"In 1988 I was engineering Gerhard Berger in the F187/88C. That was the year McLaren dominated with Honda and Bernie did all he could to help us. It was the era of turbos and pop-off valves and we had a low-pressure passage that went past the pop-off valve and would pull it open, so we could run more boost. We kept pushing that further and further, waiting to get caught, but we never were. I guess Bernie wanted somebody to try to beat McLaren, so he helped us."
|FISA Pop-off valve (drawing by Bent Sorenson, reproduced from 'The Anatomy and Development of the Formula One Racing Car from 1975', Sal Incandels, p200)|
Ian Bamsey noted the following in his monumental 1988 work, The 1000bhp Grand Prix cars, "In 1987 some engines were coaxed to run at more than 4.0 bar. With a carefully located single pop off valve merely an irritating leak in a heavily boosted system as much as 4.4 bar could be felt in the manifold. The key was in the location of the valve. It was possible to position it over a venturi in the charge plumbing system. Air gained speed through the venturi losing pressure. Either side of the venturi the flow was correct and the pressure was higher," (p29).
In fact, there appears to have been at least two distinct methods of flouting the 4.0 bar limit. If one attached the pop-off valve over a venturi, then one could keep the valve closed (contra Kimball's explanation) even if the effective boost pressure was greater than 4.0 bar. A second method simply involved inducting compressed air into the plenum chamber at a greater mass-flow rate than the open pop-off valve could vent it:
"Turbo boost was theoretically restricted to four bar via popoff valves, but there was a way around this on self-contained V6s like the Honda. They required just one pop-off valve (as opposed to those like the Porsche and Ford which effectively ran as two separate three-cylinder units and so needed two pop-off valves) by overboosting, forcing the pop-off to open and then controlling it against boost. It meant 900bhp in races, 1050bhp in qualifying," (Mark Hughes, Motorsport Magazine, January 2007, page 92).
Indeed, the general suggestion at the time is that it was Honda, rather than Ferrari, which first identified these loophole(s). Bamsey makes this point in his superb 1990 work, McLaren Honda Turbo - A Technical Appraisal: "By mid-season ...Ferrari is believed to have achieved levels of 4.1/4.2 bar through careful location of the pop off valve, a technique Honda is alleged to have pioneered," (p92).
The next question, however, concerns what happened in 1988, when the more stringent 2.5 bar limit was imposed, and a new design of pop-off valve was supplied to the teams. This valve (perhaps by deliberate design), was somewhat tardy in closing once it has been opened:
"The new pop off opened in a different manner and once opened pressure tumbled to 2.0 bar and still the valve didn't close properly...on overrun the effect of a shut throttle and a still spinning compressor (the turbine not instantly stopping, of course) could cause pressure in the plenum to overshoot 2.5 bar. In blowing the pop off open, that adversely affected the next acceleration...The answer to the problem was in the form of the so called XE2 [specification engine]...run by all four Honda cars in the San Marino Grand Prix.
"The XE2 changed the throttle position, removing the separate butterfly for each inlet tract and instead putting a butterfly in each bank's charge plumbing just ahead of the plenum inlet and thus ahead of the pop off," (ibid 1990, p91-92).
No questions of dubious legality there. However, Bamsey also explains that an XE3 version of the engine was developed by Honda, purportedly for exclusive use in the high-altitude conditions of Mexico City: "The Mexican air is thin - the pressure is around a quarter bar - so the turbine has to work harder. Back pressure [in the exhaust manifold] becomes a potential problem, affecting volumetric efficiency and hence torque. Power is a function of torque and engine speed: Honda sought higher revs to compensate. Thus the XE3 employed an 82mm bore size [compared to 79mm on the XE2] and it was apparently tuned for a higher peak power speed. It was a complete success and on occasion was tried for qualifying elsewhere thereafter (in particular, at Monza)," (ibid. 1990, p92).
What's interesting here is that the XE3 seems to have caused some scrutineering difficulties at Mexico. Road and Track magazine reported that there was "a claim that Honda had built vortex generators into its system - which would allow it to use more than 2.5 bar - and FISA scrutineers spent an unusual amount of time examining the McLarens in Mexico," (Road and Track, volume 40, p85).
Generating a vortex would offer an alternative means of keeping the pop-off valve closed. Even with a constant diameter pipe, the pressure could be lowered by transforming some of the pressure energy into the rotational energy of a vortex. One would presumably need an expanding section downstream to burst the vortex in a controlled manner, but it does offer a method of reducing the pressure without using a venturi. It's intriguing to read that an engine ostensibly developed for high-altitude conditions was used in qualifying for the rest of the season...
So perhaps it would be wrong to cast Ferrari here in their stereotypical role as regulatory bandits. Although Kimball does also suggest that their fuel-tanks carried somewhat more than the mandatory 150 litres of fuel when they won the Italian Grand Prix that year!