Grand designs 1987
Keith Howard traces two separate approaches from when active suspension seemed the answer
In 1987 both Lotus and Williams ran active suspension, starting a process that would end acrimoniously seven seasons on when all electronic ‘driver aids’ were outlawed. Remarkably, both teams won races with it (Lotus two with the 99T, Williams three with the active FW11B), yet the systems themselves could hardly have been more different.
Lotus first ran its full-active system, with springs and dampers synthesised by hydraulic actuators, in two races in ’83, after which the project was hived off to Lotus Engineering for road car development. When active returned to Team Lotus for ’87 it now worked in parallel with a conventional suspension spring, saving on hydraulic power requirements. But it was still a full active system, able to control wheel and vehicle dynamics as well as ride height.
Williams’s system was much simpler. Its aim was to precisely control ride height to maximise downforce. Paired with a passive spring and damper, it couldn’t control wheel dynamics like the Lotus system but was very effective in its narrower role.
John Davis, chief of Lotus Engineering’s active group, and Frank Dernie, designer of the Williams active suspension, talk here about the two systems, their benefits and how the drivers reacted to them.
For 1988 Team Lotus dropped active and began its decline; some blamed the active experiment. “The rumour Peter Warr put around was that ’88 was bad because the team focused on active in ’87,” says Davis. “Nonsense. I don’t remember anyone complaining of amnesia when they had to run passive again. The ’88 car just wasn’t quick enough.
“Part of my reason for swapping to Lotus Engineering in ’86 was that I felt we were being left behind. There was an inability to see where things were going. The attitude was: we don’t need a wind tunnel, or a big machining capability, or an autoclave. Other teams were putting those in place but we weren’t until it was too late.”
FD: Each active unit comprised a gas spring and a fixed damper. We started off using modified Citroen units. There was a pitifully small budget. I’d looked long and hard for a swashplate hydraulic pump instead of a gear-type because the losses were far less, and eventually came across a Vickers helicopter unit. They were about £1100 each but I bought one and it was perfect. I tried to buy more but they were on five-year delivery! I spent a fortnight phoning aeroplane repair companies across Europe until I’d rounded up the three we needed.”
JD: “We had just one DNF due to hardware failure in the active system, that’s all. Lots of other problems, but when you’re developing a technology you have the infrastructure to handle those. We had burst pipes on both cars within a lap of each other at Estoril just before qualifying, for instance, and burnt both of the left-rear actuators. But you’re geared up for that sort of thing. Essentially the reliability was good. When I’d first introduced instrumentation into the car, it was hard to keep it working because sensors would break, mechanics would cut wires through. As soon as we went active and a broken wire would stop the car, the electronics were as well treated as everything else.”
Active vs passive
JD: “We made a 10 per cent gain in downforce through controlling ride height. Testing passive against active, Senna soon had back the blistered hands he hadn’t had all season. He was quick in the passive car but said he couldn’t do that for more than a lap or two. That sums It up: over race distance the active car was much quicker. And it’s not the case that money spent on active could have gone into aerodynamics, because Lotus Engineering paid for active, not Team Lotus. Over the year we punched above our weight due to active.”
FD: “The active car was one or two seconds a lap faster than the passive, because of improved aerodynamic control. When full ground-effect was banned, the rule makers thought that flat-bottom would kill downforce. Completely wrong. You reduced peak downforce but also stability. Flat-bottom cars were 10 times more pitch-sensitive than full ground-effect cars: and if they went nose-up they generated real lift. So the range in which they worked was 10 times smaller, and the range over which they were lethal was much larger. The active system put the car in the optimum window in every corner regardless of speed.”
FD: “Our active suspension was relatively crude from a vehicle dynamics viewpoint no better in this respect than conventional suspension. But it was focused on controlling ride height, because that’s what I thought important. This is heretical, but I believe that downforce makes as much of a difference in second-gear corners as in fifth-gear corners. The benefits are always there. You might argue that in a slow corner the aerodynamic forces are smaller but you’re in the corner for longer and closer to the wheelspin limit. So although the percentage grip increase due to downforce is lower, the potential gain is still very important.”
JD: “I think the one benefit we got that Williams didn’t was dynamic control of roll-moment distribution. We were able to detect through the control algorithms whether the car was braking into or powering out of a corner and change the handling balance accordingly. You could argue whether that was worth the additional complexity, but we’d come from the full-active approach. We also used to allow the car to dip a little under braking so that it would pick up downforce to help turn-in. A passive car does the same, of course, but we were able to achieve it in such a way that we didn’t lose rear downforce, which passive cars did.”
JD: “After our Detroit win, we took out all the passive suspension geometry effects we’d left in because we didn’t want to turn everything upside-down at the start of the season. We didn’t need anti-squat or anti-dive because the active did that. Coming back from Detroit I asked Gerard [Ducarouge] if I could now take all this out, which we did. It was clear straight away that the system was purer as a result. I can’t swear it made the car any quicker, but it certainly made the numbers a lot easier. With the anti-dive out, for example, the full weight transfer under braking appeared on the load cells, which it hadn’t before.”
FD: “We left the suspension geometry exactly as it was, with anti-dive and anti-squat. There was no cash to do otherwise. It might have been preferable to take that out and have active do all the pitch control, but I’m not sure. It would have increased the amount of work the hydraulic system had to do. And it’s so difficult to test these things in isolation of each other. Suspension and steering geometry in general are very hard to test rigorously, one parameter at a time, because factors Interact. So we stuck to what we knew worked.”
JD: “Senna did all the development work on active. He wanted what active gave him but he was quicker to react than some of the system. He didn’t like the car to respond automatically when he’d already felt something and was correcting it. For instance, if the driver made a steering correction and the car didn’t respond, the system would react. He liked a bit of that but not too much. We could also dip the front of the car as it turned in. He quite liked the non-feedback controlled parts of that but if we wrote an algorithm that detected understeer and dipped the front in response he didn’t like that at all.”
FD: “Nelson [Piquet] did all the active car testing. Nigel [Mansell] wouldn’t get in it, until the active car won and then it was, ‘Where’s mine?’. Nelson loved technology, and was up for anything that made him quicker. He wanted to work hard in testing so when he went to the grand prix he knew that, without busting a gut, he could put the car on pole position and win the race by driving as slowly as he needed to keep a 15-second margin. On one occasion Nelson had gone to visit his mum in Brasilia and we wanted to test at Silverstone. Nigel refused, so I rang Nelson in Brazil. He jumped on a plane. That was the difference he was really into it.”