The Mechanic’s Tale

by Steve Matchett

  Annoyed at himself, perhaps, but determined to stick by his word, Riccardo Patrese joined Benetton, and he and I started working together as team-mates in the winter of 1992. It was a long, hard winter too; for everyone at the Enstone factory, producing our high-tech B193 resulted in a fiercely steep learning curve. To make life a little simpler, the initial idea was to retain last year’s chassis, the B192, but to remove the passive suspension and install our own in-house-designed active suspension. The 1992 gearbox was also scrapped; for the new season we were to build and develop a semi-automatic transmission. Because of both of these major modifications, as well as some additional changes to the aerodynamics, the car was sufficiently altered from its original specification to warrant it being rebadged as the B193.

  With all of the upcoming test sessions we would have more than enough work to keep everybody fully occupied throughout the winter. However, the FIA then stipulated the compulsory use of narrower tyres for the 1993 season, but in testing, regardless of the assistance offered by our new active suspension, the ’92 chassis hated the reduced grip imposed by the new regulations. The only real solution to the car’s handling problems would be to design and build a new chassis, one with the weight distribution and suspension geometry perfectly matched to the characteristics of the new tyres. But at this late stage in the year, coupled with all the development work required on the new active system and the semi-auto transmission, to then embark on such a mammoth project would, quite simply, be utterly crazy. Indeed, merely to suggest that such an undertaking was possible would be seen as an act of sheer optimism. The decision was taken – we would build a new chassis.

  The glue had barely dried on the B193 chassis plates before they were pulled back off and changed again. This time the revamped 1992 chassis were badged as B193A cars; when it came into production the brand-new ’93 chassis would be known as the B193B. It would have been impossible to start the racing season with this true 1993 car, since there just weren’t enough hours in the day, so the policy we had used in previous years was deployed once again: we would use the old car for the first two races, the intercontinental races of South Africa and Brazil; then the new car would take over once we had returned to Europe.

  I mentioned earlier that when an active car is performing well it is a dream to work with. In the pits the mechanics are still having to adjust the wing angles, check for leaks, circuit damage, and generally keep an eye on the brake wear etc. but the vast majority of suspension alterations are merely software changes. Connect car to computer, tap, tap, tap, unplug computer from car, send car back to the circuit. How long did that take, about six seconds? Well it possibly took a little longer in reality, but not much. However, the amount of work it took Benetton to get the car to that stage of efficiency was immense.

  In Estoril during the winter, where we were testing our new active suspension system shortly after its introduction, we worked on the car for three days and two nights (with no more than four hours’ sleep throughout the entire test) and it was only on the third day that we finally managed to coax the car into doing one extremely slow lap past the pits. For two frustrating days and two freezing nights the car sat in the garage doing no more than being a completely stubborn bastard. We had the thing sitting on corner-weight scales (four electronic pads, one placed under each wheel, which are used to check for even weight distribution), with the flushing-rig maintaining fluid pressure to the car while Ross Brawn and his R&D engineers tried to get the computerized electronics to communicate and work with at least some slight semblance of harmony with the system’s hydraulics. Software was installed and system pressure applied, but rather than react logically to these settings the car would just violently shudder and shake as if in the throes of some hyperactive, convulsive fit. Hydraulic pressure was cut, data was scrutinized, teeth were thoughtfully sucked.

  We tried again (several times, in fact), only to achieve slightly more disconcerting results. More sucking of teeth and then the Moog valves were blamed as being the most likely culprits (the electronically controlled flow valves, which regulate fluid movement around the car’s hydraulics). The mechanics duly changed the Moog valves, but with absolutely no improvement to the car’s attitude whatsoever.

  Working on such complex, high-pressure hydraulics was a new experience for me (it was new to us all) and knowing that the system’s operating pressure was in the region of 2500psi – certainly more than enough to blind a mechanic or cause an instant fireball in the event of an accident – I was always careful to ask the R&D engineers, the people who had designed and pioneered the concept, for advice on exactly when it was safe to disconnect a particular component. Once, one of them told me it was perfectly fine to disconnect a Moog valve from the car’s fluid manifold: ‘The system’s dead, there’s no pressure in the line now,’ he continued. I looked at the mechanic standing next to me; he seemed rather unconvinced by this news. I asked the engineer again, just to confirm that we were all happy before I started unscrewing the bolts (working on an active suspension system was a little like defusing a bomb); this time even the engineer seemed a little unsure of the validity of what he had said. ‘Yes, it should be okay to remove it; but just be very, very, careful when you do so.’ And leaving us with that he disappeared to the back of the garage. Very reassuring.

  The problems with the car continued, chins were stroked, heads were scratched and numerous telephone calls were exchanged between Portugal and England. New software was written and installed, more Moog valves were changed, then the original Moog valves were substituted for the ones we had just fitted. More software changes were followed by more coffee-drinking. Moog changes were followed by more bouts of hydraulic flushing. Long and tedious hours were followed by longer, even more tedious hours. For two days the car never turned a wheel, we just continued to exchange one possibly faulty part for another possibly faulty part. We drank coffee, we became more and more tired and we got thoroughly sick to death of each other’s company.

  It must be remembered that monumentally long hours are par for the course during the start of any Grand Prix year and, regardless of whether a team is designing an active car or not, it makes little difference to the potential workload. Pre-season testing has always been grim and the arrival of any new design is a potential all-nighter in a box. A redesigned chassis, a new transmission, a different engine, new suspension, it makes little difference; until the freshly starched creases of the new design have been thoroughly ironed out, the chances of a good night’s sleep between the months of January and April are extremely limited.

  This situation is normal and understood, but what really depressed me, right throughout the period of our active suspension troubles, was the fact that as a mechanic there was little I or any of the other mechanics could do to speed things along. We couldn’t see any fault. The parts that we removed from the car because they were apparently faulty looked as good as the parts we refitted. As mechanics, we are used to looking at a component and checking it for serviceability: too much play in a bearing or not enough lubrication on a gear; stretched threads, damaged wiring looms. Locate the fault, see or feel the problem, think of a solution, fix it and move on; but when, for example, the potential fault is caused by one wrong digit in the writing of a software program, what can we do? We feel helpless, our role gone.

  I’m all for the advancement of Grand Prix technology, but my formal training as a mechanic never allowed for such speed of advancement in electronics – not even my BMW training included lessons in correcting computer programming errors. Throughout the Estoril test the car crews would ensure that their machines were in pristine mechanical condition, but until the electricians, the lab technicians and the software specialists had worked their magic, no more could be done; the mechanics were sidelined. Terribly frustrating.

  That Estoril test was a career low point for me, not the lowest, by any means, but I hold no fond memories of it whatsoever. Well, perhaps just one: another of the mechanics, Kenny Handkammer
, came out with a wonderful phrase at some point during those three days when, tired and ashen-faced, he described our active car as being like a moose. ‘Steve, you’ve got to help me,’ he said, ‘I can’t go on anymore! I’m so knackered I can hardly walk, yet it feels like I’m having to carry a huge burden around with me. It seems just like I’ve got a massive moose flopped on my back! Its front legs draped over my shoulders, its great head lolling to one side, and I’m having to lug it about with me; it’s really starting to drag me down. The Benetton Active Moose and I’ve got it! It’s jumped on my back!’

  I thought I’d explain a little about semi-automatic transmissions at this point. I quite like them and, speaking as a mechanic, I think they are one of the most interesting things to come out of Formula One design. At the end of the 1992 season in Adelaide, I was standing below the rostrum with the other mechanics, watching Michael and Martin receive their awards for finishing in second and third places. Both drivers were delighted of course, and when Martin saw us he lifted a hand aloft, his thumb held up to thank us for the reliability of his car. As he did so I remember noticing the black tessa-tape which I’d used to bandage his right thumb and a couple of his fingers prior to the race. I remember too how odd his bandaged hand looked as he stood there smiling from the rostrum, and thinking that millions of people around the world would be wondering what had happened to his hand. In fact, this bandaging was something I used to do for him as a matter of course throughout his year with us. It was protection for the fingers of his right hand against blisters, caused by the knob of the gear-lever. Regardless of the added protection of his leather racing gloves, the constant gear changes with the Benetton manual gearbox used to blister his fingers. As he stood on the rostrum, smiling and waving to the Australian crowd, I knew he must have been in some considerable discomfort, yet his happy demeanour betrayed nothing of his pain, the adrenaline of competition keeping the soreness at bay, probably for another hour or so.

  The Australian Grand Prix of 1992 was the last time that Benetton used a manual transmission on one of their cars; and now, those old romantic days of stick-shifting the gearbox have long gone; gone forever, I assume.

  Back then, and quite unlike the current semi-auto systems, the driver had to be brave enough to take his hand from the wheel and to manually select the desired gear himself. He was also expected to remember which gear he was already using, and be able to study the rev counter to coincide his next change with the engine’s peak power output. Not only that but while down-changing he had to blip the engine with the throttle pedal to synchronize the speed of the gear shafts and be careful not to change down prematurely, over-revving the engine as a result and causing pistons and valves to over-react.

  All of these gear changes were achieved via a hand-operated control knob connected to a titanium lever, connected to a joint, connected to a carbon rod, connected to another joint, connected to the selector mechanism, mounted on the gearbox casing. Not forgetting reverse, of course, which at one time was selected by pulling a cable mounted behind the driver’s shoulder. How terribly quaint. Well, quaint the use of a gear lever may have been but it wasn’t very efficient and it wasn’t very quick – at least not when compared with the current state-of-the-art, computer-controlled, hydraulically operated systems, which are capable of changing gear within thirty milliseconds (0.03 second); doing so with perfect synchronized timing and with zero risk of over-revving the engine.

  To make such rapid gear changes the selector mechanism utilizes the same high-pressure hydraulics which are also used to control the active suspension. The greater the fluid pressure is, the faster the fluid reacts and the quicker it can shift the gear selector mechanism. Because of this extreme operating pressure (didn’t we say the active pumps worked at 2500psi?) the hydraulic fluid has to travel around the car through equally high-pressure resistant hard-pipes and flexible hoses; anything less than jet aircraft specification equipment would result in an instant and most dramatic fire. Every time these hydraulic lines are disconnected, for example, to allow the engine or gearbox to be changed, the system needs to be purged of air in exactly the same way as the active suspension. In order to reduce these delays to a minimum (remembering that track time is a very precious commodity) all the top teams utilize quick-release couplings known as a ‘dry-break’. This no-fluid-loss connector is a piece of pure aeronautical engineering magic, a push-and-click-on, click-and-pull-off device. As the dry-break is uncoupled it does so without losing a single drip of fluid, allowing components to be interchanged without the need to purge the system. Dry-breaks are enormously expensive but in valuable time saved, each one is worth double its weight in gold.

  With semi-auto transmissions, instead of the driver using a lever to manually change gear, the gears are ‘requested’ via two micro-switches mounted on the steering wheel, each switch operated by the now familiar-looking paddle levers. Pulling the right lever changes the gears up; pulling the left changes down. I think it’s interesting to note that all the teams have their up-shifts and down-shifts co-ordinated like this. For some reason it would appear illogical – as if against human conditioning – to reverse this, right-for-up, left-for-down design. Given the choice how would you prefer the paddle layout?

  When the driver pulls the up-lever all manner of exciting things happen: the micro-switch sends a signal to the computer processor located in the gearbox controller, asking if it would be possible for the selector mechanism to change up a gear. The controller hears the micro-switch and ponders the situation. The controller checks with the potentiometer mounted on the rotary selector (which is similar to a motorcycle system) as to what gear the car is in at the moment. Next the controller checks with the rear wheel-speed sensors to see how fast the car is going, and then checks the same information with the engine revs, courtesy of the crank-speed trigger. Now the controller looks at the ratio of the desired gear, and because the ratios are constantly being altered, from circuit to circuit (or due to a change in wind direction, from session to session), this ratio information is pre-programmed into the controller every time the gearbox undergoes a ratio rebuild. The controller reads the information about the desired ratio, looks at the engine speed, looks at the wheel speed and does some calculations; working out the anticipated rpm that the engine will be running at should it allow the desired gear to be engaged. Armed with all of this information, the controller is finally ready to go to work.

  The controller tells the clutch-actuator to engage the clutch slave-cylinder and disconnect the transmission from the engine. The clutch-actuator obliges. Next the controller orders the actuator working the throttle butterflies to blip the engine rpm to a figure slightly higher than it anticipates the engine to be running at once the new gear is engaged. The throttle-actuator does as it is told. Now, to bring the rpm to its correct figure, the controller asks the engine management controller if it could spark-cut the ignition system until the desired figure is reached. The engine management controller checks the operating condition of its engine and, on finding that all is well, it willingly agrees to assist the gearbox controller, and the ignition is momentarily cut. Now the controller orders the actuator working the gear-selector to pull the original gear out of mesh, rotate the selector and engage the new, desired gear. This is done. The controller seems pleased with its work but still has a final check to see that everything is as it should be and that no one has become lost or confused with what he has been asked to do. On finding only good news it then orders the clutch-actuator to re-engage the transmission with the engine. Voilà, one gear change satisfactorily completed, all in less than a third of the time it takes to click your fingers. Oh, and just to make life as easy as possible for the driver, the gearbox controller even flashes up on the dashboard which gear has been selected.

  All in all, a jolly impressive piece of engineering, don’t you think?

  Finally, after several weeks of such long and arduous tests, the car began to calm down and eventually to respond fav
ourably to treatment. More filtration and purification of the software, more delicate alterations to the hydraulics, and the Benetton active cars gradually matured into a very quick and nimble series of Formula One race car. Throughout its brief life, refinement work never stopped on the B193B, and for the last two races of the year there was a big push to fit a four-wheel-steer system. I don’t know why – the car certainly didn’t need it – but it was as if the engineers were desperate to produce the most sophisticated Grand Prix car in the pit-lane before the end of the year. We had thoroughly tested the four-wheel-steer system in England before we packed the cars and left for Japan, but when I spoke to the drivers about the effects of this rear steering, both Patrese and Schumacher could find no advantage whatsoever in using it. Riccardo told me that it certainly made the car feel different but it was a feeling he didn’t like. I couldn’t notice any improvement in lap-times either, and as far as I could see, the pre-Japan/Australia test proved that the system was not needed.

  True, the system was designed to be able to be switched on and off, so the drivers could elect to use it or leave it switched off (in which case the hydraulically operated rear steering-rack would merely work as a solid link), but it still had fluid lines connected to it, which meant that it remained a reliability risk. The theory is this: any component of a race car carries a potential risk of failure, therefore the philosophy for ultimate reliability is not to have the component on the car in the first place; if it ain’t there it can’t go wrong. Obviously, it’s not possible to do away with every component, otherwise you’d just end up with a block of carbon, but anything that doesn’t need to be on the car shouldn’t be on the car; simple common sense really.