Melt-related inclusions in titanium-alloy disk materials resulted in a greater-than-acceptable number of in-flight engine shut downs in the aircraft industry during the 1970s and 1980s[1]. The majority was due to inclusions in titanium-alloy disk materials, and a limited number of shut downs resulted from uncontained in-flight disk separations. This situation led to research at GE Aircraft Engines (GEAE), Cincinnati, Ohio and some industry efforts to significantly reduce the number of inclusions in titanium-alloy disk materials.

Fig 1 Melt-related inclusions in the titanium industry: left, hard alpha inclusion, or high interstitial defect (HID); right, high density inclusion (HDI)

Advanced-melting drivers

Inclusions of concern in premium titanium-alloy disk materials are hard alpha (Type I interstitially stabilized hard alpha) and HDIs, or high-density inclusions (Fig. 1). The frequency of these inclusions in titanium-alloy disk materials can be reduced by eliminating them during the production process (which is more cost effective) and by using billet-inspection techniques where the inclusions have a high probability of detection (POD).

Premium quality Ti-alloy disk materials were produced in the 1980s using double vacuum arc remelt (2 x VAR) and triple VAR (3 x VAR) melting practices. By the mid 1980s, most engine producers used 3 xVAR. However, 3 xVAR was limited in its ability to eliminate hard alpha inclusions, as shown in a study of a 3 x VAR Ti-17 alloy heat seeded with high N content sponge particles[2]. HDIs are known to survive the VAR process. The 3 x VAR process predominately relies on elimination of such inclusion-forming materials from the input stock prior to the melting process, so a process having a greater capability to eliminate the melt-related inclusions was needed.

Hard alpha inclusions and HDIs in titanium materials have the greatest POD using immersion ultrasonic nondestructive inspection (NDI), which can detect inclusion cracks and voids resulting from longitudinal strain during billet processing. In the 1980s, the best immersion ultrasonic detection capability available for typical billet diameters was to a #3 flat bottom hole, or #3 FBH (3/64 in., or 1.2 mm). Inspection resolution capability was lower for billet diameters greater than about 12 in. (305 mm).

While immersion ultrasonic inspection was adequate to find high N-content (8 to 15% N) hard alpha inclusions having significant cracking and void formation, it was not adequate to find the lower interstitial content (2% N, for example) hard alpha inclusions that were not cracked or significantly voided during billet processing. Enhanced ultrasonic inspection capabilities were needed to detect the considerably smaller voids created for the lower N content inclusions.

Fig 2 Hearth-melting processes: left, electron-beam cold hearth melting; right, plasma-arc cold hearth melting

Advanced melting and NDI improvements

An advanced melting process, known as HM +VAR (hearth melting + a single VAR step) was developed jointly by GEAE, Axel Johnson Metals Inc. (now Timet) and Wyman-Gordon Co. (now a Precision Cast Parts subsidiary) in the mid 1980s[2]. The HM processes included electron-beam and plasma-arc cold-hearth melting (EBM and PAM, respectively, as shown in Fig. 2). Both processes have significantly greater capabilities to eliminate hard alpha inclusions and HDIs compared with the 3 x VAR process[1-15].

Fig 3 Integrated PAM process simulation model developed by Concurrent Technologies Corp., Johnstown, Pa.

HM process modeling is helping researchers understand process capabilities and limitations and can serve as a development tool for process enhancement and as a quality control tool to review process anomalies. For example, an EBM model (Innovative Re-search Corp.'s COMPACT code) predicts molten hearth pool size, shape and temperature profiles and tracks input inclusion particle paths (for specified sizes and densities). An integrated PAM process model approach (Concurrent Technologies Corp., Johnstown, Pa., project) is illustrated in Fig. 3.

In addition, sensors have been developed to improve process control. For example, a near infrared (NIR) camera continuously monitors pool surface temperature and pool surface area, and a casting rate monitor controls residence time to ensure that adequate residence time at temperature is consistently maintained for inclusion elimination.

Currently, only Timet's "B" EBM furnace and Allvac's (an Allegheny Technologies Co.) PAM I furnace are qualified because of stringent premium quality titanium-alloy HM refining furnace requirements. However, Oremet (now Allvac) is operating an EBM furnace at International Hearth Melting, and RMI Titanium Co. is operating a small PAM furnace.

Inclusion detection was extended from the #3 FBH range using the POD capabilities of conventional immersion ultrasonic NDI to 0.75#2 FBH with the development of a multizone ultrasonic inspection (MZUI) process. GEAE began development of the process in the early 1990s and implemented the process on a production basis in 1994[16]. Other companies also are using the technique today. MZUI has successfully detected very small pores (strain-induced porosity, or SIP) in billets and some melt-related inclusions that would have been missed using conventional ultrasonic inspection. Further evaluation of MZUI through the Engine Titanium Consortium (ETC) formed under Federal Aviation Administration (FAA) funding through Iowa State University is aimed at implementing this new NDI method throughout the titanium industry.

Fig 4 Inclusion frequencies have decreased in premium-quality titanium alloys since 1990

Joint titanium-industry efforts

Joint industry programs aimed at minimizing melt-related inclusion frequencies in titanium alloy materials used in premium quality applications include the Jet Engine Titanium Quality Committee (JETQC) and the Hard Alpha Workshops. JETQC (formed under the auspices of FAA) deals with titanium-alloy melt-related inclusions, serving as an early warning system when a "flurry" of inclusions in material from a particular source is encountered. There are a limited number of suppliers providing titanium billet material to all engine manufacturers, so a problem at one manufacturer signals other engine manufacturers to look for a similar situation, and the engine manufacturers share the safety-related information. Suppliers' provide billet and bar inspection reports detailing quantity of material inspected, inclusion "finds," alloy, product size (diameter), ultrasonic signal, inclusion size and type (hard alpha or HDI) and ingot location. Annual composite ingot maps show inclusion finds by type, but do not reveal to the suppliers who had the inclusions. Suppliers can rate their own performance against the industry from these annual reports.

JETQC hard alpha inclusions data are used (with suppliers' permission) by the AIA Rotor Integrity Subcommittee in conjunction with prior data on inclusions found in engine components to establish a hard alpha inclusion size distribution curve[17] for engine-design purposes. These melt-related inclusion data have led suppliers to focus greater attention on eliminating sources of these inclusions. Thus, industry hard alpha and HDI defect rates have decreased significantly since 1990 (the first JETQC report year) as shown in Fig. 4.

The Hard Alpha Workshop (formed by suppliers and engine manufacturers) shares information on characterization of inclusions to identify the inclusion manufacturing source. Hard alpha inclusions are a safety-related issue; therefore, even the strongest competitors need to share such information. From the information presented at the first meeting in 1997 and at the second in 2000, it appears that many aspects of the hard-alpha formation and characterization technology still are not fully understood.

Production and inspection track record

The average hard alpha and HDI rates in combined (2 x VAR, 3 x VAR and HM+ VAR) billet and bar products have been significantly reduced between 1990 and 2000. However, flurry periods of higher industry defect rates have occurred. For example, an HDI flurry occurred in 1993-4 where the defect rate for 3 x VAR material considerably exceeded 1990 rates. HM+VAR heats produced in that same time frame and from the same input materials (subsequently known to contain HDIs) were totally free of HDIs. Only two HDIs (one EBM and one PAM) have ever been found in HM+VAR product, and they came from a hearth-to-ingot pour-lip weld repair, and did not traverse the hearth. Therefore, the HDI rate for HM+VAR product remains essentially at zero.

No hard alpha or HDIs were found in HM+VAR material or the early production time period. Hard alpha inclusions have been found in a few HM+VAR heats, but based on studies in 1995, there is a high probability that most of those inclusions up to that time were the result of grinding on the HM electrode prior to VAR[13]. A few post-HM inclusion introductions have been encountered from other identified sources after 1996, when grinding was no longer allowed. HM+VAR hard alpha inclusion frequency rates are similar to those of 3 x VAR. Investigations of HM+VAR heats containing hard alpha inclusions show that at least 93% of these were post-HM introductions. Also, it is important to note that results of evaluations of recent HM+VAR product using the MZUI method are being compared with results from 3 x VAR product evaluated using conventional ultrasonic inspection. Even though these results did not show HM + VAR being superior to 3 x VAR, efforts continue to eliminate peripheral post-hearth melt sources and to make the HM+VAR process the best premium process.

Future considerations

While the goal remains to develop an HM-only process to eliminate peripheral post-HM melt hard alpha sources observed in the HM+VAR process, efforts have slowed due to lack of development funding. A USAF ManTech Program demonstrated the feasibility of an HM-only process, but revealed remaining risk issues that need to be addressed before considering production scale-up[12], including chemistry control for both HM-only processes and the possibility of helium entrapment in the PAM process. Contrary to the initial belief that ingot surface finish might be a significant economic disadvantage, the program showed that the EBM ingot surface and the half-atmosphere practice PAM ingot surface were acceptable for billet processing.

Control of Ti-6-4 alloy chemical composition using the HM+VAR process is very good for both EBM and PAM, and also could be acceptable using the HM-only process. Control of the more complex Ti-17 chemical composition using the HM + VAR process was only acceptable up to about 1995, but further work in 1996-97 achieved improved composition control for the HM (both EBM and PAM) +VAR Ti-17 product. Control of the compaction step prior to HM can play a significant roll in both the refining capability and composition control of the process and also will improve the HM-only product feasibility.

It appears that the half-atmosphere practice alleviates helium entrapment-the PAM- only issue for the positive pressure practice[12]. Verification that the increased superheat of the half-atmosphere practice eliminates helium entrapment requires further work.

References

• C.E. Shamblen and D.J. Tilly, "Inclusion Free Titanium Material Efforts," Proc. Electron Beam Melting and Refining State of the Art - 1997, ed. R. Bakish, Englewood, N.J., Bakish Materials Corp.
  
• C.E. Shamblen and G.B. Hunter, "Titanium Base Alloy Clean Melt Process Development," Proc. 1989 Vacuum Metallurgy Conference, ed. L.W. Lherbier and J.T. Cordy, Warrendale, Pa., Iron & Steel Society Inc., 1989, p 3-11
  
• C.H. Entrekin, "The Removal of High Density Inclusions by the Hearth Refining Process," Proc. Electron Beam Melting and Refining State of the Art - 1985, ed. R. Bakish, Englewood, N.J., Bakish Materials Corp., p 40-47
  
• M. Ostheimer and S.R. Seagle, "Utilization of Electron Beam Melting in the Production of Defect Free Ti-6Al-4V Ingots," Sixth World Conference on Titanium, ed. P. Lacombe, R Tricot, and G. Beranger, les editions de physique, Les Ulis Cedex, France, 1988, p 579-82
  
• R.N. Jarrett, S.H. Reichman and R.G. Broadwell, "Defect Removal Mechanisms in Hearth Melted Ti-6Al-4V," Sixth World Conference on Titanium, ed. P. Lacombe, R Tricot, and G. Beranger, les editions de physique, Les Ulis Cedex, France, 1988, p 593-5
  
• C.E. Shamblen, "Ongoing Challenges for Titanium Alloy Cleanliness Improvement in Aircraft Engine Disk Materials," Proc. Electron Beam Melting and Refining State of the Art - 1990, ed. R. Bakish, Englewood, N.J., Bakish Materials Corp., p 49-54
  
• G.B. Hunter, "Process Control Considerations for Ti-Alloy Hearth Melting," Proc. Electron Beam Melting and Refining State of the Art - 1992, ed. R. Bakish, Englewood, N.J., Bakish Materials Corp., p 39-50
  
• W.R. Chinnis and W.H. Buttrill, "Production Titanium Plasma Cold Hearth Melting," Titanium '92 Science and Technology, VIII, ed. F.H. Froes and I.L. Caplan, 1992, TMS, Warrendale, Pa., p 2363-70
  
• C.E. Shamblen, W.H. Buttrill, and G.B. Hunter, "Titanium Alloy Hearth Melt Technology," Titanium '92 Science and Technology, Vol. III, ed. F.H. Froes and I.L. Caplan, 1992, TMS, Warrendale, Pa., p 2451-58
  
• R.G. Menzies, C.E. Shamblen, D.G. Backman, and G.B. Hunter, "Process Design and Control for Aircraft Engine Materials," Proc. Electron Beam Melting and Refining State of the Art - 1993, ed. R. Bakish, Englewood, N.J., Bakish Materials Corp., p 12-22
  
• W.H. Buttrill, C.E. Shamblen, and G.B. Hunter, "Hearth Melting of Titanium Alloys for Aircraft Engine Applications," Metallurgy and Technology of Practical Titanium Alloys, ed. S. Fujishiro, D. Eylon, and T. Kishi, TMS, 1994, p 209-17
  
• C.E. Shamblen, "Titanium Alloy Hearth Melt 'Only' Technology Development," Titanium '95 Science and Technology, ed. P.A. Blenkinsop, W.J. Evans, and H.M. Flower, University Press, Cambridge-UK, 1995, p 1438-45
  
• W.H. Buttrill and C.E. Shamblen, "Hearth Melt Plus Vacuum Arc Remelt: Production Status," Titanium '95 Science and Technology, ed. P. Blenkinsop, W. Evans, and H. Flower, University Press, Cambridge-UK, 1995, p 1446-53
  
• D.J. Tilly, C.E. Shamblen, and W.H. Buttrill, "Hearth Melting of Titanium Alloys for Aircraft Engine Applications Update," Proc. Electron Beam Melting and Refining State of the Art - 1996, ed. R. Bakish, Englewood, N.J., Bakish Materials Corp., p 198-205
  
• Y. Pang, private communication on "Titanium Alloy Hearth Melt Technology Enhancement Program," NAVAIR Contract No. N00140-92-C-BC49
  
• J.L. Bartos, et al., "Advanced Ultrasonic Inspection Techniques for Titanium Billet Materials," Titanium '95 Science and Technology, ed. P. Blenkinsop, W. Evans, and H. Flower, University Press, Cambridge-UK, p 1513-20
  
• J. Tschopp, et al. (Sub-Team to the Aerospace Industries Association Rotor Integrity Sub-Committee), "The Development of Anomaly Distributions for Aircraft Engine Titanium Disk Alloys," American Institute of Aeronautics and Astronautics, Inc., 1997