Introduction

Titanium alloys have many advantages compared to steels which are commonly used in the manufacturing industry, such as high specific strength and excellent corrosion resistance.1,2,3,–4 However, titanium alloys are very expensive because of their high material costs, high buy-to-fly ratio for making the components and high manufacturing costs. These factors limited titanium alloys from being widely used in the manufacturing industry, especially for civil industry, despite their properties being very attractive.5 The powder metallurgy (PM) method is able to nearly-net-shape form intricate products, leading to significantly reducing the manufacturing cost and saving materials. The drawbacks for the PM method are that it is easy to introduce impurities into the final product during the processing, and difficult to make high dense materials without post-processing, which have a potential to deteriorate the mechanical properties of PM titanium alloys. Furthermore, the cost of PM titanium alloy products is still not affordable as a result of using pre-alloyed titanium powders as starting materials. Hot processing of the blended elemental titanium powder mixtures, such as powder compact extrusion, is a potential cheap and effective method to produce high-density Ti-6Al-4 V alloy, which have comparable mechanical properties with those of ingot metallurgy counterparts.6,7 In this paper, we will discuss a possible route that combines fast heating (induction and microwave heating) and hot processing to produce complicated and non-standard titanium alloys from cheap powders, such as Ti-5Al-5V-5Mo-3Cr (Ti-5553) (wt.%) and Ti-5Fe (wt.%) alloys. This process is a possible cost-effective method to produce titanium alloy parts with comparable mechanical properties to those of ingot metallurgy titanium products, and the microstructures and mechanical properties of the Ti-5553 and Ti-5Fe alloys prepared are discussed.

Experimental Procedure

Two titanium alloys, with nominal compositions of Ti-5Al-5Mo-5V-3Cr (Ti-5553) and Ti-5Fe (wt.%), were prepared by a process combining fast heating (induction and microwave heating) and hot processing from elemental powders and master alloy powders. The starting materials were Ti powder produced by a hydride–dehydride (HDH) process (− 200 mesh), Al powder (purity: 99.9%), pure carbonyl Fe (− 45 µm) and Al35V65, Al15Mo85 and Al3070Cr (wt.%) master alloy powders (the particle size was smaller than 75 μm, commercial purity) supplied by Dalian Rongde, PR China. The powder mixture, with a designed composition, was mixed for 20 h using a roller mill at a speed of 200 rpm. Then, the mixed powder mixture was warm-compacted at 230°C under a pressure of 400 MPa into a cylindrical shape using a 100-ton hydraulic press in air. The relative density of the compacts was about 83–85%. After the compaction, the Ti-5553 powder compact was induction-heated (heating rate was about 180°C/min) to 1300°C, and holding the temperature for 10 min, and then hot-pressed, and the Ti-5Fe powder compact was consolidated by microwave sintering (heating rate was about 28°C/min) at 1300°C for 22 min. Then, the as-consolidated billets were reheated using induction furnace to the desire temperature (950°C for Ti-5553 and 1050°C for Ti-5Fe) for extrusion in air. The extrusion ratio was about 9:1, and the extrusion processes were performed by a 300-ton hydraulic press (XJ 300; Wuxi Yuanchang Machinery, PR China).

Scanning electron microscopy (SEM) (HITACHI S4700) was used to examine the microstructure of the hot-pressed, microwave-sintered and extruded titanium alloy billets and samples. The ground and polished surfaces of the samples for SEM examination were etched in a modified Kroll’s reagent consisting of 2 vol.% HF, 4 vol.% HNO3 and 94 vol.% H2O. The tensile tests were conducted at room temperature using an Instron (INSTRON 4204) universal testing machine with dog-bone-shaped specimens, having a rectangular cross-section of 2 mm × 2 mm and a gauge length of 20 mm. The strain was measured using an extensometer with a gauge length of 10 mm. The strain rate used for tensile testing was 10−4 s−1. The oxygen contents of the powder mixture and the solid parts were measured using the Inert Gas Fusion Method by the standard of ASTM-E-1019 at Durkee Testing Laboratories, USA.

Results and Discussion

Microstructure and Properties of Ti-5553 Alloy

Figure 1 shows the XRD patterns of the Ti-5553 alloys processed under different conditions. Only β phase peaks appear in Fig. 1a, which means that the Ti-5553 powder compact was fully alloyed during the process of induction heating to 1300°C, holding the temperature for 10 min and hot pressing, forming Ti-5553 alloy materials. The microstructures of the Ti-5553 alloy processed under different conditions are presented in Fig. 2. From the SEM microstructure of the hot-pressed Ti-5553 alloy billet (Fig. 2a), it can be clearly seen that the microstructure of the hot-pressed billet are mainly composed of equiaxed grain structure, with a grain size of up to 50 µm. The microstructure also suggests that the master alloy powder particles were completely dissolved into the titanium matrix and the porosity was less obvious, with a few closed pores left inside the equiaxed grains. The relative density of the hot-pressed Ti-5553 was measured to give a value of about 97%, which confirms that the hot-pressed billet had less porosity than the compact. BSE microstructure (Fig. 2b), suggesting that only a small amount of α phase, with acicular morphology, precipitated from the β matrix and located in inside the grains and at the grain boundaries. Due to the amount of α phase being very small, there is no α peak in the XRD pattern (Fig. 1a). The reason why master alloy powder particles could be dissolved into the titanium matrix in the short process of induction heating, temperature holding and hot pressing, could be explained from the pressure-assisted sintering theory. Applying pressure (about 400 MPa) on the 1300°C-induction-heated Ti-5553 compact would:6,7 (1) enlarge the contact area of the powder particles; (2) cause plastic flow through dislocation gliding; (3) promote grain boundary diffusion; and (4) accelerate volume diffusion from the grain boundary. All these factors would significantly accelerate the master alloy powder particles dissolving into the titanium matrix.

Fig. 1
figure 1

XRD patterns of the Ti-5553 alloys: (a) hot-pressed at 1300°C and with a holding time of 10 min, (b) extruded from the hot-pressed billet at 950°C, and (c) the extruded alloy double-aged (450°C/6 h + 675°C/30 min)

Full size image
Fig. 2
figure 2

Microstructures of Ti-5553 alloy: (a, b) hot-pressed at 1300°C and with the holding time of 10 min, (c–f) extruded from the hot-pressed billet at 950°C, and (g–k) the extruded alloy double-aged (450°C/6 h + 675°C/30 min)

Full size image

After extrusion at 950°C, both α and β peaks are observed in the XRD pattern (Fig. 1b); the number and intensity of the α peaks were quite small, indicating that the amount of α phase in the as-extruded Ti-5553 alloy was not very high. However, more α peaks appeared, as shown in Fig. 1c, when the extruded bar was double-aged at 450°C for 6 h and then 675°C for 30 min. The intensity of the α peaks was significantly increased, compared to that of the 950°C-extruded alloy. This was because the sample was cooled in an argon flow after extrusion and the cooling rate was faster than that of furnace cooling (close to equilibrium condition), and the precipitation of α phase from β phase was significantly suppressed during and after the extrusion, leading to the extruded bar being mainly composed of β phase. The phase transformation of β/α was triggered by the heat treatment, so that substantial α phases could precipitate from the β matrix, as a result of which many α peaks with high intensity were observed in the XRD pattern of the double-aged sample. The SEM images clearly show that the 950°C-extruded Ti-5553 alloy has a strip microstructure (Fig. 2c), in which dark strips are composed of an equiaxed grain structure with dimensions of 5–20 µm (Fig. 2d), and bright strips show many acicular α precipitations inside the original β equiaxed grain (Fig. 2e and f). After double-aging (450°C for 6 h followed by 675°C for 30 min).

The strip microstructure has almost disappeared and become more homogeneous than that of the as-extruded Ti-5553 alloy, as shown in Fig. 2g. In the high-magnification SEM images (Fig. 2h–k), α precipitations are clearly observed, which are composed of acicular α, fine particle α and grain boundary α. Since the Ti-5553 alloy was extruded at high temperature and followed by air cooling to room temperature, the stability of the as-extruded Ti-5553 was quite low, and the driving force for the nucleation and growth of α phase was high during aging. Thus, α precipitation could form more homogeneously throughout the microstructure, evidenced by the formation of large numbers of α precipitations within the β grains and discontinuous α phases at the original β boundaries.8

The mechanical properties of Ti-5553 alloys processed at different conditions are listed in Table I. It can be concluded that the hot-pressed Ti-5553 alloy had very poor mechanical properties, no yield occurred and the ultimate strength was about 800 MPa. The mechanical properties of the as-extruded Ti-5553 alloy were significantly improved compared to those of the hot-pressed alloy, with a yield strength of about 1270 MPa, an ultimate strength of about 1370 MPa and the elongation to fracture of about 3%. The double-aging heat treatment had an important impact on the as-extruded Ti-5553 alloy and made a significant contribution to increase its mechanical properties: the yield strength reached 1435 MPa, the ultimate strength was about 1530 MPa, and the elongation to fracture was increased to 8.5%. This was about a 13% increase in yield strength, about an 11% increase in ultimate strength, and about a 183% increase in the elongation, compared to those of the as-extruded Ti-5553 alloy.

Table I Mechanical properties of Ti-5553 and Ti-5Fe alloy processed at different conditions
Full size table

The low mechanical properties of the hot-pressed Ti-5553 alloy were mainly caused by the equiaxed microstructures and unstable single β phases resulting from fast cooling after hot pressing. For the as-extruded Ti-5553 alloy, the strength was significantly improved and yield occurred, because of the fine microstructure and small percentage of α particles precipitated from the β matrix, which enhanced the alloy strength through precipitation hardening. After the double-aging heat treatment, there were many α precipitations with different morphologies: fine particle α, acicular α and discontinued grain boundary α were precipitated from the β titanium matrix, which contributed to both the strength and the ductility, because fine particles and/or the acicular structure were harder than the β phase and harder to be deformed. α/β interfaces pinned the movement of the dislocations, increasing the strength,9,10 and the grain boundary α could be deformed by slipping and shearing mechanisms so that the dislocations could be activated and accumulated within them, improving the ductility.11,12 Due to the finer microstructures and α precipitations formed in the as-extruded Ti-5553 alloy compared with those of the ingot Ti-5553 alloy, the strength of the Ti-5553 alloy was much higher than that of the ingot metallurgy Ti-5553 alloy.13 The oxygen contents for both the hot-pressed and extruded Ti-5553 alloys were 0.39 wt.%, which were about a 0.06 wt.% increase compared to the starting powder mixture (0.33 wt.%). Unlike the Ti-6Al-4 V alloy, there was no critical value for the oxygen content in the β titanium alloys. It has been reported that the elongation of the β titanium alloys was not significantly affected by the contained oxygen level due to beta titanium alloys being less sensitive to the influence of oxygen on the tensile ductility than alpha titanium alloys.14 Yang et al. found that the tensile elongation was more than 8% for the as-sintered near-beta Ti-10 V-2Fe-3Al alloy with an oxygen level of 0.59 wt.%.15 Thus, the microstructure takes a more important role to render the Ti-5553 alloy with better mechanical properties compared to the oxygen.

Microstructure and Properties of Ti-5Fe Alloy

After microwaving sintering at 1300°C for 22 min, the relative density of the as-microwaved-sintered Ti-5Fe alloy was significantly improved, reaching a value of about 94%, which was slightly lower than that of the conventional vacuum-sintered titanium alloy billet.16,17 The microstructures of the Ti-5Fe alloy processed under different conditions are shown in Fig. 3. For the as-microwave-sintered Ti-5Fe alloy, no undissolved Fe powder particles were observed and the microstructure consists of a lamellar α phase which is distributed in the β-phase matrix (Fig. 3a). The lamellar are coarser with average widths of 10 µm and lengths of 20–80 µm. The porosity distribution shows large isolated pores with a spherical appearance, confirming that significant densification occurred during the microwave sintering, but the porosity level is higher than that of the hot-processed alloy. EDS analysis results indicate that the Fe distribution was not homogeneous throughout the microstructure (Fig. 3b). After extrusion, the plastic deformation led to the elimination of the large pores left after microwave sintering (Fig. 3c and d). As a result, the density of the extruded material reached almost to that of the fully solid material. Although the overall grain size was still quite coarse, with a dimension of up to 120 µm, the lamellar spacing was much finer than that of the as-microwave-sintered Ti-5Fe alloy. This was mainly caused by the large plastic deformation induced during extrusion and fast cooling (air cooling) after extrusion.

Fig. 3
figure 3

Microstructures and EDS analysis of the Ti-5Fe alloy at the different conditions: (a) as-microwaved-sintered, (b) EDS analysis of the as-microwave-sintered Ti-5Fe alloy, and (c, d) the extruded Ti-5Fe alloy from the microwave-sintered billet along the extrusion direction

Full size image

The mechanical properties of the as-microwave-sintered and the as-extruded Ti-5Fe alloy are listed in Table I. No yield occurred for the as-microwave-sintered Ti-5Fe alloy, and its ultimate tensile strength is about 860 MPa. These test results were similar to those reported in the literature for vacuum-sintered alloys with similar composition,14 while another study,15 in which controlled faster cooling was used to control the size of the α-phase lamellar, showed similar ultimate strength but significantly higher elongation to fracture of 10%. This suggests that the combination of low sintered density (about 94%), large pores, coarse lamellar microstructure and inhomogeneous Fe chemical distribution were the possible reasons for the lower tensile strength, no yield and lower ductility for the as-microwaved-sintered Ti-5Fe alloy. Furthermore, the oxygen content of the as-microwaved Ti-5Fe alloy was 0.70 wt.%, which was higher than the critical oxygen value of 0.33 wt.% for the Ti-6Al-4 V alloy.14 Yan et al.14 reported that the elongation for the Ti-6Al-4 V (α + β) alloy was significantly reduced when the oxygen content was higher than 0.33 wt.% due to the formation of specific microstructure features: (1) fine α precipitates in β-Ti, (2) an α2-type cluster in α-Ti, and (3) an α-β-α layered grain boundary. Ti-5Fe is a type of α + β titanium alloy, thus another important reason to cause the as-microwave-sintered Ti-5Fe having a lower elongation was the as-microwaved-sintered Ti-5Fe alloy having a high oxygen content. For the Ti-5Fe alloy extruded from the microwaved-sintered billet, its oxygen content was about 0.56 wt.%, and the yield strength was about 1070 MPa and the ultimate strength was about 1090 MPa, and the elongation to fracture was about 1.5%. The strength was comparable to that of the ingot Ti-6Al-4 V alloy. The higher strength for the as-extruded Ti-5Fe alloy can be mainly attributed to finer lamellar spacing and higher relative density compared to that of the as-microwave-sintered Ti-5Fe alloy. The high tensile strength also demonstrates the valuable strengthening effect of the alloying element Fe. Nevertheless, the ductility of the as-extruded Ti-5Fe alloy was low, and this can mainly be attributed to the high oxygen content contained in the as-extruded alloy.

Conclusion

Titanium alloys with homogeneous microstructures could be rapidly produced by both induction heating plus hot pressing and microwave sintering from the elemental powder mixtures. After extrusion, the Ti-5553 alloy had higher mechanical properties, with a yield strength of 1270 MPa, an ultimate strength of 1370 MPa and the elongation to fracture of 3%. Double-aging significantly improved both the strength and ductility of the extruded Ti-5553 alloy. The elongation to fracture could reach 8.5% and meanwhile the strengths were increased to 1435 MPa for the yield strength and 1530 MPa for the ultimate strength. Microwave sintering plus extrusion could produce a Ti-5Fe alloy with high strength, but the elongation was quite low because of the high oxygen content contained in the microwave-sintered and post-extruded materials.