TWI Bulletin, May 1986
1-Thermal spraying, weld surfacing and surface preparation
by 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.' The papers dealt with many aspects of this broad field, providing a useful overview for practising engineers and designers who want results without too much pre-occupation with how they are achieved. 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.
Thermal spraying
If number of papers is an indication of interest in a topic, then thermal spraying was second in popularity only to weld surfacing. Among the papers, three dealt with process and/or material study and behaviour, whilst the other five explored applications of thermal spraying which covered plasma, flame and arc spraying as well as detonation coating.
That the properties and service performance of sprayed coatings are strongly influenced by deposition conditions is acknowledged by the industry in both word and deed. The opening paper 'Component design and process considerations for plasma spraying', by Schwarz et al, emphasised not only the importance of process control in achieving a reliable performance from the coating, but also the necessity for consideration, at the design stage, of the implications of applying a coating. Although the authors referred only to plasma spraying, many of their points apply to other forms of spraying, and the philosophy is certainly applicable to all other forms of coating - thin films, thick layers, surface modification, etc. Increasing use of mechanisation, automation and robotics in spraying is visible evidence of the industry's concern with process control, but design considerations are less tangible. The paper serves as a useful reminder to the coating user that his contribution to the overall process is essential in achieving a satisfactory result.
The theme of increasingly complex mechanisation in flame spraying was continued by Britton, who described, among other things, a typical welding robot modified to suit the particular conditions encountered in flame spraying. Karpman returned to a problem that has bothered the users of thermally sprayed coatings for a long time, namely, the profusion of materials available, and the general lack of knowledge about their service performance.
Too many products are on offer and too little is known about their properties as coatings, and the number of materials sprayed in significant volume is small compared with the number in the suppliers' catalogues. This lack of knowledge on the part of the user (and of the supplier?) limits use of sprayed coatings, and Karpman argued that the development of techniques, and consequent increase in the number of materials that can be sprayed, has not been accompanied by an improvement in reliable information about their use. The importance of design aspects of thermal spraying was emphasised and the attitude of the user neatly summed up as follows: 'The engineering industry is not interested in the charms of a spraying technique, but in the advantages which can be expected from its use'.
The influence of spraying parameters on the (laboratory determined) properties of three arc-sprayed steels was investigated by Crane et al. Oxidation was demonstrated to be the most significant factor in relation to hardness and tensile properties. Thus, high hardness resulted from a high oxygen content, but the nature of a sprayed coating is such that deposits containing the least oxide have the highest tensile strength and elastic modulus. The authors found that the factor causing greatest variation during manual arc-spraying is spraying distance, which had a marked effect on oxygen content. The paper, although reporting an academic investigation, outlined the reasoning behind the steps that are necessary to optimise spraying conditions for practical applications; arc-spraying operators would benefit from its study. Heat transfer, both to the particles and to the substrate during arc-spraying of steels was studied by Harris et al who concluded that particle temperature was little changed during flight and that substrate heating was principally influenced by particle temperature and by mass-transfer-rate/area-density. Residual strains in steel coatings were generally tensile, but transformations to martensite could reduce or even reverse the strain.
High velocity combustion spraying (originally introduced as the Jetkote system) was the subject of two papers. Bomford and Donovan, in work directed eventually towards improving the oxidation and erosion resistance of gas turbine components, sprayed chrome carbide/80:20 Ni:Cr mixtures produced by three different methods on to test panels that were subsequently held at 1000°C in air for up to 1000hr. Coatings made by spraying simple physical mixtures were too heavily oxidised in the as-sprayed condition to give effective protection, but pre-alloyed and encapsulated powders gave good, dense, oxidation resistant coatings that were similar in many respects to D-Gun coatings.
Szelagowski described use of a high velocity combustion gun to spray corrosion resistant coatings under water. The author had used oxyacetylene spraying under hyperbaric conditions equivalent to 10m depth at which it was necessary to exceed the pressure safety limit for acetylene, but the high velocity gun is an oxy-propane torch and is not subject to the same pressure restrictions. An attachment to the torch allowing it to be used under water ( i.e. wet) was claimed to dry, preheat and, if necessary, post heat the substrate by the 'plasma beam' (which presumably means the flame) and the author reported that the coatings were sufficiently encouraging to warrant further tests. Existing hypersonic spraying torches are considered to have too low an output to be used economically for large area coating, but Szelagowski has a multi-nozzle gun under development.
Chandler et al described experience in the Central Electricity Generating Board with plasma sprayed coatings applied to boiler tubes. From trials conducted over several years, the authors concluded that stringent control of process parameters was essential to obtain the required degree of corrosion and wear resistance. Furthermore, their work showed that, although present practice and materials can provide promising coatings, any significant increase in coating quality can be achieved only by improving the deposition conditions; inert gas shrouding and some degree of a transferred arc were shown to be effective in this respect.
Weld surfacing
The session covered subjects that ranged from a study of high rate reclamation of Concast rolls (up to 400 kg/hr) to TIG surfacing of areas of difficult access at deposition rates down to 2kg/hr.
A paper by Killing and Thier presented data on the alloying efficiency of various flux cored wires. A high speed film showed dramatically the fierce and explosive arc/metal transfer characteristics of self-shielded flux-cored arcwires. The authors concluded that alloying occurs in both the droplet stage before transfer and in the weldpool.
Rodgers and Macleod described an electroslag cladding system which employed strips up to 100mm wide deposited in one pass using magnetic control of the weldpool. Inconel 625 and Monel 6o were clad on to various steels. The heat input, when depositing 60mm wide 0.5mm thick Inconel 625 strip, was reportedly over 10kJ/mm, and Charpy V notch tests in the substrate HAZ produced exceptionally high values of about 190J at -40°C. Where two layers of cladding had been deposited the HAZ toughness level rose to 185-241J. Some Monel deposits cracked, probably as a result of a type of solidification cracking with some contribution from a ductility minimum arising from segregation of Ti and S. With such high heat inputs, one might expect solidification cracking.
Blaskovic et al outlined the present state of electroslag surfacing of mill rolls in Czechoslovakia. Mill rolls are rebuilt with 0.2-0.3%C, 1.0-6.0%Cr, 0.1-1.6Mo type weld overlays, sometimes with the introduction of vanadium (up to1.6%) and tungsten (up to 8.0%). The rolls stand vertically whilst the electroslag cladding is built up from the base in a similar manner to continuous casting. In this way 400 kg/hr can reportedly be deposited using 25 x 45mm diameter electrodes.
A paper by Sluzalec on reclamation of a blast furnace bell was not presented, but the proceedings contain the manuscript which describes a successful submerged-arc hardfacing operation with a flux cored wire. The conditions to which the closure bells of a blast furnace are subjected are testing in the extreme and their replacement or servicing is expensive in downtime, but the reclamation and hardfacing procedure described has resulted in a considerable increase in life of the bells. There is an improvement in service life after stress relieving the deposit (paralleling the claim that 'relief checking' - or cracking - improves wear life by stress relieving the deposit and HAZ?).
Weld-cladding areas of difficult access was described by Robinson. He illustrated the use of TIG to deposit stainless steel and Inconel 625 with bores from 25-80mm diameter, but stiffness of the welding head arm governed the length of bore that could be coated (600mm long for the smallest diameter, up to 1100mm for the largest). Pulsed MIG conventionally allows bores down to 60mm diameter to be clad, but the author has developed a vertical-up technique to clad into bores of 25mm diameter which can subsequently be machined to give a hole of 6mm minimum diameter.
Svensson discussed the use of an Fe-34Cr-4.5C hardfacing alloy for high temperature applications. The main point from the paper was that below 1150°C no transformation occurs (assessed by evaluation of the equilibrium partition coefficient of Cr in austenite) and that the austenite formed is metastable up to 16-17% Cr. Liquidus projections and isothermal sections of the Fe-Cr-C ternary system were presented showing how ferrite formation would be expected below 700°C, but TEM work showed that had not occurred. The overall conclusion was that such hardfacing alloys contain adequate chromium dissolved in the austenite for use at temperatures up to 1000°C.
Fischer's paper 'On the development of Fe-Cr-C-B hardfacing alloys' described how hard, second phase, carbides of the M 7 C 3 type were replaced by harder borides of the M 3 B 2 type to improve abrasive wear resistance (assessed by a pin-on-plate tester). Two important trends appeared in the results but were not given in the paper:
- there was an increase in wear resistance with microhardness of the matrix between the second phase carbides;
- wear resistance increased with increasing volume percentage of hard phases in the microstructure. High boron (up to 3.7%) consumables performed as well as the more traditional, commercially available, material of 0.3%B; an important advantage of the 3.7%B alloys is that the chromium level can be reduced from 20% to about 9%, and the niobium from 12% to zero, thus making them less expensive.
The plasma-MIG technique, chiefly used for welding aluminium, combines the conventional consumable wire arc with a transferred arc surrounding the wire. The technique should be capable of combining the relatively low dilutionachievable by transferred arc methods with the higher deposition rates of wire fed techniques. Eichorn and Gaever described its application to deposition of stainless steel coatings on to mild steel, via cored wires. The study was experimental and intended to assess the technique for surfacing and to compare it with other weld surfacing methods. Smooth deposits of some 5mm thickness per pass with dilutions of about 10% were achieved with acceptable ferrite contents.
'Nickel alloy hardfacing of stainless steel' by McLeish illustrates the problems encountered by fabricators in their attempts to substitute nickel for cobalt in hardfacing alloys. The Ni alloy studied was Stelloy 40C which contains8Cr, 1.8B, 0.25C, 3Si, 29Fe (balance Ni). The trials were successful in producing sound welds, but the disadvantages of MIG welding with nickel alloys, such as the formation of small slag inclusions and excessive dilution in small components, became apparent. Furthermore, acceptable repair weld procedures were impossible to develop without shrinkage cavity formation under all conditions tried.
The development of solid state welding power supplies of high performance and flexibility has allowed great advances to be made in MIG welding. It is therefore logical to apply the same technology to arc weld surfacing. This reasoning was behind the paper by Nixon and Grainger which reviews the development of controlled transfer pulse welding and provides an excellent introduction to a subject that undoubtedly is going to play an increasingly important role in weld surfacing. The paper describes the characteristics of the technique and suggests possible areas of application for controlled-transfer pulse surfacing.
Surface preparation and finishing of coatings
Surface preparation is an essential part of any coating or surface treatment process, and preparation of submerged structures poses special problems. It is often impracticable to arrange a dry habitat and so the preparation of underwater structures such as ships' hulls, concrete constructions or drilling rigs must be undertaken in the wet - often at depth. Grit blasting is a versatile preparation technique that is widely used on land, and is now beingdeveloped for underwater use.
One such development for underwater grit blasting for inspection, maintenance and repair was described by Donker; equipment has been developed for working down to 50m and has been used in the North Sea on steel and concrete. The blasting nozzle delivers sufficient air to displace water from the area to be cleaned and so avoids the violent deceleration that the grit would encounter if it had to travel through water. Air consumption is consequently fairly high(11 m 3 /min for a nozzle pressure of 5bar) and the author stipulates that ultra-filtration or cryogenic drying is necessary. The type of grit is not important (sand or copper slag can safely be used) provided that the particle size is 1-2mm diameter; grit consumption is between 150 and 300 kg/m 2 . The equipment is commercially available and is presently operated manually (Szelagowski referred to the equipment in his paper on underwater spraying).
Shot peening propels a stream of metal, glass or ceramic spheres at velocities ranging from, say, 20 to 100 m/sec, on to the surface of a component. (The term 'shot peening' is to be preferred to 'shot blasting' because of possible confusion with 'grit blasting' which imparts different properties to a surface). The surface after impact develops a residual compressive stress extending from 0.1 to 5mm deep, and Al-Hassani comments that the major benefits are derived from this effect rather than from stress relief or work hardening. Shot peening is the most widely used surface treatment for improving resistance to fatigue and stress corrosion cracking, and is also used to free-form thin sheet into convex shapes suitable for aircraft wing skins (the asymmetric stresses set up by peening one side only cause the sheet to become convex to that side). The paper presents an analysis of the mechanics of the process and lists many applications in the automotive and aerospace industries.
Shot peening of a much gentler kind is one of the processes involved in deburring and finishing operations that Hignett classes under the general title of 'mechanical surface engineering'. The great majority of components that are cast, pressed, forged, machined, etc, whether it be in metal, glass, plastic or any other material require some kind of finishing operation that ranges from simple deburring to a high quality, tight tolerance surface finish. The cost of deburring or edge or surface finishing of metal components in the USA is currently averaging some 5% of the total manufacturing cost but less than 1% of capital, and less than 1% of engineering talent is devoted to the process. Decorative or functional finishing costs can be high (25% for a gas turbine compressor blade, 35% for a golf club head) but they can be mitigated to some extent by a suitable choice of design and manufacturing method. The aerospace industry in particular hasshown that such costs can quickly be recovered by the resulting improvement in operating efficiency.
Hignett describes the traditional and newer methods of finishing and points out that whilst Japan, Germany and the USA offer some form of training in the subject, there appears to be a gap in the UK in this direction. This whole subject is as much a part of surface engineering as are coatings, surface modification, etc, and deserves greater attention than it has hitherto received.
As a step towards finishing, the machining of coatings is certainly not neglected and most suppliers of materials offer detailed recommendations for grinding, turning, drilling, etc. Many engineering coatings are, however, specifically designed to resist being worn away by whatever means, and are consequently very difficult to machine. 'Self-fluxing' deposits of wear resistant alloys based onNi-Cr-Si-B are often impossible to turn with conventional tools and so must be ground, but the success of sintered cubic boron nitride (CBN) tools in cutting high hardness superalloys encouraged Inui et al to explore use of ceramic-bonded CBN tools for turning a Ni-Cr self-fluxing spray fuse coating. The paper describes the types and geometry of the tools and the cutting conditions, and concludes that the optimum combination will easily cut the coating to leave a clean surface which approaches the minimum roughness.
To be continued
