INTRODUCTION

Modern industry has increasingly rigorous demands for the structure and mechanical properties of construction and functional materials. Titanium alloys possess a unique combination of corrosion resistance and strength at room and elevated temperatures having sufficiently low density [1–12]. Due to their properties, titanium alloys have found broad application in aerospace, transport, and chemical industry, as well as in biomedicine. Two-phase titanium alloys of a martensite class are widely used, and all types of semifinished products are fabricated from them. However, one substantial disadvantage of these alloys is a considerable amount of alloying elements entering the composition, including high-cost ones: up to 6.9 Al, 4.5 V, and 5.0 Mo. The authors of [13–17] previously showed that codoping with iron and copper positively affects the structure of forged titanium alloys. However, in connection with the insufficient manufacturability of these alloys, the investigation into the deformation behavior and microstructure evolution in a broad range of strain rates and temperatures, as well as the construction of the rheological model of the relation of the flow stress with the plastic deformation parameters, are required to accelerate the development of industrial methods of pressure treatment. The goal of this study is to determine the deformation stress under compression and study the influence of thermal-deformation treatment modes on the structure of the Ti–3.8Fe–4.4Cu–0.2B titanium alloy.

EXPERIMENTAL

We selected the Ti–3.8Fe–4.4Cu–0.2B alloyFootnote 1 as the object of investigation. Boron was introduced to fabricate the fine-grained structure in a cast state. Alloy ingots 6 mm in diameter were fabricated by alloying pure components in a vacuum induction furnace and subsequent accelerated crystallization into a massive copper mold under an argon pressure of 0.3 MPa. We found three ingots of one composition 6 × 50 mm in size, of which the samples 10 mm in height were cut.

The uniaxial compression tests with true strain ε = 0.9 were performed using a Gleeble 3800 physical simulation system of thermomechanical processes (DSI, United States) at 750, 800, and 900°C and strain rates of 0.1, 1, and 10 s–1. A cylindrical sample cut from an ingot 6 mm in diameter and 10 mm in height was clamped into tungsten carbide heads, heated to the testing temperature with a rate of 5 K/s by direct current flowing, and held for 10 s. The sample temperature was monitored using a chromel–alumel thermocouple soldered to a central sample part. Graphite foil and a nickel-based lubricant were laid between the heads and sample faces to decrease friction during the test. Heating and deformation were performed under high vacuum (the residual pressure was smaller than 10–3 Pa). After the test, the sample was forcedly cooled by a compressed air jet for further microstructural analysis. The measured cooling rate in a temperature range of 900–500°C was larger than 50 K/s, which is higher than the critical rate for most titanium alloys.

In order to determine the true stress, we corrected the primary data according to the procedure [18]. This correction is necessary because of the temperature variation during the deformation (which is especially important for tests with increased rates) and also because of the presence of friction between the heads and a sample.

The alloy microstructure in the initial and deformed states was investigated by scanning electron microscopy (SEM) using a Tescan Vega 3 LMH microscope with an X-Max 80 energy dispersion detector (Tescan, Czech Republic). The chemical composition of alloys was determined by electron probe microanalysis of five microstructure segments 100 × 100 μm in size. Slices for microstructural investigations were prepared using a Struers LaboPol-5 grinding–polishing machine (Struers, Netherlands).

RESULTS AND DISCUSSIONS

The structure of the Ti–3.8Fe–4.4Cu–0.2B alloy in the cast state (Fig. 1) consists of α phase (dark segments), β phase (light segments), and titanium boride TiB2 (dark particles). The results of an analysis of the chemical composition and volume fraction of phases are presented in Table 1. It is seen that the larger part of iron and copper dissolved in the bcc lattice of β-Ti, while only a small amount of copper is dissolved in α‑Ti with the hcp lattice.

Fig. 1.
figure 1

Microstructure of the Ti–3.8Fe–4.4Cu–0.2B alloy in the cast state.

Full size image
Table 1. Volume fraction, average chemical composition, and average phase size in the cast state
Full size table

Compression curves of the samples are presented in Fig. 2. It is seen that the flow stress regularly increases with an increase in the rate and a decrease in temperature. A maximum is observed at the initial compression stage at all temperatures and strain rates, after which the flow stress decreases, which is caused by the active development of dynamic recrystallization.

Fig. 2.
figure 2

Deformation curves at a rate of (a) 0.1, (b) 1, and (c) 10 s–1. t, °C: (1) 750, (2) 800, and (3) 900.

Full size image

The relation between the flow stress at the stated stage, rate, and deformation temperature is described well by the Arrhenius equation [19]:

$$\dot {\varepsilon } = {A}{{\sigma }^{n}}{{e}^{{ - \frac{Q}{{RT}}}}},$$
((1))

where \(\dot {\varepsilon }\) is the deformation rate, s–1; T is the temperature, K; Q is the effective deformation energy, J/mol; and А, n are constants.

Unknown parameters A, n, and Q were found by minimizing the error between calculated and experimental values of the flow stress at the degree of deformation of 0.5 corresponding to the stated deformation stage. This results in A = 7.4, n = 4.1, and Q = 220 kJ/mol. An average calculation error according to the designed model was 6%. The effective activation energy of deformation lies between the activation energy of self-diffusion in α-titanium (it is from 169.1 [20] to 193 kJ/mol [21]) and in β-titanium (251.2 [20] and 282.9 kJ/mol [22]), which evidences that both phases actively participate in the deformation process. Our values of model parameters can be used when constructing finite-element models and optimizing the technology of actual pressure-treatment processes.

It follows from Fig. 3 that the structure of the samples quenched from the deformation temperature is presented by α and β phases and boride inclusions. It is seen that inclusions of titanium borides during the deformation divide into separate particles 0.5–1.5 μm in size distributed with higher uniformity than in the cast state. Herewith, the average grain size of α and β phases varies insignificantly after deformation according to almost all studied modes. The grain growth of the α phase is found only during deformation according to the mode t = 900°С and \(\dot {\varepsilon }\) = 10 s–1 (Figs. 3d, 4b). This can be associated with adiabatic sample heating at a high strain rate. In addition, the dynamic recrystallization and globularization of the α phase possibly occur at t = 900°С and \(\dot {\varepsilon }\) = 0.1 s–1 [23], while these processes have no time to proceed completely at \(\dot {\varepsilon }\) = 10 s–1. The recrystallization and grain growth in the heavier alloyed β phase can be retarded by the atoms of alloying components, which is why the grain size of the β phase does not vary considerably. The volume ratio between α and β phases decreases with an increase in the deformation temperature irrespective of the rate (Fig. 4a).

Fig. 3.
figure 3

Microstructure of the samples in the deformed state. (a) \(\dot {\varepsilon }\) = 0.1 s–1, t = 750°С; (b) \(\dot {\varepsilon }\) = 0.1 s–1, t = 900°С; (c) \(\dot {\varepsilon }\) = 10 s–1, t = 750°С; and (d) \(\dot {\varepsilon }\) = 10 s–1, t = 900°С.

Full size image
Fig. 4.
figure 4

Dependence of the (a) volume fraction and (b) grain size of α and β phases on temperature and strain rate of the Ti–3.8Fe–4.4Cu–0.2B alloy.

Full size image

CONCLUSIONS

(i) The microstructure of the Ti–3.8Fe–4.4Cu–0.2B alloy in the cast state and after hot deformation under uniaxial compression in various temperature-rate conditions is investigated. The alloy structure in the cast and deformed states contains α phase, β phase, and titanium diboride particles. It is shown that the β phase contains up to 7.1% Fe and 7.7% Cu in the cast state and is considerably more heavily alloyed when compared the α phase containing 0.7% Cu.

(ii) Compression tests of the Ti–3.8Fe–4.4Cu–0.2B alloy are performed at 750–900°C and deformation rates 0.1–10 s–1, and a model of relation of the flow stress with thermal-deformation treatment parameters is designed.

(iii) It is shown that the preferential hot-treatment pressure mode for the formation of the fine-grained structure in the studied alloy is a temperature range of 750–800°C because the grain size of the α phase increases from 2.2 to 4.5 μm with an increase in temperature to 900°C. The studied alloy has prospects of using as economically alloyed material of an increased corrosion resistance and strength.