How to build a slow Formula 1 car

As the 2019 Williams FW42 is showing, even teams with illustrious histories and good people can make a slow car.

McLaren did the same in 2013 with the MP4-28. The MP4-27 had won the last two races of 2012 in USA and Brazil, but the team plummeted to the midfield the following season. And let's not forget the car that never raced - the MP4-18 of 2003. It took McLaren two years to climb back into a winning position with the MP4-20.

It's easy to fall off the cliff of performance, all it takes is a simple mistake. The challenge is to identify that mistake and then have everyone agree that is in fact the cause of the loss of performance. This can be the hardest part. Once you've done that you can formulate a program to dig yourself out of the hole. Ironically, the loss of performance is often created by trying too hard to make a huge jump in one go.

That was the case with the MP4-18 - the design was compromised to save weight. The car wasn't stiff enough in any area to carry the loads generated on the tyres from the aerodynamics and was simply unpredictable to drive. As the driver left the circuit heading for the barrier he had had no warning of the impending spin, the car just never gave him the correct feedback.

Sometimes it's not directly a design error that makes a slow car. In the case of the 1993 Lola it was a desire to be involved in F1 but not allowing enough time to do the job properly. Eric Broadley was a smart guy. He knew the future of his company had to be in F1. Spec formula were on the horizon elsewhere in racing and that wasn't the business model he wanted. Eric was a racer who loved technology, so when Scuderia Italia approached him in '92 to design a car to replace its Dallara chassis he leapt at the chance.

Lola's involvement with Larrousse had finished at the end of 1990 and no one had anticipated a replacement project. We should have kept a small research group following F1 - tracking the regulation changes and their impact on the design and aerodynamics. Instead we went in cold.

It was mid-1992, I remember the first meeting Eric called to break the news. We were all excited about returning to F1 when Tony [Woods], the production director, put up his hand and said, "Eric which year are we entering this car for?". We had all assumed '94.

We knew you couldn't windtunnel test and design a new car from scratch in less than six months.

"1993 of course" was Eric's reply. At this point Tony read him the riot act and refused to manage the production side of the project. Not a great start to something that should have begun months earlier.

For the previous six years I'd handled Lola's F3000 cars, doing the concept, aero, design and race engineering, so I fancied stepping up to F1.

Italian race teams are special - their passion is phenomenal and Scuderia Italia was no exception. The Dallara was a fine car, and with a little winter development could easily have been used to start 1993 to buy more precious windtunnel time for our own design. But that was not an option.

When the team's engineers visited us in Huntingdon, they asked how the windtunnel model was going. In Eric's brief there was no time for a windtunnel program, the car would simply get scaled-up F3000 aerodynamics.

So, I assumed Eric had had a change of heart - after all we would need to start a program at some point. I took this as a green light to start a model and program, but got my knuckles rapped because there wasn't any budget...

The rest of the car was quite straightforward. This was to be a very honest car: the geometry was F3000 and the running gear, although bespoke designed, was based on the Lola Larrousse or Lola F3000.

When the finished car was weighed for the first time we were relieved it came out under the limit - not by a lot but enough, all down to good design. Then came the rollout test.

Years later, Alessandro Mariani, one of the Scuderia's engineers who was at the test running Michele Alboreto, told me an amusing anecdote.

Michele went out for his first run, he pitted, came to a halt, beckoned Alessandro over and whispered three words in his ear: "We are fucked".

He elaborated: "The car handles nicely in the low-speed corners but it has NO downforce". He was almost correct - it certainly didn't have anywhere near enough to be competitive.

The windtunnel program lagged the car design to the point that its first run in the windtunnel happened after all the wetted surface drawings had been issued.

I remember Eric phoning me, while I was in the windtunnel, to ask about the results. I replied that we had hit the drag target but had only half the amount of downforce our simulation predicted you needed to compete with the frontrunning teams.

That might have sounded flippant, but I was having a deja vu moment reliving my previous design nightmare: the T950, Lola's first F3000 I designed that also never saw a windtunnel until after it was drawn.

That car took 12 months and Ralph Bellamy's help to sort. This project had all the same hallmarks.

If you're going to produce a slow car, best make it an attractive one. People have more patience with a slow pretty car. That's because they don't see any obvious flaws they can point a finger at.

Sadly, the T93/30 wasn't pretty so fingers got pointed at all the wrong things. At the first test the rear bodywork was reworked to reduce its volume. To no effect, as later verified in the windtunnel.

Meanwhile, my team and I launched into a proper aerodynamic program that within four months had clawed back half the deficit in downforce. Little of this produced upgrade material for the car because the fundamental problem was the chassis shape.

That's the hardest part - to figure out the fundamental cause when you're in firefighting mode. Yet we didn't use what we learned.

I guess Eric was too embarrassed to table the windtunnel figures. Maybe he didn't believe them and thought a simple fix to an errant vortex was all that was required. It didn't stop him producing new suspension parts - geometry after geometry.

Race by race the relationship between Lola and the team became more strained. The straw that broke the camel's back, for Eric, came during a private Monza test held by Ferrari late in the season.

Ferrari baselined its car then switched in the latest engine. It went one second quicker. Eric closed the lid on his briefcase, walked out of the garage and took a cab to the airport.

At the end of the season, Eric fired me for two reasons. Well, two he verbalised: not stopping him playing with geometry when the problem lay with aerodynamics, and for not allowing enough damper length to run really soft springs.

My reply was to the effect that, with reference to the springs, I wasn't designing a Formula Ford. But ultimately it was his company and 'what Lola wants, Lola gets'.

This could never happen today. If you're starting a new team you purchase permitted knowledge from an existing one. Designing cars today is all about risk management - make an honest car that's easy to set up in the limited time available during a race weekend, bank steady aerodynamic performance and prioritise reliability.

If, in spite of this, you still drop the ball, honesty will be your only saviour.