- Email: [email protected]
An oxidation model for predicting the life of titanium alloy components in gas turbine engines
Journal of Alloys and Compounds 389 (2005) 190–197
An oxidation model for predicting the life of titanium alloy components in gas turbine engines I. Gurrappa∗ Defence Metallurgical Research Laboratory, Kanchanbagh PO, Hyderabad 500058, India Received 30 January 2004; accepted 28 May 2004
Abstract The excellent combination of lightweight and good mechanical properties makes the titanium alloys more attractive for fabrication of gas turbine engine components. However, titanium alloys readily absorb oxygen when exposed in air at elevated temperatures, leading to ␣-case formation and oxidation. It severely limits the high temperature capability of titanium alloys with respect to mechanical properties and causes failures during service. In order to extend the use of titanium alloys for gas turbine engine compressor components, it is essential to understand the properties of titanium alloys before introducing into service. In the present paper, an oxidation model has been developed based on the experimental observations, to predict the life of the components fabricated from the titanium alloys. The results show that there is a very good agreement between the measured and predicted values and no assumptions are required. The developed model is very simple, easy to apply, promising, and is found to be extremely useful. © 2004 Elsevier B.V. All rights reserved. Keywords: Titanium alloys; Oxidation model; Life prediction
1. Introduction Pure titanium undergoes an allotropic transformation at 882 ◦ C from lower temperature phase designated the ␣-phase, which has a hexagonal close packed (hcp) structure to a higher temperature phase designated the -phase, which is stable up to the melting temperature and has a body centre cubic (bcc) structure. Titanium alloys are conventionally divided as ␣type, -type or ␣ + -type alloys by virtue of the nature and amount of alloying elements present. Near alpha titanium alloys are titanium based with additions of, amongst other elements, alpha stabilising elements like aluminium and tin which promote the hexagonal close packed structure of the alpha phase. This alpha phase has excellent high temperature creep properties. These properties are achieved while still maintaining adequate low temperature strength and formability in near alpha titanium alloys.
∗
Tel.: +91 40 4586515; fax: +91 40 4440683. E-mail address: [email protected].
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2004.05.079
The excellent balance of strength, ductility, oxidation/corrosion resistance and micro-structural stability of titanium alloys as compared with their competitive materials like steels or nickel based superalloys, has resulted in near alpha titanium alloys becoming popular for fabrication of compressor components in advanced gas turbine engines. A family of subsequent near alpha titanium alloys have been developed to tolerate operating conditions involving prolonged exposure to air up to 500 ◦ C: IMI 550 (Ti–4Al–2Sn–4Mo–0.5Si), IMI 679 (Ti–11Sn– 2.25Al–5Zr–1Mo–0.25Si), IMI 685 (Ti–6Al–5Zr–0.5Mo– 0.25Si), IMI 829 (Ti–5.5Al–3.5Sn–3Zr–0.25Mo–0.3Si) and IMI 834 (Ti–5.8Al–4Sn–3.5Zr–0.7Nb–0.5Mo–0.356Si). These alloys have been commercially prepared by IMI Titanium Limited and designated as IMI and the nominal composition is in weight percent. There is a growing interest in increasing the temperature tolerance of near alpha titanium alloys with a view to increase engine efficiency by increasing the temperature. There is also the possibility that an increase in temperature tolerance might permit these titanium alloys to be used for components
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
191
tally the oxygen dissolved region and thereby to correlate the predicted data. It was shown that the depth of hardening, i.e. alpha case formation in the titanium alloy, IMI 834 could be modelled with a high degree of confidence. It was also shown that this study would be extremely useful in estimating the life of a titanium alloy component under service conditions and thereby prevent failures during service.
currently made of nickel-based superalloys with consequent reduction in component weight. Some of the conventional near alpha titanium alloys exhibit excellent creep resistance and good structural stability at temperatures up to 500 ◦ C. However, when such alloys are exposed to air at temperatures approaching 600 ◦ C, they are subject to significant and detrimental surface modification. The major factor involved in surface modification is the uptake of oxygen into solid solution by titanium. At 600 ◦ C and above, the reaction kinetics ensure rapid diffusion of oxygen into the region adjacent to the exposed surface. About 33 at.% of oxygen dissolves in titanium. The dissolved oxygen forms a hard brittle zone in the titanium alloy. The oxygen-dissolved zone is known as “alpha case” and its formation can substantially degrade the structural integrity of the affected titanium alloy by loss of tensile ductility and of fatigue resistance, even though the interior of the alloy is not subject to structural modification. The predominant factors in alpha case formation are the presence of oxygen, exposure time and the temperature. ␣-Case formation is critical to the life expectancy of titanium alloys when used in aero-gas turbine engines. With a solution to this problem, titanium alloys could be used at significantly higher temperatures. This is important as the weight reduction achieved by replacing nickel alloys in hot compressor stages will benefit engine performance. A systematic study was carried out on bare titanium alloy, IMI 834 at different temperatures and a degradation mechanism was proposed under oxidising environments [1,2]. It was established that the formation of alpha case and oxide scale are the two principal factors in degradation of the titanium alloys at elevated temperatures. Among the two, ␣-case formation is the dominant degradation mechanism particularly at higher temperatures, i.e., 800 ◦ C and above [1,2]. Application of protective coatings on the titanium alloy results in significant improvement not only in elimination of alpha case more effectively, but also the prevention of oxide scale formation [3–7]. Many researchers employed a variety of surface modification techniques [8–19] as a method to limit oxygen ingress into the titanium alloys. But, these methods are not much effective when compared to the recently developed coatings by employing combined surface engineering techniques [4–7]. With a view to relate these relatively high temperature exposures to the potential users and the mechanical property temperature limit of IMI 834, an oxidation model is proposed for the alpha case formation, i.e., oxygen ingress into the titanium alloy. Analysis of the alpha case depths as measured by the depth of hardening was undertaken after a number of exposures at different temperatures ranging from 600 to 800 ◦ C and from 3 to 200 h in order to determine experimen-
2. Experimental The chemical composition of the near ␣-titanium alloy, IMI 834 studied in the present investigation is provided in Table 1. The cyclic isothermal oxidation tests were performed in air at 600, 700 and 800 ◦ C for different time periods. For these tests, disc samples with surface area of 7–8 cm2 were machined from 20 mm diameter rods, ground up to 800 grit surface finish and ultrasonically degreased in acetone and alcohol. Then, the specimens were introduced into the furnace zone after stabilizing the required temperature. The exposed specimens were removed from the furnace after exposure of different times. Weight gain was monitored initially for every hour and subsequently for every 10 h. Scanning electron microscope (SEM) has been used to observe the alpha case formation as well as the oxide scales on the surface of the alloy due to exposure at different temperatures. Electron probe microanalyzer (EPMA) has been used to observe the elemental distribution (X-ray scans) after exposure of 100 h at 800 ◦ C in order to determine the oxygen diffusion into the titanium alloy and thereby to understand the mechanism of degradation. Microhardness measurements have been carried out in order to determine the depth of hardened zone or ␣-case in the titanium alloy during different high temperature exposures.
3. Results and discussion 3.1. As oxidized titanium alloy specimens The visual appearance of as oxidized specimens of titanium alloy, IMI 834 at 600, 700 and 800 ◦ C for 100 h in air are provided in Fig. 1. It was observed that the thickness of oxide scale is greater for the specimens exposed at 800 ◦ C when compared to the specimens oxidized at 600 and 700 ◦ C. Oxide spallation is also clearly observed from the specimens exposed at 800 ◦ C, after 80 h of exposure, particularly at the edges of the specimens. This indicates that the specimens exhibit poorest oxidation resistance at 800 ◦ C and the oxide
Table 1 Actual chemical composition of near ␣-titanium alloy, IMI 834 (all wt.%) Ti Balance
Al 5.8
Sn 4.06
Zr 3.61
Nb 0.70
Mo 0.54
Si 0.32
C 0.05
Fe 0.009
O 0.105
N 0.002
192
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
oxide growth is maximum at 800 ◦ C (Kp = 3.2 × 10−2 ) and minimum at 600 ◦ C (Kp = 3.1 × 10−4 ). On all the specimens, initially rutile is formed followed by alumina scale. It is very important to mention that the alumina scale is not uniform and at the same time rutile is also not protective at high temperatures. After reaching a certain thickness, the oxide scale started spalling for the specimens exposed at 800 ◦ C. 3.3. SEM
Fig. 1. As oxidised specimens of titanium alloy, IMI 834 after 100 h of oxidation at different temperatures showing oxide spallation at the edges.
scale is not adherent to the base alloy and thereby tended to spallation (Fig. 1). It is important to mention that no spallation of oxide scale is observed for other specimens exposed at 600 and 700 ◦ C.
The typical cross sections of specimens oxidized at 600, 700 and 800 ◦ C revealed the presence of primarily two regions (Fig. 3). The top region was covered with the oxide scale. The region just below the oxide scale is the ␣-case formed zone or oxygen dissolved/oxygen affected region. As can be seen, the depth of ␣-case is low in the case of specimen exposed at 600 ◦ C, more at 700 ◦ C and, highest at 800 ◦ C. It was also observed that the depth of ␣-case increases with increasing exposure time at a constant temperature. It indicates that ␣-case formation is a diffusion-controlled process and therefore, the depth of ␣-case increases with increasing time and temperature.
3.2. Oxidation kinetics 3.4. EPMA Fig. 2 shows the weight gain as a function of time for the titanium alloy, IMI 834 oxidized at 600, 700 and 800 ◦ C. The alloy followed parabolic kinetics at all the studied temperatures. As can be seen, the weight gain is minimal for the specimen exposed at 600 ◦ C and slightly more at 700 ◦ C, while significant weight gain is observed for the specimen exposed at 800 ◦ C. It is important to mention that the rate of
The X-ray scans of alloying elements of titanium alloy as well as oxygen after exposure of 100 h at 800 ◦ C in air is illustrated in Fig. 4. The results clearly show the presence of all alloying elements of titanium alloy in the oxide scale. Particularly, zirconium and tin were diffused outwardly and concentrated beneath the oxide scale. An important and con-
Fig. 2. Weight gain as a function of exposure time for titanium alloy, IMI 834 at various temperatures illustrating significant weight gain at 800 ◦ C.
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
193
Fig. 4. X-ray scans of alloying elements of titanium alloy and oxygen after exposure of 100 h at 800 ◦ C showing the oxygen dissolved region.
As mentioned earlier, adherence because of Therefore, it detaches essence, it is clear the in the titanium alloy.
the oxide scale does not have good the higher thickness of oxide scale. easily during sample preparation. In presence of oxygen dissolved region
3.5. Microhardness measurements
Fig. 3. Cross-sections of typical oxidised titanium alloy specimens showing ␣-case formed region and oxide scale at different temperatures.
firmed observation is the presence of dissolved oxygen zone just beneath the oxide scale. The high intensity peak represents oxygen present in the oxide scale while lower intensity one represents the oxygen dissolved in the titanium alloy. The dip observed between the two oxygen peaks is due to the detachment of oxide scale during preparation of the specimen.
Fig. 5 provides the measured microhardness profiles as a function of depth of ␣-case for the specimens exposed at 600, 700 and 800 ◦ C for a period of 100 h. The results show that the depth of ␣-case increases slowly from 600 to 700 ◦ C and then significantly at 800 ◦ C. The sub-surface of the titanium alloy affected for the specimen oxidized at 800 ◦ C is about 140 m, which is about 12 times that of the specimen exposed at 600 ◦ C (∼12 m) and about four times that of the specimen oxidized at 700 ◦ C (∼40 m). It is also very clear that the hardened zone due to dissolution of oxygen, exhibits a hardness of about 700 HV and the depth of ␣-case is about 140 m which is consistent with the SEM and EPMA results in case of the specimen exposed at 800 ◦ C. It also clearly shows that the rate of oxygen dissolution is considerably higher at 800 ◦ C in the titanium alloy and thus forms a hardened zone at a faster rate. This extent of surface embrittlement could certainly be sufficient to cause significant loss of ductility and failure by surface cracking under load [1]. It would be desirable to predict the depth of ␣-case for titanium alloys at a variety of elevated temperatures, to assess the
194
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
Fig. 5. Measured microhardness profiles of titanium alloy, IMI 834 after 100 h of oxidation at various temperatures showing the depth of ␣-case.
life of compressor components and help to eliminate failures during service. The next section describes the methodology and modeling of depth of hardening in titanium alloys through which the depth of ␣-case can be predicted with a reasonable accuracy and thereby to predict acceptable exposure conditions for titanium alloy components.
measured ␣-case depths for the specimens exposed at different elevated temperatures reflects the predicted trend (Fig. 5). It indicates that oxygen dissolution increases enormously at and above 800 ◦ C and affects the mechanical properties of the alloy. On the basis of measured parameters for IMI 834 for different times and at different temperatures, equivalent exposure conditions for an acceptable case depth would be:
4. An oxidation model for titanium alloys
4.4 h exposure at 900 ◦ C = 40 h at 800 ◦ C = 570 h of
An oxidation model has been proposed on the basis of several experiments at a variety of temperatures for different time intervals for the ␣-case formation or oxygen ingress into the titanium alloy. The Arrhenius equation, which has been modified appropriately, has been used for predicting the ␣-case formation in the present study. The exponential term used in the equation is related to a measure of activation energy for dissolution and diffusion of oxygen into the alloy [20]. Parameters taken into consideration are as follows: x: depth of hardening due to ingress of oxygen (m) T: exposure temperature (K) t: exposure time (h) Fig. 6 provides the predicted depth of ␣-case as a function of exposure time at different elevated temperatures in the titanium alloy. The depth of ␣-case is low at 600 ◦ C, increases slightly at 700 ◦ C. While at 800 ◦ C, significant depth of ␣-case is observed. For example, after exposure of 100 h, the depth of ␣-case is about 10 m at 600 ◦ C and 50 m at 700 ◦ C, which is five times more than that of the specimen exposed at 600 ◦ C. The predicted depth of ␣-case is about 20 times and 58 times more at 800 and 900 ◦ C, respectively when compared to the ␣-case predicted at 600 ◦ C. Practically
exposure at 900 ◦ C = 15000 h at 600 ◦ C In other words, if the component fabricated from the titanium alloy IMI 834 continuously operates at its maximum mechanical property temperature limit of 600 ◦ C at its designed service life of 15000 h, then the expected depth of alpha case, i.e., oxygen dissolved region, would be 109 m. This would be equivalent to an exposure of 570 h at 700 ◦ C, 40 h at 800 ◦ C or 4.4 h at 900 ◦ C. Another important factor is the correlation between the predicted and experimentally measured ␣-case formed regions (Fig. 7). The predicted values are always slightly higher than the experimentally measured values. It is due to the fact that the oxygen absorbed by titanium alloy during oxidation is used up not only for the formation of oxide scale but also for the ␣-case formation. Whereas in predicted values, it is assumed that the entire oxygen was used up for creating ␣case in the titanium alloy. This is the reason for observing slight deviation between experimental and predicted values. In other words, the marginal difference in the depth of ␣-case formation between the measured and predicted values can be considered as due to the formation of oxide scale on titanium alloys during exposure in air at elevated temperatures. It can
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
195
Fig. 6. Predicted depth of ␣-case for titanium alloy at different exposed temperatures.
be formulated as follows: ␣-casepred.
=
␣-casemeas.
+
oxide scale thickness
␣-casemeas.
=
␣-casepred.
140
=
173
− −
oxide scale thickness 33 at 800 ◦ C
40
=
45
−
5 at 700 ◦ C
Or
The above equation has been validated by substituting the data for 100 h. The equation is valid because of the fact that the titanium alloy initially oxidises to form oxide scale. Thereafter, the oxygen dissolves in the subsurface of titanium alloy and forms ␣-case. The above equation is valid including the figures. Therefore, for calculating the total affected region of the titanium alloy components in service, the
Fig. 7. Comparison between the measured and predicted ␣-case depths at 700 and 800 ◦ C.
196
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
oxide scale thickness has to be added to the measured ␣-case depth. As mentioned earlier, it is always possible to observe temperature hikes in service. Under such circumstances, the combination of ␣-case depths have to be taken into account for calculating total depth of ␣-case, i.e. the life of a component. For example, if the titanium alloy component continuously operates at 600 ◦ C for 5000 h and 15 h (in total period of operation) at 800 ◦ C, then the total depth of ␣-case would be 130 m (63 m at 600 ◦ C + 67 m at 800 ◦ C). In other words, 15 h operational hike at 800 ◦ C, reduces the life of a component by about 5000 h normal operational life, i.e. at 600 ◦ C. If the temperature hikes are observed at 700 ◦ C for some time, 800 ◦ C for some other time and normal operation at 600 ◦ C, then the total depth of ␣-case can be obtained by adding ␣-case depths at 600, 700 and 800 ◦ C. Failures of components in gas turbines most likely takes place due to non-accountability of these hikes. Therefore, it is essential to understand the consequences of hikes and the present study provides a valuable answer to these types of problems. In essence, the developed oxidation model helps to predict the component life once the temperatures at which the component operated and the period of operation at each temperature is known. It is also possible to predict the remaining life of a component by using the present oxidation model. From the present data, it is clear that there is an excellent correlation between the weight gain during oxidation process (Fig. 2) and the predicted depth of ␣-case (Fig. 6). Both the curves followed similar pattern at different temperatures studied in the present investigation. It indicates that ␣-case formation is the predominant reaction process during high temperature exposure rather than oxide scales formation though both processes contribute weight gain. This ␣case formation is a major contributing factor for the weight gain at all the exposed temperatures though the significant increase is observed above 700 ◦ C and affecting the mechanical properties of the titanium alloy components. The results clearly show that ␣-case formation is the life-limiting factor for the gas turbine engine compressor components fabricated from the titanium alloys. Therefore, it is essential to know the depth of ␣-case after exposure of definite number of hours at different elevated temperatures in order to avoid failures during service. The developed model is extremely useful to predict the depth of ␣-case with reasonably good accuracy and thus helps in avoiding failures during service. The contribution due to oxygen dissolution will be smaller than the estimated values as an outer layer of the alloy is converted to TiO2 . This aspect is clearly seen by observing smaller depth of alpha case formation for the alloy specimens exposed to lower temperatures for various time intervals (Fig. 5). This observation is consistent with the specimens exposed at different temperatures. Further, the oxygen concentration in the alloy may be smaller than the equilibrium value at low temperatures and oxygen diffusion
may be influenced by impurities. Even so the comparison clearly indicates that oxygen dissolution constitutes an important part of the parabolic oxidation. Direct estimates from studies on related specimens show that 25–30% or more of the reacted oxygen dissolves in the alloy at temperatures of 700–750 ◦ C [21,22] and more than 50% at 900–950 ◦ C [23]. From the present results, it is very clear that it is possible to predict the ␣-case formation with reasonably good accuracy for titanium alloys. It is also very clear that there is good agreement between the experimentally measured and predicted depth of ␣-case formed regions (Fig. 7). This study is extremely useful in the modern gas turbines to evaluate the life of the titanium alloy components without performing actual experiments on the alloys. It also helps to select the titanium alloy components with appropriate thickness to the designed life and thereby eliminating failures during service. The major advantage of the developed model is that it is very easy to use, no assumptions are required and time saving as no experiments can be conducted on the actual alloy components. The present study also helps in establishing that ␣-case formation plays a dominant role in observing weight gain during oxidation processes. The significant weight gain observed for the specimen exposed at 800 ◦ C when compared to the specimens exposed at 600 and 700 ◦ C, is due the fact that weight gain is observed not only due to the formation of oxide scale but also for the significant dissolution of oxygen in the subsurface zone of the titanium alloy.
5. Development of high performance coatings From the measured and predicted data, it is clear that the titanium alloys readily absorb oxygen at elevated temperatures and forms hardened phase of ␣-case even after few hours of exposure particularly at elevated temperatures and thus affecting the mechanical properties. Because of the technological importance of extending the operational temperatures of titanium alloy components primarily for aerospace applications and the danger of forming significant ␣-cases at high temperatures, protective coatings to protect them against oxidation and ␣-case formation are highly essential. With this goal, attempts were made and succeeded in developing high performance coatings for titanium alloys [4–7]. The recently developed coating produced through the combination of surface engineering techniques [5] enhances the life of the titanium alloy components considerably by successfully eliminating the oxygen uptake and thereby helps in improving the efficiency of gas turbine engines significantly.
Acknowledgements Defence Research and Development Organisation is gratefully acknowledged for providing financial assistance.
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
References [1] I. Gurrappa, Corr. Prev. Control 49 (2002) 79. [2] I. Gurrappa, A.K. Gogia, in: Proceedings of 5th National Convention on Corrosion, New Delhi, vol. I, pp. 210–219. [3] I. Gurrappa, Oxid. Met. 56 (2001) 73. [4] I. Gurrappa, Platinum Metals Rev. 45 (2001) 124. [5] I. Gurrappa, A.K. Gogia, J. Mater. Sci. Technol. 17 (2001) 581. [6] I. Gurrappa, A.K. Gogia, in: Proceedings of the 5th International Symposium on High Temperature Corrosion, May 2000, Le Embitz, France. [7] I. Gurrappa, A.K. Gogia, Surf. Coat. Technol. 139 (2001) 216. [8] R.K. Clark, J. Unnan, K.E. Wiedemann, Oxid. Met. 28 (1987) 391. [9] D.W. Mckee, K.L. Luthra, Surf. Coat. Technol. 56 (1993) 391. [10] H.L. Du, P.K. Datta, J.S. Burnell-Gray, G.B. Lewis, J. Mater. Sci. 30 (1995) 2640.
197
[11] S. Fujishiro, D. Eylon, Met. Trans. 11A (1980) 1259. [12] US Patent 4,946,749. [13] C. Leyens, M. Schmidt, M. Peters, W.A. Kaysser, J. Mater. Sci. Eng. A239/240 (1997) 680. [14] C.H. Koo, T.H. Yu, Surf. Coat. Technol. 126 (2000) 171. [15] J.R. Nicholls, JOM 52 (2000) 28. [16] C. Leyens, M. Peters, W.A. Kaysser, Mater. Sci. Forum 251–254 (1997) 769. [17] B. Cockeram, R.A. Rapp, Oxid. Met. 45 (1996) 427. [18] C. Leyens, M. Peters, W.A. Kaysser, Surf. Coat. Technol. 94/95 (1997) 34. [19] H. Liu, S. Hao, X. Wang, Z. Feng, Script. Met. 39 (1998) 1443. [20] I. Gurappa, in: Proceedings of National Symposium on Electrochemical Science and Technology, July 2002, Bangalore, p. 10. [21] T. Hurlen, J. Inst. Met. 89 (1960/61) 128. [22] J.E. Lopes Gomes, A.M. Huntz, Oxid. Met. 14 (1980) 249. [23] P. Kofstad, J. Less-Common Met. 12 (1967) 449.
An oxidation model for predicting the life of titanium alloy components in gas turbine engines I. Gurrappa∗ Defence Metallurgical Research Laboratory, Kanchanbagh PO, Hyderabad 500058, India Received 30 January 2004; accepted 28 May 2004
Abstract The excellent combination of lightweight and good mechanical properties makes the titanium alloys more attractive for fabrication of gas turbine engine components. However, titanium alloys readily absorb oxygen when exposed in air at elevated temperatures, leading to ␣-case formation and oxidation. It severely limits the high temperature capability of titanium alloys with respect to mechanical properties and causes failures during service. In order to extend the use of titanium alloys for gas turbine engine compressor components, it is essential to understand the properties of titanium alloys before introducing into service. In the present paper, an oxidation model has been developed based on the experimental observations, to predict the life of the components fabricated from the titanium alloys. The results show that there is a very good agreement between the measured and predicted values and no assumptions are required. The developed model is very simple, easy to apply, promising, and is found to be extremely useful. © 2004 Elsevier B.V. All rights reserved. Keywords: Titanium alloys; Oxidation model; Life prediction
1. Introduction Pure titanium undergoes an allotropic transformation at 882 ◦ C from lower temperature phase designated the ␣-phase, which has a hexagonal close packed (hcp) structure to a higher temperature phase designated the -phase, which is stable up to the melting temperature and has a body centre cubic (bcc) structure. Titanium alloys are conventionally divided as ␣type, -type or ␣ + -type alloys by virtue of the nature and amount of alloying elements present. Near alpha titanium alloys are titanium based with additions of, amongst other elements, alpha stabilising elements like aluminium and tin which promote the hexagonal close packed structure of the alpha phase. This alpha phase has excellent high temperature creep properties. These properties are achieved while still maintaining adequate low temperature strength and formability in near alpha titanium alloys.
∗
Tel.: +91 40 4586515; fax: +91 40 4440683. E-mail address: [email protected].
0925-8388/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2004.05.079
The excellent balance of strength, ductility, oxidation/corrosion resistance and micro-structural stability of titanium alloys as compared with their competitive materials like steels or nickel based superalloys, has resulted in near alpha titanium alloys becoming popular for fabrication of compressor components in advanced gas turbine engines. A family of subsequent near alpha titanium alloys have been developed to tolerate operating conditions involving prolonged exposure to air up to 500 ◦ C: IMI 550 (Ti–4Al–2Sn–4Mo–0.5Si), IMI 679 (Ti–11Sn– 2.25Al–5Zr–1Mo–0.25Si), IMI 685 (Ti–6Al–5Zr–0.5Mo– 0.25Si), IMI 829 (Ti–5.5Al–3.5Sn–3Zr–0.25Mo–0.3Si) and IMI 834 (Ti–5.8Al–4Sn–3.5Zr–0.7Nb–0.5Mo–0.356Si). These alloys have been commercially prepared by IMI Titanium Limited and designated as IMI and the nominal composition is in weight percent. There is a growing interest in increasing the temperature tolerance of near alpha titanium alloys with a view to increase engine efficiency by increasing the temperature. There is also the possibility that an increase in temperature tolerance might permit these titanium alloys to be used for components
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
191
tally the oxygen dissolved region and thereby to correlate the predicted data. It was shown that the depth of hardening, i.e. alpha case formation in the titanium alloy, IMI 834 could be modelled with a high degree of confidence. It was also shown that this study would be extremely useful in estimating the life of a titanium alloy component under service conditions and thereby prevent failures during service.
currently made of nickel-based superalloys with consequent reduction in component weight. Some of the conventional near alpha titanium alloys exhibit excellent creep resistance and good structural stability at temperatures up to 500 ◦ C. However, when such alloys are exposed to air at temperatures approaching 600 ◦ C, they are subject to significant and detrimental surface modification. The major factor involved in surface modification is the uptake of oxygen into solid solution by titanium. At 600 ◦ C and above, the reaction kinetics ensure rapid diffusion of oxygen into the region adjacent to the exposed surface. About 33 at.% of oxygen dissolves in titanium. The dissolved oxygen forms a hard brittle zone in the titanium alloy. The oxygen-dissolved zone is known as “alpha case” and its formation can substantially degrade the structural integrity of the affected titanium alloy by loss of tensile ductility and of fatigue resistance, even though the interior of the alloy is not subject to structural modification. The predominant factors in alpha case formation are the presence of oxygen, exposure time and the temperature. ␣-Case formation is critical to the life expectancy of titanium alloys when used in aero-gas turbine engines. With a solution to this problem, titanium alloys could be used at significantly higher temperatures. This is important as the weight reduction achieved by replacing nickel alloys in hot compressor stages will benefit engine performance. A systematic study was carried out on bare titanium alloy, IMI 834 at different temperatures and a degradation mechanism was proposed under oxidising environments [1,2]. It was established that the formation of alpha case and oxide scale are the two principal factors in degradation of the titanium alloys at elevated temperatures. Among the two, ␣-case formation is the dominant degradation mechanism particularly at higher temperatures, i.e., 800 ◦ C and above [1,2]. Application of protective coatings on the titanium alloy results in significant improvement not only in elimination of alpha case more effectively, but also the prevention of oxide scale formation [3–7]. Many researchers employed a variety of surface modification techniques [8–19] as a method to limit oxygen ingress into the titanium alloys. But, these methods are not much effective when compared to the recently developed coatings by employing combined surface engineering techniques [4–7]. With a view to relate these relatively high temperature exposures to the potential users and the mechanical property temperature limit of IMI 834, an oxidation model is proposed for the alpha case formation, i.e., oxygen ingress into the titanium alloy. Analysis of the alpha case depths as measured by the depth of hardening was undertaken after a number of exposures at different temperatures ranging from 600 to 800 ◦ C and from 3 to 200 h in order to determine experimen-
2. Experimental The chemical composition of the near ␣-titanium alloy, IMI 834 studied in the present investigation is provided in Table 1. The cyclic isothermal oxidation tests were performed in air at 600, 700 and 800 ◦ C for different time periods. For these tests, disc samples with surface area of 7–8 cm2 were machined from 20 mm diameter rods, ground up to 800 grit surface finish and ultrasonically degreased in acetone and alcohol. Then, the specimens were introduced into the furnace zone after stabilizing the required temperature. The exposed specimens were removed from the furnace after exposure of different times. Weight gain was monitored initially for every hour and subsequently for every 10 h. Scanning electron microscope (SEM) has been used to observe the alpha case formation as well as the oxide scales on the surface of the alloy due to exposure at different temperatures. Electron probe microanalyzer (EPMA) has been used to observe the elemental distribution (X-ray scans) after exposure of 100 h at 800 ◦ C in order to determine the oxygen diffusion into the titanium alloy and thereby to understand the mechanism of degradation. Microhardness measurements have been carried out in order to determine the depth of hardened zone or ␣-case in the titanium alloy during different high temperature exposures.
3. Results and discussion 3.1. As oxidized titanium alloy specimens The visual appearance of as oxidized specimens of titanium alloy, IMI 834 at 600, 700 and 800 ◦ C for 100 h in air are provided in Fig. 1. It was observed that the thickness of oxide scale is greater for the specimens exposed at 800 ◦ C when compared to the specimens oxidized at 600 and 700 ◦ C. Oxide spallation is also clearly observed from the specimens exposed at 800 ◦ C, after 80 h of exposure, particularly at the edges of the specimens. This indicates that the specimens exhibit poorest oxidation resistance at 800 ◦ C and the oxide
Table 1 Actual chemical composition of near ␣-titanium alloy, IMI 834 (all wt.%) Ti Balance
Al 5.8
Sn 4.06
Zr 3.61
Nb 0.70
Mo 0.54
Si 0.32
C 0.05
Fe 0.009
O 0.105
N 0.002
192
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
oxide growth is maximum at 800 ◦ C (Kp = 3.2 × 10−2 ) and minimum at 600 ◦ C (Kp = 3.1 × 10−4 ). On all the specimens, initially rutile is formed followed by alumina scale. It is very important to mention that the alumina scale is not uniform and at the same time rutile is also not protective at high temperatures. After reaching a certain thickness, the oxide scale started spalling for the specimens exposed at 800 ◦ C. 3.3. SEM
Fig. 1. As oxidised specimens of titanium alloy, IMI 834 after 100 h of oxidation at different temperatures showing oxide spallation at the edges.
scale is not adherent to the base alloy and thereby tended to spallation (Fig. 1). It is important to mention that no spallation of oxide scale is observed for other specimens exposed at 600 and 700 ◦ C.
The typical cross sections of specimens oxidized at 600, 700 and 800 ◦ C revealed the presence of primarily two regions (Fig. 3). The top region was covered with the oxide scale. The region just below the oxide scale is the ␣-case formed zone or oxygen dissolved/oxygen affected region. As can be seen, the depth of ␣-case is low in the case of specimen exposed at 600 ◦ C, more at 700 ◦ C and, highest at 800 ◦ C. It was also observed that the depth of ␣-case increases with increasing exposure time at a constant temperature. It indicates that ␣-case formation is a diffusion-controlled process and therefore, the depth of ␣-case increases with increasing time and temperature.
3.2. Oxidation kinetics 3.4. EPMA Fig. 2 shows the weight gain as a function of time for the titanium alloy, IMI 834 oxidized at 600, 700 and 800 ◦ C. The alloy followed parabolic kinetics at all the studied temperatures. As can be seen, the weight gain is minimal for the specimen exposed at 600 ◦ C and slightly more at 700 ◦ C, while significant weight gain is observed for the specimen exposed at 800 ◦ C. It is important to mention that the rate of
The X-ray scans of alloying elements of titanium alloy as well as oxygen after exposure of 100 h at 800 ◦ C in air is illustrated in Fig. 4. The results clearly show the presence of all alloying elements of titanium alloy in the oxide scale. Particularly, zirconium and tin were diffused outwardly and concentrated beneath the oxide scale. An important and con-
Fig. 2. Weight gain as a function of exposure time for titanium alloy, IMI 834 at various temperatures illustrating significant weight gain at 800 ◦ C.
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
193
Fig. 4. X-ray scans of alloying elements of titanium alloy and oxygen after exposure of 100 h at 800 ◦ C showing the oxygen dissolved region.
As mentioned earlier, adherence because of Therefore, it detaches essence, it is clear the in the titanium alloy.
the oxide scale does not have good the higher thickness of oxide scale. easily during sample preparation. In presence of oxygen dissolved region
3.5. Microhardness measurements
Fig. 3. Cross-sections of typical oxidised titanium alloy specimens showing ␣-case formed region and oxide scale at different temperatures.
firmed observation is the presence of dissolved oxygen zone just beneath the oxide scale. The high intensity peak represents oxygen present in the oxide scale while lower intensity one represents the oxygen dissolved in the titanium alloy. The dip observed between the two oxygen peaks is due to the detachment of oxide scale during preparation of the specimen.
Fig. 5 provides the measured microhardness profiles as a function of depth of ␣-case for the specimens exposed at 600, 700 and 800 ◦ C for a period of 100 h. The results show that the depth of ␣-case increases slowly from 600 to 700 ◦ C and then significantly at 800 ◦ C. The sub-surface of the titanium alloy affected for the specimen oxidized at 800 ◦ C is about 140 m, which is about 12 times that of the specimen exposed at 600 ◦ C (∼12 m) and about four times that of the specimen oxidized at 700 ◦ C (∼40 m). It is also very clear that the hardened zone due to dissolution of oxygen, exhibits a hardness of about 700 HV and the depth of ␣-case is about 140 m which is consistent with the SEM and EPMA results in case of the specimen exposed at 800 ◦ C. It also clearly shows that the rate of oxygen dissolution is considerably higher at 800 ◦ C in the titanium alloy and thus forms a hardened zone at a faster rate. This extent of surface embrittlement could certainly be sufficient to cause significant loss of ductility and failure by surface cracking under load [1]. It would be desirable to predict the depth of ␣-case for titanium alloys at a variety of elevated temperatures, to assess the
194
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
Fig. 5. Measured microhardness profiles of titanium alloy, IMI 834 after 100 h of oxidation at various temperatures showing the depth of ␣-case.
life of compressor components and help to eliminate failures during service. The next section describes the methodology and modeling of depth of hardening in titanium alloys through which the depth of ␣-case can be predicted with a reasonable accuracy and thereby to predict acceptable exposure conditions for titanium alloy components.
measured ␣-case depths for the specimens exposed at different elevated temperatures reflects the predicted trend (Fig. 5). It indicates that oxygen dissolution increases enormously at and above 800 ◦ C and affects the mechanical properties of the alloy. On the basis of measured parameters for IMI 834 for different times and at different temperatures, equivalent exposure conditions for an acceptable case depth would be:
4. An oxidation model for titanium alloys
4.4 h exposure at 900 ◦ C = 40 h at 800 ◦ C = 570 h of
An oxidation model has been proposed on the basis of several experiments at a variety of temperatures for different time intervals for the ␣-case formation or oxygen ingress into the titanium alloy. The Arrhenius equation, which has been modified appropriately, has been used for predicting the ␣-case formation in the present study. The exponential term used in the equation is related to a measure of activation energy for dissolution and diffusion of oxygen into the alloy [20]. Parameters taken into consideration are as follows: x: depth of hardening due to ingress of oxygen (m) T: exposure temperature (K) t: exposure time (h) Fig. 6 provides the predicted depth of ␣-case as a function of exposure time at different elevated temperatures in the titanium alloy. The depth of ␣-case is low at 600 ◦ C, increases slightly at 700 ◦ C. While at 800 ◦ C, significant depth of ␣-case is observed. For example, after exposure of 100 h, the depth of ␣-case is about 10 m at 600 ◦ C and 50 m at 700 ◦ C, which is five times more than that of the specimen exposed at 600 ◦ C. The predicted depth of ␣-case is about 20 times and 58 times more at 800 and 900 ◦ C, respectively when compared to the ␣-case predicted at 600 ◦ C. Practically
exposure at 900 ◦ C = 15000 h at 600 ◦ C In other words, if the component fabricated from the titanium alloy IMI 834 continuously operates at its maximum mechanical property temperature limit of 600 ◦ C at its designed service life of 15000 h, then the expected depth of alpha case, i.e., oxygen dissolved region, would be 109 m. This would be equivalent to an exposure of 570 h at 700 ◦ C, 40 h at 800 ◦ C or 4.4 h at 900 ◦ C. Another important factor is the correlation between the predicted and experimentally measured ␣-case formed regions (Fig. 7). The predicted values are always slightly higher than the experimentally measured values. It is due to the fact that the oxygen absorbed by titanium alloy during oxidation is used up not only for the formation of oxide scale but also for the ␣-case formation. Whereas in predicted values, it is assumed that the entire oxygen was used up for creating ␣case in the titanium alloy. This is the reason for observing slight deviation between experimental and predicted values. In other words, the marginal difference in the depth of ␣-case formation between the measured and predicted values can be considered as due to the formation of oxide scale on titanium alloys during exposure in air at elevated temperatures. It can
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
195
Fig. 6. Predicted depth of ␣-case for titanium alloy at different exposed temperatures.
be formulated as follows: ␣-casepred.
=
␣-casemeas.
+
oxide scale thickness
␣-casemeas.
=
␣-casepred.
140
=
173
− −
oxide scale thickness 33 at 800 ◦ C
40
=
45
−
5 at 700 ◦ C
Or
The above equation has been validated by substituting the data for 100 h. The equation is valid because of the fact that the titanium alloy initially oxidises to form oxide scale. Thereafter, the oxygen dissolves in the subsurface of titanium alloy and forms ␣-case. The above equation is valid including the figures. Therefore, for calculating the total affected region of the titanium alloy components in service, the
Fig. 7. Comparison between the measured and predicted ␣-case depths at 700 and 800 ◦ C.
196
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
oxide scale thickness has to be added to the measured ␣-case depth. As mentioned earlier, it is always possible to observe temperature hikes in service. Under such circumstances, the combination of ␣-case depths have to be taken into account for calculating total depth of ␣-case, i.e. the life of a component. For example, if the titanium alloy component continuously operates at 600 ◦ C for 5000 h and 15 h (in total period of operation) at 800 ◦ C, then the total depth of ␣-case would be 130 m (63 m at 600 ◦ C + 67 m at 800 ◦ C). In other words, 15 h operational hike at 800 ◦ C, reduces the life of a component by about 5000 h normal operational life, i.e. at 600 ◦ C. If the temperature hikes are observed at 700 ◦ C for some time, 800 ◦ C for some other time and normal operation at 600 ◦ C, then the total depth of ␣-case can be obtained by adding ␣-case depths at 600, 700 and 800 ◦ C. Failures of components in gas turbines most likely takes place due to non-accountability of these hikes. Therefore, it is essential to understand the consequences of hikes and the present study provides a valuable answer to these types of problems. In essence, the developed oxidation model helps to predict the component life once the temperatures at which the component operated and the period of operation at each temperature is known. It is also possible to predict the remaining life of a component by using the present oxidation model. From the present data, it is clear that there is an excellent correlation between the weight gain during oxidation process (Fig. 2) and the predicted depth of ␣-case (Fig. 6). Both the curves followed similar pattern at different temperatures studied in the present investigation. It indicates that ␣-case formation is the predominant reaction process during high temperature exposure rather than oxide scales formation though both processes contribute weight gain. This ␣case formation is a major contributing factor for the weight gain at all the exposed temperatures though the significant increase is observed above 700 ◦ C and affecting the mechanical properties of the titanium alloy components. The results clearly show that ␣-case formation is the life-limiting factor for the gas turbine engine compressor components fabricated from the titanium alloys. Therefore, it is essential to know the depth of ␣-case after exposure of definite number of hours at different elevated temperatures in order to avoid failures during service. The developed model is extremely useful to predict the depth of ␣-case with reasonably good accuracy and thus helps in avoiding failures during service. The contribution due to oxygen dissolution will be smaller than the estimated values as an outer layer of the alloy is converted to TiO2 . This aspect is clearly seen by observing smaller depth of alpha case formation for the alloy specimens exposed to lower temperatures for various time intervals (Fig. 5). This observation is consistent with the specimens exposed at different temperatures. Further, the oxygen concentration in the alloy may be smaller than the equilibrium value at low temperatures and oxygen diffusion
may be influenced by impurities. Even so the comparison clearly indicates that oxygen dissolution constitutes an important part of the parabolic oxidation. Direct estimates from studies on related specimens show that 25–30% or more of the reacted oxygen dissolves in the alloy at temperatures of 700–750 ◦ C [21,22] and more than 50% at 900–950 ◦ C [23]. From the present results, it is very clear that it is possible to predict the ␣-case formation with reasonably good accuracy for titanium alloys. It is also very clear that there is good agreement between the experimentally measured and predicted depth of ␣-case formed regions (Fig. 7). This study is extremely useful in the modern gas turbines to evaluate the life of the titanium alloy components without performing actual experiments on the alloys. It also helps to select the titanium alloy components with appropriate thickness to the designed life and thereby eliminating failures during service. The major advantage of the developed model is that it is very easy to use, no assumptions are required and time saving as no experiments can be conducted on the actual alloy components. The present study also helps in establishing that ␣-case formation plays a dominant role in observing weight gain during oxidation processes. The significant weight gain observed for the specimen exposed at 800 ◦ C when compared to the specimens exposed at 600 and 700 ◦ C, is due the fact that weight gain is observed not only due to the formation of oxide scale but also for the significant dissolution of oxygen in the subsurface zone of the titanium alloy.
5. Development of high performance coatings From the measured and predicted data, it is clear that the titanium alloys readily absorb oxygen at elevated temperatures and forms hardened phase of ␣-case even after few hours of exposure particularly at elevated temperatures and thus affecting the mechanical properties. Because of the technological importance of extending the operational temperatures of titanium alloy components primarily for aerospace applications and the danger of forming significant ␣-cases at high temperatures, protective coatings to protect them against oxidation and ␣-case formation are highly essential. With this goal, attempts were made and succeeded in developing high performance coatings for titanium alloys [4–7]. The recently developed coating produced through the combination of surface engineering techniques [5] enhances the life of the titanium alloy components considerably by successfully eliminating the oxygen uptake and thereby helps in improving the efficiency of gas turbine engines significantly.
Acknowledgements Defence Research and Development Organisation is gratefully acknowledged for providing financial assistance.
I. Gurrappa / Journal of Alloys and Compounds 389 (2005) 190–197
References [1] I. Gurrappa, Corr. Prev. Control 49 (2002) 79. [2] I. Gurrappa, A.K. Gogia, in: Proceedings of 5th National Convention on Corrosion, New Delhi, vol. I, pp. 210–219. [3] I. Gurrappa, Oxid. Met. 56 (2001) 73. [4] I. Gurrappa, Platinum Metals Rev. 45 (2001) 124. [5] I. Gurrappa, A.K. Gogia, J. Mater. Sci. Technol. 17 (2001) 581. [6] I. Gurrappa, A.K. Gogia, in: Proceedings of the 5th International Symposium on High Temperature Corrosion, May 2000, Le Embitz, France. [7] I. Gurrappa, A.K. Gogia, Surf. Coat. Technol. 139 (2001) 216. [8] R.K. Clark, J. Unnan, K.E. Wiedemann, Oxid. Met. 28 (1987) 391. [9] D.W. Mckee, K.L. Luthra, Surf. Coat. Technol. 56 (1993) 391. [10] H.L. Du, P.K. Datta, J.S. Burnell-Gray, G.B. Lewis, J. Mater. Sci. 30 (1995) 2640.
197
[11] S. Fujishiro, D. Eylon, Met. Trans. 11A (1980) 1259. [12] US Patent 4,946,749. [13] C. Leyens, M. Schmidt, M. Peters, W.A. Kaysser, J. Mater. Sci. Eng. A239/240 (1997) 680. [14] C.H. Koo, T.H. Yu, Surf. Coat. Technol. 126 (2000) 171. [15] J.R. Nicholls, JOM 52 (2000) 28. [16] C. Leyens, M. Peters, W.A. Kaysser, Mater. Sci. Forum 251–254 (1997) 769. [17] B. Cockeram, R.A. Rapp, Oxid. Met. 45 (1996) 427. [18] C. Leyens, M. Peters, W.A. Kaysser, Surf. Coat. Technol. 94/95 (1997) 34. [19] H. Liu, S. Hao, X. Wang, Z. Feng, Script. Met. 39 (1998) 1443. [20] I. Gurappa, in: Proceedings of National Symposium on Electrochemical Science and Technology, July 2002, Bangalore, p. 10. [21] T. Hurlen, J. Inst. Met. 89 (1960/61) 128. [22] J.E. Lopes Gomes, A.M. Huntz, Oxid. Met. 14 (1980) 249. [23] P. Kofstad, J. Less-Common Met. 12 (1967) 449.