Unlike a water-cooled engine, there is no safeguard: by boiling over, a water-cooled engine will force the car to stop, thereby preventing further damage to the engine. On the other hand, unless equipped with an oil temperature gauge, the driver of an air-cooled automobile will not know if the temperature suddenly goes up. Consequences of overheating is first damage to gaskets and if it gets really hot, cracked heads, dropped valve seats, and/or a warped engine case, especially if made out of magnesium (type 1). Breakdown of oil viscosity also leads to lubrication failure.
Try the dipstick test. If you can hold the dipstick right after engine shutdown without burning your fingers, then the engine is not overheating. However, the converse may not be true: if the dipstick is very hot, in some cases it may simply be because of the proximity of the exhaust system. (Could a heat shield be missing?) You could install gauges, but they are not very accurate and should only serve as relative indicators. In other words, look for changes relative to a baseline for your car/engine/gauge combo as indicators for trouble, but comparing your absolute readings with somebody else's has only limited information value. Nevertheless, one can verify a gauge's calibration by plunging the sender in boiling water (212 F, 100 C at sea level) for example. This would only verify the reading at one point, though. A better test would be to compare the readings with those of a properly calibrated lab thermometer, perhaps one borrowed from a chemistry lab. Plunge both into a pot of cold water, suspending them so they don't touch the bottom or sides of the pot. Then fire up the stove, compare both readings and write them down. Post your results to the type2 list! :)
Note that you may not find out if your engine is prone to overheating unless you drive it at cruising speed for at least 1/2 hour. That's the time it took my 1800 T4 to reach the steady state temp.
As I had mentioned earlier, since most gauges are not very good, readings may not accurately reflect the true temperature. Besides, there are many "true" temperatures, depending on where it is measured: right after exiting the oil cooler, under the oil screen, above the oil screen, etc. Ideally we should measure the temperature at the first location, as is done on the Porsche 911, but that is difficult to do on the VW engine. After listening to years of discussions on the bus lists, I can perhaps attempt to offer an answer that, I think, reflects list members' average opinions. These reading are typically made in the oil sump. (Readings reported on the Porsche list are typically lower, reflecting the fact that they're measured after the oil cooler.)
Below 180 F is too cool; is the thermostat still there?
210-250: you can probably live with it; many busses w/ type4 engines get that hot under load in the summer. (For a beetle, I'd be concerned, though.)
250-280: something is wrong and must soon be fixed; however, the viscosity of a good synthetic oil should still be fine. If it's after dark and I'm in East St. Louis, I keep on driving.
Above 280 F: better stop; unless the oil pressure light is on (in which case you must shut the engine off immediately!), let the engine run. Glen Buhlman recommends that it not be left idling, but that revving it will allow the fan to do a better job of cooling the engine.
Problems can roughly be classified as either cooling problems or as heat generation problem.
Hot summer day? (d'uh! :)
Driving in the wrong gear? Remember that for a given speed, higher engine rpm => faster cooling fan => cooler engine.
Open engine lid? Do you keep your engine lid open to "let the heat escape"? Don't. For better cooling and to help the fan (hence the engine), you need positive air pressure.
Missing or damaged seal between engine and engine compartment? Don't "recycle" hot air!
Holes in tin? Are all seals and covers there? At holes for sparkplug wires, oil pressure switch, CHT sensor (on FI models), or adjusting screw for alternator belt (type 4 engines)? Is all the tin there, correctly installed and tight? No fastening screws missing? A list member found that missing the seals for the spark plug holes increased oil temp by 10-15 F. I've heard that Corvair spark plug seals are superior to VW seals.
Missing ducts? Kim Kariotis wrote: "Be sure the auxiliary electric fan for the heater is properly ducted to the engine. There are flapper valves in the fan housing that prevent your cooling air from escaping through the auxiliary fan via the ducts. No ducts, big leak, engine runs hotter than it should."
Unobstructed oil cooler? Leaves and debris can be sucked in through the fan and cover up the oil cooler; if the cooler is oily, dust will stick to it, impeding air flow and acting as an insulator.
Chrome valve covers? Valve covers are in contact with oil and work as oil coolers, *provided they're black*. Chrome valve cover were meant for arctic climates.
Dirty engine? The goop under your engine acts as an insulator; scrape it off.
Type of oil: air-cooled engines are also oil-cooled and some oil brands are better than others at picking up heat in the engine and shedding it in the oil cooler. A credible source (Ed Hackett) told me that synthetics are better in that regard, although some list members did not share that belief based on their personal experience. Matthew Chapman writes: "Never Use Synthetic Oil in an air-cooled motor. The oil temp goes down but head temp goes up. Synthetic oil rejects heat. The air-cooled engine is air and oil cooled. if the oil rejects the heat heads will not cool properly. This was tested and proved by gene berg ent. in orange ca." I have also heard that Berg's findings might no longer hold for newer formulations. Besides, synthetics are much better at withstanding high temps, hence should provide added protection for overheating engines.
Amount of oil: too little ain't good; too much ain't good either. Dennis Haynes says to keep it halfway between the marks on the dipstick. If there's too much, it gets beaten into foam, causing a drop in pressure and less cooling.
Timing: Check advance at idle and check total advance, too. Compare with the specs given in Bentley. (It is still not clear to me if heat goes up steadily with advanced timing, or if the only significant increase in heat is the jump at the detonation threshold.) It is generally said that 29 deg total advance is safe and that you start pushing your luck at 32 degrees. This only applies to type 1 engines, though. Stock distributors on the Type 4 engines will occasionally advance past 40 degrees.
Running lean: Adjust carbs or, if you have fuel injection, check for false air. False air is air that sneaks into the intake manifold without first passing through the air-flow meter; hence the computer underestimates the amount of fuel to be injected. A possible source of false air is an EGR valve that was not disabled, say, after the EGR filter fell victim to rust perforations; other sources are old, leaky rubber boots at the air intake distributor, disconnected or cracked vacuum hoses, loose valve covers, etc.
Engine friction: General condition of the long-block; crankshaft endplay is said to be a good indicator (provided, of course, that it was not reset since the last rebuild). A smooth, well-balanced engine will be more efficient and generate less heat. Retorque the heads. If you suspect a bearing is going bad, have a lab analyze your oil.
Non-engine friction: Poor aerodynamics exacerbate the fact that required work is proportional to the square of speed. Hence you need about 40% more hp to go 65 mph instead of 55 mph. Reduce loss to friction in transmission by replacing transmission fluid every 2 years. Do the wheels turn easily? Check wheel bearings, look for binding brakes. Check wheel alignment.
Valves adjusted? A tight exhaust valve, for example, will never completely close and will let partially unburned fuel escape into the exhaust system, which in turn gets extremely hot. Moreover the loss of compression in that cylinder translates into loss of power, hence requires more work from the other 3 cylinders to keep up initial speed.
Please send corrections or amendments to Sami Dakhlia