Features of densification, structure formation, and properties of powder titanium under hot die forging

Introduction

The unique properties of titanium – its high specific strength, corrosion resistance, and good biocompatibility – determine the wide application of this metal and its alloys in aerospace, automotive engineering, medicine, and other industries [1]. The prospects of powder metallurgy technologies for manufacturing titanium and titanium alloy products are largely driven by the high cost of alternative casting technologies [2; 3]. The production of cast titanium is further complicated by significant losses during machining. The average material utilization factor in manufacturing titanium products from wrought stock does not exceed 18 % and is often much lower [4].

Research in titanium powder metallurgy has been ongoing for more than 60 years. Nevertheless, there are still relatively few examples of its practical application. In many cases, consumers prefer cast titanium over powder-based titanium due to unsatisfactory mechanical properties and/or high cost of the latter [1]. Reducing the cost of powder titanium products requires the development of economically efficient technologies both for producing titanium powder and for fabricating finished parts [5].

One promising approach to lowering the cost of titanium powder is the use of by-products from sponge titanium production as the raw material. Another promising route is the hydride–dehydride (HDH) process, which involves mechanical milling of pre-hydrided sponge titanium, turnings, scrap, and other machining waste, followed by dehydrogenation of the milled material [6–8].

Improving the mechanical properties of powder titanium products during processing is also of critical importance, although success in this area often directly depends on the quality of the starting powder [9]. The ability of titanium products to resist static and dynamic loads is determined by the presence of residual porosity, non-metallic inclusions, and microstructural characteristics [10]. The residual porosity of powder titanium depends on the production technology. In recent decades, the press–sinter method has become the most widespread [1; 11; 12]. Porosity in sintered (undeformed) titanium typically ranges from 3 to 15 %. Since pores act as stress concentrators, they reduce the effective cross-section of the sample and diminish its load-bearing capacity. Therefore, improving the physical and mechanical properties of powder titanium requires the development of effective methods to reduce porosity [1].

The use of fine, amorphous, and nanostructured powders, thermocycling in the α → β phase transformation range (t = 800–1100 °C), thermomechanical treatment, and spark plasma sintering are effective methods of sintering activation [11; 13–15]. In addition, the use of activating additives also provides positive results; these are conventionally divided into two types [16].

Additives of the first type activate self-diffusion of the base element, thereby improving densification conditions during sintering. The diffusion rates of transition metals and phosphorus in α-Ti are 3–5 orders of magnitude higher than its self-diffusion [17]. Self-diffusion of titanium is also enhanced by elements that reduce the solidus temperature of the alloy [16].

The second type of additives includes elements that promote the formation of a transient liquid phase, which disappears during sintering and provides effective diffusion pathways for mass transport, thereby increasing the density of sintered products. Moreover, liquid-phase formation may be related to eutectic reactions between two precipitated phases (solid–liquid sintering). Significant sintering activation is observed during liquid-phase sintering of amorphous titanium powders [13]. Powders of iron, nickel, silicon, cobalt, and copper are used as activators [18–26]. However, liquid-phase sintering technology has several drawbacks: distortion of the shape of preforms, segregation of solid and liquid phases, and rapid grain growth, all of which negatively affect the mechanical and service properties of the products.

High diffusion rates are also observed during sintering of titanium powder produced by the hydride–dehydride process [27]. This is due to the formation of a large number of lattice defects during dehydrogenation, which promote diffusion processes [28].

Among the more recent technologies of sintering activation for powder titanium are induction vacuum sintering methods and FAST (Field-assisted sintering technology), which employs strong electromagnetic fields [29; 30]. A major limitation of these methods is the need for expensive specialized equipment, which significantly increases production costs [31].

Limiting factors also include the low ductility and toughness of powder titanium and its alloys, which adversely affect their fracture toughness and fatigue strength [32; 33]. In this regard, it is worth recalling a recent event. At the 2021 Tokyo Olympics, an accident occurred involving the bicycle of an Australian rider, caused by the fatigue failure of handlebars manufactured by additive technology from Ti–6Al–4V powder alloy [34; 35].

Although the specific strength of powder titanium is about four times higher than that of steels, its resistance to cyclic loading is often low due to its limited ductility. In addition to porosity, ductility, fracture toughness, and fatigue strength are negatively affected by interstitial impurities (O, H, N, C) and microstructural characteristics [32]. The presence of impurities is associated with the high chemical affinity of titanium for these elements, necessitating heating operations in vacuum or inert atmospheres.

The dependence of fatigue life on microstructural characteristics is determined by the slip length during plastic shear of the crystal lattice. In titanium-based materials, this corresponds to the α-phase grain diameter in equiaxed structures, the α-phase plate width in basket-weave structures, or the colony size in lamellar structures [33]. A reduction in slip length due to microstructural refinement contributes to improved fatigue strength. It should be noted that prolonged high-temperature exposure during sintering leads to grain growth. This highlights the relevance of employing methods in powder titanium processing that enable simultaneous reduction of porosity and refinement of structural constituents.

A promising approach to addressing this challenge lies in hot deformation methods. The first studies in this area were carried out in the late 1950s to early 1960s. At Novocherkassk Polytechnic Institute, I.N. Goncharov proposed a method of hammer forging titanium powder particles or crushed titanium sponge to achieve welding [36]. Densification was performed under isothermal conditions in a preheated die at 900–920 °C. The resulting samples exhibited high density (4.46–4.48 g/cm³), along with high ductility and deformability of forged powder titanium. Since the starting material subjected to hot deformation was powder or sponge particles, this method developed by I.N. Goncharov can, in modern terminology, be classified as direct powder forging (DPF).

In industrial practice, hot forging of loose powder prepacked into cans was introduced later, in the 1970s, for the production of high-alloy tool steels [37]. More recently, studies have investigated the feasibility of applying DPF technology to titanium powder processing [31; 38]. A high-density Ti–6Al–4V powder material with a homogeneous lamellar two-phase (α + β) structure was obtained, characterized by high ductility, low impurity content both on the surface and in the bulk of the experimental samples, and only minor deviations from the chemical composition of the starting powder. Using this technology, a femoral implant was fabricated that meets the requirements of ASTM F136-13 (2021) [39].

One advantage of the DPF process is that the high stresses and strains generated during forging at the powder – can interface promote the fracture of intermetallics and oxides formed there, facilitating subsequent can removal. However, this advantage may present a problem: after can removal, the powder material surface exhibits a fracture-like relief, necessitating additional machining.

Research on hot forging of porous titanium preforms (HFPTP) was conducted by S.S. Kiparisov and co-authors in the late 1960s to early 1970s [11, pp. 47, 48]. These studies established the optimal parameters for HFPTP: heating temperature of 900 °C and impact energy of 2000–2500 kJ/m². However, the upper limit of the heating temperature range did not exceed 900 °C, and the impact energy values were expressed per unit area rather than per volume of the sample in the compact state, which complicates their practical application for densification.

Subsequently, V.A. Pavlov and co-authors carried out research on hot deformation of titanium powders and other non-ferrous metals [40; 41]. They examined the forging behavior of porous titanium preforms in open and closed dies, heating under various conditions (unprotected, protected with glass coatings, in argon, and in vacuum), as well as deformability characteristics and energy consumption during cold hydrostatic pressing of titanium powder. However, the energy requirements for densifying powder titanium to a pore-free state by hot forging were not determined. Such data are also absent in the works of Yu.G. Dorofeev and his scientific school. The lack of information on the parameter W_max (maximum specific densification work, i.e., the work required to reach monolithic density [42; 43]) for HFPTP complicates the selection of technological modes, since densification energy is one of the key control parameters in dynamic consolidation of powder materials [44].

The aim of this work was to study the densification behavior, microstructure formation, and properties of powder titanium during hot die forging (HDF).

Experimental procedure

The experimental samples were fabricated from electrolytic titanium powder grade PTES-1 according to TU 48–10–22–85. A powder fraction of –0.63 + 0.18 mm with a bulk density of 1.5·10³ kg/m³ was used. The processing flowcharts for sample fabrication are shown in Fig. 1. At the initial stage of the study, the energy characteristics of hot densification were determined using cylindrical samples with dimensions diameter 20 × 8 mm (scheme 1). The ratio of the final height of the sample (h_f) to its diameter (d) was kept constant (h_f/d = 0.4 = const). The porosity of the cold-pressed billets was in the range of 22–25 %. After heating to various temperatures (800–1050 °C), the billets were further compacted on a laboratory drop hammer with a falling weight of 50 kg. The densification work (W) was varied in the range of 30–150 MJ/m³. Preheating of the billets before further compaction was carried out in a portable heat-resistant steel container into which helium was supplied. The container with the sample was then placed into a laboratory muffle furnace with silicon carbide heating elements, which was also purged with helium.

The value of W_max was determined by a graphical–analytical method, processing the plots of lgW = f (lgβ) according to Dorofeev’s methodology (β – relative volume) [42; 43]. The density of the samples was measured by hydrostatic weighing in accordance with GOST 18898–89.

At the next stage of the study, prismatic samples with dimensions 130×26×15 mm were fabricated to determine the mechanical properties and conduct microstructural analysis. To reduce the probability of oxidation, the billets for prismatic samples were made bimetallic. The outer layer (sheath) of the billets was produced from high-pressibility iron powder grade ABC100.30 (Höganäs AB).

After HDF, the billets were subjected to machining. The fracture toughness (K_1с) was determined on type 4 samples (GOST 25.506–85) with dimensions 12.5×25×112.5 mm containing a pre-induced crack using the three-point bending method. The crack was initiated under cyclic loading. The values of K₁ were calculated according to the procedure [45]. Tensile tests were carried out on type II Gagarin samples (GOST 1497–84) with a working section diameter d₀ = 5 mm and a gauge length l₀ = 25 mm.

Based on the results of the first stage of HDF testing, the mechanical properties and fracture toughness were determined at a specific densification work of W = 195 MJ/m³ Since fully dense samples were not obtained in the first stage, the pre-deformation heating temperature of the billets was varied in the range of t_HDF = 1000÷1200 °С.

During the fabrication of samples according to technological scheme 3, porous cold-pressed billets were subjected to hydrogenation in hydrogen, followed by dehydrogenation during vacuum sintering. This step was intended to reduce impurity content and activate interparticle surfaces for subsequent hot forging, similar to the previously observed effect of hydrogenation–dehydrogenation of initial titanium powders during sintering [27]. For comparative analysis, control samples were produced according to technological scheme 2, which did not include the hydrogenation step of cold-pressed billets.

The fracture surfaces of the samples were examined using a Quanta 200 i 3D scanning electron microscope–microanalyzer and an MBS 9 binocular microscope. The fraction of fracture surface components was determined by grid and area methods applied to electron microscopic images [46], with 10–12 fields analyzed. Metallographic studies were performed with an AltamiMET 1M optical microscope (Altami Ltd., Russia) on both etched and unetched polished sections. Kroll’s reagent (2 ml HNO₃ + 2 ml HF + 96 ml H₂O) was used as the etchant, with an etching time of 20 s.

Results and discussion

Fig. 2 presents the lgW = f (lgβ) dependences for porous titanium billets subjected to different pre-deformation heating temperatures. The relationships between the relative volume of powder titanium and the specific densification work, plotted in logarithmic coordinates, exhibit a linear form, which is characteristic of ductile materials [42]. The W_max were determined by extrapolating curves 1–4 to their intersection with the ordinate axis (dashed lines).

The dependence of the W_max values determined in this way on the hot die forging temperature is shown in Fig. 3. It is evident that the dependence is nonmonotonic. Increasing t_HDF in the 800–950 °C range leads to a decrease in W_max due to the enhanced plasticity of the α-phase. At t_HDF = 950–1000 °C, the W_max value increases, which is attributed to the α → β phase transformation and the formation of the β-phase, characterized at these temperatures by reduced plasticity and higher strength. At t_HDF > 1000 °C, the plasticity of the β-phase increases, resulting in a decrease in W_max.

The effect of phase transformations on densification behavior and deformability has been previously reported in studies of dynamic and explosive hot pressing of porous iron-based billets [42; 43]. With regard to titanium and its alloys, the literature data remain contradictory [11]. Some authors indicate a monotonic increase in plasticity with temperature across all titanium alloys [47], whereas others note a nonmonotonic trend [48]. The reduction in plasticity in the α → β transformation range has been linked to the development of a coarse-grained structure [49].

In our experiments, the grain size of samples fabricated at t_HDF = 950 °C, was 10–20 µm (Fig. 4, a), corresponding to grades 2–3 on the titanium alloy microstructure scale used for metallurgical quality control [49]. At t_HDF = 1000 °C, the grain size increased to 25–35 µm (grades 5–6) (Fig. 4, b). This provides sufficient grounds to assume that the nonmonotonic W_max = f (t_HDF) dependence is also associated with grain growth during phase transformation. The samples exhibited a lamellar intragranular α-phase structure with areas of basket-weave microstructure. This morphology is related to the temperature distribution within the billet during deformation and subsequent cooling.

A similar lamellar α-phase intragranular structure is observed in titanium castings under slow cooling conditions [48]. In our experiments, cooling was relatively slow: after removal from the die preheated to 600 °C, further cooling proceeded under a layer of potassium chloride. Consequently, the formation of the lamellar structure can be attributed to β → α recrystallization. In this process, α-phase lamellae nucleate at β-grain boundaries and grow inward. The formation of basket-weave regions is apparently associated with cooling of the billet surface layers in contact with the die walls during hot forging. Deformation of these zones begins in the β-region and ends in the α + β region, which provides favorable conditions for basket-weave formation [50].

The W_max values obtained in the investigated pre-deformation heating range were 150–200 MJ/m³, whereas the heat of fusion of titanium in comparable units is significantly higher – 1411 MJ/m³ [51]. This indicates that in powder titanium processing by hot die forging, as in other metals, the energy expenditure is considerably lower than during melting, since only part of the particle material undergoes plastic deformation, determined by the material’s ductility [43]. The measured W_max values are also much lower than the reported activation energy of powder titanium sintering (~15,000 MJ/m³) [52], which approximately corresponds to the activation energy of self-diffusion in β-Ti.

Fig. 5 presents the dependence of fracture toughness and mechanical properties of powder metallurgy titanium on the hot die forging temperature. Samples were fabricated according to technological schemes 2 and 3. The K_1c = f (t_HDF) dependences are nonmonotonic. Increasing t_HDF to 1150 °C leads to higher K_1c due to improved deformability of the porous billet material. A further increase in t_HDF to 1200 °C results in a decrease in K_1c, which is attributed to grain refinement (see Fig. 4, c).

Grain refinement in samples fabricated at t_HDF > 1150 °С, is associated with the higher post-deformation cooling rate caused by the increasing temperature gradient between the heated porous billet and the die. Throughout the studied t_HDF range, the fracture toughness of samples fabricated according to scheme 3 was consistently higher than that of control samples fabricated according to scheme 2 (cf. curves 1 and 2 in Fig. 5, a). This difference can be explained by distinct fracture mechanisms in the compared types of samples.

Raising the hot die forging temperature also altered the fracture characteristics of the samples. In control samples fabricated at t_HDF = 1000–1100 °C, the fracture surface of the initial fatigue crack contained secondary cracks in the failure origin (Fig. 6, a). In the final fracture zone, which formed under static loading, the fracture was of a ductile fine-dimple type with areas of interparticle fracture (Fig. 6, b, c). Dimple sizes ranged from 4 to 6 µm. At t_HDF > 1100 °С, secondary cracks in the fracture origin were not observed; however, interparticle fracture areas remained present in the final fracture zone. The dimples of ductile fracture in this zone were larger – 10–12 µm – than in control samples fabricated at 1000–1100 °C, indicating higher material ductility (Fig. 6, d).

The fracture surfaces of samples fabricated according to scheme 3 were characterized by the absence of secondary cracks in the fracture origin at all studied t_HDF values (Fig. 6, e). In the final fracture zone, interparticle fracture was not detected, which indicates higher fracture energy compared to control samples. The trend in dimple size variation with t_HDF was similar to that observed for control samples (see Fig. 6, f). Thus, the predominant factor influencing the fracture toughness characteristics of scheme 3 samples is the hydrogenation–dehydrogenation treatment. An analysis of impurity composition in the initial titanium powder and in the fabricated samples showed that under scheme 2, the oxygen concentration in the resulting material increased by about two fold (see Table).

In samples obtained according to scheme 3, the oxygen content was significantly reduced, while no substantial changes were observed in the amount of other impurities. This is explained by the fact that hydrogen released from titanium hydrides during vacuum sintering of porous billets after hydrogenation promotes the reduction and activation of oxides localized on the surfaces of open pores. A similar sintering activation effect when using hydrogenated titanium powder was reported earlier [27]. In our experiments, the activation of pore surfaces and their refinement from oxides improved the conditions for interparticle bonding during subsequent hot die forging [53]. In contrast, the relatively high oxygen content in control samples (scheme 2), associated with the presence of oxides on the collapsing pore surfaces during hot die forging, impaired interparticle contact conditions and led to lower fracture toughness compared to scheme 3 samples (Fig. 5, a).

Strength and ductility characteristics were less sensitive to the quality of interparticle bonding and were mainly determined by grain size. Accordingly, the dependences of σ_u and δ on t_HDF were monotonic: increasing temperature led to growth of these properties throughout the studied range (see Fig. 5, b). In this case, the dominant factor was grain refinement with increasing t_HDF (see Fig. 4, c). Improved interparticle bonding conditions during hot die forging of billets subjected to hydrogenation–dehydrogenation resulted in higher ductility of scheme 3 samples compared to control samples (cf. curves 3 and 4 in Fig. 5, b).

Conclusions

1. The maximum specific densification work required to achieve monolithic density in porous powder titanium was determined for different pre-deformation heating temperatures of billets. The nonmonotonic character of the W_max dependence on t_HDF is associated with the formation of a coarse-grained structure and reduced ductility of the deforming material in the α → β phase transformation temperature range.

2. A processing technology was proposed for producing hot die forged powder titanium, involving hydrogenation–dehydrogenation of porous billets. This treatment promotes hydrogen-assisted reduction by hydrogen of oxides localized on the surfaces of open pores and their activation, thereby improving interparticle bonding during subsequent hot die forging and enhancing the fracture toughness and ductility of the resulting samples compared to control samples.

3. Increasing the hot die forging temperature of porous titanium billets contributes to higher fracture energy and improved ductility of the samples. At the same time, the likelihood of secondary crack formation in the origin of the initial fatigue crack is reduced, and the size of dimples in the final ductile fracture zone increases.