Quantitative relationship between microstructural factors and fatigue life of Ti-5Al-2Sn-2Zr-4Cr-4 Mo (Ti-17) fabricated using a 1500-ton forging simulator

The microstructures, tensile properties, and fatigue lives of the forged Ti-17 using a 1500-ton forging simulator subjected to different solution treatments and a common aging treatment under both loadand strain-controlled conditions to evaluate high cycle fatigue and low cycle fatigue lives, respectively were examined. Then, the tensile properties, microstructures, and relationships between fatigue lives and the microstructural factors were discussed. The fatigue limit under load-controlled conditions increases with increasing solution treatment temperature up to 1143 K, which is in the (α + β) region. However, it decreases with further increase in the solution treatment temperature to 1203 K in the b region. The fatigue ratio at fatigue limit is increasing with decreasing solution treatment temperature, namely increasing the volume fraction of the primary α phase, and it relates well qualitatively with the volume fraction of the primary α phase when the solution treatment temperature is less than the b transus temperature. The fatigue life under strain-controlled conditions to evaluate the low cycle fatigue life increases with decreasing solution treatment temperature, namely increasing the volume fraction of the primary α phase. The fatigue life under strain-controlled conditions to evaluate the low cycle fatigue life relates well quantitatively with the tensile true strain at breaking of the specimen and the volume fraction of the primary α phase for each total strain range.


Introduction
The amount of usage for commercial aircrafts is recently rapidly increased because CFRP (carbon fiber reinforced plastics) is used for air frame structures [1]. On the other hand, titanium alloys are used not only in air frames, but also in jet engine components such as fans and compressor disks, which function at relatively low temperatures up to 673 K. Ti-5Al-2Sn-2Zr-4Cr-4 Mo (Ti-17), which is a age hardenable β-rich two-phase (α+ β) -titanium alloy, was developed to reduce the titanium components in the early 1970s in the USA and registered as AMS 4955; it exhibits greater strength, crack propagation resistance, and creep resistance compared with those of Ti-6Al-4V at intermediate temperatures. Ti-17 is β-forged to achieve high fracture toughness [2,3,4]. Notably, fatigue endurance is one of the important factors for the aforementioned engine components. The investigation of fatigue properties of Ti-17 by focusing on their relation with microstructural factors is highly significant. In particular, it is important to consider the quantitative relationship between fatigue properties and microstructural factors during the fatigue life estimation of engine components made of Ti-17. Therefore, in this study, the fatigue properties and microstructures of hot-forged disk-like Ti-17 samples were investigated to define the quantitative relationship between the fatigue properties and the microstructural factors in this study.

Experimental
Disk-like Ti-17 samples were fabricated using a 1500-ton forging simulator [5]. First, a Ti-17 ingot with a diameter of 120 mm and a height of 240 mm was forged at 1073 K. Then, a quarter of the forged disk-like sample was cut to prepare the tensile and fatigue specimens. For the preparation, the cut pieces were subjected to solution treatment (ST) at four different temperatures: 1063, 1123, 1143, and 1203 K followed by water quenching (WQ). The β transus temperature (T β ) of Ti-17 is around 1153 K; therefore, the ST temperature of 1203 K was slightly higher than T β , but within the b region. However, the other three ST temperatures were lower than T β , and in the (α + β) region. Each sample was then subjected to an aging treatment at 893 K for 8h, followed by air cooling (AC). The heat treatment processes are schematically shown in Fig. 1 [6]. The samples subjected to aging after the different STs were denoted as ST/1063 K, ST/1123 K, ST/1143 K, and ST/1203 K, according to the ST temperatures. Smooth round bar-shaped tensile specimens with a gage diameter of 7.0 mm, a gage length of 14 mm, and a total length of 34 mm, hourglass shaped fatigue specimens with a minimum diameter of 3 mm and a length (without the gripping part) of 28 mm, and a total length of 52 mm for load-controlled fatigue tests, and smooth round bar fatigue specimens with a diameter of 3 mm, a gage length of 15 mm, and a total length of 52 mm for straincontrolled fatigue tests were machined from the aged samples. Tensile tests were carried out at a strain rate of 0.5%/min up to a strain of 5% and after that, a strain rate of 1.5%/min using an Instron-type testing machine with a capacity of 10 kN, at ambient temperature. Fatigue tests under load-controlled conditions to evaluate the high cycle fatigue life were carried out at a stress ratio (R) of 0.1 using an electro-hydraulic servo fatigue testing machine with a capacity of 5 kN, at ambient temperature. The fatigue tests under strain-controlled conditions to evaluate the low cycle fatigue life were carried out at representatively selected total strain ranges of 1.05, 1.1, and 1.2, respectively using an electro-hydraulic servo fatigue testing machine with a capacity of 100 kN, at ambient temperature. Microstructural observations of the sample were carried out using an optical microscopy. Fatigue fracture surfaces were observed using a scanning electron microscopy (SEM).

Microstructure
The microstructure of ST/1203 K exhibited equiaxted prior β grains composed of fine acicular α because it was subjected to ST at a temperature higher than T β . On the other hand, microstructures of ST/1063 K, ST/1123 K, and ST/1143 K exhibited elongated prior β grains composed of two different microstructural feature regions: primary acicular α and fine spheroidal α regions as shown in Fig.  2. The primary acicular α region was predominat as compared with that of the fine spheroidal α region.Therefore, the authors focused on the primary acicular α region. The volume fraction of the primary α in the primary α region is shown in Table 1.

Tensile Properties
Notably, the 0.2% proof stress and tensile strength increased with increasing ST temperature up to 1143 K (within the (α + β) region), but decreased with further increase in the ST temperature to 1203 K (within the β region). In addition, the elongation and reduction of area decreased with increasing ST temperature, and it became nearly 0 % corresponding to a ST temperature of 1203 K. Interestingly, the 0.2% proof stress and tensile strength of ST/1203 K were equivalent. Therefore, ST/1203 K was considered to be fractured in a very brittle manner. The R value at the fatigue limit was found to be related with the volume fraction of the primary α phase, V(%) as shown in Fig. 5 [6]. Therefore, the relationship between the R value and V(%) can be expressed by the following equation [6]. N fs , was found to be well related to the tensile true strain at the breaking of the specimen (ε tt ) as shown in Fig. 7 [6]. Therefore, the relationship between N fs and ε tt can be expressed by the following equation for each Dε t [6]. ε tt was also found to be quantitably related to the volume fraction of the primary a phase. Therefore, N fs for each Δε t can be also well quantitatively related to the volume fraction of the primary α phase.

The fatigue life under load-controlled
conditions to evaluate the high cycle fatigue life increases with increasing solution treatment temperature up to 1143 K, which is in the (α + β) region. However, it decreases with further increase in the solution treatment temperature to 1203 K in the β region.
2. The fatigue ratio at fatigue limit under load-controlled conditions increases with decreasing solution treatment temperature, namely increasing volume fraction of the primary α phase, and relates well qualitatively with the volume fraction of the primary α phase when the solution treatment temperature is less than the β transus temperature.