Formation of Alpha Case Mechanism on Titanium Investment Cast Parts

1. Introduction

At present the practical utilization of titanium (Ti) and its alloys ranges from sports equipment to the aerospace, power generation and chemical processing industries, and automotive, marine and medical engineering [1]. Yet, Ti alloy forming processes are much more laborious than one thinks, which is mainly due to the fact that in solid solution Ti combines with carbon, nitrogen and especially oxygen and additionally Ti is constrained to be melted in inert environments or in a vacuum [2]. For the intended applications, Ti and its alloy forming processes should incorporate forging, powder metallurgy and casting. A center bulkhead of the F-22 Raptor fighter aircraft is a good example of the difficulty in Ti alloy forging. Although the final Ti6Al4V alloy component weighs only about 150 kg, it should be forged initially from a single cast ingot of nearly 3,000 kg, which clearly exhibits extremely high machining losses in the Ti alloy forging process [2]. By the way, the powder metallurgy of Ti alloys offers a practicable way of producing complex components with less machining losses than in forging processes. However, the applications of the powder metallurgy have been limited by the freedom of size and shape. In addition, in order to ensure the soundness of the powder metallurgy products without porosity defects, hot isostatic pressing (HIP) is essentially required, which makes the powder metallurgy even more expensive [3].

Casting is a typical net shape forming technique in which molten metals are poured into a mold to produce an object of desired shape. However, the casting of titanium alloys is considered as only a near net shape forming technique. That is because regardless of mold type, titanium alloys casting has a drawback, called alpha-case (α-case), which makes it difficult to machine them and can lend themselves to crack initiation and propagation, due to their enormous reactivity in molten states [4]. Thus, the depth of α-case must be taken into consideration in the initial design for casting since the brittle α-case must be removed by chemical milling. For the reason, the wide use of titanium alloys casting has been limited, although titanium alloys castings are comparable, and quite often superior, to wrought products in all respects [3].

In order to avoid the -case problem, the expensive ceramics, such as CaO, ZrO₂ and Y₂O₃ have been adopted as mold materials since the standard free energy changes of the formation of their oxides are more negative than that of TiO₂. Regardless of thermodynamic approaches, the -case formation reaction still remains to be eliminated in the chemical milling processes. In order to develop the economic net shape forming technique of titanium alloys by casting process, it is necessary to take a much closer look into the -case formation mechanism. Therefore, the exact α-case formation mechanism must be examined and the development of -case controlled mold materials is required for practical applications of Ti alloys.

2. Evaluation of the alpha α-case formation mechanism

2.1. Conventional alpha α-case formation mechanism

The investment casting of titanium has a drawback, called ‘α-case’, which makes it difficult to machine and can lead to crack initiation and propagation, due to their enormous reactivity in molten states [4]. The α case is generally known to be developed by the interstitials such as carbon, nitrogen and especially oxygen dissolved from mold materials. In order to avoid this problem, the expensive ceramics, such as ZrO₂, ZrSiO₄, CaZrO₃, CaO and Y₂O₃ have been adopted for mold materials because their standard free energy changes of the formation of oxides are more negative than that of TiO₂ as shown in figure 1.

Figure 1.

Standard free energy changes of the formation of various oxides.

Regardless of the thermodynamic approaches, some amount of α-case still remains to be eliminated with the complex chemical milling processes except for Y₂O₃ mold. However, in the case of Y₂O₃ mold, there is not enough strength to handling due to the silica-free binder and there occur metal-mold reactions to some extent, too.

2.2. Evaluation of alpha α-case reaction

The wax patterns for the examination of α-case reactions were made by pouring molten wax into a simple cylindrical silicon rubber mold (Ø15 × 70 mm). Subsequently, the patterns were inspected and dressed to eliminate any imperfection or contamination, and coated with Al₂O₃ slurry. The Al₂O₃ shell molds were dried at a controlled temperature (298 K ± 1 K) and a relative humidity (40% ±1%) for 4 hrs. Dipping, stuccoing and drying procedures were repeated three times. After the primary layer coating, the patterns were coated with the back-up layers by the chamotte. To prevent the shell cracks, the dewaxing process of the shell molds were carried out at around 423 K and 0.5 MPa in a steam autoclave. Finally, the shell molds were fired at 1,223 K for 2 hrs.

Figure 2.

Schematic diagram of drop casting procedure of titanium with a plasma arc melting furnace.

In order to prevent any contamination from refractory crucibles, the specimens for the of α-case formation examination were prepared in a plasma arc melting (PAM) furnace with drop casting procedure. The pressure in the PAM was controlled at about 1.33×10^-1Pa by a rotary pump and then in order to minimize the effect of oxygen contamination, high purity argon was backfilled to the pressure of 4.9×10³ Pa.

After the drop casting, Ti castings were taken out of the molds, sectioned, polished and etched using a Keller solution [5]. The microstructure in the reaction region of the castings was observed using Olympus PME3 microscopy, and its hardness was measured using a Mitutoyo MVK-H2 microvickers hardness tester with the condition of 100 g load and 50 µm intervals. For a closer examination of the alpha-case formation, the distribution state of the elements analysis at the alpha-case region was performed by SHIMADZU EPMA 1600 and the phase structures of reaction products were identified by JEOL JEM-3011 TEM.

3. Thermodynamic calculation of alpha-case formation mechanism

3.1. Evaluation of the alpha-case reaction

In this research, Al₂O₃ was selected for the mold material because the standard free energy changes of the formation of its oxides are more negative than that of TiO₂. In addition, Al₂O₃ features suitable strength, permeability and collapse-ability, which can ensure dimensional accuracy of castings. Fig. 3 shows the microstructure and hardness profile on the surface of pure titanium investment castings poured into Al₂O₃ mold.

Figure 3.

Microstructure of the interface between Ti and Al₂O₃ mold, and hardness profile.

The microstructure of a distinct reaction layer is about 200 µm thick on the interface between Ti and Al₂O₃ mold. In addition to the reaction layer, there shows a hardened layer about 300 µm thick. The reaction layer and the hardened layer together are called the α-case of titanium castings. However, when the molten titanium (around 2,000 K) is poured into the Al₂O₃ investment mold, and if the α-case results between Ti and the interstitial oxygen, the reaction could be described as follows:

$$\text{Ti}\left(l\right)+{\text{O}}_{\text{2}}\left(\text{g}\right)={\text{TiO}}_{\text{2}}\left(\text{s}\right) \Delta {\text{G}}^{\text{o}}{\text{TiO}}_{\text{2}}=-\text{585}.\text{83}0\text{ kJ}/\text{mol}$$ E1

$$\text{4}/\text{3Al}\left(l\right)+{\text{O}}_{\text{2}}\left(\text{g}\right)=\text{2}/{\text{3Al}}_{\text{2}}{\text{O}}_{\text{3}}\left(\text{s}\right)\Delta {\text{G}}^{\text{o}}{}_{\text{Al2O3}}=\text{ 1},0\text{28}.\text{367 kJ}/\text{mol}$$ E2

$$\text{Ti}\left(l\right)+\text{2}/{\text{3Al}}_{\text{2}}{\text{O}}_{\text{3}}\left(\text{s}\right)={\text{TiO}}_{\text{2}}\left(\text{s}\right)+\text{4}/\text{3Al}\left(l\right)\Delta {\text{G}}^{\text{o}}{}_{\text{F}}=+\text{99}.\text{748 kJ}/\text{mol}$$ E3

The above calculations utilized the joint of army-navy-air force (JANAF) thermochemical tables [6]. According to the equation (3), the α-case formation by interstitial oxygen cannot occur spontaneously. Therefore, the reason why the α-case reaction is generated cannot be explained by the conventional α-case formation mechanism.

3.2. Alpha-case formation mechanism

For the clear examination of the α-case formation mechanism, the distribution of the elements of the α-case and its chemical composition were investigated by EPMA elemental mapping as shown in Fig. 4. On the α-case region, the oxygen element was uniformly distributed. Also, the Si element originating from the colloidal silica binder was scarcely detected on the surface.

Figure 4.

Comparison of elemental mapping images of O, Al and Si in Ti castings into Al₂O₃ mold and BEI image by EPMA.

However, the concentrated Al contamination layer about 30 µm thick was detected on the interface. The EPMA mapping result shows that not only the interstitial oxygen elements but also the substitutional Al elements dissolved from the mold material affect the metal-mold reactions. Until recently, the effect of substitutional metallic element dissolved from mold materials has been ignored as negligibly small [7]. However, Fig. 3 and 4 indicate that the effect of metallic elements cannot be overlooked any more.

The phase identification of the detected Al element will be the very core of α-case formation mechanism. The phases of the detected Al on the interface were examined by transmission electron microscopy (TEM). In order to observe α-case region precisely, the titanium castings specimen was grinded from inside to the surface until about 30 µm and final thinning was carried out using ion milling. Fig. 5 is the cross-sectional bright field TEM image of the α-case region.

Figure 5.

Bright field TEM image of the α-case region.

To examine what kinds of phases are presented in the α-case region, C2 aperture was temporarily removed. Fig. 6 (a) shows ring and spot patterns on the TEM image without C2 aperture. The definite contrast of continuous ring pattern could not be found on the TEM image since the ring pattern was an extremely small size polycrystalline TiO₂ phase. In the case of spot patterns on Fig. 6 (b), the contrast could be distinguished on the TEM image. And the indexed pattern was a hexagonal close-packed (HCP) phase in the [2īī0] beam direction as shown Fig. 6 (b).

Figure 6.

Results of analytical transmission electron microscopy of (a) ring and spot patterns on the TEM image without C2 aperture on the bright field image, and (b) spot pattern on the TEM image was HCP phase in the [2īī0] beam direction.

The convergent beam electron diffraction (CBED) analysis was carried out for the verification of the phase of spot patterns with the primitive cell volume method [7]. The primitive cell volume is characteristic material constant. So, the phase of interest can be easily identified by comparing the measured primitive cell volume of an unknown phase using CBED patterns with the known values of the possible phases. This method is accurate (error range of <10%) for the phase identification. In the CBED pattern where zero order Laue zone (ZOLZ) disk and high order Laue zone (HOLZ) ring are present, by measuring the distance of ZOLZ disk from the transmitted beam to the diffracted beam (D₁, D₂), an internal angle (ANG) between D1 and D2, and HOLZ ring's radius (CRAD), the primitive cell volume of unknown phase can be easily determined using the equation (4). Camera Length (CL) was calibrated with Au standard sample at 200 keV.

$$Cellvolume=\frac{C{L}^2\cdot {\lambda }^3}{D_1\cdot D_2\cdot \sin \left(ANG\right)\left[1-\cos \left\{{\tan }^{-1}\left(\frac{CRAD}{CL}\right)\right\}\right]}$$ E4

Fig. 7 (a) shows the CBED pattern of the HCP phase. There are two possible HCP phases of Ti and Ti₃Al. The measured primitive cell volume (61.08 ang³, CL=521.4 mm, D₁=D₂=6.5 mm, ANG=60º, CRAD=23 mm) from CBED pattern corresponded to the theoretic value of Ti₃Al (66.60 ang³). This phase identification was in accordance with EDS spot analysis as shown in Fig. 6 (b). The phase of the detected Al from EPMA mapping was Ti₃Al phase. Thus, considering the microstructure, hardness profile, EPMA mapping and TEM, the α-case is not wholly TiO₂, but TiO₂ and Ti₃Al between Ti and Al₂O₃ mold.

Figure 7.

Convergent beam electron diffraction analysis with the primitive unit cell volume method (a) The measured primitive unit cell volume, 61.08 ang³ (CL=521.4 mm, D₁=D₂=6.5 mm, ANG=60º, CRAD=23 mm, λ=0.0251 Å) and (b) EDS spot analysis result.

In this study, from the synthesis of the microstructure, hardness profile, EPMA mapping and TEM, it could be confirmed that the α-case is formed by not only interstitial oxygen element but also substitutional metallic elements dissolved from mold materials as shown in Fig. 8.

Figure 8.

Schematic diagram of the interstitial and substitutional α-case formation mechanism of titanium castings.

4. New alpha-case formation mechanism with alpha-case controlled mold

4.1. Alpha-case controlled stable mold fabrication

The α-case formation mechanism was applied to the development of α-case controlled stable mold and verified by titanium alloy castings with the α-case controlled mold materials. To synthesize the α-case formation products, 10 wt% to 50 wt% of about 45 µm titanium powders were added into Al₂O₃ for the purpose of eliminating the α-case. After blending, the slurry was prepared with the Al₂O₃ based powder. The α-case controlled stable mold manufacturing and titanium casting procedures were carried out at the same condition as mentioned above. The synthesized phases between titanium and Al₂O₃ were identified by RIGAKU X-ray diffractometer.

4.2. New alpha-case formation mechanism

Although the above mentioned Al₂O₃ has the proper strength, permeability and collapse-ability enough to ensure dimensional accuracy of castings, it could not be applied for mold due to its severe α-case formation. However, in this study, for the maximization of performance and cost-effectiveness of the investment mold, the α-case controlled mold was designed on Al₂O₃ base.

In consideration of the interstitial and substitutional elements in α-case formation mechanism, 10 wt% to 50 wt% of titanium powders were blended with Al₂O₃ for the previous synthesis of α -case reaction products into the mold. Then the blended powders were mixed with the colloidal silica and agitated.

Fig. 9 shows the X-ray diffraction result of 50 wt% titanium mixed Al₂O₃ powders before curing. The composition of the starting titanium and Al₂O₃ remained unchanged after being mixed with colloidal silica. After the mixing and agitation, the mold materials were cured at 1,223 K for 2 hrs. According to XRD analysis, TiO₂ and Ti₃Al phases were in-situ synthesized on the Al₂O₃ base between titanium and Al₂O₃ powders as shown in Fig. 10.

Figure 9.

Illustration of the diffraction patterns and replots for identifying phases in 50 wt% titanium and Al₂O₃ mixed with colloidal silica before curing.

Figure 10.

X-ray diffraction pattern of synthesized α-case controlled stable mold materials.

The scan angles were from 20 to 80 and the pattern was verified with JCPDS card numbers 88-0826 (Al₂O₃), 84-1284 (TiO₂) and 14-0451 (Ti₃Al) [8]. No trace of the starting titanium reagent was detected by XRD analysis. Fig. 11 shows the variation of XRD patterns with the amount of titanium powders. The component ratio of TiO₂ and Ti₃Al phases increased gradually with the addition of titanium. Consequently, the synthesis of the α-case formation products TiO₂ and Ti₃Al phases into Al₂O₃ based mold could be possible by the addition of titanium powders.

The effect of α-case controlled stable mold which contained the interstitial metal-mold reaction product TiO₂ and the substitutional metal-mold reaction product Ti₃Al was verified with titanium casting.

Figure 11.

Illustration of the diffraction patterns and replots for identifying phases from 10 wt% to 50 wt% titanium and Al₂O₃ after curing at 1,223 K for 2 hrs.

Fig. 12 (a) indicates that in the case of CaO stabilized ZrO₂, the expensive and thermally stable mold, the externals of titanium castings lost metallic luster as a result of the α-case formation reactions. However, in the α-case controlled stable mold, its characteristic luster of titanium was well preserved as shown in Fig. 12 (b).

Figure 12.

Comparison of externals of titanium castings (a) CaO stabilized ZrO₂ mold and (b) the α-case controlled stable mold.

Figure 13.

Microstructure and hardness profile between pure titanium and α-case controlled stable mold.

This external examination result of the α-case controlled stable mold is in good accordance with the microstructure and hardness profile as shown in Fig. 13. The effect of prevention of α-case formation can be obtained after addition of more than 2 wt% titanium.

4.3. Noble synthesis of the alpha-case controlled stable mold

The α -case controlled stable mold is much less expensive than the conventional mold materials for titanium such as ZrO₂, CaO stabilized ZrO₂ and Y₂O₃, and the complete control of α-case is possible. However, in this study, the noble synthesis route was conceived, which is more cost-effective than the above route by the inexpensive TiO₂ and aluminum raw materials.

The homogeneous powders containing 30 mol% TiO₂ powders (anatase, 45 µm, 99% pure) and 70 mol% aluminum powders were prepared by blending. After blending, the powders were mixed with colloidal silica and agitated. The α-case controlled stable mold manufacturing and titanium casting procedures were carried out at the same condition as mentioned above. The 3:7 molar ratio of TiO₂ and aluminum was chosen to yield the Al₂O₃ and TiAl final product after in-situ synthesis according the following reaction.

$${\text{3TiO}}_{\text{2}}+\text{ 7Al}\to {\text{2Al}}_{\text{2}}{\text{O}}_{\text{3}}+\text{ 3TiAl}$$ E5

In order to synthesize the Al₂O₃ and TiAl phases, the mixed TiO₂ and aluminum powders were cured at the 1,323 K for 2 hrs. However, the equation (5) is an ideal reaction, and thus it is very difficult to synthesize the ideal product by the simple blending and after agitation.

Figure 14.

Illustration of the diffraction patterns and replots for identifying the synthesized phases between TiO₂ and aluminum powders.

Figure 15.

Microstructure and hardness profile between pure titanium and noble route α-case controlled stable mold.

According to the XRD analysis as shown in Fig. 14, the major phases are anatase TiO₂ and rutile TiO₂ as well as the in-situ synthesized Al₂O₃, Ti₃Al and TiAl phases since the partially non-reacted anatase TiO₂ remained and the TiAl, Ti₃Al intermetallic compounds and rutile TiO₂ were synthesized between anatase TiO₂ and aluminum powders. As a result, the synthesis which is similar to the titanium powders added -case controlled stable mold can be obtained by the economical route. The α-case free titanium casting can be possible by the noble route synthesized α-case controlled stable mold which contained the α-case formation products such as Al₂O₃, TiO₂, TiAl and Ti₃Al phases as shown in Fig. 15.

5. Conclusion

By the conventional α-case formation mechanism regardless of thermodynamic approaches, the alpha-case generation cannot be explained. In order to ascertain the reason, α-case formation mechanism was closely examined and from the mechanism, two kinds of α-case controlled stable molds were developed.

1. Regardless of thermodynamic approaches, about 500 µm thick α-case was generated between titanium and Al₂O₃ mold. The reason why the α-case generated cannot be explained by the conventional α-case formation mechanism, which is known to be formed by the interstitials, especially oxygen dissolved from mold materials.

2. In spite of having used pure titanium, the concentrated aluminum contamination layer about 30 µm thick was detected on the interface by the EPMA elemental mapping. From the results of the TEM phase identification, the phase of the detected aluminum from EPMA mapping was identified as Ti₃Al phase.

3. 3. Considering the microstructure, hardness profile, EPMA mapping and TEM analysis, it could be confirmed that the α-case is formed not only by interstitial oxygen element but also by substitutional metallic elements dissolved from mold materials.

4. The synthesis of the TiO₂ and Ti₃Al phases which are the α-case reaction products between titanium and Al₂O₃ mold can be obtained by the simple curing of Al₂O₃ mold added titanium powders. The complete control of α-case formation can be possible with the α-case controlled stable mold.

5. The α-case free titanium casting can be possible by the noble route synthesized α-case controlled stable mold which is in-situ synthesized between TiO₂ and aluminum powders to obtain the α-case formation product such as Al₂O₃, TiO₂, TiAl and Ti₃Al phases.

Consequently, the economical net-shape forming of titanium alloys without α-case formation and the verification of a newly established α-case formation mechanism, can be possible using the α-case controlled stable molds.