ALFAS AND PERFORMANCE IN THE NINETIES

by Yves Boulanger

Performance improvement has always been a popular topic. Before going any further, I'd like to take some time to discuss a few basic concepts that are widely misunderstood. Here are a few definitions:

Force: is an action done by one body on another body in order to do work. Force units are pounds or Newtons.

Work: not the 9 to 5 kind but rather the mathematical quantity that physicists use for dynamics calculations, is defined as follows: if a force F is applied on an object, and the object is moved over a distance L, then the work done is equal to F times L. Work units are foot-pounds or Newton-meters.

Torque: is a force applied on a rotating lever. It follows that the work done by such a mechanism is equal to the torque times the angle of rotation. Torque is then also an action in order to do work. Torque and work have the same units because angles are defined as non-dimensional.

A horse pulling a sleigh along a slope is doing work; a horse pushing against a stone wall is not doing any work (unless the wall is moving). Also, please note that time, or speed, have nothing to do with the concept of work; whether the horse has pulled the sleigh along a given stretch of road in 5 minutes or in an hour, the same work has been done.

Energy: an object acquires kinetic energy as it gets moving. The amount of kinetic energy that an object possesses is proportional to its mass, and to its velocity squared. The SI unit of energy is the Joule, but the Kilowatt-hour (such as on your utility bill) is another unit of energy.

Power: is the rate at which energy is supplied to an object: energy divided by time. The SI unit for power is the Watt. 776 Watt = 1 hp.

In a car, the engine supplies both power and torque every second the car is moving. Both quantities, at a given rpm, are linked mathematically: power is equal to torque times engine rpm. In traditional units, lb-ft and hp, the exact formula is:

Power = Torque times rpm / 5252

Since power is the rate at which mechanical energy is supplied through the transmission to the tires and then converted in kinetic energy (or speed), power is thus necessary for the car to accelerate. The more power you use, the stronger the acceleration will be.

But wait! I didn't say peak horsepower. During the acceleration process, unless you have a continuously variable transmission that keeps the engine at peak rpm all the time, the engine speed will vary from low rpm as you get underway up to the shift point, then down somewhat depending on your gear ratios, up again to the shift point and so on... It follows that the total amount of energy supplied to the car is the sum of the little amounts of energy supplied during each second of the acceleration process. For those who remember something about calculus, this is the integral of the power vs. time curve. In other terms, the area underneath the power vs. time curve represents the energy supplied. The broader the area, the stronger the acceleration.

We can already draw some observations:
- Peak horsepower is only one point of the curve. It only matters if you use it. Say your engine produces 150 hp at 5800 rpm; if you never shift above 4000 rpm, you are never using your 150 hp. Midrange power has a lot more importance for everyday driving.
- Gearbox ratios are important to make full use of the available power. While the ratios in a racing box can be tailor-suited for each curve of any specific race track for any other application you must depend on midrange power.

However there is no direct relationship between peak power and midrange power. This depends on engine design.

What about torque? Earlier, we said that power and torque at a given rpm are linked by a mathematical equation. The torque curve is fully dependant on the power curve, and conversely. In general, car makers publish data for peak power and peak torque, along with the corresponding rpm's. What the peak torque figure is, along with the corresponding rpm, is really a measure of midrange power.

Torque at low rpm's is the measurement that will indicate if a given engine will accelerate a car smoothly from low rpms, independent from the strength of the accelerations. I briefly owned a 2.5 liter Chevrolet Celebrity that wouldn't accelerate in second gear from anywhere below 10 km/h; it would just kick, buck and stall. In traffic, it was enough of a pain to disgust anybody from driving a car with a manual gearbox. The 1.4 liter, 65 hp Renault that replaced it was a wonder in comparison.

If you had two engines with equal peak horsepower, at the same rpm, but one had more peak torque than the other, also at the same rpm, then the first engine would have the broader midrange power curve. This would also be true if both engines had the same peak torque value, but at a lower rpm on the first engine. In either case, the engine with the most midrange power will provide better all-out acceleration, and will require less gear-rowing to keep up with everyday traffic.

Alfa designers of the sixties and seventies fully understood this. In fact, the Alfa 2000 engine had the most torque of any 4 cylinder on the market when it was introduced in 1972. Even a 1300 of the same period is more flexible than a lot of modern cars.

In practice, a street engine should provide peak torque in the 3000 rpm range. Modern car manufacturers seem to have gone crazy in a race for peak horsepower, and even your father's Buick can have an engine that provides peak torque at over 4000 rpm. Road testers of the latest full size Buick, for example, complain about "lack of power" and "noisy engine". People who drive that kind of car hate to hear the engine revving.

So far, we've been discussing power, and acceleration. The performance driver is also concerned with two more issues: top speed and handling.

Top speed is a function of aerodynamics, gearing and peak power. As the car gets underway, rolling resistance builds up and eats the power supplied by the engine. This comes from friction in the wheel bearings, drag from the brake pads, tire distortion, but mostly, air drag. Power required to overcome air drag increases to the third power of velocity. Top speed occurs when the total resistance equals available power. Ideally, this should happen at a speed where the engine runs close to peak power rpm. However, in some cases where the designer incorporated a ,fuel-saver" 5th gear the engine will be revving so low that peak power rpm will correspond to an absurdly high speed - my 1.4 Renault ran 2400 rpm at 100 km/h. which means that at the 5300 rpm peak, it would have done 221 km/h! Of course, it never did, but this car (and many others) reached a higher top speed in 4th than in 5th.

I do not intend to go into the details of how to increase the power output of your engine. For the time being, I will direct you to Jim Kartalamakis' book, "How to power tune Alfa Romeo twin cam 4 cylinder engines", available from Classic Motorbooks and elsewhere. However, John Hoard once told me a very interesting story about Alfa engines, and I'd like to repeat it here. For those who don't know him, John is an engineer who works in the Ford engine labs. He is also an Alfa Romeo Owners Club member and owns a GTA Junior.

Several years ago, Ford wanted an evaluation of some production four cylinder engines (Manufacturers call that "benchmarking"). Engines of all makes and sizes were bought and tested for power per unit displacement, thermal efficiency, efficiency of the filling process, and so on. John threw in the test the engine from his GTA, in stock, 110 hp street tune. This engine, which was at the time a 15 year old design, got first place for power per liter, filling efficiency and more. For each engine, the test engineers then tried to identify how much power was lost to the peripherals such as the air cleaner, muffler, etc. Each time they tampered with anything on the GTA engine, power went down. Straight out of the factory, the package was already optimal for street use. The sort of changes that would have increased its power were not compatible with its street vocation.

That's a fact of life. An Alfa engine is already a fairly sophisticated piece of machinery, and extracting more power out of it is going to be a lot harder (and more expensive) than getting more power out of a Datsun or MG engine of the same period. Besides, many budget racers have found out that the bottom end of an Alfa 2000 engine doesn't take too kindly to great increases of power, the engine block grew out of a 1954, 1300cc design and major bucks will have to be spent there to avoid ending up with a "ventilated block".

There is one idea that will work 100% of the time: If you're not racing in a particular category, and your Alfa is less than 2000cc, changing your engine to the next larger size is a guaranteed way of increasing your power throughout the range. It's the cheapest hp per dollar modification you can make.

About tampering with smog controls:
A pair of proper cams in a Spica injected 2000 GTV will do wonders for engine breathing at 5000-6000 rpm. Likewise, replacing the poorly engineered 4 in 1 exhaust manifold of mid-70's Spiders and Alfettas, and installing the Euro-spec. manifolds (or tube headers) will give your engine the power it should have had all along. But while Canadian enthusiasts have escaped the problem of periodical smog control inspections so far, they're coming, and at this stage I heavily recommend that you leave your post-1972 car in stock form if you want to be able to drive it into the next century. This is especially true of the later Bosch injected cars: this system replaced the Spica injection in order to cope with more stringent regulations. If you replace your Bosch injection with some other Neolithic induction system (such as Weber carbs) you are guaranteed to flunk a smog test in a big way.

I briefly mentioned the word "handling", and you must have wondered what this had to do with engine power. In fact, it has nothing at all to do with it, but handling is directly related to a car's mass (measured as its weight) and of course, acceleration is also directly related to mass. (Newton's law, anyone?)

Extra mass directly affects about every aspect of a car's performance: acceleration, handling and braking (however, top speed is almost independent from mass). Getting rid of extra mass is a guaranteed way of improving performance. Of course, if you're racing this means gutting out every possible piece of trim and useless equipment (such as heater, air conditioning, electric windows); otherwise, while I know you're not going to gut out your concours GTV, there are still ways to spare useless mass: get rid of your 50 lbs of tools, stop carrying your mother-in-law around, don't carry your squash equipment when you don't need to, go on a diet, etc. There is no point spending money trying to increase your power to get better acceleration when you're 10% too heavy, which is harming your acceleration, braking and handling.

GTV's and Spiders used to weigh about 2200 lbs dry. Later Spiders got really fat: 2500 1bs for a mid-80's Graduate, and (my estimate) around 2800 lbs for a dressed-up Quadrifoglio with air, power steering, electric windows, running boards, spoiler ... All show, no go. GTV6's, Milanos and l64s are also around the 3000 lbs mark. Weight, there is the enemy (1).

Which brings me to a conclusion: a (stock) Civic can run circles around a (stock) GTV. Yes, I agree that the performance of a GTV (1963 technology) can be improved, given enough time and money, and made to match that of a Civic (1995 technology). But an equal amount of effort spent to improve a Civic will produce something that will still ran circles around your improved GTV.

I think Alfas were terrific performers in their time, but are now obsolete technology. Kids with Hondas and GTI's are now looking at us the same way we were looking at MGB owners 15 years ago. Alfas should be enjoyed for what they are. If it doesn't suit you, save yourself a lot of trouble, buy something else.

About "cheap Hondas"
The value of non-Spider Alfas has collapsed to ridiculous levels in Montreal. An excellent GTV would hardly bring $6000, if a buyer came up. A decent Alfetta GT took a couple months to sell at $1500. I once mentioned to the owner of a non-running Alfetta that I wasn't sure anybody would pick up his car if he advertised it for free. And I'll spare you the running, $ 1500 Maserati.

This doesn't buy much Honda, does it? (Before anybody takes this seriously, note that 1've never owned a Japanese car and probably never will.)

(l): Ettore Bugatti

© VEA

1932 Alfa Romeo 8C Gran Sport
1950 Alfa Romeo 6C 2500 Freccia d'Oro
1962 Alfa Romeo Giulietta Spider
1963 Alfa Romeo Giulia 1600 Sprint
1969 Alfa Romeo Spider
1971 Alfa Romeo GTV 1750
1972 Alfa Romeo 2000 GTV Bertone
1972 Alfa Romeo GTV Competition
1972 Alfa Romeo Montreal
1976 Alfa Romeo Spider Veloce
1978 Alfa Romeo Spider
1979 Alfa Romeo GTV
1979 Alfa Romeo Alfetta GT
1983 Alfa Romeo GTV 6
1984 Alfa Romeo GTV 6
1984 Alfa Romeo Spider Veloce
1986 Alfa Romeo Spider Veloce
1988 Alfa Romeo Spider Graduate
1993 Alfa Romeo Spider
1994 Alfa Romeo 164 LS