If you want to know something about turbos, you ask Geoff Kershaw. So we did. Questions from Jon Lawson
Few could claim to have had as much influence within their profession as Geoff Kershaw. After graduating from Queen Mary College, London University, followed by stints at Rolls Royce aero and Rolls Royce diesel, in 1974 he found himself working at Garrett AiResearch, where he managed the company’s Scandinavian customers.
Over the next four years, Kershaw’s primary role was the development of the turbo system for the Saab 99, which was launched in early 1978 and became the first successful mass production turbo car.
Despite the limitations of early electronics, it was fuel injected from day one.
Kershaw says, “They were already using the Bosch K-Jetronic system as it’s a natural marriage with a turbo. The change of density isn’t always well metered by a carburettor, and you have to pressurise the whole thing, so using carbs is a bit of a fudge. With injection, you get better presentation of fuel to the cylinders and it runs a lot cleaner, with better emissions. At that time, Saab were very keen on this, particularly for the American version, where they used the KE-Jetronic which had a more precise metering system.”
There were no dramas designing the turbo. But after the aerodynamic matching and bench testing was completed, a problem emerged. Kershaw continues, “Saab noticed a noise under certain circumstances – they referred to it as a wolf whistle. We only discovered this when we delivered the first half-dozen turbos for durability testing. So I went to Sweden with a view to investigating the source, expecting to stay around three days.”
It actually took seven weeks to find a fix. “To begin with, we didn’t know what it was. An aerodynamic noise? Balance? So I did the logical thing and selected three of the turbos: a quiet one, a noisy one and a very noisy one, and I visited the Garrett test facilities in Los Angeles. We set them up on a gas stand, which essentially adds a combustion chamber to it with a bit of an air feed so we can run it almost like a gas turbine. The whole thing is instrumented and within a few weeks we realised it was down to balance at the compressor end.”
Tightening up manufacturing tolerances helped, although for cost reasons this could only go so far. The solution was to high-speed balance the assembly which required the development of a new machine and a new process.
Understandably, Kershaw became attached to the project and after launch made an offer for one of the two pre-production prototypes. He says, “Due to a quirk of Swedish law, they would have had to remove the turbo before they could sell it on the domestic market, and as I’d spent so much time on the thing they accepted my cheeky offer and I drove it straight out of the country.”
He still owns it today. “It’s been in storage for four decades,” he notes with a chuckle, though restoration is close to complete.
Flying Solo
Although he was happy at Garrett, Kershaw had observed with interest the commercial success of the aftermarket. It made him want to launch his own company. “Telling them I wanted to go was one of the worst days of my life,” he recalls. “I told my boss, ‘I’m quitting, but by the way I want a special relationship, please.’ It took about three weeks to find the right moment.” Fortunately, the response was a positive one, and the two have enjoyed friendly – and commercially successful – relations ever since.
So, in 1981, Turbo Technics was formed. Its first major project was turbocharging the MG Metro. Unlike with the Saab, Kershaw retained the original SU carb with a suck-through arrangement, noting, “It’s not wonderful for emissions, but actually works quite well from the driver’s point of view. We took it to British Leyland and the company was enthusiastic, giving it the centrefold in the dealer magazine before TT put it in production. We even ended up working with the factory race car.”
Just like the 99 Turbo, the car has a loyal following to this day.
Back then, the focus of OEMs was to use turbos to improve performance. Kershaw notes that over the last decade or so the focus has changed to downsizing, for both petrol and diesel, to improve emissions and fuel economy. It’s not the only change he’s witnessed.
“Back in the early days there was very little FEA or modelling in general. What existed was pretty basic. It’s come on tremendously, so much so that, although there may be small changes in the future such as reductions in cost and ease of use, the results today are so close to reality it’s incredible. Having said that, we do still use old school basic calculations.”
Testing Times
One of Turbo Technics’ early examples of Kershaw’s ‘special relationship’ with Garrett was with VSR testing machinery. The original manufacturer-level test machines were expensive, cumbersome and needed two operators. Kershaw says, “The machine we devised was around 15% of the price, needed only one operator and when we benchmarked it in the Garrett factory, it was really accurate.”
In fact, machinery manufacturing has become a key part of the Turbo Technics story. The company make two types: core balancing and flow testers. Kershaw says, “With the former, we make a wide range of different sizes. We’re seeing huge growth in China where we’ve supplied more than 250 machines over the last 20 years. It’s a really high-growth market and it’s been a huge help in the pandemic. We’ve been flattened by demand!
“The trends are that customers want machines that are more user friendly, with greater test capability, like checking for leaks – basically a wider QC check of the core assembly. We’re testing the whole turbo without the end housings, which means we can do a pretty good job on the core – it’s what the OEMs are looking for. Also, I predict there will be more automation to reduce cycle time. Holset, Schwitzer and Borg Warner all use these machines.”
With the flow testers, which are more associated with turbo diesels, Kershaw is not so optimistic. “Given the dieselgate emissions saga, I think the trend is towards more sophisticated petrol engines. The diesels that will survive will be increasingly complex, with multiple turbo stages and more variable geometry. There may be some new applications in the pipeline, but the emphasis is not where it was a few years ago.”
Fuelling The Future
So what could replace fossil fuels? After a pause for thought, he says, “Engines in general will use whatever fuel is available. Should that become, for example, bioethanol of some kind, we will adapt the engines to run on that. It can be done – look at Brazil, they run high percentages of ethanol and have done so for years.”
And where do hybrids fit in? “There is an argument that smaller batteries should be used with range-extending engines. Here, turbocharging does improve efficiency and emissions, and can allow the engine to run leaner. With recirculation systems, the temperature can be kept down which helps keep the NOx down. Assuming 100 miles is a reasonable range for most drivers, when longer journeys are required range-extenders are more practical than pure EVs.”
On the subject of the latter, Kershaw says, “I’m not backing it. I feel either pure hydrogen or manufactured hydrocarbons is the way to go. I’m not sure whether IC or fuel cells will win in the end; they both have pluses and minuses, and there’s research money going in both directions, although lately it’s gone a bit quiet on the fuel cell front. Storing liquid hydrogen in smaller vehicles isn’t easy, and while mixing green hydrogen from renewables with carbon from the atmosphere would solve this problem, it’s not yet clear how practical that will be.”
Making The Grade
Kershaw has experience with all sorts of manufacturing techniques. “3D printing may never be used for mass production but it’s still useful for certain components; things you just couldn’t make any other way, like forms with odd, curved passages for example. At the moment it’s still a bit slow, and therefore expensive, but we do use it on a small scale in particular for plastic parts – connectors and the like. We like it for prototyping, both SLS and filament methods. Direct metal sintering usually doesn’t work economically for us, though we are currently investigating using it to make an exhaust manifold.
“We’ve got loads of CNC multi-axis machines on site whereas, when I started out 40 years ago, even one would be a rare sight. This technology speeds things up and allows improved accuracy – we now fairly routinely work to tolerances down to 3 microns on a production basis.”
Materials research is another key area for Kershaw. Due to the low density, there was intensive research into ceramics for the rotors in the 1980s and 1990s. “A few people tried it in production,” he recalls, “but the stuff is quite brittle and it can’t tolerate any kind of abuse. Even the entry of a relatively small particle can cause a chip and in a few milliseconds you’ve got a cloud of powder coming out of the exhaust.
“The vast majority of current turbos will remain nickel-based, with a tiny amount of cobalt alloy in specialist applications. I also know that Borg Warner is experimenting with some competition units made of titanium aluminide because of the low inertia.”
Housings will also remain nickel-based, with a preference for water-cooling of the bearing housings, a design feature which has mitigated an issue with early turbos. Kershaw says, “When I started years ago we did have a problem with hot shut-down. The heat soaks through the shaft and into the bearings, carbonising the oil. It could choke the bearing and oil-feed channels. Plumbing the whole lot into the engine coolant partly solved this, aided by improvements in engine oil formulation.”
Over Kershaw’s career, perhaps the area that has changed most for engines is the advance of electronics, technology he uses on his own racecar, a 600HP Focus. “These modern wastegate controls are far better than the old pneumatic versions, enabling greater precision so you can tailor the boost. Also, an area of active research right now is electric assistance for the turbo, in particular at low speed. Another big trend is downsizing, accompanied by speed increases, up to around 270,000RPM. I’m also seeing an increased interest in variable geometry in the petrol market. With the pressure to increase fuel efficiency, I see a healthy future in the turbo market, with continuing OEM interest assured.”