Just hot air?

There can be no doubt that there have been considerable problems with turbochargers, particularly compressor wheels. Christian Lützen from AP M?ller and Curt Christensen from the FORCE Institute highlighted a number of these problems at this year?s CIMAC Congress in their paper Turbocharger Failures, a Law of Nature or Neglect. As they explained, “turbochargers rank high on the scoreboard of non-planned outages of marine diesel engines”. Most of the main turbocharger manufactures are aware of the problems and have adopted different methods for solving them. The alloy used for compressor wheels is AA 2618 T6, which is a high strength, precipitation-hardening alloy. The alternative to using this alloy is to go for a titanium compressor wheel, but this involves considerable extra cost. However, the UK Government?s Department of Trade and Industry (DTI), in conjunction with the Engineering and Physical Sciences Research Council (EPSRC), is carrying out research on the microstructure/process modelling of forged conventional titanium alloys. The research is intended to produce a burn resistant titanium alloy that would continue to allow for lower cost manufacture, for use in aero-engines. Further work in Europe [projects such as BRITE-EURAM, COST 501, 513, 522] has concentrated on the development of a series of cast and wrought g -TiAl alloys and into an orthorhombic variant [BRITE-EURAM EPROTAC]. This range of compositions offers the possibility of using g -TiAl components from the compressor, through to the combustor, turbine and exhaust, so there is the possibility of reducing costs in marine turbochargers. MAN B&W engines and turbochargers are designed for both marine use and stationary applications. It explains that while shore-based engines may run at constant engine speed, the turbocharger speed will vary with the load ? “in some cases the conditions for shore-based engines running at high load, high ambient temperature and high altitude are even more severe than for marine engines,” it says. Dr Alexander Rippl, senior technical manager of MAN B&W?s turbocharger department, says that “today there is no real alternative material to aluminium available. For use as a compressor wheel it has the advantage of low specific weight and therefore good acceleration behaviour.” The company does offer titanium compressor wheels in its new TCL range, but only for specialist applications where a pressure ratio above 4.7:1 is required (see page 8). Mitsubishi also avoids the titanium route. Keiichi Shiraishi, team manager for turbocharger design at Mitsubishi Heavy Industries says: “Corrosion of AA 2618 T61 alloy is not avoidable and, therefore, the impeller wheel should be designed to have enough margin of mechanical strength on the surface, even in corroded condition, after many years.” He explains: “We consider that corrosion of the impeller is not a critical issue on impeller life time as we have very few experiences of failure related to corrosion.” Volkmar Hauseisen, design team leader for the TPL..-B at ABB Turbo Systems says that ABB?s approach has been to carry out a study into the mechanical integrity of the compressor wheel over a lifetime of operation in severe conditions and to add in a safety margin to give a safe lifetime for a compressor wheel. It calls this safety concept SiKo and is about to publish a new brochure based on the concept. It?s all in the coating Corrosion can be a particular problem with the AA 2618 T6 alloy. The FORCE Institute has identified that fatigue cracks can be initiated by Sodium chloride (NaCl) attack. The answer is to protect the surface from attack. Mitsubishi resorts to anodising for this purpose. “Every MET type turbocharger impeller has an anodising treatment applied on its surface,” says Shiraishi. ABB also offers a protective coating as an option. MAN B&W uses something called Kepla-Coat, which Jakob Bucher, senior manager for turbocharger applications at MAN B&W, describes as “a patented process of our subsupplier”. He explains that it is a plasma-chemical process, “with the Kepla-Coat layer created by transforming the basic aluminium of the wheel into a corundum-like skin of equal thickness on the wheel surface”. He says that the hardness and thickness can be controlled by the parameters of the process. Sergey Usov, spokesman for Keronite Ltd ? a UK-based company that was born out of research into the development of a metal surface treatment with exceptional properties on light alloys, carried out by Russian scientists in the 1980s ? spoke to The Motor Ship about another possible coating, Keronite. This work led to the development of the Plasma Electrolytic Oxidation (PEO) process, which was further refined by Dr Alexander Shatrov and Dr Victor Samsonov who worked separately to develop two forms of PEO, which both gave light metals a super hard, wear resistant surface. At the beginning of the 1990s a private company was formed in Moscow to develop the technology further and to patent it internationally and in 1998 a company was set up in the UK. Both the technology and the coating were given the trade name Keronite and the developers relocated to the UK to scale up the process from a laboratory to a pilot plant facility. Dr Samsonov recently joined the company. While the company admits that it has only carried out coatings on auto compressor wheels so far, Usov tells The Motor Ship: “We have had a very positive experience in dealing with compressor wheels for turbochargers ? but until now we were working with small and medium size parts. Principally, there is no limit on the size of the component that can be treated with the Keronite process, because it is a scalable process ? the bigger the surface you need to coat, the bigger the bath and power supply you need to use.” Usov adds: “At the moment we have a medium size machine, but soon, if it is necessary, the bigger machine can be easily produced.” The technology allows light alloys to be coated with a hard ceramic layer, which the company claims, has excellent adhesion to the substrate. It cannot be achieved by any other form of technology, such as plasma-sprayed ceramics, anodising or electroplating, says Usov. The coating is produced by transforming the metal surface using plasma discharge in a liquid electrolyte, to form a layer of hard and dense ceramic material. Keronite says it can be applied to aluminium, magnesium and titanium and their alloys and inter-metallides. The company says that unlike anodising, the presence of copper and other alloying elements does not affect the quality of the coating. Usov says, “since our process creates a ceramic layer on the surface of aluminium, the coating can work in several different ways: l Being a ceramic, it reduces the temperature of the substrate material (for example the thermo barrier coating on top of a diesel piston allows an increase in the combustion chamber temperature while maintaining the same temperature of the piston crown material, or to prolong the life of the piston) l By the creation of a ceramic ?shell? around the part, it increases its overall strength.” Such coatings sound like the ideal solution to the problem of corrosion on AA 2618 T6 alloy turbocharger compressor wheels. The alternative The one alternative to using aluminium wheels is to use titanium, but the cost can be high. ABB is unwilling to say exactly how much more a titanium wheel costs only that “titanium wheels are considerably higher in cost, which is influenced by the material price and the production process”. Mitsubishi is a bit more forthcoming, although it has not estimated the cost of a turbocharger with titanium impeller wheel because it hasn?t previously applied a titanium impeller on a turbocharger. Shiraishi explains: “The problem is not only high material cost but also very long machining time that occupies a five-axis milling machine for many hours.” MAN B&W agrees. Rippl says, “For higher thermal and mechanical loading only titanium could be an alternative.” But he explains that “the extra costs are considerably high ? for prototype applications we have the experience that the costs for the unmachined part are six times higher than the aluminium blank” and “the machining costs had approximately the same factor”. He told The Motor Ship that “with advanced investigations into the milling process a cost reduction was possible, but the improved material parameters are not justified by the higher cost of titanium.” A question of creep The FORCE Institute has continued its investigations into compressor wheel failures and has found that there have been several failures in the past five years. Christensen tells The Motor Ship: “Failure investigations performed so far by the manufacturers indicate the failure mechanism to be overload due to overspeeding.” He says that creep damage is normally ruled out as the design temperature and the design stresses are claimed to be too low. However, it has recently been recognised by some manufacturers that the aluminium used for compressor wheels may suffer creep if the inlet air temperature is in the high end. AP M?ller and the FORCE Institute are conducting an investigation into compressor wheels retired from service for any reason other than complete wheel failure. Samples are cut out of the wheels from various positions near the shaft bore and near the periphery. “The bore samples have experienced higher stresses than the periphery samples, while the latter has experienced higher temperatures in service,” explains Christensen. In tests at the Institute, the test samples are pulled at constant uniaxially load at high temperatures and the time to failure is recorded. The loads used ranges from 50 to 80% of nominal yield and temperatures range from 140 to 180OC. The test results are plotted as load versus the Larsson-Miller Parameter and compared with tests results from virgin material. The Larson-Miller parameter is a function of time and temperature and it allows a comparison between tests at various temperatures thereby making it possible to compare test results obtained at different temperatures. The test results are shown in the accompanying diagram. The master curve is obtained with virgin material. The single test points are obtained with samples cut from a retired compressor wheel. Christensen says, “the creep results shows that some of the useful life of the compressor wheel has been consumed during service.” The confirmed service record of this wheel is as follows: Air inlet: outside air intake (about 25OC); Engine rating: 85% during service. “If the air intake had been from the machinery room and if the engine rating had been more severe the creep damage could have been even more significant,” says Christensen. “The results so far confirm that creep should be given proper consideration in the design and materials selection for turbocharger compressor wheels.” Rippl explains that “the creeping of aluminium depends mainly on the load and load cycles,” which “must be considered in the development and lay out stage of the compressor wheel”. He says: “During wheel production, the heat treatment and cold spinning of the wheel are additional steps taken to avoid creep in operation. “In some special applications with high compression ratios and high air inlet temperatures, the thermal loading of the compressor wheel outer diameter reaches the material limits. For this alternative, aluminium alloys with increased temperature resistance are under development.” Mitsubishi says an answer to the problem is impeller air cooling at the rear. “This idea has been employed on some of our MET type turbochargers operated on generating engines in tropical conditions,” says Shiraishi. “Alternative materials that have high creep strength is always our key issue.” Double-edged sword Another problem highlighted in the paper presented at CIMAC was that of the difference in strength between cast and forged turbine blades and the problem of scatter when it came to the strength of these parts. ABB says that forged blades are only being used for some of its previous turbocharger types and that “today?s generation of TPL are equipped with cast blades”. ABB says, “The strength calculations are based on our own values tested with similar components ? note that these are not based on theoretical values but on our specific material data. These are requirements our suppliers have to fulfil and the compliance is tested on the components ? any product release is based on a vast and defined qualification tests, such as low and high-cycle fatigue tests.” MAN B&W also uses cast blades “because of the cost and production advantage”. Rippl says, “Our investigations show clearly that the fatigue limit for the cast blades is sufficient. The material damping properties of cast blades are higher than those of forged material, resulting in lower vibration loads.” The casting process has been developed with the foundry to improve the material characteristics. Mitsubishi differs from the other two in that it uses turbine blades milled from chromium steel alloy bar. All the manufacturers told The Motor Ship that they are careful to control the materials from suppliers to ensure repeatable quality and consistent fatigue strength. n