TWI Bulletin, August 1986
2 - Heat treatment and surface modification, PVD, nonmetallics, wear resistant coatings, testing
Ian Bucklow and David Noble
Ian Bucklow, MA, PhD, is Head of Surface Technology, and David Noble, BSc(Eng), ARSM, is a Research Metallurgist, both in the Materials Department.
Tailoring of surfaces, by a variety of means, to meet particular engineering requirements was the theme of the first international conference on 'Surface engineering.' As an introduction to the subject, this two part article highlights the main areas of interest covered by the conference papers, which will shortly be published in full by the Institute.
Novel techniques
Build-up and repair of surface damage by diffusion brazing, reported by Lesgourgues of SNECMA (their 'RBD process'), is an example of a technique that has been successfully used for many years in the brazing business. Although the process is limited in application, it is nevertheless useful and often cheap; the SNECMA paper describes the particular advantages of the technique for turbine blade repair, and the metallurgical characteristics of the brazing alloys that were developed both for repair and for wear resistance. The cost of RBD coating compares favourably with the manual and automated TIG, and with laser and plasma spraying.
All sprayed coatings have a certain degree of porosity, ranging from, say, 25 vol% down to 2% or less (claimed). In the majority of applications porosity can be a major drawback and although much effort has been devoted to reducing it or rendering it harmless, the technique of closing internal voids, etc, by hot isostatic pressure (HIP) has hitherto not been commercially applied to sprayed coatings. There are two reasons: the first is the high cost of HIPing, and the second the presence of interconnected porosity in asprayed coating which necessitates encapsulation of the coating in a gas tight envelope (itself an expensive process).
The unit cost of HIPing has been steadily dropping with the advent of bigger furnaces available for jobbing work, but the problem of interconnected porosity remains. However, Anderson and Rickinson, via their paper on the Hi-Pac process, now claim that plasma sprayed coatings can be deposited with isolated, not interconnected, porosity and that they can therefore be HIPed without going through the expensive encapsulation process. The secret lies in thespraying technique (on which the paper is wholly silent) but, if the claims are substantiated, the combination of spraying followed by HIPing could result in coatings of considerably improved properties - at a price. It remains to be seen if the process can become economically viable.
Professor Singer, Swansea University, described his simultaneous spraying and shot peening (SSP) process for making dense sprayed coatings. Essentially the process consists of plasma spraying on to a component and at the same timepeening it with steel shot to consolidate the coating. The shot and the plasma spray are directed on to the same area and hence their paths cross, but because the shot is smooth, cold and spinning rapidly, the spray does not stick; the shot bounces off the component, and is collected and re-used. The process appears to offer many advantages over conventional plasma spraying and produces dense deposits. Also, the coating is in compression (another feature of shotpeening) which improves fatigue life. The disadvantages of the process are:
- the need to perform the operation in a chamber;
- the inability to coat the inside of bores because of access problems arising from the large, unwieldy, shot slinger.
Two papers were devoted to the subject of friction surfacing. Friction welding - in which the two components to be joined are heated by friction and are then, in effect, forge welded - has been practised for many years, but the application of the technique to surfacing is new to many in the engineering business (although the concept was first proposed in a British patent applied for in 1941). The coating material is in rod form, typically 20mm diameter and is rotated under load against the surface to be coated. Plastic flow and forge welding begin when the conditions are correct and then the substrate is translated beneath the rotating rod which is continuously fed forward at an appropriaterate. The result is that a coating of rod material, some mm in depth, is smeared across the surface and metallurgically bonded to it with virtually no dilution.
Thomas, The Welding Institute, gave an introductory paper which also tackled the problem of multiple deposits consisting of overlapping runs, and Bedford, of Portsmouth Polytechnic, expanded on some of the properties of the coatingsand the advantages of the technique. The two authors between them cited the following examples of coatings: stainless steel (s/s) to a mild steel (m/s) substrate, Stellite to m/s, Hastelloy C to s/s, Monel to s/s, Inconel 625 to m/sand s/s, Al alloy to Al alloy, and Al-bronze to m/s and s/s.
Heat treatment and surface modification
Hick, Wolfson Heat Treatment Centre, presented a broadly based paper that described advances made in the many areas covered by the session title. Selective surface hardening employing such intense high-energy sources as laser and electron beams was outlined; gas nitriding and nitrocarburising below 600°C were treated in rather more detail. A description of high temperature ( i.e. above 800°C) carburising and carbonitriding in various atmospheres - including fluidised beds - also covered low pressure techniques with and without glow discharge ('plasma') energising, and a short section on thin film coatings by chemical vapour and physical vapour deposition techniques concluded with a description of the Toyota diffusion process for production of carbide coatings. Although the paper is brief in its treatment of each topic, it nevertheless forms a comprehensive survey of the subject and has an excellent bibliography.
Friction surfacing into a recessed groove
Improvement of the wear and oxidation resistance of a surface by changing its chemical composition ('surface modification') is at least as old as the carburising of steels, but there is still plenty of scope for improvement in the processes used and in their fields of application. Marijnissen described the development of a theoretical model - and its translation into practice - of a single-step pack process for forming an MCrAlHf diffusion coating on nickelturbine blade alloys. In addition MCrAlY coatings deposited by thermal spraying or PVD are widely used, but the processes are expensive and a single-step pack process promises to be much cheaper, provided that the complex chemistry of the controlled co-deposition of three elements can be sorted out. The paper demonstrates that this is possible, but the control required is strict, and the economics are not yet known.
Carburising, nitriding and carbonitriding of steels have long been in use, and the different ways of applying the treatment (gas, pack, salt-bath, etc) are all chemical in operation. The reactions are thus driven solely by chemical potentials over which the operator has very little control once the reactive formulation has been chosen. An entirely different approach liesin the use of glow discharge techniques whereby the reactive species are ionised in a low pressure plasma and accelerated to bombard the surfaces to be treated. The driving energies are therefore controlled by the plasma conditions which can easily be changed to suit the situation and which allow reactions to be performed which would otherwise be virtually impossible (for example, a bombarding energy of 100eV - easily achieved - is equivalent to a temperature in the region of 4 x 10 4 K). Plasma nitriding, especially, is widely used in the major engineering countries (with the exception of the UK) and it is a measure of the importance of plasma treatments that three papers were devoted to the subject, one (encouragingly) from the UK.
A differential pinion gear undergoing plasma carburising under laboratory conditions. Carbon, derived from hydrocarbon additions, is only generated in the thin uniform glow seam which surrounds the workpiece surface, thereby promoting uniform carburising
Jacobs and Law summarised the principles of plasma nitriding and illustrated them with reference to equipment in industrial use. They concentrated on the metallurgical aspects of the technique and demonstrated that control of thephysical conditions within the treatment vessel led to excellent control of the metallurgical and chemical reactions taking place during treatment. Plasma nitrided surfaces were compared with, and shown to be superior to, gas and saltbath nitrided components.
Rie et al described investigations into methods of plasma carburising (in hydrogen/methane mixtures) and plasma carbonitriding (in hydrogen/methane/nitrogen mixtures) at austenitic temperatures. These two treatments are not aswell developed as plasma nitriding and so the paper describes some of the basic work done to establish the operating parameters using pure iron and a low carbon steel.
Boronising can give a harder surface to steels than carburising or nitriding, it is commercially realised only via a pack or paste technique which frequently forms two distinct layers that may subsequently crack. An alternative technique is to boronise via a glow discharge in gaseous BCl 3 (vide plasma nitriding) with the object of controlling the boron chemical potential so as to avoid formation of duplex layers. A paper by Bloyce et al describes progress to date and also reports on the possibility of using laser remelting of a conventionally boronised surface, with the aim of diffusing the boron more deeply into the substrate and thus avoiding sharp concentration gradients.
A second paper on boronising by Matsuda and Nakata reported a grapeshot approach to the process by heating various metals and alloys in boron powder. Metals in the Ti group (IV A), V group (V A), Cr group (VI A) and Fe group (VIII)were easy to treat, but those in the Cu (I B) and AI (III B) groups were not although their reactivity could be improved by alloying with the 'easy' metals.
Two papers covered application of lasers to transformation hardening. Steen discussed use of a mathematical model designed to analyse the hardening process. Among other uses, the model was applied to determining correlations between processing variables and the size and shape of the hardened zone, with the ultimate objective of predicting the zone hardness values.
The versatility of lasers was emphasised in a survey of laser surfacing applications by Oakley. The various types of surface treatment require different process parameters, and lasers are particularly suitable for such treatments. Thus, laser glazing requires short interaction times (milliseconds to microseconds) and high power densities (megawatts per square cm), whereas transformation hardening requires the opposite characteristics. Although industrial lasers are capable of operating over these ranges, there is as yet little information to relate the corrosion and wear resistance of treated surfaces to the laser operating parameters.
Physical vapour deposition
Bergmann and Vogel presented a paper on the use of ion plated coatings for wear resistance, and emphasised the necessity to consider the whole system (including the possibility of some re-design) when choosing to use a coating. They considered the chalcogenides ( e.g. molybenum disulphide), nitrites ( e.g. TiN), i-carbon (diamond structure) films, hard metal systems, and soft metals, each used to combat specific wear conditions. They also warned against the consequences of attempting to use coated parts that had not beende-burred before coating.
Titanium nitride is used extensively for coating machine cutting tools of all descriptions, dies, bearings, and even for wear resistant decorative coatings. There are several variations in the techniques used for deposition but one,'sputter ion plating' (SIP), which was originally developed at Harwell, is now in commercial use. A paper by Jacobs and Newbery describes the particular advantages of SIP, one of which - the ability to coat simultaneously components ofwidely differing size and shape - goes a long way to offset the inherent slowness of the process (1µm/hr for coatings of e.g. 5µm thickness). The authors present a useful discussion on the relative merits of the various factors affecting tool life and behaviour and how they are influenced by TiN coatings.
Titanium nitride is one of the major engineering successes of thin films produced by ionisation-assisted processes, and other compounds of Ti are being investigated for the same use. It is therefore rather surprising that theparallel zirconium compounds appear to have been relatively neglected. Duckworth, however, reports that reactively sputtered stoichiometric ZrN films, in addition to having useful wear resistance characteristics, have an excellent resistance to oxidation (possibly up to 600°C) that far surpasses that of TiN. These films, plus multiple graded composition layers of zirconium-nitrogen-oxygen alloy have been successfully used for some years.
The success of TiN has stimulated research into similar compounds in addition to the ZrN already mentioned. Knotek et al have investigated CrN, TiC, Ti(C,N) and TiAl x N coatings produced by reactive sputtering at rates from 7-30 µm/hr. The coatings were applied to high speed steel, cemented carbide and Stellite, and machining tests showed them to be effective in prolonging tool life and, in some cases, to be markedly superior to TiN. As with ZrN, the oxidation resistance of (Ti,Al)N and CrN was much better than that of TiN.
Non-metallic coatings and design considerations
Much of the impetus to develop flame sprayed plastics coatings came from a torch nozzle design which accelerated and melted plastics powder in an oxy-propane flame without thermal decomposition of the material. It was found that certain thermoplastics such as polythenes, nylons and vinyl acetates could be successfully sprayed. Subsequently, Nylon 11 and a saponified ethylene vinyl acetate became the principal materials applied for corrosion and erosion resistance, with some use of epoxies for heavy duty. Coatings from 200µm to 3mm in thickness are now applied at rates up to 10 kg/hr. Hearn, in a paper that included many case histories to illustrate the diversity of applications, summarised the background to sprayed plastics coatings and predicted that they would come to have an increasing role as protective and decorative finishes.
Korotchenko discussed the increasing importance of ceramics and cermets in surface engineering. Initially, bulk forms were seen as the appropriate solution to the need for surfaces having ceramic or cermet properties, but now thetrend is towards use of coating and treatment technology to achieve this aim. In particular, the plasma assisted surfacing methods are attracting attention. Using pressure as the differentiating process variable, the range of available techniques from ion implantation through plasma assisted vapour deposition to plasma spraying was listed. Various factors relating to the current status of these technologies were reviewed.
Mulder used the repair of marine diesel engine components to illustrate how the application of many different coating processes - each suitable for treating a particular component or material - can result in significant economies in the life of an engine. He emphasised that repair and recovery should be regarded as aspects to be taken into consideration at the design stage if the full economic benefit was to be obtained.
Wear resistant coatings
Papers were assembled under this heading on the basis of the performance of coating rather than the techniques used to produce them, and therefore only four were included under this title although possibly some 75% of the total had some reference to wear resistance.
The first paper, by Garrett et al, yet again made the point that the multiplicity of coating materials available to combat heavy abrasive wear is a source of much confusion. The mining and mineral processing industries usually base their selection of coatings on initial cost and historical prejudice, rather than on a sound scientific judgement. The authors therefore surveyed the relevant South African industries to identify the coatings principally used and the characteristics ofthe most widely used hardfacing materials in terms of their composition, microstructure, and abrasive wear performance. The work subsequently sought to lay the foundation for a more effective approach to hardfacing selection, and the paper summarised the complex conclusions that were eventually derived. In a paper describing erosion of the blades in a waste-gas turbine, Zu reported how waste gases, at some 600°C and containing fine dust particles in clustersup to 400µm across, were used to drive a turbine fitted with blades sprayed with a cobalt hardfacing alloy. A mathematical model predicted that the velocity, path and area of impingement of dust clusters on the blade was dependent upon the cluster size, and subsequent morphological studies of the eroded surfaces confirmed the model's predictions.
Electrodeposits filled with entrained carbides and oxides are used as wear resistant coatings, but Tulsi described an application of the entrainment technique to produce low friction coatings by incorporation of PTFE particles. Composite coatings of PTFE in electroless nickel contain some 25 vol% of PTFE as sub-micron particles in deposits of 5-50µm thickness. The coatings can be heat treated at up to 300°C to increase the hardness of the metal matrix, and the resultant coefficient of friction against chrome steel in a pin-on-ring tester is between 30 and 50% of that of a conventional electroless nickel coating. When run against itself, however, a PTFE/Ni composite shows even lower friction, and dramatic falls in the wear rate are seen. The material is used, e.g. in butterfly valves, as mould release coatings for rubber and plastics components, and on nuts, bolts, bearings, etc.
Mahdi and Eyre, working with simple aluminium alloys, reported results of wear experiments designed to separate effects of wear initiation and propagation. A single point diamond abrader was traversed across specimens anodised or coated with electroless nickel, and examination of the resultant wear tracks showed that damage was predominantly caused by plastic deformation and brittle fracture of the coating. The best of the hard anodised layers showed the least evidence of brittle fracture, and there was no direct correlation between hardness and wear resistance. Eyre strongly emphasised the danger of relying upon hardness as a criterion of wear resistance.
Testing of coatings
The 'bond strength' of a coating is always a contentious issue because no test yet devised can wholly distinguish between the various types of failure that operate when a coating is parted from the substrate; moreover the 'strength' values derived from tests can have no meaning in design. However, destructive tests can play a useful role in quality control, and Böhme and Bültmann briefly described some results from an extensive programme of tests appliedto weld surfaced steels. They produced artificial defects of known shape and position at the interface between a nickel based coating and a steel substrate, and investigated bend tests of various kinds, also direct tensile tests, andtests designed to separate an annulus coating from a rod. The latter test was held to be the most successful on grounds of simplicity, large tested area, and response to interface defects.
A suitable non-destructive testing system is urgently required to detect the presence of adhesive or cohesive defects beneath coatings or surface treatments, and to measure the thickness of the coating or the depth of the defect. Tog ain general acceptance, the method should be applicable to components of complex geometry and of many materials and should, ideally, be non-contacting and easily interpretable. No such method presently exists (or is likely to exist)but advances in techniques based across the spectrum of physical properties are being achieved in several centres, and there is hope that one or two comparative, if not absolute, methods of non-destructive evaluation for coatings, etc, will be successfully developed before too long. The remaining three papers on testing were devoted to non-destructive methods and included magnetic, optical and thermal techniques - all three being non-contacting.
Photothermal image of an adhesion defect at the interface between an alumina coating and a mild steel substrate. A modulation frequency of 16HZ was used and the area scanned had dimensions 3 x 5cm
Any change in magnetisation of a ferromagnetic substance is effected in a series of small jumps as a consequence of the irregular movement of domain walls (the Barkhausen effect), and such wall movement is, in turn, influenced by the microstructure of the material and residual stresses therein. In consequence, an alternating magnetic field generates a signal (magnetic Barkhausen noise) which is characteristic of the microstructure and stress within a ferromagnetic alloy. Theiner et al employed the magnetic Barkhausen noise effect to investigate the hardness and hardened depth of laser treated steels by applying magnetising frequencies from 2-100kHz. Because penetration depth varies inversely with frequency, they were able to explore a depth range from 1 mm to less than 0.1 mm; a spatial resolution of about 1 mm was achieved by using a tape recorder head as the detector. Magnetic results correlated well with metallographic observations. The same equipment was used to measure variations in the magnetic permeability of a steel surface after grinding, and again the magnetic measurements could be correlated with microstructural changes and the residual stress distribution beneath the surface. However, both Barkhausen and permeability techniques are applicable only to ferromagnetic materials or coatings.
An 'optically rough' surface is one on which the roughness is of the same order as the wavelength of light, thus a ground surface of, say, 5µm CLA is within the range. When such a surface is illuminated by laser light, each high spot acts as a source of reflected light to produce a random interference effect called 'speckle', and any minute change in the shape or position of the surface can be detected as a change in the speckle pattern. Grant and Smith applied the technique to examination of stress distribution in a T weld under load, and to detection of a known unbonded area beneath a PVC film bonded to metal. Their paper briefly reported the results of their work, and they suggested various areas and methods of application to coatings and to surface preparation.
Loss of heat from a surface by conduction to the bulk has been used to detect defects beneath a coating; defects result in poor thermal contact and thus give rise to surface hot spots. One can envisage several ways of detecting those hot spots, and one - a thermal interferometric technique - was described in a paper by Almond et al. A 5W laser, modulated by a rotating chopper, was used to pulse-heat an area of about 1mm diameter on a sprayed coating of alumina on mild steel. Thickness changes in the coating, and interface defects, caused changes in both the surface temperature and in the phase of the temperature variation with respect to that of the incoming beam. These changes were detected as an interference effect, and the specimen was scanned beneath the equipment to build upa thermal image of the surface. The paper gives no details of the scanning times involved nor of the real size of a detected defect. Furthermore, the authors warn that the technique shows an edge effect and that worn coatings, although sound, will give an image that differs from that of an as-sprayed coating. In spite of these problems, however, they point out that the equipment is rugged and capable of being adapted for industrial use.
In conclusion
The technical content of the whole meeting was, inevitably, rather unbalanced, with a bias towards thicker coatings. It is unfortunate that no papers on ion implantation were included, but was good to see some papers devoted to preparation and finishing.
One further area that deserves greater attention than it received is the whole question of 'design for surfacing' but, as anyone who has tried to write on this subject knows, it is a difficult one to tackle.
Another surface engineering conference is planned for next year, when the approach will be directed more towards achieving the required properties in a surface - and maybe by then the availability of design data will be less of a problem.
