2021 Volume 62 Issue 7 Pages 935-942
TiB-reinforced Ti–3Al–2.5V, in which the TiB whiskers are oriented parallel to the direction of heat extrusion, was fabricated via powder metallurgy. To investigate the effects of TiB whisker volume fraction on fatigue properties in Ti–3Al–2.5V, four-point bending fatigue tests were conducted on plate-type specimens having TiB whiskers of varying volume fraction and orientation. The fatigue limit and fatigue life of Ti–3Al–2.5V having TiB whiskers that are oriented parallel to the loading direction tended to be increase with increasing TiB volume fraction, whereas the fatigue limit and fatigue life of specimens having TiB whiskers oriented along the short transverse direction did not depend on the TiB volume fraction. The fatigue limit of specimens with TiB whiskers oriented along the long transverse direction tended to decrease with increasing TiB volume fraction, because fatigue cracks were initiated and propagated along the TiB whiskers. Thus, the resistances of fatigue crack initiation and propagation behavior in Ti–3Al–2.5V changed with both the TiB volume fraction and orientation.
Fig. 9 (a) Optical micrograph, (b) boron map obtained by EPMA analysis, and (c) SEM micrograph of the surface of a 10.0 vol% TiB-reinforced T-type specimen tested at σa = 370 MPa near the crack initiation site (Nf = 7.3 × 103 cycles). White arrows represent the crack initiation site and red arrows represent the crack path.
With the development of various engineering fields and products, titanium-based materials have become important in the biomedical, aerospace, and automotive industries owing to their excellent corrosion resistance,1) low density,2) high specific strength, and high heat resistance.3) Recent years have seen an increase in the demand for improvements in the mechanical properties of metallic materials, including the Ti–3Al–2.5V featured in this study, and for a reduction in the cost of titanium materials, as a way of addressing the weight saving and downsizing of the aerospace and automotive industries. Thus, increasing the mechanical properties of metallic materials in operation has become an important research goal.
Previous studies have reported that the mechanical properties of various materials, such as iron-based composites,4,5) copper-based bulk metallic glass composites,6) and aluminum alloys,7) can all be improved by the addition of TiBw. In particular, TiB whiskers are chemically stable in titanium alloys,8–10) because titanium boride is formed as pristine single-crystal whiskers in titanium alloys8) by the combination of titanium and boron through the reaction Ti + TiB2 → 2TiB. Morsi et al.10) have reported that TiB whiskers have excellent thermodynamic stability, thermal expansion coefficients comparable to that of the titanium matrix, a clean interface, and excellent interface bonding between the TiB whiskers and the titanium matrix. Thus, titanium alloys are expected to show a high strengthening efficiency. As such, the abovementioned properties of TiB make it a prime candidate as a reinforcement material for titanium alloys, especially since whisker-shaped reinforcements for titanium alloys cannot be obtained from any other material.
Furthermore, many researchers have examined the mechanical properties of TiB-reinforced titanium alloys, such as their yield strength,11–13) tensile strength,13,14) hardness,15,16) elastic modulus,15,17,18) fatigue properties,19–23) fatigue crack propagation,24) strength of TiB/Ti matrix interface,25) and fracture toughness.26) For example, Yang et al.19) found that the fatigue strength of a TiB-modified titanium alloy containing 6.5 mass% randomly oriented TiB whiskers was 40% higher than that of the TiB-free alloy. Thus, the addition of TiB whiskers is effective in improving the fatigue properties of titanium-based composites. The present authors20) have demonstrated that the orientation of TiB whiskers also influences the fatigue properties of titanium alloys. Therefore, we do not have a full understanding of the effects of TiB volume fraction on the fatigue fracture mechanism, the mechanisms of fatigue crack initiation and propagation, or the high-cycle fatigue properties of titanium alloys having different TiB orientations.
The present study examined the combined effects of TiB volume fraction and orientation on the fatigue properties of a TiB-reinforced Ti–3Al–2.5V composite to achieve sufficient strength for practical engineering applications. The purpose of this study was to establish the relationship between the TiB volume fraction and the fatigue fracture mechanism under four-point bending in a Ti–3Al–2.5V composite incorporating TiB on the basis of fractography and fracture mechanics.
A powdered Ti–3Al–2.5V alloy having the composition (in mass%) of 3.28Al, 2.48V, 0.04Fe, 0.069C, 0.122O, 0.008N, and balance Ti was employed. This material was fabricated by argon atomization and comprised spherical particles 58 µm in diameter. The TiB2 used was a ceramic powder with a particle diameter of 4.8 µm and the composition (in mass%) 30.0B, 0.1Fe, 0.5C, 1.1O, 0.6N, and balance Ti.
The experimental alloy was obtained by combining the Ti–3Al–2.5V powder with TiB2, such that the mixture contained 7.5, 10.0, and 12.5 vol% TiB. Details of the initial microstructure and the fabrication of sintered compacts have been described elsewhere.9,27) The resulting alloy was extruded at 1273 K under an applied force of 5 MN by using a press, at a 15:1 extrusion ratio to adjust the orientation of the TiB whiskers. Also, additional heating was performed at 1473 K for 14.4 ks in a vacuum tube furnace to ensure the complete conversion of the original ceramic to TiB, because TiB2 particles in titanium alloys completely transformed into TiB whiskers on being heated above 1173 K.13) The fully consolidated composites were prepared via hot isostatic pressing and heat extrusion.27) Mechanical properties of the 10.0 vol% TiB reinforced Ti–3Al–2.5V treated with heat extrusion27) are shown in Table 1. For comparison, TiB-free compacts fabricated from only the Ti–3Al–2.5V powder were prepared as well.
To determine the TiB whisker size distributions, the microstructure of the TiB-reinforced titanium alloy was characterized across the entire cross-section of each sample over a 40,000 µm2 region using electron probe micro analysis (EPMA) under a magnification of 500× and at an acceleration voltage of 15 kV.
2.2 TestingFour-point bending fatigue tests were conducted in an electrodynamic fatigue testing apparatus under a stress ratio R of 0.1 using plate-type specimens in air at room temperature. The sintered compact was machined by wire electric discharge machining to produce specimens of this type with the dimensions shown in Fig. 1. Specimens having different volume fractions and orientations of TiB (7.5, 10.0, or 12.5 vol%) whiskers were employed in the present study. Specimens labeled S, T, and L have TiB whiskers oriented parallel to the short transverse, long transverse, and longitudinal directions of the plate-type specimen, respectively. The specimen surfaces were polished to a mirror finish using emery papers and a SiO2 suspension.
Schematics of plate-type specimens having TiB whiskers, and TiB orientations for four-point bending fatigue tests.
After testing, the fracture surfaces of the failed specimens were observed using scanning electron microscopy (SEM), and crack initiation sites were analyzed by EPMA. To determine the mechanisms of fatigue crack initiation and propagation in the TiB-reinforced alloy, optical microscopy analysis was conducted on the surfaces of some specimens under fatigue testing.20)
First, the microstructures of the TiB-reinforced composites treated with heat extrusion were characterized. Figure 2 shows maps of the TiB-reinforced Ti–3Al–2.5V treated with heat extrusion obtained by EPMA, showing the distribution of boron throughout the materials. In general, TiB whiskers formed in titanium alloys have random orientations,8,9) whereas the compact fabricated in this study had TiB whiskers oriented parallel to the direction of heat extrusion. This suggests that the locations of TiB whiskers could be adjusted by varying the direction of heat extrusion regardless of TiB volume fraction. Also, Fig. 2 revealed that the area with high boron intensity tended to increase with increasing TiB volume fraction and that the degree of TiB orientation did not depend on the TiB volume fraction. Figure 3 shows SEM micrograph and boron intensity determined by EPMA for the 10.0 vol% TiB-reinforced Ti–3Al–2.5V treated with heat extrusion. Figure 3(a) revealed that equiaxed α grains surrounded by a small amount of β phase and TiB whiskers were observed in the titanium matrix. In addition, the boron concentration increased through the TiB whisker, as shown in Fig. 3(b).
Boron maps obtained by EPMA analysis for TiB-reinforced Ti–3Al–2.5V alloys having (a) 7.5, (b) 10.0, (c) 12.5 vol% TiB.
(a) SEM micrograph and (b) boron intensity determined by EPMA analysis for TiB-reinforced Ti–3Al–2.5V alloy having 10.0 vol% TiB.
Table 2 shows the distributions of the thickness and length of TiB whiskers determined from boron maps for the TiB-reinforced Ti–3Al–2.5V composite. Table 2 revealed that the lengths of the TiB whiskers were much higher than their thicknesses, which indicates that the aspect ratios of the TiB whiskers formed in the Ti–3Al–2.5V were increased by heat extrusion. In addition, the TiB length did not vary with increasing TiB volume fraction. Thus, the TiB volume fraction did not influence the TiB length, but rather the number of TiB whiskers, as shown in Fig. 2.
Figure 4 shows the results of four-point bending fatigue tests for the (a) S, (b) T, and (c) L series. A stress amplitude, σa, was applied to the specimen surface. Figure 4 plots σa as a function of the number of cycles to failure, Nf, where data points with an arrow represent run-out specimens without failure. In this study, the fatigue limit, σw, was defined as the average of the maximum stress amplitude without specimen failure at numbers of cycles larger than 5.0 × 106 and the minimum stress amplitude at which the specimens failed. Figure 4 revealed that the Nf for the entire series tended to increase with decreasing σa regardless of TiB volume fraction and orientation. The σw for the whole series was higher than that for the TiB-free Ti–3Al–2.5V.
Results of four-point bending fatigue tests for (a) S, (b) T, and (c) L series having different TiB volume fractions.
Figure 5 plots the relationship between σw and TiB volume fraction for specimens having different TiB orientations. As can be seen, the fatigue limit of the Ti–3Al–2.5V in which the TiB whiskers were oriented parallel to the loading direction (L series) was higher than the S and T series, regardless of TiB volume fraction. Figure 5 also revealed that the σw for the S series did not vary with increasing TiB volume fraction, while the σw of the L series increased and then kept a constant value with increasing TiB volume fraction. In contrast, the σw of the T series for the 7.5 vol% TiB reinforced Ti–3Al–2.5 V was higher than those for the 10.0 vol% and 12.5 vol% TiB-reinforced Ti alloys. Thus, the σw of the T series decreased and then kept a constant value with increasing TiB volume fraction. The fatigue properties of TiB-reinforced Ti–3Al–2.5V after heat extrusion depended on the TiB volume fraction.
Relationship between fatigue limit and TiB volume fraction in TiB-reinforced Ti–3Al–2.5V having different TiB orientations.
Thus, it was confirmed that the fatigue properties of Ti–3Al–2.5V composites having different TiB orientation after heat extrusion were influenced by TiB volume fraction. The T specimens, in which TiB whiskers were oriented parallel to the transverse direction, exhibited a lower fatigue limit with increasing volume fraction of TiB reinforcement, although the addition of TiB generally increases the strength of titanium alloys.
The σw of the T series decreased under four-point bending fatigue testing, although the volume fraction of the TiB reinforcement increased. To elucidate the mechanism for fatigue fracture in the T series, the fracture surfaces of specimens that failed during four-point bending fatigue testing were examined by SEM. Figure 6 shows SEM fractographs of the T series having different TiB volume fractions at various magnifications. In the present study, every specimen failed in the surface fracture mode. Figure 6 reveals that the morphology of each fracture surface was different and varied with TiB volume fraction. TiB-facet-like patterns were observed on the fracture surfaces, and the number of TiB-facet-like patterns tended to increase with increasing TiB volume fraction. This suggests that the TiB volume fraction influenced fatigue crack propagation behavior in TiB-reinforced Ti–3Al–2.5V.
SEM fractographs of T series with (a) 7.5 vol% (σa = 370 MPa, Nf = 1.1 × 104), (b) 10.0 vol% (σa = 370 MPa, Nf = 7.3 × 104), and (c) 12.5 vol% TiB (σa = 300 MPa, Nf = 2.75 × 103).
To examine the fatigue crack initiation and small fatigue crack propagation behavior in TiB-reinforced Ti–3Al–2.5V composites, surface crack initiation was observed for the T series (in which the TiB whiskers were parallel to the transverse direction of the specimen) during four-point bending fatigue tests using an optical microscope. Figure 7 revealed that there were no fatigue cracks in the T specimens with 10.0 vol% TiB at a stress amplitude, σa, of 370 MPa at 0 cycle (Fig. 7(a)), but a crack 1-a appeared abruptly after 1.0 × 102 cycles (Fig. 7(b)). It was found that a crack 2-a was formed at 2.0 × 103, as shown in Fig. 7(d). The crack 2-a gradually propagated and, cracks 1-b and 1-c were formed nearby crack 1-a (Fig. 7(e)), which merged to form a single crack (crack 1-a+b+c), as indicated Fig. 7(f). Also, cracks 2-b and 2-c were formed nearby crack 2-a, which merged to form a crack 2-b+c between 6.3 × 103 and 7.0 × 103, as shown Fig. 7(g). Then, cracks 2-b+c and 2-a merged to form a single crack (crack 2-a+b+c) between 7.0 × 103 and 7.2 × 103 cycles, as indicated in Figs. 7(g) and (h), leading to final fracture at 7.3 × 103 cycles. Figure 8 plots crack length against number of cycles in 10.0 vol% TiB-reinforced Ti–3Al–2.5V. It can be seen that several fatigue cracks were formed in the T series having 10.0 vol% TiB whiskers under four-point bending fatigue testing.
(a)–(h) Optical micrographs of surfaces of T series with 10.0 vol% TiB tested at σa = 370 MPa (Nf = 7.3 × 103 cycles).
Relationship between surface crack length and number of cycles in 10.0 vol% TiB-reinforced Ti–3Al–2.5V.
Fatigue crack propagation was examined in more detail in the T series with 10.0 vol% TiB whiskers. Figure 9 shows (a) an optical micrograph, (b) boron map, and (c) SEM micrograph of the 10.0 vol% TiB-reinforced Ti–3Al–2.5V surface tested at σa = 370 MPa. The crack 1-a was initiated near the specimen edge and gradually propagated on the specimen surface. In Fig. 9(b), the crack profile is represented by the pink line, and the crack initiation sites of cracks 1-a, 1-b, and 1-c are indicated by white arrows. This figure revealed that the fatigue crack grew along the TiB whisker/matrix interface in the T series. Figure 9(c) shows an SEM micrograph taken near the crack 1-a initiation site and reveals that fatigue cracks in the T series propagated along the TiB/Ti matrix interface and through TiB whiskers in some cases, as indicated by red arrows. Thus, the crack profile of the T series having 10.0 vol% TiB was influenced by the TiB whiskers.
(a) Optical micrograph, (b) boron map obtained by EPMA analysis, and (c) SEM micrograph of the surface of a 10.0 vol% TiB-reinforced T-type specimen tested at σa = 370 MPa near the crack initiation site (Nf = 7.3 × 103 cycles). White arrows represent the crack initiation site and red arrows represent the crack path.
In contrast, a single surface crack was observed for the T series having 7.5 vol% TiB during four-point bending fatigue tests using an optical microscope. Figure 10 revealed that there were no fatigue cracks in this series at a stress amplitude, σa, of 370 MPa at 0 cycle (Fig. 10(a)). Then, a single fatigue crack appeared abruptly after 1.00 × 102 cycles (Fig. 10(b)). This crack gradually propagated (Figs. 10(b)–(g)), leading to final fracture at 1.10 × 103 cycles, which is longer than the Nf of the T series having 10.0 vol% TiB. Figure 11 shows the crack length as a function of the number of cycles to final failure, Nf, for the T series having 7.5 and 10.0 vol% TiB. The crack initiation life was almost the same regardless of the TiB volume fraction; however, the several fatigue cracks for the T series having 10.0 vol% TiB were initiated and merged to form single cracks, as shown in Figs. 7 and 8. The present authors20) have reported that the fatigue cracks of specimens in which TiB whiskers were oriented parallel to the transverse direction (T series) were initiated at the TiB whisker/matrix interface or because of the fracture of TiB whiskers. These results indicate that in the T series, the number of fatigue cracks initiated from the surface increases with increasing TiB volume fraction under four-point bending fatigue testing, which results in a reduction in fatigue crack initiation resistance and σw with increasing TiB volume fraction, as shown in Fig. 5. Thus, the fatigue limit of TiB-reinforced Ti–3Al–2.5V was determined by the threshold stress for fatigue crack initiation at TiB/Ti matrix interface.
Optical micrographs of 7.5 vol% TiB-reinforced T series tested under σa = 370 MPa at N values of (a) 0, (b) 1.00 × 102, (c) 2.00 × 103, (d) 8.00 × 103, (e) 1.00 × 104, (f) 1.04 × 104, and (g) 1.09 × 104 (Nf = 1.10 × 104 cycles).
Relationship between surface crack length and number of cycles for T series having different TiB volume fractions.
Figure 12 plots the crack growth rate, da/dN, for the T series having 7.5 and 10.0 vol% TiB against stress intensity range, ΔK, under four-point bending fatigue tests at 370 MPa. The stress intensity factor, K, was calculated using the formula proposed by Newman and Raju,28) and the aspect ratio of the fatigue crack was estimated on the basis of a previous study.29) The da/dN for the T series tended to increase with increasing ΔK. Also, the da/dN values of specimens with 10.0 vol% TiB were higher than those of 7.5 vol% TiB reinforced samples at a comparable ΔK (see Fig. 12). This was because the several fatigue cracks in the T series with 10.0 vol% TiB merged to form single cracks, whereas a single fatigue crack propagated gradually in the specimens with 7.5 vol% TiB under fatigue testing.
Crack growth behavior in T series having different TiB volume fractions.
It is common for the correlation between the parameters C and m in the Paris law, i.e., da/dN = C(ΔK)m, is typically satisfied for stage IIb fatigue crack growth. Therefore, the average crack growth rate was estimated for the T series having 7.5 vol% TiB whiskers as follows:
| \begin{equation} \text{d}a/\text{d}N = 9.00 \times 10^{-12}(\Delta K)^{3.65}. \end{equation} | (1) |
The m values for the T series having 7.5 vol% TiB whiskers (3.65) is in the range of those for conventional materials: from 2 to 4. This means that the T series with 7.5 vol% TiB does not differ from conventional materials in terms of macroscopic crack propagation behavior. Thus, the fatigue crack propagation behavior of the T series with 10.0 vol% TiB, which exhibits a higher crack propagation rate, is unique and differs from that of conventional homogenous materials.
In summary, the fatigue crack initiation resistance is reduced and fatigue crack propagation rate increases with increasing TiB volume fraction in the Ti–3Al–2.5V alloy having TiB whiskers oriented parallel to the transverse direction of the specimen (the T series) during fatigue testing, which reduced the fatigue limit with increasing TiB volume fraction.
This work examined the combined effects of the volume fraction and orientation of TiB whiskers on fatigue properties in TiB-reinforced Ti–3Al–2.5V alloys. Surface analyses conducted with SEM, EPMA, and optical microscopy identified the fatigue crack initiation and propagation behaviors under four-point bending on the basis of fractography and fracture mechanics. The main conclusions of this work are as follows:
The authors would like to acknowledge financial support for this work by The Light Metal Educational Foundation, Inc. and Toukai Foundation for Technology.
