…or simply a case of arrested development? In the first of a three-part series, we focus on cars that perform better in historic racing today than ever they did in period. To kick things off, we look at the 1982 Arrows A4
Marc Surer, Mauro Baldi (and, briefly, Brian Henton) suffered a dismal 1982 in their Arrows A4s. In 27 attempts the Cosworth DFV-powered cars failed to qualify six times, rarely started higher than 20th and never better than 16th. The best race results all year were fifth and sixth places for Surer and two sixths for Baldi.
Arrows was 11th in the championship for constructors, Surer and Baldi 21st and 25th in that for drivers – not much for the British team to show for its efforts. Henton was there only as an early-season stand-in for the injured Surer. Keke Rosberg won the 1982 world championship in a Williams FW08 – the final such success for the Cosworth DFV.
Fast forward to 2014. Driving an Arrows A4 Steve Hartley won eight of the 12 races he contested on his way to a dominant victory in the FIA Masters Historic Formula 1 Championship. Competition included several Williams FW07s and FW08s, a Brabham BT49 and a Lotus 91.
This was no fluke.
Hartley’s Arrows A4 has been a regular winner since 2010, including an FIA victory at Brands Hatch in May 2017 – impressive for a genuinely amateur driver who races in no more than eight meetings per year, very rarely tests and is otherwsie extremely busy running a large engineering company.
So why is the Arrows A4 competitive in 2017 when it wasn’t in 1982? There are a number of factors, including regulatory differences between the F1 world championship back then and the current FIA Historic F1 regulations. But perhaps the most revealing aspects lie deep in the design and set-up of the car itself. On the surface, there are few differences between the Arrows A4 and Rosberg’s Williams FW08 but dive a little deeper – and apply GSD RaceDyn’s modern vehicle dynamics and kinematics software – and more is revealed.
THE DAVE Wass-designed Arrows A4 has a narrow aluminium honeycomb monocoque chassis that ends immediately behind the fuel tank. Torsional stiffness is 8400 Newton metres/degree (6200 lb ft/degree), which means the A4 was stiff for period, albeit less so than the carbon-fibre McLaren MP4 and Lotus 91. To give context, the A4 tub is five times stiffer than the 1964 spaceframe Brabham BT11 and slightly more than twice as stiff as a Lotus 25 – the first F1 car with a fully-stressed monocoque. A current F1 machine would be three to four times stiffer than the A4.
The Cosworth DFV V8 engine is bolted in four places to the rear of the monocoque and acts as a fully-stressed element, as does the transaxle. Both front and rear suspension are by lower wishbones and upper rockers, operating vertical inboard spring/damper units. Koni 8212 double adjustable dampers are used with shear plates [light, thin plates, often aluminium, used to absorb forces in a single direction] only to support the rocker pivots front and rear. The rear suspension is mounted on the transaxle casing.
Wheel to spring/damper motion ratios [the ratio of vertical movement of the wheel to change in length of the spring and damper] are good – 1.408:1 front, 1.436:1 rear. Suspension geometry and kinematics [an analysis of the movement, velocity and acceleration of parts of a multi-link mechanism] are also generally very good. Front and rear anti-roll bars are also fitted.
The car has full venturi sidepods [a device involving a curved inlet, flat ‘throat’ and an angled exit ramp known as a diffuser, creating a low pressure area at the throat]. Various materials were used for the fixed skirts and these were fitted with ceramic rubbing strips. At static ride height, the skirts were set with just a few millimetres of ground clearance.
The neat fabricated rear suspension uprights do not protrude into the diffusers. Underbody and upper body panels are smooth and neatly faired. A dual-plane rear wing, with an adjustable second element, was fitted.
During 1982, several different front wing types were used, all single-plane. At some circuits the venturis provided sufficient downforce and front wings were not fitted. On other occasions, very small front wings were set at negative incidence [the angle between a wing section and the free stream airflow] to act as flow guides into the venturi tunnels and to trim the handling balance. During airfield testing at Kemble in 2010, the A4’s maximum downforce was measured at more than 1260kg, with 453kg of drag at 167mph. This resulted in a lift/drag efficiency figure of 2.78.
In terms of overall aerodynamic and mechanical design, the Arrows A4 is actually very good – comparable with the Williams FW07 but not quite as efficient as the FW08. So as the plot thickens, now’s the time to take a closer look at Rosberg’s Williams FW08…
IN GENERAL the FW08 is similar to the Arrows A4 (it also has an aluminium honeycomb chassis), but it is rather more refined in detail. Compared with the Arrows the key differences are mentioned in the separate box, but it’s the detail differences we’re looking for here. They include the front suspension – which uses a pullrod and bellcrank arrangement to operate vertical inboard spring/damper units. This avoids the undamped flexing that can occur in rocker systems such as those used on the A4. However, there is a penalty – a poor (1.818:1) wheel to spring/damper motion ratio that reduces damper efficiency. The A4’s front motion ratio is 1.408:1.
Secondly, the FW08 is lighter than the Arrows, needing at least 20kg ballast to reach the 585kg minimum weight limit. It was probably even lighter in the early half of 1982 as the non-turbocharged teams were using a rules loophole to qualify and race underweight. It was permissible to top up lubricant and coolant levels post-race and qualifying, before being weighed. Teams used 40-50kg of ‘brake coolant’ that they rapidly jettisoned, topping up later to meet the weight limit. This gave a gain of up to 1.7sec per lap. Brabham, Lotus, Williams and McLaren could do this, but Arrows probably couldn’t. This affected early-season results, before the FIA swiftly put a stop to it.
Finally, finances. More money generally means more speed – and we must be aware that Arrows was constrained by a much smaller budget than those enjoyed by Williams. However, in Frank’s defence, McLaren and Brabham operated on similar budgets in 1982 – and Williams took the title.
As the 2010 airfield testing and torsional measurements demonstrate, the Arrows team produced a good basic design in the A4, but financial constraints meant less wind tunnel time, less track testing and therefore much less analysis. In particular, this would have limited its ability to explore ride height sensitivity, pitch sensitivity and yaw sensitivity – aerodynamic and mechanical set-up refinements and the vital integration between the two that would undoubtedly have generated more speed.
Now some data. Like the Arrows, the Williams FW08 has been subjected to modern airfield testing – this time at North Weald in 2012. It was discovered that the car delivers approximately 16 per cent more downforce than the Arrows A4 – but with the penalty of 11-15 per cent more drag.
To understand these differences, it’s worth looking at comparative performance for the two cars on the current Brands Hatch circuit. But rather than bolting in a driver and attempting to achieve consistent lap times for analysis, this is the point where GSD steps into the virtual world and uses simulation.
IN 1982, Rosberg qualified his Williams FW08 on pole for the British Grand Prix at Brands Hatch on 1min 9.54sec. This was 3.64sec (5.14 per cent) faster than Surer in the faster of the two Arrows. If you simulate lap times of the FW08 and the A4 on the current Brands Hatch circuit, the absolute times are not relevant because of circuit changes but the comparative times are valid. Assuming equal tyre performance and equal drivers, the Arrows is only 1.43sec (2 per cent) slower than the Williams. This must mean that an additional 2.21sec was lost due to a combination of set-up, aerodynamic efficiency, driving skill and tyre performance.
Surer was very capable, so it is reasonable to assume that the majority of time was lost due to tyres and set-up. We’ve covered why the team was unlikely to have achieved a perfect set-up (that ‘b’ word), but the tyres?
Simulated data points to poorer qualifying tyre performance on the Arrows – with anecdotal evidence from period suggesting the same. The Arrows ran on Pirellis and the Williams on Goodyears. Detailed data is hard to come by, but it appears that the Arrows was considerably closer to the pace in race trim than it was in qualifying – which suggests that the Pirelli race tyres were closer in performance to the Goodyears.
With regards to the comparative performance of the 1982 cars, GSD reached three conclusions. Firstly, the Williams FW08 had better aerodynamics – but only slightly. If you combine all the aero data, including simulations, the difference in lap time would have amounted to just 2 per cent.
Secondly, Surer, Baldi and (briefly) Henton had little chance in qualifying – the Goodyear qualifying tyres used on the Williams, Brabham, Tyrrell, Lotus and Ferrari, were far superior to the Arrows’ Pirellis.
Lastly, set-up optimisation was limited by Arrows’ budget restrictions – but this effect was probably smaller than the aerodynamic and tyre deficiencies. And the nut behind the wheel? Surer and Baldi were not journeymen. Both were distinguished single-seater drivers, so it can be safely assumed that given better qualifying tyres and a fully developed aerodynamic package they would have been contenders for podiums in 1982 – and maybe even a race win. And remember, world champ Rosberg scored only a single victory that year.
The Arrows A4 was generally a good, sound car in period – but its potential lay untapped. Today, in current FIA Historic F1 racing, it is a winner. Let’s explore the reasons why.
FIRSTLY, FIA Historic F1 regulations for cars built and raced in 1982 dictate a four-centimetre minimum ride height and 585kg minimum weight, with the rear wing limited in height and width. Avon A11 slicks are the mandatory ‘control’ tyre – no other tyre brand is allowed.
For ground-effect cars built between 1977 and 1982, a single front wing chord [the length of a wing element measured from leading to trailing edge] and profile are mandatory. This is basically a straightforward NACA [National Advisory Committee for Aeronautics, a US agency that later became NASA] low-speed section, similar to that used on the Williams FW08 in 1982. In other words, all ground-effect cars built between 1977 and 1982 must run with a front wing of the same shape.
Sliding skirts are not permitted. Fixed skirts are allowed, but only one set is allowed per race and since 2015 these have carried FIA seals to prevent changes being made. Rubbing strips, such as the ceramic ones used in 1982, are outlawed.
During the 2010 season, A4 owners David Abbott and Steve Hartley were struggling with the performance of their cars. With little time in-season for credible aerodynamic testing, a full geometry, kinematics and stability analysis was carried out using GSD vehicle dynamics simulation software, resulting in a conventional, interim mechanical set-up change. Revised anti-roll bars, springs, ride heights and geometry settings were implemented – and Steve finished third in the next race at Spa.
Lap times were still some way off the well-developed Williams FW07, Brabham BT49 and McLaren MP4s, however. Given the absence of aero data on the A4, this suggested that it was indeed aerodynamic performance that was lacking. The final piece of the puzzle? There was only one way to find out.
Wind-tunnel testing can be prohibitively costly – a rolling road and expensive models are needed, but credible aerodynamic data can be achieved at a very low cost by using straightline airfield testing. Enter the aforementioned Kemble airfield in Gloucestershire, England.
Temporary damper potentiometers were fitted (they are not permitted in competition) and a full, very intense day of straightline testing was carried out together with Mirage Engineering, the race team that run the Arrows. Basic downforce and drag data was gathered at various ride heights, rakes and wing settings, resulting in six balanced and efficient aerodynamic configurations, from low to high downforce. The budget for the day was £3000. A day in a suitable moving-ground wind tunnel would cost massively more, not to mention the huge cost of building a quarter scale or 60 per cent model.
The tests confirmed and quantified what was known in principle – that the underbody only generates downforce when the skirts are in reasonably close proximity to the ground, preventing ambient air from ‘filling’ the low pressure area under the car. However, if ride height falls too low, flow separation in the diffuser tunnels – or diffuser ‘stall’ – can cause a dramatic reduction in downforce. In turn, this can induce ‘porpoising’ – a very unpleasant vertical oscillation for the driver.
The solution? It involves developing relatively soft springs and carefully calibrated bump rubbers, which meet the regulations but significantly improve underbody aerodynamic performance at speeds above 85mph. This fix costs less than £200 in hardware, but a lot of maths and physics to optimise ride height versus speed and downforce – no mean feat when you have to consider things like tyre compression, which can shrink the overall radius by more than 11mm at 150mph.
With the bump rubbers and springs in place the A4 became competitive in the Masters series – however, the skirt material began to cause some problems resulting in a switch to polypropylene. It all came good at the 2010 Silverstone Classic, which Steve won outright. The cars have been generally competitive since then, including a pole for David at Donington, a slightly fortunate win for Steve at the 2012 British GP support race and Steve’s dominant championship win in 2014.
One particular problem still occurs intermittently – porpoising. The solution to this rather debilitating phenomenon? More software – this time focusing on more sophisticated damper analysis and a low-cost (£2000) CFD analysis of underbody ride height sensitivity.
Today in Historic F1, most teams run some type of ride height optimisation arrangement which means competition in the Masters F1 series is very intense. Moreover, there are several very quick young drivers in the championship with Michael Lyons, Sam and Ollie Hancock, Nick Padmore and Jonathan Kennard competing alongside top-line professionals such as Martin Stretton, ex-F1 driver Paulo Barilla and Andy Wolfe. Despite this, Steve Hartley in the Arrows A4 again made the top step of the podium at Brands Hatch in May 2017.
The skirt regulation in FIA Masters F1 makes it difficult to extract the aero advantages inherent in Williams, Brabham and McLaren designs and the Avon control tyre helps level the playing field. Furthermore, if you sprinkle some modern science on a car like the A4 – and provided it is well driven and carefully engineered – an unlikely contender can become fully competitive.
Overall there’s little doubt the Arrows A4 was fundamentally a good car. Jackie Oliver and Dave Wass should be proud – although there’s likely a sense of frustration that they didn’t have the budget, the tyres, the computing power or the software to extract its latent performance potential.
Nigel Rees lectures in Vehicle Dynamics at Oxford Brookes University and is founder of GSD RaceDyn – a vehicle dynamics software and analysis firm. More than 560 cars have been analysed by Nigel and GSD, with many notching up victories and championships in historic and modern categories.
OUTBRAKING THE RULES
The FIA tried to tame ground effects, but F1’s engineers were a step ahead
Turbocharged engines and ground-effect aerodynamics were changing the face of F1 by 1982. Eight races were won by turbocharged Renaults and Ferraris, eight by normally aspirated Ford-Cosworth DFV- powered cars. The turbos were still hampered by pre-ignition/detonation and throttle lag, but the writing was on the wall.
During 1979 and 1980, Williams, Brabham, McLaren and Ligier evolved the ground-effect theme to new levels through intensive wind tunnel testing. By 1980, downforce levels reached more than 1400kg at 160mph – at least twice the figure achieved by the first ground-effect F1 car, the Lotus 78, introduced just three years earlier. Tyre development continued apace with the introduction of radial-ply tyres. Lateral acceleration during cornering approached 4g.
Cornering speeds were extremely high, giving rise to serious concerns over safety, so the FIA banned sliding skirts and introduced a 6cm minimum ride height regulation for 1981. The British teams complained bitterly that 65 per cent of total downforce would be lost, handing a huge advantage to the ‘grandees’ – Renault and Ferrari – with their more powerful turbocharged engines.
However, Brabham immediately introduced an electro-hydraulic system that kept the car at the regulation ride height limit in the pits, but lowered it until the skirts touched when on track. Other teams immediately followed suit, circumventing the regulations and re-introducing ground effect downforce. However, the cars were even stiffer, more physically demanding and arguably more dangerous.
For 1982, the FIA conceded that it was impossible to police a ride-height regulation so the 6cm restriction was removed. Fixed skirts were allowed, but the ban on sliding skirts remained. At the end of that year the FIA introduced a flat-bottom rule – drastically reducing ground-effect downforce – but turbocharging would remain until 1988.