Track test -- Unipart March 783 Dolomite

Something different in Formula Three

Formula Three in whatever guise has a good pedigree for producing the next generation of Grand Prix drivers. Occasionally the top Formula Three man will go straight from what is now a 2-litre category into Formula One, the latest example being the career of Nelson Piquet, who dominated the British championship scene before making his debut with the Brabham GP team for the last round of the World Championship in Canada. In fact so many drivers have leapt to prominence straight from F3 that the validity of F2 is occasionally questioned: it is fair to say that the aspiring driver really must do F3, but he may get away with either a minimal or complete absence of F2. Most progress in logical fashion and derive benefit from the 300 b.h.p. stepping stone that F2 provides.

Formula Three in its present guise has been with us since January 1st, 1974 and is to continue next year, though wider bodywork will be permitted. The formula specifies a minimum weight of 440 kg., maximum wheel diameter of 13" at the rear and no more than 10"-wide wheels at either end, plus an engine derived from a unit produced at no less than 5,000 units a year, the same requirement as for a production Group 1 touring car.

The engine must be of 2-litres or less and is "strangled" by a 24 mm. diameter passage of no less than 3 mm. in length which controls the ingress of all air to the engine (or should legally!). The policing of this restrictor rule is now rather more efficient than in the early days of the formula, and the small cannon-like devices have themselves become almost standardised as Toyota's Formula Three engine by Novamotor in Italy dominates the category.

In 1974 a number of engines emerged as likely candidates for glory in F3 including Holbay's version of the Ford Pinto SOHC design, BMW's SOHC four (no more than four cylinders are allowed), but by the end of that first season there was little doubt in anyone's mind what the winning unit would continue to be. Brian Henton started off the season with a Holbay Pinto, but by the end of the year he was equipped with the Novamotor Toyota which is a twin overhead camshaft design equipped with a conventional two-valves-per-cylinder.

Unipart and the Dolomite Sprint four-valve-per-cylinder, SOHC motor appeared first in the 1976 season and have been through a fair number of permutations in pursuit of a first victory. For 1978 the team settled around David Price Racing of Twickenham in Middlesex for preparation of its three March cars (two 783s and one updated 773), the engine preparation carried out by Swindon Racing Engines in liaison with Leyland ST at Abingdon.

The team has suffered from a lack of cohesive management over the years. Witness the passage of seven drivers and preparation specialists like Benny Rowland, Holbay and Neil Brown before the present arrangement. It is estimated that Unipart have invested over half a million pounds themselves in Formula Three. This season alone, with three cars turned out on occasion and two forays abroad (Dijon and Osterreichring) will have put away the best part of £100,000 on running the cars and the engine development programme comes on top of that.

This does not mean that Unipart have been profligately wasting money. Let's also get it clear that it is not taxpayers' shekels either, for Unipart is a very profitable Cowley-based division of the Leyland empire. The facts are that all forms of single-seater racing are even more expensive than the mere presence of inflation would indicate. The teams have followed the examples of those in Formula One: spare cars; constant practice sessions midweek; far safer cars (the life support oxygen supply to the helmet for example, that is part of the Unipart and most other modern formula cars' specification); and constant development.

Even with tyres restricted to Goodyear G54 for dry use and G45s for the wet in Britain, there is plenty of work to be done if you are to find an edge over the opposition. In Unipart's case there was even more to do as the engine had to be developed to the point where it was competitive with the Toyota.

It has been an uphill struggle.  Tony Dron was the sole driver and inspiration behind the project in 1976, but the car did not start to appear in the results at the right end until 1977, when Grovewood Award winner Tiff Needell and Ian Taylor, the Ford 2000 Champion, were fielded in Holbay-engined Marches. Needell is the only link between 1977 and 1978: last year he started 25 races, finished 18 times and recorded seven top six placings plus a second overall. This year, when I spoke to him before the team's last race of the year at Thruxton, Needell had started 17 events, finished 14 and scored eight places in the top six, including two second places. Team-mate Brett Riley from New Zealand raced only five times and took a couple of third positions. Reliability is generally good, especially of the engine now that it has been modified to overcome its early season cylinder-head water overheating problems. Needell retired once with this problem, a second time while fighting for the lead in the Austrian GP supporting event with suspected plug failure, and a third time when the master switch proved faulty. Riley had a gearbox failure -- the Hewland Mk. 9 five-speed is near the limit of its capacity with this engine. Riley also suffered water on the electrics and a collision in his three non-finishes.

I  said earlier that Unipart were not wasting their money. I say this for two reasons: 1, they have derived an enormous amount of publicity from the project (and publicity has been their objective for they will frequently favour a GP supporting race over any championship considerations); and 2, because the starting price for a ride in a works F3 March-Toyota is liable to be £70,000 next year. Needell says resignedly (having driven in the F2 event at Hockenheim), "With a spare car, doing F2 the way it is done these days, you can reckon to want £160,000. If you are a good driver you can go along in either Formula Two or Three and start talking with roughly a third of' the budget -- i.e. £50,000 or so in F2 -- but the team will need a good sponsor if you are to stand a chance of getting in."

You could say that Needell has had all the breaks, winning his first FF as he did in a magazine competition and going on to win a Grovewood Award. Though he proved fast at Hockenheim in an F2 March (he was lying third when the BMW engine caught fire) Needell is unlikely to join Hunt, Watson and Keegan in Formula One unless he finds the kind of backer that has appeared in each of those men's careers.

Theoretically the provision of an orifice to limit the amount of air coming into an engine should provide a large number of units with the chance of being competitive. In the restricted 1600 Formula Three days that preceded the present category the pushrod Renault was developed into a winner and today Renault are developing their 2-litre SOHC unit, but with only the beginnings of success so far.

Looking at the Dolomite Sprint engine in the back of the long March chassis, a monocoque with a tubular frame engine bay starting behind the cockpit, the immediate impression is of the sheer hard work that goes into making 165-168 b.h.p. from an engine that will give well over 200 b.h.p in carburated British saloon car championship trim!  Presumably the "hole" was adopted with the best of intentions, running along the lines of reducing cost along with engine r.p.m.

However, the competitive spirit lives on and the result is a series of restricted engines that peak between 5,000 and 6,000 r.p.m.!  Those who have driven both Toyota and Triumph comment that it is now hard to tell the difference between the two. "The Toyota just gets there quicker, that's all," was the consensus of opinion. Now the Triumph engine seems superior at the top end, pulling better from its 5,800 r.p.m. power peak, but it still slightly lacks acceleration.

The work that goes into the Triumph motor at SRE includes installing a fuel injection system with cast magnesium inlet manifolds (now cut to optimum length in the light of track experience) using a Bosch-Kugelfischer injection pump and the mandatory air restrictor.

The cylinder head has the inlet ports opened by 1 mm. and a lot more metal removed from the exhaust ports. The valve head sizes are as standard but the valves themselves are specially manufactured. SRE machine specially valve-pocketed versions of Cosworth piston forgings leaving the compression ratio at 12.7:1.  I thought that was high until I gazed at the FIA Yellow Book for 1978 and found that Holbay had been talking of 13.6:1!

The cylinder block has over 11 lb. removed in external metal but carries a standard bore and stroke of 90.3 mm. by 77.8 mm.  The connecting rods are the production pieces, which are lighter and have been heat treated for extra strength, as is the Tuftrided crankshaft which is polished and balanced. The crankshaft bearing type cannot be altered.

The engine mounts via three points at the front and via a four-bolt attachment to the gearbox, the installation designed by March with the motor vertically mounted rather than at the production slant. This created some water circulation problems in the cylinder head. Extensive work went on to try to solve the trouble (there have been three water feed positions on the head alone) and the provision of a new head gasket with revised water channelling to ensure that the number four cylinder closest to the driver (and therefore receiving the least air) was adequately cooled.

The valve gear and camshaft profiling are the black magic areas of F3 still, and the more so when you have a single camshaft operating 16 valves. I was told that the timing and full lift points were very much as for standard racing practice. Though the orifice alters things so that peak power is developed at very low speed for a racing engine, valve lift, and plenty of it, is still important: so development continues at SRE on this subject.

The car I tried was Tiff Needell's regular mount.  A March 783 cost £7,450 plus VAT at the time of delivery, which was Febuary of this year. This included a complete rolling chassis with Hewland gearbox. Because there was some question of the car flexing with the side-engine mountings of last season, March did the engine mounting themselves, very much along Toyota lines though the Cosworth oil pump mounts in a different position and clearance has to be provided.

Literally thousands of test miles have been completed (usually in a spare 773 which has the current rear half attached)  testing both the chassis and the engine. On the engine side the problem was the usual one of good power given on the dyno which was not translated in track performance. Sessions at Goodwood provided the optimum exhaust and inlet lengths.

The team also worked through a 2,000 mile 24 point programme with March on the car itself adopting eight ideas including the so-called Schenkel rear wing design. This looks much the same as usual but offers more downforce with less drag than the production arrangement. Looking at the kind of things this team and others have done with their production Marches, one gets the impression that you only start off with the basic car and then work toward an ultimate, rather than simply race the ex-factory product.

During the year changes to the car have included: lightweight and thinner gauge glassfibre for the top of the cockpit section and the nose; deleting the rear engine cowling; using aircraft specification K-nuts (at nearly £1 each!); utilising a lighter battery cable; a 22 gauge steel exhaust tailpipe system and the use of lighter but  stronger roll-over hoops. These moves took some 30 lb. off the car, which is now some 5 lb. or so above the minimum specified weight. Weight has not dictated the use of Speedline wheels (8" wide at the front to" rear) which are used for their strength in a category where collisions with kerbs and other competitors are commonplace. Handling requirements also were behind the use of a new U-piece short rear anti-roll bar and the change from Bilstein to Koni shock-absorbers, easy adjustment being the key here.

During all those development miles most types of radiator layout have been tried; now they use the standard front-mounted item with a bleed to the header tank. No oil cooler is fitted at all for the one-gallon dry sump layout, so it is not surprising that the oil normally runs at some 90 degrees while the water is more usually at 75 degrees.

Also tried during development were the fashionable skirts, but since no difference could be detected those were abandoned, though Chevron did use them in the lesser formulae. Next season both March and Chevron have ground-effect wing cars scheduled for the category, the March expected to be some 6" shorter in wheelbase than the rather lengthy 783 configuration.

The 6' plus Needell was not expected to leave a very comfortable seating position for one some 4" shorter, but by wedging the wrap-around Needell de-luxe bucket seat (finished in fashionable racer's tape and luxurious matt black glassfibre) into a more upright position we were able to enjoy the best seating position encountered for many months.

Conditions were good for the test, an overcast but dry day at Goodwood.  Sensation of the day was that the old wooden pits were giving way to a new brickwork ediface financed to the tune of £2,500 by Esso.

Taking care not to crack the smart tinted perspex around the cockpit, I found the March easier to get into than expected.  The three-spoke steering wheel carried a master ignition on/off switch, so starting was a simple matter of clicking that down, pushing the button and depressing the throttle slightly until the warm engine caught. The older type of Smiths chronometric tachometer flicked up the reset telltale (a maximum of 6,000 r.p.m is recommended), information supported by two smaller combined dials. That on the left provided fuel pressure (40 lb.) and water temperature (some 70 degrees on a November day). On the right we were informed of oil pressure, which was as mentioned earlier, and lubricant temperature, which was well down at 65 degrees.

The right-hand gearchange controls five ratios arranged thus:  R-2-4 (on the upper plane), 1-3-5 (on the lower plane).  Working with David Price's gear ratio chart afterwards I found that the first four ratios provided 54 m.p.h.; 80 m.p.h.; 98 m.p.h. and 115 m.p.h. Needell was reaching 138 m.p.h. at 6,100 r.p.m. in fifth gear going into Forwater with the wind behind the car.

Keeping the engine running and selecting a gear without stalling is just as difficult as expected in a conventionally highly tuned car. The clutch is very sudden, but on my first departure beginner's luck took me away cleanly, and through the subsequent changes. Later on I did miss gears, but generally it was a delightful change that responded better and better as you grew more confident and demanded more speed. Goodwood is a good test track in many respects: you find you have to use the gearbox from second to fifth; there are many different types of corner and there's a lot of space for learner journalists!

The view ahead is dominated by the large roll over hoop and an orange oil pressure warning light, but you soon grow used to their intrusion on the sight line, and the disconcerting flapping from the lightweight cockpit top-section.

From the vantage point of a single-seater Goodwood has quite distinct crests and bumps on the corners. These are quickly assimilated and then the true nature of the course and the car can be assessed.

On a flying lap you emerge from the chicane and (if you are Needell) the car is accelerating from 70 m.p.h. It's quickly into third, fourth and fifth before Madgwick, which is taken on full throttle in fifth, exiting at some 5,400 r.p.m., which represents fractionally over 120 m.p.h. Then there's the long run to Fordwater past the fast right-hand kink. Fordwater is also a fifth gear corner, with the instant demand that you should re-cross the track to be lined up for St. Mary's in fourth gear. This left falls away sharply, but the Sprint engine should be pulling 5,300 r.p.m. (112 m.p.h.) to rush you uphill briefly searching for the left of the track in preparation for the downchange to third and the 90 m.p.h. plus righthander of Levant with its two apex layout. On the test day there was a stronger headwind growling down the Levant straight which restricted maximum r.p.m. to some 5,900 r.p.m. before changing down from fifth to take Woodcote (also over 90 m.p.h.). You then sprint to the chicane (built in 1949 so it was an early example of concern about speeds past the pits) which demands second gear.

That is how it should be done. I used the same gears and had twice the excitement at half the speed! The slower section I could relate to and here we found that my r.p.m-readings were very similar to those expected, but on the fifth gear corners it takes a long time to acclimatise to the sheer cornering power offered by a current F3 car. The steering is quite heavy for such a small capacity racing car with its skimpy body, but the responses fed back through the wheel monitor every nook and cranny of the track. The Goodyears hang on beyond belief for what look to be such a slim outline by today's standards. When they do let go it is usually progressive. I did spin the car once coming out of the chicane, but that purely because I applied too much throttle too soon, having enjoyed the gently sliding sensation of previous laps.

The car turns into corners very well indeed and the long wheelbase (or was it just psychological?) seems to allow great margins for error in line without reacting sharply. On the faster corners the car settles in very steadily, but the speed at which it hums along the straight and deep into corners before the brakes are needed are the hard things to describe. Over 130 m.p.h. does not sound much today, but when you are so close to the ground and the wind is trying to pluck off your head, it's more than enough. 

The brakes are simply superb. Such an all-disc system (inboard at the rear) is well proven today and in such a light car with such grippy rubber they are the most impressive attribute. They are far better than my nerves, though on one occasion I did manage to hit the bump on the way into St. Mary's and brake at the same time, which destroyed any semblance of chassis balance and gave me a good look at the verges before I was able to steer the car gently back on course at much reduced speed.

The engine does need to be kept between 5,000 and 6,000 r.p.m., but it will pull well enough from 4,500 r.p.m if needs be. The acceleration from rest is very good indeed and with such a narrow power band you just keep changing gears, savouring the sideways squirming that marks a 3,000 standing start. Once you get up to 5,500 in top progress tails off, and the fastest reading I had was some 5,900. The conventional wishbone front suspension and parallel link and arms provides both the stable roadholding and handling described and a ride that would have been considered luxurious by the standard of sports cars some years ago. Even the engine noise hardly intrudes on your conscious thought stream. Mine, at least, was pre-occupied with tackling the next corner in the right gear.

Although I settled into consistent lap times for the last four of my eighteen laps (which actually included two to the same hundredth: as Tiff said, "You made the same mistake but in a different place!") the gap between "us" and them grows at this point. My best was 1 min. 25.23 sec. for the 2.4-mile circuit while on that day Needell set the car up at 1 min. 16.65 sec. The point is illustrated by the double wins scored with convincing style by former Unipart F3 driver Ian Taylor when he ventured into Sports 2000 during 1978. Formula Three is still breeding fast young men and fulfilling its purpose albeit at a price far beyond that conceived by what were quite a sensible set of engine restrictions.

It remains only for me to thank David Price for allowing me so many laps in a car that was destined to race in the televised Thruxton meeting a few days later. Unipart, Price John Dunn's SRE concern ought to stay together to reap the benefit of the fourth season of the 2-litre Triumph engine. There have been enough changes over the years, though success will be even harder to achieve next year as the Europeans will be coming in greater force to swell the somewhat thin grids that characterised F3 in Britain this season.

However those grids did contain quality, and that is still the name of the game in Formula Three. -- JW