Enhancing the Surface Integrity of Ti-6Al-4V Alloy through Cryogenic Burnishing

Available online at www.sciencedirect.com

ScienceDirect Procedia CIRP 13 (2014) 243 – 248

2nd CIRP 2nd CIRP Conference on Surface Integrity (CSI)

Enhancing the surface integrity of Ti-6Al-4V alloy through cryogenic burnishing J. Caudilla, B. Huanga, C. Arvina, J. Schoopa, K. Meyerb, I.S. Jawahira* a

Institute for Sustainable Manufacturing, University of Kentucky, Lexington, Kentucky 40506, USA

b

GE Aviation, Cincinnati, Ohio 40506, USA * Corresponding author. Tel.: 859-323-3239; fax: 859-257-1071.E-mail address: [email protected].

Abstract Burnishing is a chipless machining process that modifies the surface integrity by severe plastic deformation (SPD) of the burnished material. The application of cryogenics during the burnishing process acts to rapidly cool the burnished work material; thus, leading to surface modification. In this research work cryogenic burnishing was performed to produce the SPD layer to obtain improved properties in Ti-6Al-4V alloy. In the SPD layer produced by cryogenic burnishing, increased hardness, refined grain structure, and a drastically improved surface finish, along with compressive residual stresses are achieved. The cryogenic burnishing of Ti-6Al-4V alloy was shown to be superior to flood-cooled and dry burnishing in terms of surface integrity and the likely functional performance of components during the service life of such components. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2014 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of The International Scientific Committee of the “2nd Conference on Surface Selection and peer-review under responsibility of The International Scientific Committee of the “2nd Conference on Surface Integrity” Integrity” in the person of the Conference Chair Prof Dragos Axinte [email protected] in the person of the Conference Chair Prof Dragos Axinte [email protected] Keywords: Cryogenic Burnishing; Surface Integrity; Titanium

1. Introduction Ti-6Al-4V is the most widely employed Titanium alloy used in the aerospace industry. Its unique combination of high strength, low weight (relative to steel), thermo-mechanical properties, formability, and corrosion resistance make it an ideal material choice for aerospace airframe and engine components [1-3]. Severe plastic deformation (SPD) processes have been commonly used to enhance the surface integrity properties of materials by creating a layer of grain refinement in the surface and subsurface regions that exhibit enhanced mechanical properties [4-6]. In particular, SPD processes have also been used to modify the surface integrity of Ti-6Al-4V [7-9]. Burnishing is one such process for SPD generation, and it has been shown in the current work that by applying cryogenics, which rapidly cools the workpiece during workhardening, surface integrity can be significantly modified as compared to burnishing conducted under dry and flood cooling conditions.

2. Experimental work 2.1 Material Information The Ti-6Al-4V alloy under investigation was supplied by GE Aviation, Cincinnati, Ohio, with a composition (wt. %) of approximately 6.31% Al, 4.06% V, 1.16 % Fe, less than 0.25% of O, Si, C, Mo, and the remainder Ti [10]. The material was prepared as 145.4 mm long cylindrical bars with a diameter of 38.6 mm. The microstructure was composed of equiaxed primary αgrains and lamellar α+β [11] with a measured mean hardness of 34.1 HRC. 2.2 Burnishing Process Burnishing experiments were conducted on a HAAS TL-2 CNC lathe with an Air Products ICEFLYTM cryogenic delivery system. Liquid Nitrogen was applied at the contact interface between the tool and work piece. The tool used was a UniversalTM burnishing tools UBT-

2212-8271 © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Selection and peer-review under responsibility of The International Scientific Committee of the “2nd Conference on Surface Integrity” in the person of the Conference Chair Prof Dragos Axinte [email protected] doi:10.1016/j.procir.2014.04.042

244

J. Caudill et al. / Procedia CIRP 13 (2014) 243 – 248

T2 with a 3.04 mm diameter tungsten carbide roller. The following modifications were made to the tool to improve its accuracy and its functionality in the presence of liquid nitrogen: The stock gauge was replaced with a more accurate version, the mounting bracket was redesigned in order to move the gauge to the top of the tool for accurate real time force observation, the tolerance of the mounting holes were tightened in order to remove slack from the system, a polycarbonate sled was added to reduce the relative sliding friction of the parts, a Teflon shim was added to the system to reduce friction between the gauge bracket and the base, springs were added to reduce vibration and to hold tension where necessary, and low temperature grease was used in order for the tool to function properly in the -196 °C environment created by liquid nitrogen. The final assembled burnishing tool and the experimental setup are shown in figure 1, and the sub-components of the tooling assembly are shown in figure 2.

c b

a Figure 1: Experimental setup: (a) modified burnishing tool, (b) workpiece, (c) nozzle for liquid nitrogen application

Surface roughness measurements were made by using a Zygo New View 7300 scanning white light interferometer. Hardness measurements were obtained using a Vickers indentation profile on a CLARK Digital micro-hardness tester with an applied load of 100 gf and a 15 second dwell time. Microhardness indentation measurement and microstructure analysis were performed using a Nikon EPIPHOT 300 optical microscope. 3. Results and Discussion 3.1 Surface Roughness As-burnished surface roughness showed significant reduction in the Ra values of the finished surface as compared to the pre-burnished samples, which were turned to roughness Ra = 1.31 μm. After processing, dry, flood-cooled, and cryogenic burnishing showed maximum reductions of 63.4%, 58.2%, and 56.7% respectively. These results are shown in Figure 3. Since Ti-6Al-4V possesses an extremely low thermal conductivity, the local temperature rise in the burnishing zone can be quite significant. If there is no mechanism of heat removal, as is the case in dry burnishing, surface asperities can more readily be deformed during burnishing. This increase in temperature-induced plasticity effect offers a possible explanation for the more uniform and slightly smoother surface finish in dry burnishing. The effect of preload upon surface roughness was found to be minimal, regardless of burnishing condition.

c

b

1.6

d

Surface roughness Ra (μm)

a

2.3 Measuring Equipment

Surface roughness vs. Preload

1.4 1.2 1 0.8 0.6 0.4 0.2 0

Figure 2: Sub-components of tooling assembly: (a) force gauge, (b) modified mounting bracket, (c) tungsten carbide roller, (d) Teflon shim

The cylindrical bars were burnished under dry, flood-cooled, and cryogenic conditions at preloads of 1000 N, 1500 N, 2000 N, and 2500 N. The following parameters were held constant: pre-burnishing feed rate = 0.18 mm/rev, burnishing feed rate = 0.05 mm/rev, and burnishing speed = 150 m/min.

1000

1500

2000

2500

Preload (N) Cryogenic Dry

Flood Pre-Burnished

Figure 3: Surface roughness comparison for dry, flood-cooled, and cryogenic burnishing conditions at varying preload

245

J. Caudill et al. / Procedia CIRP 13 (2014) 243 – 248

Investigation of surface hardness reveals significant increases in hardness from the virgin material. As can be seen in Figure 4, all tested burnishing conditions exhibited elevated surface hardness values. Burnishing performed under cryogenic and flood-cooled conditions were capable of producing maximum surface hardness increases of 64.2% and 48.9% respectively, while dry burnishing only led to a maximum surface hardness increase of 32.8%. Moreover, both cryogenic and floodcooled burnishing showed a trend of increasing surface hardness with increasing preload, whereas in dry burnishing, increased preload led to lower surface This observation offers further hardness values. evidence for justifying the thermal softening resulting from increased temperature with no cooling in dry burnishing.

to fully elucidate the complex interplay of strain hardening and thermal softening in Ti-6Al-4V during burnishing under various conditions. Subsurface microhardness variation for cryogenic, dry, and flood cooling conditions are illustrated in Figures 5-7 for a selected preload of 2500 N. Measurements were taken to a depth of 2 mm in order to establish the extent of the burnished layer depth (BLD), which is illustrated in Figure 7. Hardness vs. Depth: 2500 N Preload

60 55 Hardness (HRC)

3.2 Microhardness

Cryogenic Dry Flood Bulk

50

45 40 35 30 0

70

40

1000

1500

2000

Depth from surface (μm)

54.2 HRC

60 50

500

56.0 HRC 52.7 HRC

51.2 HRC

Figure 5: Subsurface microhardness variation for a selected preload of 2500 N for cryogenic, flood-cooled, and dry burnishing conditions

50.8 HRC

50.2 HRC

49.9 HRC 46.6 HRC 45.3 HRC

30

43.2 HRC

42.3 HRC

42.3 HRC

20 10 0 1000

1500

2000

2500

Preload (N) Cryogenic

Flood

Hardness vs. Depth: 2500 N Preload (Inlet)

60 Hardness (HRC)

Surface Hardness Increase (%)

Surface Hardness Increase vs. Preload

Dry

Cryogenic Flood Dry Bulk

55 50

45 40 35 30

Figure 4: Surface Hardness Increase vs. Preload for cryogenic, floodcooled, and dry burnishing conditions

50

100

150

200

Depth from surface (μm) Figure 6: Inlet of subsurface microhardness variation

Hardness vs. Depth: 2500 N Preload (Outlet) BLD

Hardness (HRC)

The presence of a cooling mechanism in flood and cryogenic burnishing offsets the effect of thermal softening seen in dry burnishing. Coolant removes heat generated during the burnishing process, and thus allows the material to be compressed at higher preloads. When preload is increased, the dislocation density in the titanium workpiece increases, leading to strain Therefore, there are two competing hardening. mechanisms during the burnishing of Ti-6Al-4V, namely strain hardening and thermal softening. Without coolant, thermal softening is more dominant than strain hardening because heat accumulates near the surface of the workpiece, essentially annealing the dislocation network and thus lowering hardness. Conventional flood cooling carries away some heat but liquid nitrogen is an even more capable coolant. Consequently, cryogenic burnishing is the most effective method of increasing surface hardness due to the strong cooling effect of liquid nitrogen. Further study will be necessary

0

40 39 38 37 36 35 34 33 32 1000

Cryogenic Dry Flood Bulk

1200

1400

1600

1800

2000

Depth from Surface (μm) Figure 7: Outlet of subsurface microhardness variation and illustration of burnished layer depth (BLD)

J. Caudill et al. / Procedia CIRP 13 (2014) 243 – 248

Microhardness depth profiles in as-burnished Ti-6Al4V varied, and were distinct for each burnishing condition. Cryogenic burnishing showed higher hardness values throughout the burnished layer when compared with flood-cooled and dry burnishing. As Figure 7 illustrates, each burnishing condition was associated with a different BLD. The value of the BLD was taken as the depth at which hardness reached and maintained an absolute value of ±1.1 HRC from the bulk hardness. This represents the range of a 95% confidence interval obtained when measuring the hardness of the virgin material.

Depth from Surface (μm)

The BLD values for all tested experimental preloads and burnishing conditions are summarized in Figure 8. It was observed that for all burnishing conditions increased preload led to a higher BLD. Also, cryogenic experiments produced a deeper burnished layer than flood-cooled or dry burnishing for all tested preloads (Maximum differences of 25% and 16.7% for dry and flood-cooled respectively). Likewise, flood cooling produced a deeper BLD than dry burnishing for all conditions. From a 1000 N preload to 2500 N, cryogenic, flood-cooled, and dry burnishing show an increase in their BLD by 25%, 21.4%, and 16.7% respectively. Higher preloads generate higher surface pressure assuming constant cross-sectional contact area. This in turn implies higher contact stress, and consequently increased deformation and a larger BLD. The differences observed between burnishing conditions can again be explained by the effect of temperature on the workpiece plasticity. Similar to forging, increased temperature as in dry burnishing reduces the effect of strain hardening and reduces the change in mechanical properties due to processing. The rapid temperature decrease of cryogenic cooling effectively increases the amount of permanent strain hardening due to a reduction in thermal softening.

Figure 9 illustrates the mean increase in hardness over the entire burnished layer from the workpiece surface to the BLD. The mean is meant to present a measure for the overall change in sub-surface hardness as a result of burnishing. Mean Hardness Increase (%)

246

30

Mean Hardness Increase in Burnished Layer

25 20

15 10 5 0 1000

1500

2000

2500

Preload (N) Cryogenic

Flood

Dry

Figure 9: Mean Hardness Increase in Burnished layer vs. Preload for cryogenic, flood-cooled, and dry conditions

Just as the surface hardness trends observed in Figure 4, the mean hardness increase displays a similar profile for the various burnishing conditions. Consequently, it is assumed that the general mechanisms giving rise to these trends in surface hardness, i.e. thermal softening and strain hardening also control hardness in the sub-surface. The mean hardness in flood cooling is essentially unchanged across the spectrum of tested preloads. In terms of sub-surface hardness, cryogenic burnishing exhibited higher hardness values than flood-cooled or dry burnishing throughout the BLD for all tested preloads. Again, this observation is readily explained by the ability of liquid nitrogen to mitigate the negative effects of thermal softening, thus preserving the dominance of strain hardening as the mechanism controlling (sub-) surface hardness.

Burnished Layer Depth vs. Preload

1800 1600 1400 1200 1000 800 600 400 200 0

3.3 Microstructural Analysis

1000

1500

2000

2500

Preload (N) Cryogenic

Flood

Dry

Figure 8: Burnished Layer Depth vs. Preload for cryogenic, floodcooled, and dry conditions

A novel approach was used to determine the correlation between changes in surface integrity parameters due to burnishing and observed changes in the microstructure of Ti-6Al-4V. It is hypothesized that by carefully over-etching a given sample, the prevalent compressive residual stresses in the SPD layer would induce a layer of corrosion resistance that would resist the chemical etching more than the bulk. The result of this would be an optical reflectivity gradient from the burnished surface of the material into its bulk regions that could be measured and quantified. Figure 10 shows 2D cross sectional views of the microstructures at 25x

247

J. Caudill et al. / Procedia CIRP 13 (2014) 243 – 248

Cryogenic

magnification for cryogenic, flood-cooled, and dry burnishing conditions for 2500 N and 1000 N preloads.

2500 N Preload

Quantification of the observed optical gradient was performed by measuring the surface reflectivity in terms of intensity using an image analysis software package. By determining at what depth the reflectivity fell below the mean and no longer displayed a trend of decreasing intensity, allowed for the establishment of the SPD layer depth for a given burnishing condition. These results are illustrated in Figures 11-13. Intensity vs. Depth: Cryogenic 2500 N Preload

Flood-cooled

1000 N Preload

Intensity (cd)

55000

Cryogenic Mean

95 μm

50000 45000 40000 35000 30000

0

2500 N Preload

100

200

300

Depth from surface (μm) Figure 11: Intensity vs. Depth for cryogenic burnishing at 2500 N preload

Intensity vs. Depth: Flood 2500 N Preload

1000 N Preload

56 μm

Dry

2500 N Preload

Intensity (cd)

55000 50000

Flood Mean

45000 40000 35000 30000 0

1000 N Preload

300

Figure 12: Intensity vs. Depth for flood-cooled burnishing at 2500 N preload

100 μm

Intensity vs. Depth: Dry 2500 N Preload 14 μm

55000

Figure 10: 25x Microstructure images at 2500 and 1000 N preloads for cryogenic, flood-cooled, and dry burnishing

Dry Mean

50000 Intensity (cd)

From Figure 10 it is readily observable that a layer of enhanced surface reflectivity is indeed present in the surface and near surface regions of samples burnished with a 2500 N preload. For samples burnished under cryogenic and flood-cooled conditions this optical gradient is particularly prominent. For samples tested at 1000 N, the reflectivity gradient is not easily discerned even though it was etched under identical parameters. This observation is consistent with our hypothesis, since this preload also showed the lowest hardness values. Furthermore, a more significant optical gradient is expected to be present at higher preloads for cryogenic and flood-cooled samples but not for dry burnishing, where thermal softening led to reduced hardness values.

100 200 Depth from surface (μm)

45000 40000 35000 30000 0

100 200 Depth from surface (μm)

300

Figure 13: Intensity vs. Depth for dry burnishing at 2500 N preload

From this intensity analysis it was determined that cryogenic burnishing produced a SPD layer that was 69.6% thicker than flood-cooled and 578% thicker than dry burnishing at a given preload. In the future, grain size and volume fraction of the individual phases will be

248

J. Caudill et al. / Procedia CIRP 13 (2014) 243 – 248

measured, and X-ray diffraction methods will be used to measure the residual stress depth profile, which will allow further evaluation of this novel technique. A comprehensive picture of the surface integrity after burnishing is shown in Figure 14 for Ti-6Al-4V cryogenically burnished with a 2500 N preload. It illustrates the presence of a highly compressive and refined SPD layer near the surface which correlates with higher hardness values and likely the most significant levels of compressive residual stresses. Beneath the SPD layer a transition zone is observed that shows decreasing hardness and compression until a unique BLD is reached and the mechanical properties exhibit unmodified characteristics.

SPD

Transition Zone BLD

1mm 25x Figure 14: Post-burnished structure of Ti-6Al-4V cryogenically burnished at 2500 N

4. Conclusions Based on the preliminary experimental observations presented in this paper, it can be concluded that burnishing of Ti-6Al-4V is controlled by two major mechanisms namely strain hardening and thermal softening. This hypothesis was substantiated by the experimental observations that show lower surface roughness, i.e., higher plasticity was found in dry burnishing over burnishing conducted under cryogenic and flood cooling. Moreover, the same thermal softening effect led to lower hardness values in dry and flood burnishing compared to cryogenic burnishing. Cryogenics controls the detrimental effect of thermal softening, thus allowing for increased hardness throughout the burnished layer depth. Acknowledgements The authors would like to sincerely thank GE Aviation for their sponsorship of the burnishing research presented in this paper.

References [1] Froes, F.H., Eylon, D., Bomberger, H.B., 1985. Titanium Technology: Present Status and Future Trends, ITA. Denver, Co. [2] Froes, F.H., Yau, T-L, Weidenger, G.G., 1996. Titanium, Zirconium, and Hafnium, Weinheim FRG: VCH Materials Science and Technology. [3] Peters M., Leyens, C., 2003. Titanium and Titanium Alloys, Wiley-VCH, Weinheim. [4] Yang, S., Dillon Jr., O.W., Puleo, D.A., Jawahir, I.S., 2013. Effect of cryogenic burnishing on surface integrity modifications of Co-Cr-Mo biomedical alloy, J Biomed Mater Res Part B 101B, p. 139-152. [5] Vailev, R.Z., Alexandrov, I.V., Enikeev, N.A., Murashkin, M.Y., Semenova, I.P., 2010. Towards enhancement of properties of UFG metals and alloys by grain boundary engineering using SPD processing, Rev Adv Mater Sci 25, p. 1-10. [6] Sato, M., Tsuji, N., Minamino, Y., Koizumi, Y., 2004. Formation of nanocrystalline surface layers in various metallic materials by near surface severe plastic deformation, Sci Technol Adv Mater 5, p.145-152. [7] Zherebtsova, S.V., Salishcheva, G.A., Galeyeya, R.M., Valiakhmetova, O.R., Mironova, S.Yu., Semiatinb, S.L., 2004. Production of submicrocrystalline structure in large-scale Ti-6Al4V billet by warm severe deformation processing, Scripta Materialia 51 (12), p. 1147-1151. [8] Yapici, G.G., Karaman, I., Luo, Z.P., 2006. Mechanical twinning and texture evolution in severely deformed Ti-6Al-4V at high temperatures, Acta Materialia 54, p. 3755-3771. [9] Semenova, I.P., Raab, G.I., Ssaitova, L.R., Valiev, R.Z., 2004. The effect of equal-channel angular pressing on the structure and mechanical behavior of Ti-6Al-4V alloy, Materials Science and Engineering A 387-389, p. 805-808. [10] Majorell, A., Srivatsa, S., Picu, R.C., 2002. Mechanical behavior of Ti-6Al-4V at high and moderate temperatures-Part I: Experimental results, Materials Science and Engineering A 326, p. 297-305. [11] Altenberger, I., Nalla, R.K., Noster, U., Scholtes, B., Ritchie, R.O., 2007. On the Fatigue Behavior and Associated Effect of Residual Stresses in Deep-Rolled and Laser Shock Peened Ti-6Al-4V Alloys at Ambient and Elevated Temperatures, National Turbine Engine High Cycle Fatigue Conference.