Cooling fans.

Fan FAQs.

General Operation.

Let's cover how to keep the stock system together. You should be able to keep the belt on up to 6k RPMs with the following setup:

HiPo Operation.

For higher RPM use, you start to run into the limits of the stock system, but there are some things that can help:

There have been a number of very informative posts on the technical specs of the Corvair's main cooling fan on Virtual Vairs. I've condensed a couple from Rad Davis and Ray Sedman to give you an idea of what the requirements are:

The Corvair engine design requires approximately 18 cubic feet of air per minute (CFM) per indicated horsepower and a system pressure of 7 inches of water at 4,000 rpm (see SAE 140C "The Chevrolet Corvair"). This works out to 80 hp and 1800 cfm on the '60, which represents a 20% safety factor. Note, that the temperature curve is quite close to the indicated mean effective pressure (IMEP) curve. The peak is about 2,600 rpm. At this rpm, temperature will be 30 degrees F hotter than at maximum speed. If the power output of the engine is increased by shifting the IMEP to a higher speed, the same temperature levels will follow. No change in fan or speed ratio is required to cool the higher output engine, however there is the problem that the fan starts to stall above 4K rpms (oil temperature is a different beast).

The GM Stock Engine Test Reports (reprints available from Clark's, "GM Restricted") show fan power consuption curves as part of Standard Test 1. Airflow over the engine was also measured, and is cited, as were delta-t and plenum pressure. These data are shown on the panel for Standard Test 9a. Note that these are observed values, not theoretical.

 1960 fan 1964-69 fan
RPMhpcfmcfm/hphp cfmcfm/hp
1k.44501125.1 3803800
2k1.09509501.1 740673
3k4.013503373.5 1115319
4k9.018002008.2 1500183
5k    15.21680*111

* = Note: the airflow curve is starting to flatten out on the 140 test. The HP demand curve isn't, though.

Allowing for graphical error, the 1960 fan is always more efficient than the magnesium fan, at least to 4K rpm. The one number I have for the 61-63 fan is 1460 cfm at 4k rpm. This would make it pretty close to identical to the magnesium fan as an air pump. At 2-3K rpm, where most 2-carb engines live, the '60 fan is functionally the best. It does have shortcomings of noise and belt jumping problems, however.

But I think these data make it pretty clear that a) any stock fan starts to suck up a LOT of hp at rpm above about 3500, b) an axial flow fan's linear power demand curve is really much better for an engine with a wide rpm range like a 140 or turbo.

In another post, Andrew Berg asked:

I'm not much of an expert, but do Corvair engines really generate that much more heat than another engine? I know that mine (when it's running) didn't seem to.

And Rad replied:

No. They generate the same amount of heat as any other engine of similar volumetric efficiency, combustion efficiency, and horsepower output. The critical difference is that the medium used to carry the heat away from its point of generation is much less dense than the water/glycol mixtures commonly used. Water pumpers actually have a worse time getting rid of the heat to air, because the difference in temperature between air and heated surface at the heat exchanger is much less than it is with a Corvair.

The major difference in requirement as far as airflow is concerned is not the total mass airflow, but the pressure required to push that mass of air over the cooling surface in the required amount of time. Water pumpers use a radiator with a pretty low pressure drop at rated airflows, which means that ram air is sufficient above about 30 mph, and an auxiliary cooling fan need only move the air against a slight pressure gradient at speeds below this. Corvair engines are in fact pretty much self-cooling at speeds above 30 if properly ducted (4-inch dryer hose from the headlight buckets works nicely). Unfortunately, there are circumstances in car operation where the car simply doesn't go fast enough long enough to use this "free" cooling. The stock early model air ducting is adequate; the stock late-model is much more optimized, but both are compromises between the need for interior space, clean dry air, lack of noise, and as much pressure differential across the engine as possible from vehicle motion.

At steady state cruise at, say, 45 mph, a Corvair engine in a coupe or sedan is pretty much wasting the power used to turn the fan, provided that the various fins, shrouds, and seals are where they should be and doing as they should do. But who actually spends a majority of their drive cycle at 45 without speeding or slowing? Convertibles have a less optimized plumbing arrangement, and suffer considerably with the top down. Forward controls are actually pretty well plumbed for cooling air, but generate more engine heat because they're pushing a heavier vehicle with more frontal area.

After a question about spring loaded idlers vibrating, Craig Nicol wrote this:

Here's what causes the belt to deflect and causes the idler to react: The crankshaft does not rotate at a constant speed. Three times during every rotation of the crankshaft, one of the pistons radically accelerates the crankshaft and pulley as it passes through its powerstroke.

The belt is somewhat elastic and the combined mass and load from the fan and generator don't react immediately. Instead, the belt stretches tight on the driver's side as the crank tries to accelerate the fan and generator and it correspondingly goes slack on the passenger side of the engine. That's why GM installed deep groove pulleys and belt guides to keep the belt from jumping out of the pulley on the slack side.

Engines with strong idle power pulses, lighter flywheels, heavier fans, higher generator loads, and more elastic belts will all contribute to increased "flapping" of the belt at idle speeds.

Fan variations.

Here's a thread that started when Mark Sutton asked a question about lowering the power used by the fan:

Mark -> Powerdisc installed under fan, this is a round thin disc of aluminum placed directly below the fan. It restricts the airflow at higher RPM, airflow at idle isn't affected. I have used one for 3 years without any problems, it works well and doesn't add much mass to the rotating assembly.

Bryan -> Pro: Simplest installation, easy to change your mind. Stock belts fit. Does the job, probably the best setup for a true dual purpose car that sees a lot of street miles (no worries about being able to get the right belt).

Con: Doesn't help the belt tracking (tho the stock setup can be made to stay on unless you get over 6k a lot). No tuneability.

I recently read a variation of this, from Bruce Carlton of the FastVair list:

In fact if the car has a chance to cool down between runs, most autocross courses do not even require fan cooling. Fans are frequently required by the rules however, but such things as an aluminum plate with only a small opening over the fan to block most of the air to the fan during competition will substantially increase available power to the drive wheels. You can remove part or all of the plate after the run to quickly cool down the engine.

Mark -> Variable ratio fan pully to slow fan at all engine speeds. I have one of these on the shelf but haven't tried it.

Bryan -> Pro: Good solution, especially with a shorter geared car that spends a lot of time above 2500 rpms. Should help some with belt tracking if installed properly, since the fan has less inertia, but this may be offset by the greater pulley weight. Most tuneable setup. Reasonable installation.

Con: Requires a non-stock belt length. Some folks have had problems with getting the pulleys aligned, and poorly aligned pulleys eat belts.

Bruce Schug -> I have run the Otto pulley for sometime and have never had any trouble with it. I use the washers they supply to shim the alternator and the idler. I have used it with a stock idler and with various spring-loaded idler pulleys. I have never had any trouble with belts coming off.

I have also heard others say they had trouble keeping belts on and don't know why. I use Otto belts usually, long or standard. I have used a standard Clarks belt. The only critical thing I have found is the following:

I use the Otto "long" belt when the pulley is on the "slower" settings. This works on the two or three slowest settings only. To use the "fastest" setting, you can use a stock-length belt, but the idler is awfully far forward. It would be good to get a belt with a length between the stock belt and the "long" belt. This should work better in the "fast" settings.

I have found that in the summer you MUST use a stock pulley and belt when using AC. In the cold weather, I use the "long" belt and either the "slowest" or "next- to-slowest" settings. Faster warm-up, more power. You have to be a little careful, it's one thing to set it for use around town, another to set it for highway use. For instance, I may have it set on the "next-to-slowest" setting for use back-and-forth to work and around town in cold weather. Then, I may be going out on the highway for 20-30 miles at 70mph, perhaps it may even be an unusually warm day. The head temp guage starts to climb, then the oil temp. I have never had a serious problem, I have enough cooling I guess, but you do have to watch the gauges.

Mark -> Cutdown cooling fan, this as the advantage of also reducing weight and mass the engine spin up and down. Probably increases belt life too.

Bryan -> Pro: No doubt the best setup for a dedicated competition car, lightest rotating weight. Uses stock belts, guides, etc.

Con: Most hassle to install or remove, no tuneability, probably most expensive.

Al Kidd -> Prior to switching to a water pumper I ran a cut down fan on my 140 for autocrossing. Never got hot and you could feel the extra horses in the seat of your pants where it matters. I mistakenly left it on at Lake Placid when I ran the rally and the corsa temp gauge showed Low 400's while climbing the mountains.

The car was a daily driver and I was nervous running the cut down fan in case I got stuck in some of Toronto's great traffic jams. I didn't run it during the heat of the summer so in order to make fan changes easy I took an old top shroud and ran a jigsaw around the existing hole until it was large enough to pass a full size fan through it. I than took another top shroud and sliced the sides off. The two pieces looked like a regular top shroud when they were together. Fan changes were easy as all I had to do was remove my customized top to gain access to the fan.

Road Racing Setups.

Alternator warning light.

Brian O'Neill wrote a really good question that made the point that on a race car, you might not notice the stock Gen/Fan idiot light, or perhaps you don't have the stock dash any more. Since you should shut the engine down right now if the belt comes off, here's a way of setting up a brighter light, courtesy of Larry Claypool:

gen light? not too much rocket science here. brown wire from the voltage regulator goes to one terminal of your new 'big' light. the other side of the circuit is power from the ignition switch, typically pink or black with a pink stripe. the trouble is that most marker lights have only 1 lead, the base is connected to ground. so you need to isolate the base so it does not ground, then connect the remaining wire to where the ground would normally be made on the light. or you could make own light, go to the parts store and get a socket with a base configuration like a corvair back up lamp bulb (1156), with dual pigtails. use a #1004 bulb. it has 15 CP, and uses the two leads to operate (does not use the base of the bulb for ground). larry claypool

BTW, this is a late model with an alternator. If you have a generator, the wiring may be different.

Warren's Fan Experience.

Warren Leveque wrote this post to FastVairs about his experience cooling his race car:

When I was road racing in D/P I tried many cooling innovations. One was directing headlight air into the fan via 8" ducts. This overloaded the fan and shredded the belt. Then I ducted the same air into shoe like ducts directly over the heads, this worked well at speeds over 60 mph (2" H2O press.) and released much power for straight a way speed. Two very slow pace laps hot soaked the engine and ruined it. Some racing Porsches don't even put fins on the cylinders, so I didn't worry about air to them. The centrifugal fan does a very good job of flinging air to the extremities which is where the head fins are. Electric fans and Porsche fans only build up a uniform pressure head and are less efficient in air distribution.

My most successful cooling was a half speed full size fan to make the most efficient fan speed occur at an engine speed of 7500 rpm. The only reason fans were ever cut down was because they cavitated at these rpms, not to reduce horsepower loss. Cavitation releases lots of horsepower. The half speed was accomplished by reducing the drive pulley diameter and increasing the driven pulley to get the same overall circumference and use the stock belt. This required realignment of everything, and of course at slow speeds it was very inefficient. 450 degrees F. was considered good at race speeds. By slowing down the belt by half speed pulleys additional power was saved by running the belt at half speed as well as the fan. I discovered the importance of the belt horsepower while fooling with electric motors to drive the stock fan. Even a single plane belt would stop the drive motor if tightened much at all. Imagine if it had to make two right angle turns.

I got a photo of Warren's engine compartment at the 2001 Corsa convention, he's got the half speed fan on it here:

Note the flange on the top shroud, this mates to the duct to the flaps on the Yenko decklid Warren refers to below.

The hp disc was an autocross only item designed to virtually block air flow at high rpms (enforced cavitation?) and flow naturally at low rpms.

My current system on my 6000 rpm Turbo (15 psi high heat producer) engine is to use the full size, full speed fan, a silicone coated [use silicone spray lube] wrapped belt and a spring loaded idler with a 1" wide pivot. Since the side sheet metal is removed the hot air recirculation is prevented by ducting from the Yenko flaps directly to the fan inlet. This 2 pound item replaces all of the side sheet metal. All belts with the same part number are not the same length. The belt must leave the damper to the idler in a straight line. There is 0.5" H2O press at the flaps. When boost exceeds 8 psi, I inject water mist into the fan inlet and the carburetor. I have run for hours at 350 F. I have a huge AC evaporator as a front oil cooler, never exceeds 220 F.

I tried electric fans on my mid engined cars. The air came up faster than the two fans could push it down. As these mid engine cars are autocross only cars, I use the fans for cool down between runs.