Heh, and you thought Lotus turbo kits are expensive??? Try $8K-$16K just for a WRC spec'd turbo... read on...
Boost! Group N and WRC Turbos | Rally car | Racecar Engineering
Boost! Group N and WRC Turbos
Thursday, 9 October 2008 Peter Knivett 0 Comments
Print Comment Mail RSS Feed The demands of rallying at both Group N and WRC levels are producing increasingly specialised turbocharger solutions, as Brian Owen of Owen Developments explains
Article originally published in RCE V17N8
Ever since the demise of the FIA's Group B formula at the end of 1986, 2.0-litre turbocharged engines have dominated the global sport of rallying, spread across Group N, Group A and WRC categories. Two decades on, the frenetic nature of the competition and the demands at the very highest levels of the sport have pushed the boundaries of turbocharger development to achieve improvements in engine power, driveability, reliability and ease of service. Indeed, at the very pinnacle of the sport - the World Rally Championship - turbocharger designs are now highly specialised and far removed from road going production units. But despite the handicap of onerous restrictions to minimise performance modifications, practices have been developed that realise performance and durability gains on Group N turbos, as the lessons of rallying feed back into production turbocharger designs.
To discover more, Racecar talked to Brian Owen, managing director of Abingdon, Oxfordshire-based Owen Developments, one of the leading UK-based turbocharger specialists. The company has a heavy involvement in supplying and servicing Garrett, IHI and MHI units to Group N and WRC customers. As an approved Prodrive supplier, Owen Developments liaises with the Banbury concern's privateer WRC customers regarding IHI turbocharger units for pre-2002 World Rally Cars and pre-2003 Group N machines. In addition, Owen Developments deals directly with Garrett's motorsport facility in California, the epicentre of the huge company's motorsport involvement for more contemporary WRC fitments for Ford, Skoda and Citroën World Rally Car turbo applications.
Aside from these two companies, the third major player in the modern rallying turbocharger market is Mitsubishi Heavy Industries (MHI). Unsurprisingly, MHI produces the turbocharger units for Mitsubishi's four-wheel drive turbocharged Lancer Turbo Evolution, the only car currently capable of challenging the Subaru Impreza in contemporary Group N rallying. Given the popularity of the Mitsubishi as a Group N contender, Lancer-destined MHI units make up a considerable proportion of Owen Development's workload, so this is the first focus of our attention.
THE GROUP N CHALLENGE
‘On a normal road car, conditions don't allow you to hold full boost for long enough to cause a problem,' Owen explains, ‘but with rallying it's different. It's very arduous use, constantly on and off the power.' Add in the detrimental effects of modern, aggressive anti-lag systems and it's not hard to see why the production-based internals of a Group N turbocharger have such a hard existence. The key challenge in this area is the design of the turbine shaft bearings, which have to withstand rotational forces of over 160,000rpm as the compressor and turbine wheel go about their work.
The conventional design - favoured by both MHI and IHI - is to use lead bronze ‘hydro-dynamic' journal bearings at either end of the steel turbine shaft, which rely on a film of engine oil just a few microns thick. These work in concert with a centre ‘thrust' bearing produced from the same material, which again is reliant on engine lubrication.
As a low-cost, relatively low-tech solution, it isn't without shortcomings. In particular this design can allow lateral movement on the turbine shaft, caused by the force of the waste gasses acting upon the exhaust turbine wheel's angled blades. Much like a helical gear, this pressure attempts to force the turbine shaft at an angle to its centreline, as Owen explains: ‘This means when the engine is on and off boost the exhaust gas is trying to force the shaft in and out. You don't want any actual movement through the rotating assembly of a turbocharger because the oil seals, crude as they are, will start to pump oil.'
In practice, hydrodynamic bearing-equipped Group N turbochargers can be modified to mitigate against the worst shortcomings of the basic design, as Owen highlights with the MHI Mitsubishi Evo unit. ‘The centre thrust bearing on an Evo turbocharger has two thrust faces either side of it and they're very small, so the movement under boost load wears them out.' Increasing the surface area of the outer thrust bearings addresses the issue, as does careful CNC machining of the main thrust's oil channel to spread lubricant onto the larger surface.
Ultimately though, there's no escaping the fact that a hydrodynamic bearing-equipped turbo is a compromise, hence why Mitsubishi's arch Group N rival, Subaru, has experimented with ball bearing IHI units in recent years, as it's a design that offers a number of advantages. Owen: ‘Ball bearings reduce the drag of conventional hydrodynamic (thrust) bearings... so the turbo is quicker to spool up onto boost. Plus they're not affected by the amount of boost the unit runs, because they hold the turbine shaft rigidly in place. Importantly, they don't require the same amount of lubrication as hydrodynamic bearings, so you're not heating the engine oil as much as on a conventional thrust bearing design.'
As an alternative to the adoption of costly, close tolerance, ball bearing cartridge turbochargers, MHI has attempted in recent years to improve the spool-up qualities of its Group N-destined Lancer Evolution units by shaving the weight of its compressor wheels to reduce rotational resistance. ‘They've used a magnesium compressor wheel which was half the weight of a pure, heat-treated aluminium equivalent,' Owen concurs, while the quest for superior turbo response has also seen the adoption of titanium alloy items on the turbine wheel side. The challenge of bonding a titanium wheel to a steel turbine shaft with sufficient endurance to tolerate the punishment of a modern Group N rally car's anti-lag systems has seen a number of failures so, where possible, Inconel is now the favoured option. Easier to bond, lightweight, durable and with excellent heat resistance, it's a material that offers significant advantages.
Significant attention has been paid to the design of production turbo casings in recent years, with both Subaru and Mitsubishi opting for what's become known as a ‘twin-scroll' turbine housing design, cast in either SG iron or, in the case of the MHI units, stainless steel. The concept is a simple one, as rather than relying on a single, large exhaust feed to the turbine housing, the twin-scroll design splits the ports in an attempt to increase incoming gas velocity and improve low-speed turbo response. On a road car this practice is well proven and becoming increasingly popular, but at extremes of use such as Group N, it's not without compromises. ‘A twin-scroll turbo is not good for high power outputs because, when you split the ports in the turbo, the wall between the two can distort,' says Owen. ‘At high speeds and high temperatures, that can start to effectively blank off one of the ports. It's a problem on Subaru and it's a problem on Mitsubishi - you get one particular port not functioning as it should.' Casing durability can also be an issue, particularly on Group N Evo turbine housings, which are prone to cracking.
Surprisingly, the heart of any turbocharger - the centre housing rotational assembly (or CHRA for short) is still produced from good old cast iron. ‘It's very good at dissipating heat and it doesn't distort,' Owen says, hence its widespread use as a CHRA material right up to the WRC, in combination with production-style water cooling systems to help with thermal management. Likewise, less exotic materials are selected for the compressor housing too, which has a much easier existence on the inlet side of the turbocharger, with most production-based Group N units using aluminium alloy designs.
And it's here that gains can be made, because careful selection of the appropriate castings during a Group N turbocharger build enables specialists like Owen Developments to effectively ‘blueprint' a unit to manufacturer tolerances. ‘When there are thousands of turbochargers going down a production line, no one is looking at things like internal phrases being left in turbine housings and nobody's looking at what the maximum diameter of the wastegate orifice is so it could dump more gasses,' Owen explains. Such attention to detail can potentially make the difference between a Group N turbo's boost characteristics, primarily in response and spool-up performance.
On the same theme, great attention is paid to balancing the reciprocating parts of a Group N turbo, much like a full dynamic balance of a motorsport engine. ‘We take great care balancing a core on a Group N turbo. The balance is all derived from the centreline, so you need to balance the parts on their own, on a machine that does vector balancing, then they get built into the CHRA.' The bearing, centre thrusts and washers are all subject to the same practice, and once the assembly is bolted together a process called ‘trim balancing' takes place. ‘We use compressed air to drive the turbo at high speed in order to measure out of balance,' Owen explains, ‘then the key thing is identifying that out of balance to an angular or radial position, then we trim this by taking a miniscule amount from the compressor nose nut.' Any out of balance can cause a slight eccentric wandering of the compressor wheel within its housing, potentially causing a less than perfect seal that could sap the turbo's ability to produce boost effectively.
WORLD RALLY CAR TURBOCHARGERS
A 34mm inlet restrictor is a comparatively small entry tract, being considerably narrower than a normal unrestricted compressor inlet port on a high power output 2.0-litre engine. In practice, the FIA restrictor has the unwanted effect of shrouding a significant part of the compressor wheel's surface from the incoming air which, given the boost pressures required in the WRC, would cause the unit to overspeed. The solution sees WRC turbo designers clawing back the blade area in a different way, as Owen reveals: ‘To ensure we're getting all of this air compressed by the wheel on a WRC car turbo, the designers at Garrett or IHI will produce a wheel that gives additional blade area in height, rather than width. Ultimately, that gives a WRC unit the same total blade area as a unit with a larger restrictor.' This enables the turbo to produce relatively high boost levels at lower rotational speeds, all in concert with a matching compressor housing design. As you'd expect, given the specialist nature of the use and larger budgets of the WRC, turbine wheels are usually produced from pure Inconel, with heat-treated alloy being favoured on the compressor side.
Arguably, the casings are the main difference between a WRC turbocharger and a production-based unit, with the focus on weight reduction and thermal stability. ‘WRC turbos use some pretty precious metals, because a turbine housing used on a production car will be three or four times the weight of a thin wall, cast Inconel WRC version,' states Owen. Even so, a water-cooled WRC turbo CHRA will still be produced from cast iron, whereas magnesium alloy remains the material of choice for WRC turbocharger compressor housings. In practice, these methods allow a Garrett TR30R Skoda WRC unit to tip the scales at a featherweight 4kg, whereas a Group N MHI Evo 9 GSR turbo represents a 10kg mass. That's a huge difference, particularly given the location of this weight on a rally car engine.
Mindful of the shortcomings of twin-port designs, all WRC turbochargers use conventional, single entry port, exhaust turbine housings, with Garrett favouring an external wastegate to improve gas flow characteristics. Conversely, even on contemporary WRC units, rival manufacturer IHI retains a potentially restrictive production-style integral wastegate, but both companies utilise low inertia cartridge ball bearing designs in an attempt to maximise spool-up characteristics and reduce lag. This has seen the introduction of ceramic, rather than steel balls, which are less affected by the tremendous 1100degC heat of the exhaust turbine housing, affording significant benefits in dimensional stability at all stages of the thermal cycle.
Meanwhile, with rapid serviceability in mind, ‘V-bands' allow for quick unfastening of WRC turbochargers casings to access the vital innards. All this extra attention paid to their design and manufacture explains why a WRC turbo can comfortably run 2.0bar of boost all day long. But it comes at a price - WRC units cost between £6000 and £12,000, compared to around £1200 for a fully prepared Subaru Impreza IHI or Lancer Evolution MHI Group N unit. But then this is hardly a surprise for, as with all things in motorsport, the best is never cheap.