Scholarly article on topic 'Characterising the Integrity of Machined Surfaces in a Powder Nickel Alloy used in Aircraft Engines'

Characterising the Integrity of Machined Surfaces in a Powder Nickel Alloy used in Aircraft Engines Academic research paper on "Materials engineering"

Share paper
Academic journal
Procedia CIRP
OECD Field of science
{"Surface integrity" / RR1000 / "Nickel alloy" / "Scanning electron microscope" / "X-ray diffraction" / "Low cycle fatigue"}

Abstract of research paper on Materials engineering, author of scientific article — M.C. Hardy, C.R.J. Herbert, J. Kwong, W. Li, D.A. Axinte, et al.

Abstract This paper describes methods for evaluating and characterising the integrity of machined surfaces in a powder nickel alloy that is being used for disc applications in aircraft engines. It initially reviews techniques for inspecting the effects of process parameters on surface integrity for hole making and finish turning and then presents the findings of work that has been conducted to understand the influence of machining anomalies on fatigue life. The techniques considered for characterising surface integrity include surface inspection, surface roughness measurement, metallographic assessment of etched surfaces using light microscopy and micro-hardness measurement. More novel techniques are then discussed, exploiting advanced electron microscopy, nano-indentation and x-ray diffraction methods. These are capable of understanding the effects of machining processes on microstructure and quantifying the depth to which machining processes can change the material microstructure.

Academic research paper on topic "Characterising the Integrity of Machined Surfaces in a Powder Nickel Alloy used in Aircraft Engines"

Available online at


Procedia CIRP 13 (2014) 411 - 416

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

Characterising the Integrity of Machined Surfaces in a Powder Nickel

Alloy used in Aircraft Engines

M.C. Hardya*, C.R.J. Herberta, J. Kwonga, W. Lia, D.A. Axinteb A.R.C. Sharmanc,

A. Encinas-Oropesad and P.J. Witherse

aRolls-Royce plc, PO Box 31, Moor Lane, Derby, DE24 8BJ, UK. bThe University of Nottingham, School of Mechanical, Materials and Manufacturing Engineering, University Park, Nottingham, NG7 2RD, UK. cAdvancedManufacturing Research Centre with Boeing, Advanced Manufacturing Park,Wallis Way, Catcliffe, Rotherham, S60 5TZ, UK. dCranfield University, School of Industrial and Manufacturing Science, Cranfield, MK43 0AL, UK. eThe University of Manchester, School of Materials, Grosvenor Street, Manchester, M13 9PL, UK. * Corresponding author. Tel.: +44-1332-240412; E-mail address:


This paper describes methods for evaluating and characterising the integrity of machined surfaces in a powder nickel alloy that is being used for disc applications in aircraft engines. It initially reviews techniques for inspecting the effects of process parameters on surface integrity for hole making and finish turning and then presents the findings of work that has been conducted to understand the influence of machining anomalies on fatigue life. The techniques considered for characterising surface integrity include surface inspection, surface roughness measurement, metallographic assessment of etched surfaces using light microscopy and micro-hardness measurement. More novel techniques are then discussed, exploiting advanced electron microscopy, nano-indentation and x-ray diffraction methods. These are capable of understanding the effects of machining processes on microstructure and quantifying the depth to which machining processes can change the material microstructure.

© 2014 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/).

Selectionandpeer-reviewunderresponsibility ofThe InternationalScientific Committeeofthe"2ndConference onSurfacelntegrity" in theperson ofthe

Keywords: Surface integrity, RR1000, Nickel alloy, Scanning electron microscope, X-ray diffraction, Low cycle fatigue

1. Introduction

The primary purpose of discs in gas turbine engines is to hold the blades in position and rotate at high speeds by means of a shaft that passes through the centre of the disc. There can also be a number of secondary functions that require complex features. To realise the design intent, manufactured components must be produced economically to specified dimensions, tolerances (which can be particularly demanding for blade attachments and mating surfaces) and a cost target, and to ensure that the required service life is achieved. A summary of the challenges and solutions for manufacturing such high integrity components can be found in [1].

Nickel alloy disc rotors in modern aircraft engines experience high stresses from centripetal and thermal loading and are required to operate safely over a significant number of flights. Given the kinematic energy of these large components, they cannot be contained within the engine nacelle if failure occurs during service operation. Disc rotors are therefore termed safety critical as release of a failed section could threaten the safety of the aircraft and passengers. The majority of features in these components are designed to operate at stresses below the macroscopic yield stress of the material, with small levels of plasticity being confined to local areas around stress concentrations. As such, low cycle fatigue is the main source of damage that could lead to component failure through the nucleation and growth of fatigue cracks.

2212-8271 © 2014 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


The results of laboratory fatigue tests on nickel disc alloys have shown that significant changes in material microstructure from test piece manufacture, which are referred to as machining anomalies, can reduce fatigue life from high cyclic stresses/strains due to the premature nucleation of fatigue cracks [2], [3]. Regrettably, such events have occurred in service and have led to uncontained disc failures. Consequently, to ensure that the expected service life is achieved consistently, the effect of manufacturing process parameters and tools on surface microstructure and integrity must be understood, as must the effect of surface microstructure and integrity on fatigue life. Such understanding is necessary to specify a minimum quality standard for machined surfaces in disc rotors.

This paper examines methods that have been used for evaluating and characterising the integrity of machined surfaces in a powder nickel disc alloy.

2. Material

The machined surfaces examined in this paper are for nickel disc alloy RR1000 (Ni-18.5Co-15Cr-5Mo-3Al-3.6Ti-2Ta-0.5 Hf-0.027C-0.015B-0.055Zr in weight %). Billet material for isothermally forging this alloy is produced by powder metallurgy, i.e. hot isostatic pressing of argon gas atomised powder, followed by extrusion. Forgings are solution heat treated either below or above the gamma prime (y') solvus temperature of the alloy to develop a fine (average of 4-8 |im) or coarse (average of 23-64 |im) grain microstructure. The fine grain microstructure optimises yield strength and resistance to fatigue crack nucleation whilst the coarse grain structure offers optimised resistance to dwell crack growth and creep.

Irregular shaped primary y' particles (1-5 |im in size) are present in the fine grain microstructure at grain boundaries but not in the coarse grain variant. Compressed or fan air cooling is used to quench forgings after solution heat treatment. This produces near spherical secondary y' particles, 100-300 nm in size, in fine grain RR1000. Larger, octo-dendritic shaped secondary y' are produced in coarse grain RR1000. Forgings are then aged at 760°C for 16 hours to precipitate tertiary y', having an average size of 15-20 nm.

3. Examination of machined surfaces

The first integrity assessment of a machined surface is typically a low magnification visual inspection by binocular microscope. With the component in the as-machined condition, such inspection will detect

significant deviations in the expected surface finish such as gross material redeposition, smearing or tool vibration issues such as chatter or fish-scale. After etching, visual inspection can detect anomalies in material microstructure and handling features such as minor scratches and impact marks. Evidence of gross surface flaws, which are greater than 0.75 mm in surface length can be detected by fluorescent penetrant inspection. Other techniques for non-destructive evaluation may be able to detect smaller size of flaws but such methods are beyond the scope of this paper.

Fig. 1. Backscattered electron image of a white etch layer in coarse grain RR1000. Sample was produced from dry drilling and was electrolytically etched with 10% phosphoric acid in water.

The surface roughness of machined disc rotors is specified as it is understood that the resistance to fatigue crack nucleation improves with reduced surface roughness. Reference [4] shows an example of this effect for Type 304 stainless steel although there is significant reported evidence in the literature. This occurs because the "valleys" of the machined profile behave as micro-notches that concentrate stress and become locations of crack nucleation. Surface roughness is specified in terms of either an arithmetic average of absolute values (Ra) or as a maximum valley depth (Rv) or a maximum peak height (Rp) of a machined profile. Such specifications are necessary as (1) the component may have features that are not shot peened and (2) subsequent etching for inspection and shot peening should not significantly increase the surface roughness. Measurement can be undertaken by contact or contact free methods on machined surfaces or replicas. Whilst surface roughness can be used as a measure of process compliance and control, a low Ra value does not necessarily equate to an acceptable surface since small depths of cut, for example, can smear rather than cut material, leading to an excellent surface finish but potentially a highly deformed microstructure. The response of the machined surface to etching can be an indicator of the quality of the machining process.

Experience has shown that an area that exhibits a poor etch response may be the result of local strain hardening.

Serrated swarf can be produced under certain turning conditions, which can accumulate, if not broken into chips, and become trapped by the tool, leading to re-deposited swarf or "pick-up" [5]. Such re-deposited material can be detrimental to fatigue behaviour if it is undetected and shot peened. Detailed surface inspection (using light microscopy or white light interferometry) or examination of polished samples can detect these features at the development stage of process optimisation so that amendments to tools, process parameters or coolant application can be made. Examination of swarf morphology can also provide an indicator that changes to process conditions are necessary.

4. Examination of polished surfaces

The influence of the machining process on surface integrity is often examined by preparing polished samples. These can be taken at various stages of tool life from both the cutting direction and perpendicular to the cutting direction. Typically samples are then etched and viewed initially using light microscopy. The etch is chosen to reveal changes in the material microstructure as a result of the hot deformation processes involved in the formation of chips. The list below shows changes in microstructure that can occur:-

• White etch layer (Fig. 1), with associated cracking,

• Surface tearing, "plucking", or loss of material,

• Deformation of primary and secondary y' or "swept

grain" (Fig. 2),

• Strain traces, evidence of strain hardening (Fig. 3),

• Evidence of foreign material, from tools for example,

The presence of a white etch layer, cracking and "plucking" can be detected in surfaces that have been etched with solutions that remove either the y phase or the y' particles. However, deformed primary y' is most readily detected if the y phase is removed using, for example, the electrolytic etch of 10% phosphoric acid in water, particularly when examined in a scanning electron microscope. Conversely, slip traces, which are indicative of strain hardening are most visible if y' particles are removed, often to emphasise grain boundaries (Fig. 3). Kalling's reagent (5 g CuCl2, 100 ml HCl, 100 ml methanol) is a good example of a grain boundary etch.

The structure of the white etch layer is discussed in more detail below in the section concerning electron and ion beam microscopy. For now, it is sufficient to say

that a significant change in microstructure is likely to result from a combination of rapid heating, hot plastic deformation and rapid quenching.

Fig. 2 Light microscopy image of fine grain RR1000 showing minor deformation of primary y1 (large white particles) from turning. Sample electrolytically etched with 10% phosphoric acid in water.

Fig.3 Light microscopy image of coarse grain RR1000 showing strain traces from turning. Sample was etched with Kalling's reagent.

As mentioned earlier, the strategy for achieving the expected fatigue behaviour in service is based on producing acceptable machined surfaces, which for RR1000 typically requires a surface roughness less than 0.8 |im, no cracks, no white layer, no re-deposited swarf, no extrinsic (tool) material and allows a depth of strain hardening, deformed y1 or surface tearing (plucking) that is no greater than 10 |im.

5. Use of scanning electron or ion beam microscopy

Light microscopy of etched samples has been used exclusively in the past as equipment with the required resolution is widely available. Assessment of machined surfaces in fine grain microstructures is possible though assisted by the presence of primary y' particles in RR1000, or delta (8) phase in Inconel 718. However, detection of deformed secondary y' particles is less convincing for coarse grain RR1000 using light microscopy. In this case, and for all alloys during the development of optimised process conditions, scanning electron microscopy provides greater clarity of images at higher magnifications.

Protective Pt layer

Machining produced microstructure

Fig. 4 A secondary electron image of fine grain RR1000 produced from a focused ion beam microscope (FIB). The image shows changes in surface microstructure as a result of grinding to Ra of 0.25 ^m.

Other advanced microscopy techniques can be used to gain further insight and understanding of the effect of machining on microstructure. For example, dual beam focused ion beam (FIB) microscopes are now widely available and are used in many fields of materials science [5]. At low ion beam currents, FIB microscopes can provide similar resolution, in terms of topographic imaging, to scanning electron microscopes but offer advantages in as much as they have two imaging modes that show grain orientation contrast (secondary electron) and chemical difference contrast (secondary or total positive ion). The former can be used for visually assessing the depth of plastic deformation as machining can produce changes in grain size and morphology (Fig.

The FIB facility can also be used to mill out small micro-samples from specific locations for transmission electron microscopy (TEM). In this technique, a beam of electrons is passed through an ultra-thin sample, which interacts with the sample, forming an image, which is magnified and focused onto an imaging device. The advantage of TEM is that it offers high resolution at very high magnifications, as well as information regarding the crystallography and composition of the sample. In a recent study on coarse grain RR1000 [6], a 0.3 |im thick section was milled, by FIB, from a white layer of 3 |im in depth, which was produced by abusive drilling conditions without coolant. A TEM foil was then prepared, revealing a very fine grain size of circa 50 nm. As a result of Bragg scattering of electrons from the ultra-thin sample, TEM also provides diffraction, such that electrons are dispersed to discrete locations, which form an image of a diffraction pattern by adjusting the magnetic lenses of the microscope. For fine grained polycrystalline materials, the diffraction pattern consists of a series of rings. It was found [6] that the diffraction pattern for the white layer showed similar rings to the bulk material. By identifying (indexing) these rings, it was confirmed that the white layer has a face-centred

cubic structure. It can therefore be inferred that the very fine grain size is a result of dynamic recrystallisation, which occurred from the rapid heating and high strain rate plastic deformation imposed by the drilling process.

The nano-structure of the white-layer in coarse grain RR1000 has since been examined using electron backscattered diffraction (EBSD), a technique that allows the extent of the surface plastic deformation, from machining, to be visualised [7]. Further details of this technique are given in the section concerning strain hardening.

6. Evidence of strain hardening

Evidence of strain hardening in machined surfaces can be found by characterising the level of plastic deformation from the surface of a polished sample sample, either by hardness testing or by EBSD or from a cut, as-received sample using X-ray diffraction.

Hardness is typically measured by indentation methods and can be applied on macro-scopic, microscopic or nano-scopic length scales, depending on the forces applied and the size of the indentors. Point to point scatter in hardness data often prevents the detection of plastic deformation from machining. Sources of variation and experimental error can be minimised by following recognised procedures on polished samples, performing multiple measurements and by using the largest possible load in micro-hardness testing. Micro-hardness profiles do provide a method of estimating the depth of strain hardening from machining. Nano-indention has been used to understand the difference in hardness of white etch layer [6] in coarse grain RR1000 compared to the bulk material. It was found that the very fine grains in the white etch layer increased the hardness by 45% over bulk material.

There are many portable products available that use the Leeb rebound test method for indicating the hardness of surfaces. These devices calculate a hardness value by comparing the energy of a test body before and after it impacts on a sample, based on the impact and rebound velocities of the impact body [8]. In simple terms, the impact body rebounds faster from harder surfaces than from softer ones, resulting in a greater energy. However, as quoted in ASTM A956-12 [9], the Leeb hardness test indicates "the condition of the surface contacted", and "the results generated at that location do not represent the part at any other surface location and yield no information about the material at subsurface locations". Clearly, a better understanding can be achieved by sampling many areas of the machined surface. The technique is useful for indicating

differences in surface behaviour over batches of machined surfaces.

EBSD is another widely available technique that can be undertaken on flat polished samples in a scanning electron microscope (SEM) that is fitted with the appropriate detector. It is able to identify crystallographic orientation in a polycrystalline material through the indexing of Kikuchi bands in an electron backscatter diffraction pattern. EBSD can also detect local differences in orientation within deformed grains, which occur as a result of plastic deformation. As such, data from EBSD of machined [10] and shot peened surfaces [11], [12] have been used to characterise the extent of plastic deformation via calculated misorientation values. Whilst different values were considered in these pieces of work, they show that the magnitude of the misorientation decreases with distance away from the surface. Once the background level for the bulk material is established, the depth of the plastically deformed material can be estimated. It has been reported that the apparent depth of plastic deformation in shot peened surfaces, which were determined from EBSD were approximately half of the total internal alloy strain measured from hardness profiles or from X-ray diffraction [11], [12].

X-ray diffraction is an established technique for the determination of surface and near surface residual stresses. These stresses are calculated by measuring strain in the crystal lattice, assuming elastic distortion of the lattice structure. By rotating the sample of interest by known amounts relative to the X-ray beam, the change in angular position of the chosen diffraction peak can be measured, which occurs as a result of an increase or reduction in stress. Using Bragg's law (n^ = 2dSin0), the spacing of the crystal lattice (d) can be calculated from the angle (20) by which the incident beam of wavelength is scattered to form a diffraction peak, where n is an integer denoting the order of diffraction. For RR1000, the (311) lattice plane, and diffraction peak, has been considered for residual stress determination [13]. Broadening of the diffraction peak width can be a measure of plastic deformation, provided other material sources of peak broadening can be deemed insignificant and the practical difficulties in determining diffraction peak position and width can be overcome [14]. The latter is achieved by fitting an appropriate function to the raw data and measuring the width of the diffraction peak (in degrees) at an intensity value corresponding to half of the maximum height, hence the full width half maximum (FWHM) description. This information can be extracted when profiles of residual stress with distance from the machined surface are generated. To measure residual

stresses at distances below the surface requires that material be removed by electro-polishing. Figure 5 shows FWHM data that were determined from X-ray diffraction experiments on face finish turning in fine grain RR1000 [13].

-*-Coated 3982rpm -•-Coated 2282rpm

\ -»-Uncoated 2282rpm

Depth fim

Fig. 5 Full width half maximum (FWHM) depth profiles for finish face turning on fine grain RR1000 using coated and uncoated tools [13].

They show an increase in FWHM values at the surface (up to a depth of circa 40 |im), which it is assumed, results from the plastic deformation sustained during turning. The level of plastic deformation can be quantified if FWHM values are determined from material that has received measured amounts of strain during compression and tensile tests [13]. Such work has indicated that the plastic deformation during turning [13] and drilling [15] exceeds 20%.

Fig. 6. Endurance data from 600°C fatigue tests on single hole test pieces, made from coarse grain RR1000. These test pieces showed "damage-free" (base) and 2 heavily deformed surfaces, a white etch layer (1 = depth of 10 ^m, 2 = depth of 5 ^m) and severely distorted y1 structures to a depth of 15 ^m (mat drag 1) and 10 ^m (mat drag 2).

7. Effect of surface integrity on fatigue performance

The results of low-cycle fatigue tests at 600°C have recently been reported [16] for coarse grain RR1000 surfaces that were produced from hole making processes (drilling and plunge milling). Tests were conducted to

understand the effects of machining induced changes in material microstructure on fatigue life. As-machined surfaces included a reference "damage-free" surface, and two distorted microstructures, (i) a white etch layer, produced to a depth of 5 and 10 ^m, and (ii) a heavily distorted Y structure, produced to a depth of 10 and 15 ^m but with a greater degree of severity of surface distortion in the latter. The effect of shot peening on damage-free and 10 ^m deep white etch layer surfaces was also examined. It was found that the presence of a white etch layer significantly reduced fatigue performance, compared to that shown by the damage-free surface (Fig. 6). In contrast, surfaces showing distorted Y structures produced much less debit in fatigue life and only from a depth of 15 ^m. Shot peening the damage-free surfaces was found to increase the fatigue life by as much as 100% presumably by improving the resistance of the material to crack nucleation, through strain hardening, and by leaving highly compressive surfaces residual stresses.

8. Closing remarks

Based on experience with powder nickel disc alloy RR1000, this paper has provided a summary of the techniques that are available for characterising surface integrity. It is proposed that these techniques are employed during the development of manufacturing processes to determine the quality standard for machined surfaces so that expected fatigue lives are achieved in service. Without this understanding, machined surfaces could be produced that show significant changes in material microstructure, which may result in premature nucleation of surface fatigue cracks and uncontained component failure.

The demand to reduce fuel consumption and improve engine efficiency is promoting higher engine temperatures, and high climb ratings to move large aircraft more quickly to altitude. Firstly, these factors give rise to fatigue cycles with long dwell periods at elevated temperatures. Under such conditions, oxidation and creep damage accelerate the rates of crack growth, which can significantly reduce the number of cycles to component failure if stresses are sufficiently high to nucleate cracks from surface anomalies. Secondly, they require alloys that are capable of operating at higher temperatures. Such alloys will contain higher volume fractions of Y, which are likely to produce faster tool wear unless work is undertaken to develop improved tools and processes.


The authors would like to thank Rolls-Royce plc for permission to publish this work. We also acknowledge financial support from the EPSRC and TSB in the UK as well as advice from Rolls-Royce colleagues Jon Penny, Yue (YG) Li, Gregor Kappmeyer.


[1] Kappmeyer, G., Hubig, C., Hardy, M., Witty, M. and Busch, M., 2012, Modern machining of advanced aerospace alloys - enabler for quality and performance, Procedia CIRP, 1, pp. 28.

[2] Connolley, T., M.J. Starink, M.J. and Reed, P.A.S., 2004, Effect of broaching on high-temperature fatigue behaviour in notched specimens of Inconel 718,Met. Trans., 35A, 3, pp. 771.

[3] Wusatowska-Sarnek, A.M., Dubiel, B., Czyrska-Filemonowicz, A., Bhowal, P.R, Ben Salah, N., Klemberg-Sapieha, J.E., 2011, Microstructural characterization of the white etching layer in nickelbased superalloy, Met. Trans., 42A, 12, pp. 3813.

[3] Maiya, P.S., and Busch, D.E., 1975, Effect of surface roughness on low-cycle fatigue behaviour of type 304 stainless steel, Met. Trans., 6A, 9, pp. 1761.

[4] Axinte, D.A., Andrews, P., Li, W., Gindy, N., and Withers, P.J., 2006, Turning of advanced Ni based alloys obtained via powder metallurgy route, Annals of the CIRP, 55, 1, pp. 117.

[5]. Phaneuf, M.W., 1999, Applications of focused ion beam microscopy to materials science specimens, Micron, 30, pp. 277.

[6] Herbert, C.R.J, Axinte, D.A., Hardy, M.C., and Brown, P.D., 2011, Investigation into the characteristics of white layers produced in a nickel-based superalloy from drilling operations, Procedia Engineering, 19, pp. 138.

[7] M'Saoubi, R., Axinte, D.A., Herbert, C.R.J., Hardy, M.C., and Salmon, P.D.G., 2014, Surface integrity of nickel-based alloys subjected to severe plastic deformation by abusive drilling, CIRP Annals Manufacturing Technology, in press.


[9] ASTM A956 - 12, Standard Test Method for Leeb Hardness Testing of Steel Products.

[10] M'Saoubi, R., Larsson, T., Outeiro, J., Guo, Y., Suslov, S., Saldana, C., Chandrasekar, S., 2012, Surface integrity analysis of machined Inconel 718 over multiple length scales, CIRP Annals — Manufacturing Technology, 61, 1, pp. 99.

[11] Child, D.J., West, G.D., and Thomson, R.C., 2011, Assessment of surface hardening effects from shot peening on a Ni-based alloy using electron backscatter diffraction techniques, Acta Materialia, 59, pp. 4825.

[12] Foss, B.J., Gray, S., Hardy, M.C., Stekovic, S., McPhail, D.S., and Shollock, B.A., 2013, Analysis of shot-peening and residual stress relaxation in the nickel-based superalloy RR1000, Acta Materialia, 61, pp. 2548.

[13] Li, W., Withers, P.J., Axinte, D., Preuss, M., and Andrews, P., 2009, Residual stresses in face finish turning of high strength nickel-based superalloy, Journal of Materials Processing Technology, 209, pp. 4896.

[14] Prevey, P.S., (1987), The measurement of subsurface residual stress and cold work distributions in nickel base alloys, Residuals Stress in Design, Process and Material Selection, (Ed. by W.B. Young), ASM, Metals Park, Ohio, USA, pp. 11.

[15]Kwong, J., Axinte, D.A., and Withers, P.J., 2009, The sensitivity of Ni-based superalloy to hole making operations: influence of process parameters on subsurface damage and residual stress, Journal of Materials Processing Technology, 209, pp. 3968.

[16]Herbert, C.R.J., Axinte, D.A., Hardy, M.C., and Withers, P.J., 2014, Influence of surface anomalies following hole making operations on the fatigue performance for a Nickel-based superalloy, Journal of Materials Processing Technology, in press.