examined for large through-thickness cracks >4 mm in length), short through-thickness cracks (~200 μm in length), and microstructurally-small surface cracks (10–50 μm in diameter) under combined mode I and mode II loading at load ratios (ratio of minimum to maximum load) ranging from 0.1 to 0.8. For mode-mixities ranging from pure mode I to predominantly mode II, large crack, mode I ΔKI,TH thresholds were found to decrease substantially with increasing phase angle. However, by characterizing in terms of the range in strain energy release rate, ΔGTH, incorporating both mode I and mode II contributions, it was observed that the pure mode I threshold could be regarded as a ‘worst case’ under mixed-mode loading in this alloy. By estimating the effective crack-driving force actually experienced at the crack tip, the observed increase in the mixed-mode ΔGTH threshold with mode-mixity was attributed to an increasing influence of crack-tip shielding due to crack closure and crack-surface interference. Equivalent thresholds for through-thickness short cracks, where the crack wake and hence the effect of such shielding is minimized, were consequently far less sensitive to mode-mixity and corresponded in magnitude to the shielding-corrected large-crack thresholds. This effect was accentuated for the measured thresholds of microstructurally-small surface cracks; such small-crack, mixed-mode ΔGTH thresholds not only displayed a minimal influence of mode-mixity but were up to two orders of magnitude smaller than those for corresponding large cracks at the same load ratio and mode-mixity conditions.

Thermal fatigue crack propagation and failure life evaluation on CrMoV and SUS304 steels : Nitta, A., Ogata, T. and Kuwabara, K.J. Soc. Mater. Sci., Jpn. Apr. 1989 38, (427), 423–429 (in Japanese)

The role of crack-tip shielding in influencing mixed-mode (mode I+II) fatigue-crack growth thresholds for large, through-thickness cracks in a Ti–6Al–4V turbine blade alloy is examined under high-cycle fatigue loading conditions, i.e., at a loading frequency of 1000 Hz in ambient temperature air for load ratios (Kmin/Kmax) of R=0.1–0.8. Techniques are developed to quantify crack-tip shielding with respect to both the mode I and mode II applied loading, enabling an estimation of the shielding-corrected, crack-driving forces actually experienced at the crack tip (ΔKI,TH,eff and ΔKII,TH,eff or ΔGTH,eff). In Part I, it was shown that when the crack-driving force is characterized in terms of the range in strain-energy release rate, ΔG, which incorporates contributions from both the applied tensile and shear loading, the mixed-mode (I+II) fatigue-crack growth resistance increases monotonically with the ratio ΔKII/ΔKI. When the fatigue-crack growth thresholds are expressed in terms of the near-tip (shielding-corrected) crack-driving force, this increase in crack-growth resistance with increasing mode mixity is markedly reduced. Moreover, for all mode mixities investigated, the near-tip mixed-mode fatigue threshold is lower than the applied (global) value, with the effect being particularly pronounced under shear-dominant loading conditions. These observations illustrate the prominent role of crack-tip shielding for the mixed-mode loading of fatigue cracks with crack-wake dimensions large compared with microstructural size scales; specifically, they indicate that the elevation of the ΔGTH fatigue-crack growth threshold with increasing applied mode mixity is largely due to a shear-induced enhancement of crack-tip shielding.