Introduction

Titanium alloy Ti6Al4V is widely used in aerospace, power generation, automotive, applications due to high specific strength and high corrosion resistance in aggressive environments and in biomedical applications due to its good biocompatibility (Chong et al. 2019; Shaikh et al. 2019; Prasad and Kumar 2011; Lin et al. 2018). The microstructure of the materials plays an important role in governing the mechanical properties such as ductility, tensile strength, fracture toughness and crack propagation resistance in the materials (Kim and Boyer 1991; Lütjering et al. 1994). The microstructure and resultant mechanical properties depend primarily on the chemical composition, history of thermomechanical processing and thermal treatment (Ding et al. 2002). Thermomechanical processing is a very useful method for refinement of the microstructure, e.g. controlling the size of the α, optimizing the ratio of the α and β phases, and controlling the morphology of α and β phases (Ding and Guo 2004; Chao et al. 2016). The Ti6Al4V alloy has been extensively studied by a large number of researchers. Previous studies have shown that thermomechanical processing of the (α + β) titanium alloy in the β-phase field leads to a lamellar microstructure after cooling to room temperature (Flower 1990; Lütjering 1998). Subsequent deformation in (α + β) phase field and heat treatment by varying cooling rate and heat treatment temperature can result in microstructure ranging from acicular to equiaxed. In addition to processing temperature, other hot working parameters such as strain also influence the microstructure, e.g. the Vf of α and β phases and the size of α (Gerhard et al. 1993; Nieh et al. 2005; Seshacharyulu et al. 2000).

Thus, microstructure evolution during hot working is influenced by the initial microstructure, hot working temperature, strain, strain rate and subsequent cooling rate (Guo et al. 2005; Gupta et al. 2016; Welsch et al. 1988; Seetharaman et al. 1991; Gupta et al. 2018a, b; Smith 1993). The morphology of α phase can vary between equiaxed, plate-like, basket weave, Widmanstätten or acicular with different hot working temperatures and cooling rates (Shaikh et al. 2019; Flower 1990). Desired microstructure can be obtained by using suitable parameters. However, problems still persist at industrial scale in hot deformation to obtain defect-free forged/ rolled products of ultrasonic quality class A1 as per AMS 2631-2017 standard (1.2 mm flat bottom hole or equivalent size indications) for aerospace applications (Gupta et al. 2007). Such products are used to realize very thin section components, where defect tolerances are very low. To obtain such sound ultrasonic quality forgings, the deformation temperature is suggested to be in the vicinity of 900 °C, which is quite difficult to maintain especially for large-sized products (Gupta et al. 2018a, b). Also, obtaining uniform mechanical properties in large-sized products (with very less scatter) is another challenge faced by titanium alloy product mills. It is important to understand the mechanical properties of titanium alloy Ti6Al4V in various conditions, which assumes great significance to design a particular forging operation and/ or heat treatment cycle. It may be noted that a systematic study simulating industrial practice of forging followed by heat treatment and its effect on tensile as well as impact properties of the Ti6Al4V alloy has not been reported. Thus, to bring out a better understanding on the effect of deformation temperature and heat treatment on the microstructural evolution and mechanical properties of Ti6Al4V, the present study is carried out.

The objective of this study was to investigate the effect of different forging reduction percentages and different annealing processes on the microstructure and mechanical properties of Ti6Al4V simulating its industrial scale processing. The influence of hot working parameters such as strain and temperature on the changes in microstructure and changes in mechanical properties has been investigated in the present work.

The β transus temperature (Tβ) of the Ti6Al4V alloy under study is 992 °C, which is similar to calculated value of ~ 995 °C (Guo et al. 2005). In order to achieve the above mentioned objectives, the forging has been done to different % reductions of 25, 50 and 75% at different temperatures varying from the lower (α + β) phase field (i.e. 800 and 850 °C), intermediate (α + β) phase field (e.g. 900 °C), higher (α + β) phase field (i.e. 950 °C) and in the single-phase β field (i.e. 1000 °C). There are components in aerospace applications such as liners for composite over-wrapped pressure vessels and other hostile applications in power sector, chemical industry as well as biomaterials sector where uniform mechanical properties are desired. In such cases recrystallization annealing (RA) is more appropriate to achieve a uniform and refined microstructure (Gupta et al. 2016). Considering all these aspects, a systematic investigation has been conducted in the present work. Similar works mentioned above have been reported with lab-scale studies in isolation. However, a systematic approach of hot working with different % reduction and with different post work annealing on industrial scale is the prime focus of this work, which is not reported earlier.

Material and Methods

The chemical composition of the as-received (AR) material is Ti-6.2Al-4.2 V-0.19O-0.05Fe. The as-received material was in hot forged + annealed at 730 °C- 2 h-A/C to RT in the form of rectangular cross-section blocks of size 70 × 70 × 100 mm3. The blocks were preheated in an electric furnace for 2 h at different temperatures (i.e. 800 °C, 850 °C, 900 °C, 950 °C and 1000 °C). Preheating temperature was set to 10 °C higher than the deformation temperature for duration of 10 min just before taking out the job from the furnace so as to compensate for the estimated heat loss due to handling during forging. The blocks were upset hot forged along the highest dimension direction (100 mm) using a hydraulic forging press of 1000 MT capacity to different percentage reductions (i.e. 25%, 50%, 75%). The forging operations were carried out without any reheating. Strain rate during deformation was ~ 0.1 s−1.

After the forging operation, the blocks were air cooled to room temperature. Subsequently, forged blocks were cut into two halves. One half was taken for mill annealing (MA) at 730 °C for 2 h in a muffle furnace followed by furnace cooling to room temperature. The second half was taken for recrystallization annealing (RA) at 930 °C for 1 h in the same muffle furnace followed by furnace cooling to room temperature. After completion of both types of annealing heat treatments, the microstructure evaluation and mechanical testing were carried out. Additionally, another set of samples in as-received (AR) condition was heat treated at different temperatures (e.g. 800, 850, 900, 950 and 1000 °C) for 1-h duration followed by air cooling to room temperature to compare the effect of heat treatment alone with that of coupons subjected to forging at these temperatures followed by RA and MA heat treatments. The identification details of samples denoted in the manuscript are given in Table 1. Subsequently, ‘M’ for mill annealing and ‘R’ for recrystallization annealing have been put as suffixes in sample designation. Grain size and volume fraction (Vf) referred throughout the paper are for the α phase. The alloy samples were characterized for microstructure in different conditions viz. as-received (AR), AR + heat treated (HT), as-forged + heat treated (AF + MA and AF + RA). Microstructural observations have been carried out using Olympus-make GX71 optical microscope (OM) and fractography analysis of tensile-tested and impact-tested samples was carried out using Carl Zeiss Sigma HD Field Emission Scanning Electron Microscope (FESEM).

Table 1 Identification of heat-treated and forged samples of Ti6Al4V used in present study
Full size table

The samples for electron back scattered diffraction (EBSD) analysis were prepared by conventional metallographic polishing followed by fine colloidal silica polishing. The samples were then subjected to electro polishing using 80% methanol, 20% perchloric acid mixture at a temperature of − 25 °C, voltage of 11 V for 30 s in Struers-make Lectropol-5 model electro polishing machine. EBSD measurements were performed in a Carl Zeiss field emission scanning electron microscope, model Gemini 500. The EBSD maps were recorded on a scan area of 300 μm × 200 μm and step size of 0.3 μm. For further analysis step size of 0.05 μm was also used for selected samples. The EBSD data were analyzed using TSL OIM software, version 8.

Tensile testing was carried out using Zwick Roell universal testing machine (UTM) and impact strength was evaluated using INSTRON-make impact testing machine. All the test samples were taken from the core of the forging and along the direction perpendicular to forging axis. Three tensile and three impact test specimens representative of each condition were tested. The size of α and its Vf was estimated from the optical microstructures using ImageJ™ software as well as through EBSD. The input material, forging process and blocks subjected to different reductions imparted on blocks are shown in Fig. 1.

Fig. 1
figure 1

Photographs showing a as-received Ti6Al4V material, b forging in a hydraulic press and c as-forged Ti6Al4V blocks forged to different % reduction

Full size image

Results and Discussion

Microstructure Analysis

The microstructure of the as-received (AR) Ti6Al4V alloy presented in Fig. 2a consists of equiaxed α phase in transformed β matrix. The phases were uniformly distributed in the microstructure (bright phase indicates α grains and dark phase indicates transformed β phase). The average grain size and Vf of equiaxed α phase was 26 ± 3 µm and 42 ± 5%, respectively.

Fig. 2
figure 2

OM Images of Ti6Al4V forgings a representing as-received (AR) sample in mill-annealed condition, b heat treated at 950 °C, c heat treated at 1000 °C and d variation in α grain size and Vf of α with heat treatment temperature

Full size image

Effect of Heat Treatment

The samples which had undergone heat treatment below Tβ showed that both phases (α and β) were uniformly distributed. The grain size distribution of α was almost of same with ± 3 μm variation for the heat treatment conditions up to 900 °C. The Vf of α was also within ± 5% variation for the heat treatment conditions up to 900 °C. The representative photomicrograph of ‘AR’ sample in forged and milled annealed condition is presented in Fig. 2a. The microstructure of the sample heat treated at 950 °C (near Tβ) as shown in Fig. 2b shows lower amount of α. In the samples which had undergone soaking at temperature above Tβ (1000 °C), the microstructure was found to be of Widmanstätten structure with isolated acicular α phase (Fig. 2c) and almost no primary alpha. The variation in α grain size (section of acicular grain) and Vf of α phase with heat treatment temperature is shown in Fig. 2d. Acicular α grains were found to have aspect ratios in the range of 5–6. The α grain size and Vf both were found to decrease with increase in heat treatment temperature near to the Tβ of the alloy.

Effect of % Reduction and Heat Treatment

The as-forged samples were subjected to mill annealing and recrystallization annealing heat treatment cycles and analyzed for microstructure as shown in Figs. 3, 4 in all the three forged conditions (25%, 50%, 75% reduction). Mill annealing is not expected to alter the microstructure of as-forged (AF) condition samples significantly, but for reducing the residual stresses, since it is well below the Tβ (~ 995 °C). Only minor changes are expected in Vf of α according to % reduction imparted along with different reduction temperature. Whereas recrystallization heat treatment is carried out ~ 50 °C below the Tβ, which can alter the microstructure significantly. The same is observed in the present work as well.

Fig. 3
figure 3

OM images of Ti6Al4V forgings subjected to different % reductions followed by MA heat treatment- a 25% reduction at 850 °C, b 25% reduction at 1000 °C, c 50% reduction at 1000 °C, d 75% reduction at 800 °C and e 75% reduction at 1000 °C, Variation in grain size and Vf of α in MA heat-treated samples f 25% reduction, g 50% reduction, h 75% reduction at different prior forging temperatures

Full size image
Fig. 4
figure 4

OM images and corresponding variation in grain size and Vf of α with % reduction at different temperatures followed by RA heat treatment - a microstructure representing below Tβ forging (at 950 °C) to 25% reduction and b variation in α grain size and Vf of α with 25% reduction, c representing sample forged at 1000 °C to 50% reduction and d variation in α grain size and Vf of α with 50% reduction, e representing sample forged at 1000 °C to 75% reduction and f variation in α grain size and Vf of α with 75% reduction

Full size image

Effect of Combination of Forging and Mill Annealing Heat Treatment

In the case of 25% reduction samples subjected to mill annealing, it is observed that when forging was carried out below Tβ, the α grain size (Fig. 3a) was 30 ± 3 µm and Vf of α phase was 40 ± 2% with respect to different forging temperatures. In case of sample forged above Tβ followed by MA (Fig. 3b, c), larger transformed β grains containing α acicular/ Widmansttaten type of laths of finer alpha within them with lower Vf of α are observed. However, when finish forging temperature is below Tβ, although the start temperature is above Tβ, the β grain size is reduced and equiaxed α evolves. In the sample shown in Fig. 3b, c, the Vf of α phase was also higher as compared to sample subjected to heat treatment alone at the corresponding temperature. Microstructure was Widmanstätten laths of α in transformed β matrix with very little amount of α phase in Fig. 3b, c which is due to finish temperature being below Tβ during which α phase evolves.

The samples forged above Tβ (Fig. 3b, c) and mill annealed exhibited coarser α grain size of ~ 28 ± 2 µm, as compared to that of heat-treated sample as shown in Fig. 2c in which α grain size is ~ 15 µm. The microstructure of samples forged below Tβ to 50% reduction and mill annealed (Fig. 3c) consisted of primary α as well as fine secondary α phase in β matrix. The primary α phase was having grain size of 28 ± 2 µm and Vf of 40 ± 3% with respect to different forging temperatures. However, above Tβ, the microstructure consists of Widmanstätten laths of α in a β matrix with nearly globular morphology (Fig. 3c). The corresponding  Vf of equiaxed α is ~ 18 ± 2% due to finishing temperature below Tβ.

With increased % reduction, a change in morphology of α can be clearly seen from as-heat treated as well as with 25% reduction sample at the same temperature. This can be attributed to adequate deformation energy available for the globularization of α phase. In 50% deformed samples, with decrease in deformation temperature, the Vf of equiaxed α was found to be marginally increasing up to 850 °C. This could be due to the combined effect of moderate reduction (50%) and forging temperature. Above Tβ the Widmanstätten structure of α phase was present in β matrix where the Vf of α and grain size of α was slightly lower as compared to that of samples deformed below Tβ.

The samples subjected to 75% reduction below Tβ followed by mill annealing exhibited primary α grain size of 38 ± 2 µm and Vf of α phase was 62 ± 3% at various forging temperatures up to 900 °C. From forging temperature of 900 °C onwards, the Vf of α phase decreases sharply, indicating large deformation might be helping in early dissolution of α phase. Similar findings have been reported during tensile deformation of another α + β titanium alloy (Anil Kumar et al. 2019). Forging above Tβ followed by mill annealing resulted in Widmanstätten lath morphology along with globular α in transformed β matrix (Fig. 3e) with a Vf of α increasing up to ~ 24%. This confirms the increasing trend of recrystallized α (secondary) with increase in amount of % reduction and finish temperature below Tβ, although the deformation start temperature is above Tβ. With increase in forging temperature the Vf of α phase and α grain size are found to be decreasing with minor variations (Fig. 3f–h). Minor changes in α grain size and Vf of α have been observed in as-heat treated samples also (section “Effect of heat treatment”). However, morphology of α is found to be more irregular in as-forged + mill annealed condition as compared to as-heat treated samples. The samples forged above Tβ exhibited α phase colonies with acicular morphology (Fig. 3e). This is distinctly different as compared to as-heat treated sampedles where acicular α is found to be present in isolated pockets with Widmanstätten structure (Fig. 2c).

With an increase in amount of reduction, Vf of α is also found to be increasing especially from 50 to 75% reduction. This clearly indicates role of deformation in modifying the microstructure at the similar temperature due to availability of energy through deformation. Increase in Vf of equiaxed α with higher amount of reduction (75%) is observed. This is possible in higher % reduction samples due to the large deformation strain imparted to it. This also indicates that large deformation (% reduction) can lower the recrystallization temperature of alloy, which may help in recrystallization of α during forging itself (Gupta et al. 2016). The mechanism of globularization or recrystallization of α phase during deformation is also elaborated in literature on a similar titanium alloy (Anil Kumar et al. 2019).

Effect of Combination of Forging and Recrystallization Annealing Heat Treatment

In recrystallization-annealed (RA) condition (Fig. 4), the microstructure is found to be distinctly different compared to the AF + MA samples (Fig. 3). The forged microstructure transformed to microstructure consisting of recrystallized α in transformed β matrix. With 25% reduction below Tβ and RA heat treatment, the grain size of α was 30 ± 2 µm in all the forged + RA conditions with respect to forging temperatures (Figs. 4a, b). The Vf of equiaxed α was almost same with respect to forging temperatures plus RA (70 ± 3). However, for the samples forged above Tβ and RA heat treatment, the Vf of recrystallized equiaxed α phase was ~ 30%. Widmanstätten microstructure seen in AF condition is found to be completely modified with fine lamellar α plus basket weave microstructure along with presence of globular α.

For 50% reduction below the Tβ and RA heat treatment, grain size of equiaxed α phase was almost same, i.e. 30 ± 2 µm in all the conditions with respect to forging temperatures (Fig. 4c, d). The Vf of equiaxed recrystallized α phase was (70%) also almost same in 900–800 °C forging temperatures. At 950 °C, the Vf of recrystallized equiaxed α phase reduced to ~ 58%. Above Tβ, the Vf of recrystallized α phase was ~ 42% and α phase precipitation is observed in the form of short elongated α in transformed β matrix. The reduction in Vf of recrystallized α at 950 °C and above could be due to availability of stored energy in the samples subjected to large deformation (reduction %) when approaching Tβ. This indicates samples deformed at relatively lower temperature result in large amount of recrystallized grains.

75% reduction below the Tβ- followed by RA heat-treatment resulted in similar grain size of α (30 ± 3 µm) irrespective of forging start temperatures (Fig. 4e, f). The Vf of equiaxed α phase was also found almost the same in all the conditions with respect to temperatures (75 ± 3%). Samples forged above Tβ showed reduction in Vf of recrystallized α. It is due to similar reasons as mentioned for 50% reduction sample. However, the extent of reduction in Vf of recrystallized α in 75% reduction samples is slightly lower as compared to 50% reduction sample. A very unique observation is noted here in RA treated samples where Vf of recrystallized α is ~ 70–75 with 25–75% reduction, indicating important role of RA treatment.

Simplifying further, α grain size and its Vf estimation obtained through optical microscopic studies are summarized into two categories on the basis of forging temperature with respect to Tβ and is presented in Table 2. It can be observed that the α grain size exhibited minor variation with change in % deformation or post deformation annealing heat treatments. However, Vf of α, which is a combination of refined primary α and recrystallized α has significant change with MA/ RA treatment. Large deformation (75%) followed by MA results in higher Vf of α, which is achievable with 25% deformation in RA treated samples (Luo and Li 2010). But further change in Vf in RA treated samples is not seen significantly with increase in % reduction. This indicates higher Vf is achievable with low to moderate deformation (25–50% reduction below Tβ) followed by RA. From the results of MA and RA treated samples, higher Vf of recrystallized α and marginal change in size of α (Figs. 3 and 4) for forging below Tβ can be helping to obtain uniform mechanical properties in 75% reduction + MA treated or 25–75% reduction + RA treated samples. The findings are similar to that reported in closed die forgings reported earlier (Gupta et al. 2016).

Table 2 Estimation of α grain size and its volume fraction through optical microscopy
Full size table

For the samples forged and annealed above Tβ, similar increasing trend in Vf of primary α is seen for MA/ RA treated samples from 25 to 75% reduction. The change in slope of Vf of primary α (Figs. 3 and 4) with % reduction is seen where Vf increases with increasing % reduction imparted during forging. In fact, this change is noticed even at 950 °C (Fig. 3) for 75% reduction sample and MA treated. This can be attributed to larger reduction resulting in higher amount of fine recrystallized α in subsequent cooling from forging temperature assisted by MA/ RA. A significant change in Vf of primary α with % reduction is more pronounced with RA treatment due to added recrystallization process.

Microstructural Analysis Through EBSD

Selected representative samples of different conditions were analyzed using EBSD. Image quality (IQ) maps and Inverse Pole Figure (IPF) maps along with grain size charts are presented in Figs. 5 and 6. Distinct differences can be clearly seen between the AF + MA and AF + RA samples (which were prior forged at different temperatures). Samples forged at two extreme ends of temperature range and one intermediate temperature with maximum amount of deformation were analyzed. Comparing Fig. 5a1, b1, c1 with Figs. 5a2, b2, c2 reveals that samples deformed above Tβ showed limited amount of α recrystallization. Close observation of AF + MA samples shows that they consist of relatively lower fraction of recrystallized grains and distribution of grains is found to be non-uniform as compared to AF + RA samples. It indicates that recrystallized fine grains of α observed in Fig. 5a1, b1, c1 could be the result of large deformation and subsequent MA, which could have contributed to some recrystallization. However, in case of a2, b2, c2 samples, temperature drop during forging, post deformation recrystallization annealing contributes a large fraction of recrystallized grains with relatively uniform distribution of α grains. Though the deformation temperature has been just above Tβ, the temperature drop occurring during deformation, post deformation cooling and subsequent annealing have certainly contributed to microstructure evolution and hence the resultant microstructure.

Fig. 5
figure 5

EBSD maps a1-a5: IQ maps, b1–b5: IPF maps, c1–c5: Grain size area fraction plots, (a1, b1, c1: A7M, a2, b2, c2: A7R, a3, b3, c3:C7R, a4, b4, c4: E7M, a5, b5, c5: E7R. The EBSD maps in these figures were scanned with a step size of 0.3 µm

Full size image
Fig. 6
figure 6

GS maps (a1–a5) and GOS charts ≤ 0–0.75° (b1–b5) showing recrystallized grains (with size greater than 3 µm), (a1, b1: A7M, a2, b2: A7R, a3, b3:C7R, a4, b4: E7M, a5, b5: E7R), scanned with 0.3 µm step size (in b1–b5, the colored regions are recrystallized grains, where colors refer to the grain orientation spread, and not the crystallographic orientation)

Full size image

Further, when the forging temperature is below the Tβ (Fig. 5a3, b3, c3) along with post deformation RA, almost complete recrystallized grains can be observed with higher extent of uniform distribution of grains. It also indicates that a lower temperature deformation provides adequate energy for static recrystallization. Further lowering the deformation temperature shows presence of very fine grains in AF + MA condition (a4, b4, c4) as compared to higher temperature deformed + MA samples (a1, b1, c1). Similar observations were noted in optical microscopy as well. Distribution of α grains is found to be relatively non uniform in AF + MA samples. It is due to large deformation at lower temperature resulting in large amount of strain, but primary α grains are still not converted to fine grains indicating inadequacy of driving force for static recrystallization. Mill annealing normally results in reduction in residual stresses but no significant changes in the microstructure.

Further when the same sample is subjected to RA, it fully converts to recrystallized α grains (Fig. 5a5, b5, c5). But certainly, uniformity in grain size is not as observed in deformed samples at intermediate temperature (Fig. 5a3, b3, c3). It indicates that though the fraction of smaller grains may be high for lower temperature deformed + RA samples, the uniformity in grain size can be obtained only at intermediate deformation temperature.

When recrystallization occurs, the new grains are smaller in size compared to the original ones and are also strain-free. To confirm and quantify these recrystallized grains from EBSD data, two parameters are used, viz. grain size (GS) and grain orientation spread (GOS). The criteria are: GS being greater than 3 µm (since there will be no impetus for recrystallization in very small grains, typically 1 µm or less in size) and GOS less than or equal to 0.75°. Thus, grains with size > 3 µm and GOS ≤ 0.75° are treated as recrystallized (Raveendra et al. 2008). GS maps with grain of size greater than 3 µm, and GOS less than or equal to 0.75° represent the recrystallized grains and are shown in Fig. 6. Typically, if the value of GOS is more, it indicates that there is strain within the grain, and hence may not be a recrystallized grain. It can be seen clearly that samples C7R and E7R have a higher fraction of recrystallized grains.

Further analysis was conducted with smaller step size of 0.05 µm on specimens having maximum reduction (75%), i.e. forged at higher temperature (above Tβ) and at lower temperature (800 °C, below Tβ) (Fig. 7). Both the samples were selected in mill-annealed condition to analyze the extent of recrystallization. It has been clearly observed that very fine recrystallized grain nucleated in the sample forged at lower temperature (E7M) along the grain boundaries of large primary α grains. The same is not observed in the sample forged at higher temperature and mill annealed (A7M). This confirms that lower temperature reduction provides sufficient stored strain energy in the material, which helps in generation of recrystallized grains during subsequent mill annealing. Also, since amount of deformation is high, lowering of actual recrystallization temperature is possible, and the same is seen as initiation of recrystallization and resulting in partially recrystallized grains. Optical microscopy also has shown presence of equiaxed recrystallized grains in sample forged at lower temperature.

Fig. 7
figure 7

EBSD maps a1, a2: IQ maps, b1, b2: IPF maps, c1, c2: Grain size area fraction plots (a1, b1, c1: A7M, a2, b2, c2: E7M, with 0.05 µm step size)

Full size image

Α grain size analysis is also carried out through EBSD and is presented in Figs. 5, 6, 7. Summary of the grain size estimate for major area fraction obtained through EBSD is presented in Table 3. It clearly shows that recrystallization-annealed samples have relatively larger size of α grains, as compared to mill annealed samples. This is due to availability of higher driving force in the form of relatively higher temperature in recrystallization annealing. In EBSD, lower size of α grains are revealed as compared to observation made through optical microscopy due to higher resolution images in EBSD. Also, it is analyzed in relatively very small area in EBSD. However, both optical microscopy and EBSD results are found to be complementing each other.

Table 3 α Grain size estimate obtained through EBSD from Fig. 6
Full size table

Mechanical Property Evaluation

Tensile and Impact Strength

Tensile testing of as-heat treated samples as well as forged + annealed samples was carried out. The nature of stress–strain curves for as-heat treated samples and forged plus annealed samples at respective temperatures is found to be similar. Typical examples of samples tested in as-heat treated, forged + annealed conditions are presented in Fig. 8.

Fig. 8
figure 8

Representative plots of engineering stress–strain curves of Ti6Al4V in a As received + Heat Treated at different temperatures, b 25% reduction + MA heat-treated condition, c 25% reduction + RA heat treatment condition. Note: All testing has been done at room temperature

Full size image

Mechanical properties of different heat-treated samples are given in Table 4. The 0.2% offset yield stress (YS) of higher temperature forged + MA sample shows marginal increase as compared to heat-treated samples at respective temperatures, possibly due to α refinement in microstructure during forging. However, the same is not seen in UTS. At the same time, impact strength is found to be consistently low for the samples forged at all the temperatures as compared to the as-heat treated samples. This also indicates that microstructural refinement of the α lamellae structure, which normally results in higher fracture toughness and impact strength gets fragmented during forging and the microstructure is refined. Further, Vf of α also plays an important role. In as-heat treated condition, transformed β phase in the matrix shall be higher, which contributes to higher impact strength.

Table 4 Room temperature mechanical properties of Ti6Al4V in different heat-treated conditions without and with different prior forging reductions
Full size table

In similar lines, 50% reduction + MA samples exhibited higher 0.2% offset yield stress in all the samples except those forged at 1000 °C. This can be attributed to occurrence of large amount of microstructure refinement of α. Unlike 25% reduction + MA samples, UTS is found to be higher than that of heat-treated as well as 25% reduction + MA samples. This indicates that higher amount of % reduction contributes towards generation of more lattice defects and α phase refinement. It is also found that impact strength is in similar lines of 25% reduction + MA samples, indicating microstructure refinement of α is affecting impact properties as explained earlier. 1000 °C forged + MA sample exhibited no significant change as compared to heat-treated samples. Minor reduction in 0.2% offset yield stress is observed as compared to 25% reduction + MA samples. It indicates that two or more opposite factors are working at higher temperatures. It may be possible that benefit of higher amount of reduction through microstructure refinement of α is balanced by high temperature forging especially above Tβ.

In 75% reduction + MA samples, the trend is similar to 50% reduction + MA samples. Marginally higher 0.2% offset YS in 75% reduction + MA sample are in expected lines due to presence of higher amount fine grains. However, marginal reduction in 0.2% offset YS and UTS is seen for the samples forged at lower temperature (E7M). This can be attributed to large extent of recrystallization in subsequent annealing, as observed through optical microscopy and EBSD. Also, this has been clearly indicated by the Vf of α. Impact strength in 75% deformed + MA condition is seen to be marginally lower than other three conditions indicating refinement of α microstructure resulting in mainly equiaxed morphology of microstructure or minimum amount of lamellar structure. However, low-temperature deformation (E7M) resulted in higher impact strength indicating role of recrystallization.

In RA condition, as the alloy is subjected to recrystallization annealing at 930 °C, the previous history of deformation up to RA temperature modifies significantly in recrystallization phenomenon changing the microstructure to fine recrystallized microstructure (Figs. 5, 7, 8). The mechanical properties are thereby governed by such microstructural changes. With 25–75% reduction and RA, 0.2% offset YS of alloy exhibits marginal change in higher temperature forged samples (1000 °C). However, in case of lower temperature forged samples 0.2% offset YS were similar to that obtained in as-heat treated samples at respective temperatures but it has reduced strength as compared to forged + MA condition at respective temperatures with same amount of % reduction, respectively. Minor change in strength for high temperature worked structure (1000 °C) and RA treated can be attributed to limited α refinement in microstructure. However, impact strength is found to be marginally higher as compared to deformed + MA samples, but it is marginally lower than that of as-heat treated condition sample. This indicates higher recrystallization may be helping to improve impact strength.

Minor variation in trend is noted in the mechanical properties with respect to % reduction and deformation temperature. But it is certainly clear that, RA is playing a significant role in bringing uniformity in mechanical properties as observed in the Vf of α, that is almost similar in forged samples (25–75%) after RA treatment. Marginal increase in impact strength for deformed and RA heat-treated samples indicate more recrystallization though it reduces strength, a marginal increase in impact strength is observed. It is due to increased grain boundary area resultant fine recrystallized grains and generation of large amount of low angle boundary (Balasubramanian et al. 2008; Wen et al. 2019).

Three important factors are found to be playing a significant role in deciding the mechanical properties, viz. presence of acicular α from near Tβ exposure/deformation, refinement of primary α grains during deformation and extent of recrystallization (generation of new recrystallized α grains). First factor plays a predominant role in high-temperature heat treatment/ deformation. The second factor plays a predominant role at temperature below Tβ and higher amount of deformation with MA resulting in refined primary α. The third factor is predominant in large deformation combined with MA or RA resulting in mainly recrystallized α. Accordingly, variation in mechanical properties is observed. Moderate cooling from higher temperature will result in formation of large amount of acicular α whereas large deformation at lower temperature provides opportunity of grain refinement/ recrystallization. Analyzing the results in this study, it can be inferred that large amount of deformation, which results in refinement of primary α grains provides improvement in tensile strength with marginal reduction in impact strength. Further higher amount of deformation, which is associated with recrystallization, showed marginal reduction in tensile strength and minor improvement in impact strength.

% Elongation to Failure

% Elongation to failure is found to be consistent in heat-treated conditions below the Tβ (17–18%). The sample heat treated above Tβ exhibited lower % elongation due to the fully transformed β microstructure. However, the same is not seen in 25% reduction + annealed sample, where elongation is consistently higher (20–21%) with marginal improvement compared to the as-heat treated samples indicating effect of α microstructure refinement. However, at 50% as well as at 75% reduction + MA conditions, % elongation exhibited a marginal drop (14.5–16%). Elongation is found to be consistently high in forged + RA samples. This could be due to extensive deformation and its effect on extent of recrystallization during RA according to forging temperature.

Fractography

Fractography of impact-tested and tensile-tested samples was carried out through SEM and the analysis is presented in subsequent sections.

As-Heat Treated Condition

The presence of large grains in samples heat treated above Tβ resulted in Widmanstätten structure with α plates. In impact testing, it helped to absorb large amount of energy thereby resulting in higher impact strength. The same is seen in fractography (Fig. 9a) showing larger fragmentation of α plates as compared to other sample heat-treated below Tβ (Fig. 9b), where impact strength is marginally low.

Fig. 9
figure 9

Representative fractographs of Ti6Al4V impact samples heat treated at a 1000 °C and b 800 °C; fractographs of Ti6Al4V tensile samples heat treated at c 1000 °C and d 800 °C. Note: All testing has been done at room temperature

Full size image

However, tensile elongation is found to be lower in samples heat treated above Tβ as compared to the samples heat treated below the Tβ. Similarly, quasi-cleavage features are observed in fractographs of impact specimen as shown in Fig. 9c which indicates the microstructure consists of significant fraction of transformed β. The sample heat treated at 800 °C as shown in Fig. 9d, exhibited substantially higher amount of dimples as compared to Fig. 9c. This indicates importance of nearly equiaxed primary α in transformed β microstructure resultant of heat treatment below the Tβ to result in ductile failure in tensile or impact modes of deformation.

Forged + Heat Treated Condition

The effect of 25% reduction during forging on tensile and impact properties is found to be not significant. The impact strength is higher for the sample forged above Tβ and subjected to MA heat treatment as compared to the sample forged below the Tβ and subjected to MA heat treatment. In fractography, significant differences are not observed among different conditions. Fine dimples could be observed in most of the samples subjected to tensile testing as shown in Fig. 10a indicating presence of microstructure resultant of higher temperature deformation, which assisted in formation of acicular α in Widmanstätten structure. This can result in higher impact energy absorption through crack propagation as shown in Fig. 10b.

Fig. 10
figure 10

Representative fractographs of Ti6Al4V samples subjected to 25% reduction at 1000 °C + MA heat treatment- a tensile-tested and b impact-tested specimens, 25% reduction at 1000 °C + RA heat treatment- c tensile-tested and d impact-tested specimens. Note: All testing has been done at room temperature

Full size image

Hot forging followed by RA heat treatment modified the microstructure and corresponding mechanical properties significantly. Large reduction at lower temperature has resulted in marginally high impact strength as compared to higher temperature forged sample (above Tβ). Very fine dimples along with flat cleavage type, i.e. mixed mode of fracture could be observed in sample forged at above Tβ + RA heat treatment followed by tensile testing indicating highly refined Widmanstätten microstructure due to large % reduction (Fig. 10c). Sample subjected to impact testing showed large amount of deformation, i.e. deeper dimples (Fig. 10d) compared to Fig. 10b resulting in marginally higher impact strength through more energy absorption. So it can be concluded that higher grain boundary area either through lamellar structure or through large volume of very fine recrystallized α grains shall be helping in improving the impact strength.

For samples subjected to 50% reduction, significant effect of RA was observed with more consistent elongation in tensile loading. In case of impact testing, though the sample forged above Tβ may contain some remnants of transformed β microstructure, due to large deformation and RA heat treatment, significant effect of higher temperature deformation is not dominant. Similar is the case for tensile testing. In all these cases, the mechanical properties are more consistent and corresponding fractographs were having mostly similar features.

At large reduction of 75%, the impact strength of samples forged above Tβ are similar to the samples forged below Tβ. The fracture surfaces are mostly similar with presence of fine dimples. Here, the benefit of large deformation and RA is highly effective in bringing uniformity in microstructure towards equiaxed morphology and correspondingly the fracture surface.

Conclusions

The titanium alloy Ti6Al4V was hot forged at different temperatures and with different % reduction followed by mill annealing (MA) and recrystallization annealing (RA) heat treatments. The following conclusions are drawn from the present study:

  1. 1.

    Heat treatment/ forging above Tβ led to significant presence of Widmanstätten structure, which lowered the strength, but led to marginal improvement in the impact strength.

  2. 2.

    Samples forged below Tβ showed improvement in strength in mill-annealed (MA) condition due to α grain refinement during deformation and further partial recrystallization during heat treatment.

  3. 3.

    Forging with 25–50% reduction at intermediate temperatures (850–900 °C) + MA heat treatment resulted in higher strength as compared to either lower temperature deformation or deformation above Tβ. However, with further increase in % reduction to 75%, improvement in strength is not seen due to large amount of recrystallization of α phase due to combination of reduction and subsequent MA heat treatment.

  4. 4.

    The α phase refinement as well as Vf of primary α is found to be lower in samples deformed at high temperatures ~ Tβ., whereas in samples forged at lower temperature, the refinement of α is observed and its Vf is high.

  5. 5.

    Impact strength is found to be moderate to low with increased % reduction followed by MA heat treatment. However minor improvement is observed in samples subjected to large deformation + MA heat treatment, implying that very fine recrystallized α contributes to the improvement in impact strength. Impact strength is also found to be relatively higher in RA condition further indicating completely recrystallized α grains improve the impact strength.

  6. 6.

    Similar trend in strength with lesser scatter is observed in forged + RA condition at different amounts of deformation. Marginally lower strength in RA condition is attributed to presence of large extent of recrystallized α.

  7. 7.

    Fraction of smaller grains is higher for lower temperature deformed + RA heat-treated samples, but the uniformity in α grain size is obtained at intermediate deformation temperature (850–900 °C).

  8. 8.

    Large deformation results in refined primary α grains and results in improvement in tensile strength with marginal reduction in impact strength. Deformation associated with α recrystallization resulted in minor reduction in tensile strength and improvement in impact strength obtained through low-temperature and large deformation reduction + RA or MA heat treatment.

  9. 9.

    % Elongation is found to be varying (14–18%) in MA conditions and further higher elongation was observed in samples subjected to RA heat treatment.

  10. 10.

    Samples forged below Tβ followed by heat treatment (MA/ RA) exhibited predominantly ductile features, whereas samples forged above Tβ exhibited cleavage-type fracture features.