Fair warning. This will be a long article spanning two web pages (several print pages), perhaps hard to read (without nodding off) or hard to follow at times. It may also be tough to write, as some of it borders on mediating a religious war, similar to brake fluid arguments. There are some misconceptions about how a cooling system works, and some of that is commonly touted in public as fact. For instance, it is often claimed that a lower temperature thermostat may eliminate overheating, which is bunk. It is often said that coolant flowing too fast through a radiator will reduce efficiency of the radiator and cause overheating, which is also bunk. While explaining how a cooling system works, I intend to de-bunk or at least de-mystify these ideas, and I hope to do it in layman's terms that anyone can understand. If at the end of this lesson you think "I knew that", then this would be a successful effort (at least if you remember it).

The basic concept of the liquid cooling system is generally understood, that there is excess heat generated in the engine that must be expelled through the radiator, and that heat is carried from the engine to the radiator by the flowing liquid inside. But from there on the details are often neglected in favor of quick and easy impressions and opinions. Understanding the details should allow a person to deal more effectively with cooling problems, finding workable solutions in minimal time with minimal expenditure of funds. This first page will deal with how engine operating temperature is controlled within a fairly wide but acceptable range. The second page will deal with how the system copes with more accurate regulation within a much narrower temperature range.

Let us start first with an irrefutable law of physics, the conservation of energy, as it applies to a car engine. Energy is not created or destroyed, but is converted to various forms and is moved from place to place, and it always exists somewhere collectively in a constant amount.

An internal combustion engine is very good at converting chemical energy to heat energy. When gasoline and air is burned with proper mixture ratio, about 14.5 parts of air for each part of fuel by mass, the result is very near to 100% efficiency of the conversion in making all possible heat with minimal fuel remaining. Unfortunately the same internal combustion engine is not so good at converting heat energy into mechanical energy. Typical thermal to mechanical conversion efficiency is in the range of 25% to 35%, with 50% being the current world record (but not in a car engine).

The remaining 65% or more of the available heat which is not converted to mechanical output has to be expelled or disposed of in some manner. Much of it goes out the tail pipe as heated exhaust gas. Some of it is radiated away from the exhaust system and other engine parts as infrared energy, while some more is carried away by thermal conduction to surrounding air, with thermal convection or forced air ventilation moving cool air in and heated air out. For an air cooled engine that may account for all of the waste heat being expelled. For a liquid cooled engine any excess heat still remaining in the engine must be carried out to remote radiators by working fluids, there to be transferred into the surrounding environment. For some cars an oil cooler may do some of this work, but that is similar in nature to the primary liquid coolant radiator. For this article I will discuss only the primary liquid cooling system, and leave the relationship of an oil cooler for some other day.

An engine running at idle speed with no load will be consuming the least fuel and generating the least heat of any mode of operation. Modern engines with a coast down fuel cutoff mode will be ignored (very short term condition anyway). Any operational mode with increased throttle or engine speed will consume more fuel and will generate more heat. The most heat will usually be generated with highest throttle and highest engine speed (the condition of greatest fuel consumption). Thermal to mechanical conversion efficiency may be highest near the peak of the torque curve, but variations in efficiency through other operating conditions (although substantial) are nearly irrelevant compared to the total heat being generated.

Here we are dealing only with the problem of excess heat that has to be expelled through the radiator. In fact 100% of all heat not converted to mechanical energy must be expelled somewhere, and exactly that amount, no more and no less. If too much heat is expelled the engine would eventually become quite cold. If too little heat is expelled the engine would eventually overheat. So the function of the liquid cooling system as a whole is to precisely control and expel only the exact amount of heat necessary to keep the engine operating in thermal equilibrium, at or near a fixed desired temperature.

We like to think that this temperature control function is performed by the thermostat, and to some extent it is, but only within a very narrow temperature range roughly plus or minus five degrees Fahrenheit (or plus/minus three degrees Celsius) of the desired target coolant temperature. Below that range the thermostat is completely closed, and its only active participation is to completely stop coolant flow until the fluid temperature might happen to come up within the active control range. In extremely cold weather that may never happen, and the thermostat might never open. Above that ideal range the thermostat is wide open, and its only active participation is to permit free flow of coolant until the fluid temperature might happen to come down within the active control range. In very hot weather or hard working conditions that may not happen, and the thermostat may remain wide open for long periods of time. Where a thermostat is removed to allow maximum flow for maximum cooling in very hot weather or hard racing conditions, the system will run constantly in the free flow mode, and yet may never fall to be too cool. At such times the cooling system is self regulating at thermal equilibrium without the intervention or assistance of a thermostat.

As a whole, the combined engine and cooling system will always expel exactly 100% of the (excess) heat being generated (with some short term fluctuations). In the absence of a thermostat, an engine expelling heat too quickly will drop to a lower temperature until it expels less heat to come into equilibrium. As long as it is not too cold outside, a thermostat can stop coolant flow through the external radiator to reduce heat loss until the coolant temperature comes up to the desired operating minimum. In this manner a thermostat can control the minimum engine operating temperature. An engine with radiator expelling heat too slowly will rise to a higher temperature until it will expel more heat to come into equilibrium. If the coolant temperature rises above that narrow operating range of the thermostat, it is beyond control of the thermostat, and the only way to bring the temperature down (without reducing heat generation) would be to increase the cooling capacity of the radiator. That problem is covered in CO-101 to the extent that you can optimize the existing hardware. Beyond that effort, the only remaining solution may be to increase the size of the radiator. When the radiator is large (larger) enough to bring the coolant temperature down to within the operating range of the thermostat, then and only then can the thermostat once again control the minimum coolant temperature.

So for the engine to maintain the coolant temperature within its desired narrow control range the engine (without radiator) must expel less heat than it generates (or generate more heat than it expels), and the engine with radiator combined must be capable of expelling more heat than is ever generates (over cooled). In reality many of our Little British Cars [obligatory MG content] can easily accomplish the first task (no problem), but often fail in the second case (yes, the radiator is too small for that). This is where the first proposition above is debunked. Installing a lower temperature thermostat will not prevent temperature rise in this case, because any thermostat would be constantly wide open (coolant temperature beyond the control range of the thermostat) regardless of its temperature rating. In essence, the thermostat only controls minimum operating temperature within a narrow range, and has nothing to do with maximum operating temperature. The only way to keep the maximum temperature down in the same small range as the controlled minimum is to have such a large radiator as to be constantly over cooled, even with the hottest running conditions.

This higher running temperature is not necessarily a problem with our LBC's, but should be viewed as a "design feature". In other words, it's just the nature of the beast, so get used to it. Lots of cars used to be built this way, keeping the radiator as small as practical to reduce weight (and cost), especially in a sports car or race car. As the engine runs hard and generates more heat the coolant temperature rises above the controlled minimum. As long as there is liquid coolant inside which does not boil, this is generally not a problem. The "normal" or acceptable operating temperature range then is anything between the minimum thermostat rating and the top end of the temperature gauge scale, as long as the coolant never boils. Fortunately our vintage MGs run at quite moderate temperatures compared to much newer cars. Note that the MG (A, B or Midget) temperature gauge scale only goes up to 230dF (110dC), so running between 180dF and 230dF (50dF total range) might be reasonable.

Elevated coolant temperature at slow speed and low load is more of an air flow management problem (which can usually be corrected). Vapor lock in the carburetors at very low speed on a hot day is a carburetor cooling problem, not a problem with the liquid engine cooling system.

Newer cars on the other hand commonly push the temperature much higher by intention, partly for fuel economy and partly for emissions control. Through use of coolant fluid with high boiling temperature and a radiator pressure cap with higher pressure capability it is possible for some newer cars to have coolant temperature controlled to around 250dF or even up to 270dF MINIMUM. But when it runs that high, it is more important to not allow it to go much higher, so most newer cars will also be over cooled (relatively speaking). As such, the maximum operating temperature might never rise much higher than the controlled minimum under any circumstances, resulting in a very narrow total temperature range to be considered "normal". Once you get used to that, a 220dF temperature gauge reading on an MG might finally impress you as being nicely cool.

Please turn page for continued cooling theory.