Effect of different drilling techniques on high-cycle fatigue behavior of nickel-based single-crystal superalloy with film cooling hole

2.1 Experiment material

The second-generation single-crystal superalloy DD6 is used in the experiment. The single-crystal superalloy specimens with [001] orientation were cast by the helical crystal selection method and directionally solidified in the furnace of high-temperature gradient. The crystal orientations of the specimens were determined by the Laue X-ray back-reflection method. All the plates tested in this experiment had deviations within 5° from the perfect [001] direction.

Plate specimens with 14 holes are designed according to the complicated structure of the turbine blades. The schematic diagram of the plate is shown in Figure 1. The specimens with a single hole were introduced to study the effects of temperature on fatigue life. The schematic diagram of the plate is shown in Figure 2. The unit in Figures 1 and 2 is mm.

Figure 1

Specimen with 14 film cooling holes.

Figure 2

Specimen with a single hole.

2.2 Experiment method

All the HCF tests were carried out on the servo-hydraulic material testing system (GPS-100), and the temperature was limited in ±5°C. The stress ratio R of minimum to maximum stress is 0.1, and the loading frequency is 90 Hz. The tests of specimens with 14 film cooling holes were conducted at 980°C in an ambient atmosphere. The tests of specimens with a single film cooling hole were conducted at 900°C, 980°C, and 1,050°C in an ambient atmosphere. The experimental program was stress-controlled with trapezoidal load waveforms which is shown in Figure 3. All the test data were recorded by the computer automatically. Fracture surfaces and micrographs of the holes were observed using scanning electric microscopy (SEM; Hitachi S-4800) after ultrasonic cleanout.

Figure 3

Load waveforms of the HCF test.

3.1 Microstructure of the film cooling hole

Micrographs of film cooling hole in different drilling techniques are shown in Figure 4. The micrographs of MSLD, EDM, and ESM are shown in Figure 4a–f, respectively. It can be seen from Figure 4 that both MSLD and EDM have recast layers around the film cooling hole whereas ESM does not. The reason is that the MSLD and EDM are all belonged to hot working, but the ESM is cold working. Figure 4a shows that the recast layer thickness of MSLD is about 43 µm. The high temperature in the processing process separates a large number of elements around the hole, and the final major residual component is carbides and oxides of cobalt and nickel. Consequently, the out-layer materials of the recast layer are loose, there are several microcracks such as Letter A in the recast layer as shown in Figure 4b, and Letter B is a magnify micrograph of Letter A. It is obvious that there are microcracks existing in the recast layer during the MSLD process. The microcracks are likely to become the original positions of the high-cycle fatigue (HCF) cracks of the material and gradually expand under the continuous action of the cyclic load. As for EDM, the recast layer is about 35 µm, and it is thinner than MSLD. No obvious microcracks are observed in the inner wall of the film cooling hole.

Figure 4

Micrographs of film cooling holes (a and b) MSLD, (c and d) EDM, and (e and f) ESM.

The micrograph of ESM is significantly different from the MSLD and EDM. The film cooling hole of ESM shows regular round, and the inner wall is smoother. There is no recast layer, and the heat-affected zone can be found around the hole. As ESM is mainly through the dissolution of the anode metal to achieve the hole, the recast layer and microcracks can be avoided.

3.2 Effect of different drilling techniques on fatigue strength

S–N curves of fatigue testing for MSLD, EDM, and ESM specimens with 14 film cooling holes at 980°C are shown in Figure 5. The triangles, squares, and circles are representing MSLD, EDM, and ESM, respectively. The arrows indicate that the specimens were no failures during the tests. It can be seen from Figure 5 that S–N curves of different drilling techniques show a general trend that the fatigue strength decreases with the stress increasing. Moreover, at the same stress amplitude, the fatigue life of the ESM specimens is longer than that of the MSLD and EDM. The fatigue limit of MSLD is 353 MPa, while EDM is 359 MPa and ESM is 378 MPa. The fatigue limit of ESM is higher than other drilling techniques. The fatigue limit of ESM is higher than that of MSLD 25 MPa that is 7.1%. Meanwhile, the fatigue limit of ESM is higher than that of EDM 19 MPa that is 5.3%.

Figure 5

S–N curve of different drilling techniques specimens at 980°C.

The fatigue limit of the material is determined by the up-and-down fatigue test method. The tests carried out in the present study are in accordance with the test standard (metal material axial loading fatigue test method) [22]. The fatigue limit σ _D can be written as follows:

where m is the total number of valid tests, N is the stress level series, V _i is the number of tests under the i stress level, and σ _i is the stress level i.

The S–N curves were obtained by non-linear fitting with the Basquin equation [23], which can be written as follows:

where σ _a is the alternating stress, σ f ′ is the fatigue strength coefficient, N _f is the number of cycles to failure, and b is the Basquin exponent.

The parameters of the Basquin equations and the fatigue limits of different drilling techniques are listed in Table1.

3.3 Fractography

To understand the behavior of fatigue crack initiation and propagation clearly, the failed specimens fracture surfaces with 14 holes of different drilling techniques were examined.

Figure 6 shows the macroscopic fracture morphology of the MSLD specimen. Figure 6a is a fracture profile of the specimen. It can be seen that under the action of high temperature and alternating load, crack initiate from the edge of the middle hole at first and then propagate along the path between the hole. Figure 6b is a fracture surface near the hole. We can see that the fracture surface of the hole is made up of several inclined planes, and the normal direction of A plane (001) is parallel to the loading direction. The angle between the normal direction of B plane and loading direction is 45°–50°; it can be confirmed from the crystal theory that B plane is (111) plane. Generally speaking, if there is no interaction between porous, the cracks will propagate along (001). But the stress is complex because of the several holes under interaction, then the propagation mechanism may be different.

Figure 6

Macroscopic fracture morphology of MSLD specimen: (a) fracture profile and (b) fracture surface near the hole.

The fracture morphologies of MSLD, EDM, and ESM are shown in Figures 7–9, respectively. From the fracture morphologies, it can be concluded that different drilling techniques have similar morphologies. And, there are multiple fatigue sources for the fracture morphology of the specimens after HCF tests at 980°C under different stress amplitudes. During the crack propagation, fatigue cracks expand from outmost surfaces deep into interior along the primary slip plane under the alternative stress. Figure 7a is the macroscopic fracture morphology of MSLD, and there are two different areas indicated by 1 and 2. Figure 7b is the magnify micrograph of hole 1 in Figure 7a. It can be seen that the cracks propagate along the {001} planes and the bright shell area can be found, but no obvious river strip and ladder fracture characteristics exist. Thus, we can infer that the area surrounds the film cooling hole is the crack extension area. Figure 7c is the magnify micrograph of hole 2 in Figure 7a. As can be seen from Figure 7c that the letter A which has similar morphologies with hole 1 is the crack propagation area of hole 2. However, the length of the crack propagation area is only 0.4335 mm. The letter B in Figure 7c is an instant rupture zone which is along the {111} planes and has the ladder fracture characteristics. Figure 7d is the magnify micrograph of the shell area in Figure 7b. It can be seen from Figure 7d that the crystal grain is complete and slightly oxidized. The EDS analysis of oxides in Figure 7d is shown in Figure 7e. Figure 8 shows the macroscopic fracture morphology of the EDM specimen. As can be seen, it is obviously different from the fracture morphology of MSLD. There is no shell area and it si typical crisp fracture. Figure 9 shows the macroscopic fracture morphology of the ESM specimen. We can see that there are many tear ridges along the hole edge, and the crack propagates from the hole edge to both sides of the hole. In summary, typical fatigue morphology includes fatigue source area, crack propagation area, and instant rupture zone. There is no recast layer and a heat-affected zone can be found around the hole of ESM. As ESM is mainly through the dissolution of the anode metal to achieve the hole, the recast layer and heat-affected zone can be avoided. Fewer microcracks can be found around film holes. Fatigue is most sensitive to defects such as cracks; therefore, ESM is better than the others.

Figure 7

Fracture morphology of MSLD: (a) macroscopic fracture morphology, (b) middle hole fracture morphology, (c) edge hole fracture morphology, (d) shell area, and (e) the EDS analysis of oxides.

Figure 8

Fracture morphology of EDM: (a) Macroscopic fracture morphology, (b) Microscopic fracture morphology.

Figure 9

Fracture morphology of ESM: (a) Macroscopic fracture morphology, (b) Microscopic fracture morphology.

3.4 Effect of temperature on fatigue strength

S–N curves of fatigue testing for EDM specimens with single film cooling holes at 900, 980, and 1,050°C are shown in Figure 10. According to Figure 10, the temperature makes a significant influence on the fatigue limit. As the temperature increases, the fatigue life decreases gradually. The fatigue limit of the specimen is 472 MPa at 900°C. The fatigue limit of specimen decreased to 42 MPa, i.e., 430 MPa with an attenuation rate of 2.25 MPa/10°C while the temperature increased from 900 to 980°C. But the fatigue limit of specimen sharply decreased when the temperature is 1,000°C. The fatigue limit of the specimen is 293 MPa at 1,050°C. Comparing the fatigue limit of the specimen at 1,050 to 980°C, it can be found that the fatigue limit decreased to 137 MPa with an attenuation rate of 19.57 MPa/10°C. Consequently, the attenuation rate of the fatigue limit increases with the temperature rising.

Figure 10

S–N curve of single hole specimens HCF tested at different temperatures.

The parameters of the Basquin equations and the fatigue limits of different drilling techniques are listed in Table 2.

As for the specimens with a single hole, the film cooling hole destroyed the integrity of the structure, and as a result, the microcracks will occur around the hole. The HCF tests of specimens without any hole were carried out, and the fracture morphology is shown in Figure 11. It can be seen that the severe stress concentration is first generated at the casting defect, which is the primary location of crack initiation. For specimens with a film cooling hole, the cracks occur around the hole. For specimens without any hole, the cracks occur at the defects.

Figure 11

Fracture morphology of specimens without any hole.

The crack initiation has a close relationship with temperature. In engineering practice, q is commonly used to characterize the notch sensitivity of structures. Usually, the notch sensitivity q was defined as follows [24]:

where K _f is the fatigue notch factor and K _t is the theoretical stress concentration factor, which defined as follows:

where σ R is the fatigue limit of the smooth specimen, σ R H is the fatigue limit of the specimens with film cooling holes, σ max is the maximum local stress, and σ 0 is the principle stress.

The value of q varies from q = 0 or no notch effect to q = 1 or full theoretical effects. The notch sensitivity of the specimens with a single hole under different temperatures is listed in Table 3. It can be seen from Table 3, the notch sensitivity increases with the temperature increases, and the notch sensitivity at 900, 980 and 1,050°C is 0.012, 0.085, and 0.027, respectively. The higher the notch sensitivity is, the easier the crack initiation around the film cooling hole and the lower of the fatigue life.

To investigate the microstructure changes after HCF deformation under different temperatures, the γ/γ′ microstructure of the specimens with a single hole was observed by SEM. The microstructure of the specimens at 980°C is shown in Figure 12. The microstructure of the original specimen is shown in Figure 12a, which has the typical phase of γ and γ′ [25]. The microstructure of the specimen after the HCF test is shown in Figure 12b. The images are taken from the position shown in Figure 13. The change of γ′ shape is related to time and temperature. Generally, the longer the time, the higher the temperature, and the more obvious the change of γ′ shape is. The change of γ′ shape generally occurs in the creep process. For high-cycle fatigue, due to the short time and low stress, no obvious plastic deformation can be observed. Although there is stress concentration, it has little effect on the change of γ′ shape. Comparing Figure 12a and b, it is interesting to notice that the shape of γ′ changed severely after the HCF test. The shape of γ′ changed from regular cube to irregular cube. The γ′ particles show a tendency to dissolve into the γ matrix. The change of γ′ shape only occurred in some localized areas closed to the specimen surface near the hole where the maximum stress amplitude was implemented.

Figure 12

Microstructures of the specimens: (a) original specimen and (b) after the HCF test.

Figure 13

The position where the images are taken from.