The MGA With An Attitude
THEORY OF OPERATION of the Cooling System, Part 2 - CO-200B

I suppose reading Part 1 should be a prerequisite to this page, so you know what I mean while chatting about the narrow temperature control range of the thermostat, how the law of conservation of energy is applied to a cooling system, and something about the speed of water through the radiator (which we didn't actually get to yet). Also put another concept on the table, the fact that heat can be conducted through material from a point of high temperature toward a point of lower temperature. On this page I want to discuss only what happens to heat flow in the engine and cooling system with constant operating condition within the narrow temperature range of thermostat control. Generally speaking, this is equivalent to cruising down the highway at a steady road speed with coolant temperature just a few degrees above the threshold required to open the thermostat. Then I will get into what some variations in the conditions might do, good or bad.

We start this discussion with your engine running under constant load at "normal" operating temperature, and the thermostat perhaps half way open and controlling the rate of flow of the coolant. The water pump is doing its best to move coolant through the system, drawing fluid from the bottom of the radiator and pushing the fluid into the front of the engine block. Fluid flows from front to back in the block and also somewhat from bottom to top to enter the cylinder head via multiple small holes (through the head gasket). Fluid entering the head from below flows from back to front to exit upward through the thermostat at front, flowing then through the top radiator hose, across the top tank, down through the radiator core, across the bottom tank, out the bottom hose and back up to the water pump. Also assume there is a blanking sleeve or moving ring on the thermostat that blocks the bypass port in the head, so we can ignore any bypass flow for now.

The engine is purring along under normal load generating some waste heat, mostly in the combustion chamber, around the exhaust valve and in the exhaust port of each cylinder. The ceiling of the combustion chamber may easily be warm enough to boil water, but in fact it is water cooled from the top, and the thickness of the metal causes a temperature gradient to end with lower temperature in the water jacket above. The top of the piston may be a little warmer, while it is being cooled by oil mist and air turbulence underneath. The head and stem of the intake valve is also warm, but not terribly hot, as it is being cooled by the incoming fresh air and fuel vapor with every intake stroke. The hottest part of combustion (and highest pressure) happens with the piston near top of stroke shortly after the spark ignites the vapor into a controlled explosion. The spark occurs early, but with luck the piston should be just past top dead center and on its way back down before combustion is complete and combustion pressure peaks, at which time it all sees the highest combustion temperature. Combustion temperature at that moment is hot enough the it might melt the aluminum piston, except that it doesn't persist for very long, and the heat can't transfer into the piston that fast. As the piston is forced rapidly away by the pressure, the gas temperature drops dramatically with continued expansion. As the piston drops, progressively more of the cylinder wall is exposed to the heat of combustion. The higher part of the cylinder wall saw the higher initial temperature and is exposed longer to the hot gas. The lower part of the wall is exposed for much shorter period of time and sees only lower gas temperature after much expansion. So the upper cylinder wall will absorb the larger amount of heat and conduct it through the wall into the water jacket.

As the piston is approaching bottom of stroke the exhaust valve begins to open, and hot gasses are proceeding out past the valve. The exhaust valve seat is being heated by the combustion gasses, but at least that part has a heat sink through the metal casting toward the water jacket. Bad enough that the bottom side of the exhaust valve head had to tolerate the highest temperatures of combustion. But now the hot exhaust gases also swarm all around the valve head heating it even more, as well as the immediate area of the valve stem, the exposed bottom end of the valve guide, and the whole internal surface of the exhaust port as the gasses move toward the exhaust manifold. The walls of the exhaust port have a heat sink toward the water jacket and will run with a temperature considerably lower than the exhaust manifold, which is only cooled by external ambient air.

Alas, the poor exhaust valve is hung out to cook in those hot exhaust gases throughout the full exhaust stroke and a short time into the intake stroke. During this time some heat is being conducted up the valve stem and into the valve guide, but it is a small heat channel with a long way to travel, not very good thermal contact between the valve stem and guide, and the guide still has to conduct that heat through the casting into the water jacket. Finally as the piston progresses a little way down the intake stroke the exhaust valve closes against the valve seat. Through the four stroke sequence (two full revolutions of the crankshaft), the exhaust valve was hanging in the flow of very hot exhaust gas for about 30% of that time. Now for the remaining 70% of the four stroke cycle time the head of the exhaust valve gets to sit on the valve seat with good thermal contact and will conduct heat rapidly into the seat (it was quite hot you know) where the heat will in turn be conducted through the head casting into the water jacket. But a small fraction of a second and less than two strokes later the exhaust valve is again being hit with peak combustion temperature at the beginning of the next power stroke.

For the head of the exhaust valve this is a dog's life in hell as it settles in to constant life being the hottest part of the engine by a large margin, perhaps hot enough to melt solder. The exhaust seat is running a frantic but distant second, as it has to conduct heat quickly through the casting into the water jacket to remain cool enough to keep taking the heat from the exhaust valve. The exhaust port from seat to manifold is also hot from the exhaust gases and is busy passing heat through the casting to the water jacket side. Because of its large surface area the exhaust port will be passing more heat in to the water jacket than the exhaust valve, even though the valve is hotter.

Meanwhile all this heat from combustion and exhaust soaking through the head casting is trying to boil the coolant on the other side. The momentary temperature of the coolant right at the wall of the water jacket near the exhaust port (before mixing with nearby liquid) is the highest temperature of liquid anywhere in the cooling system, AND it MUST NOT be allowed to boil at that point. If the coolant was to boil at the casting surface it would form a thin layer of vapor (part liquid and part gas), which is a much poorer thermal conductor than the liquid. With the vapor acting as a thermal insulator (relatively speaking), the surface of the casting will heat faster and the vapor may quickly turn to steam (a gas with no liquid), which is an even worse thermal conductor. The inner surface of the water jacket at that point may then rise rapidly in temperature, causing increased local boiling of the coolant. All this time the radiator might hardly notice any change in pressure or expansion, and it may not even blow out much coolant, as the locally boiling coolant may condense rapidly back to liquid as it mixes with the nearby fluid flow. This may work for a short while, but the thermal shock from large local temperature variations in the cylinder head may cause stress cracks to form in the casting, usually starting near the exhaust valve seat.

The only thing keeping the coolant from boiling is the fact that it is in constant motion, thanks to the work of the water pump. The liquid is constantly washing the inside of the water jacket to absorb the waste heat and carry it away. Each drop of coolant picks up a small amount of heat and goes on its way before it gets hot enough to boil. With enough coolant flow the temperature variations in the fluid will be minimized. Higher pressure due to having a pressure cap on the radiator will raise the boiling temperature of the fluid and help to prevent this local boiling. Permanent antifreeze and water will also have a higher boiling point than plain water. But the antifreeze also carries less heat than plain water, so the local coolant temperature will rise quicker and a little higher with the same heat transfer rate. Racers know that with a given available flow something close to plain water will carry away the most heat with the smallest temperature rise. But that only works well if you have a relatively high flow rate. When you stop the hot engine suddenly and allow it to soak for a minute, the stagnant fluid in the cylinder head will make a rapid momentary rise in temperature as it absorbs the latent heat remaining in the head from the hot valves and ports. This is why a race car needs a casual cool down lap after a long race, to help reduce the temperature around the combustion chamber before shut down.

Okay, lets get back to the steady state engine condition. Start with a reference point at the thermostat, which is partly open and may be slightly throttling coolant flow at a temperature of say 185dF with static conditions. This is reasonable for a thermostat designed to start opening at 180dF. Coolant temperature in the head near the exhaust ports will be slightly higher. Coolant temperature entering the top of the radiator will be very close to thermostat temperature. Coolant temperature exiting bottom of radiator (down flow radiator) will be much cooler, say 165dF. The absolute temperature at exit from radiator depends on the liquid flow rate inside the radiator and the air flow rate over the outside of the radiator core. But in simplistic form, the liquid temperature drops as it gives up heat to the air, and the air temperature rises as it takes on heat. The temperature of the air cannot exceed the temperature of the radiator core at any given point of contact, but the liquid and the core and the air will all be warmer near the top of the radiator and gradually less warm farther down. On average the radiator core temperature is about 175dF.

Now the coolant temperature at top of radiator is the same as at top of engine (thermostat exit). Coolant temperature at bottom of radiator is same as at bottom of engine (return to water pump). Also the coolant flow is the same at top and bottom hoses (conservation of material). That 20dF drop in coolant temperature through the radiator implies also a 20dF rise in coolant temperature through the engine. Funny how that works. The same temperature difference and same flow rate means the radiator is dumping to the air exactly as much heat as the engine is expelling into the coolant (conservation of energy with steady state running).

Now suppose we have a little larger water pump, or we change a belt pulley turn it a little faster, so it will be moving say 11% more coolant, but nothing else changes (flow rate = 10/9 of prior flow). With the coolant moving 11% faster each little bit of coolant will spend 10% less time inside of the radiator core (9/10 of the time). With a single pass of coolant having 10% less time to transfer heat into the air, the coolant temperature may only drop a proportionate amount or 9/10 of the previous 20dF drop, resulting in only 18dF drop in coolant temperature through the radiator. This is what keys off the quick jumpers to erroneously think that it's less efficient. But wait. Coolant entering the radiator at 185dF would then exit at 167dF. When the bottom part of the radiator core is now 2 degrees warmer than before, it would transfer more heat into the air with the same air flow rate. That means the radiator is actually more efficient at transferring heat with the faster coolant flow (and yes, it really is).

But this can't possibly be right, because the engine is not generating or expelling any more heat than is was before. Steady state, remember? To make this come out with the same total heat movement the average temperature across the area of the radiator core has to be about the same as before, around 175dF. So with only 18dF temperature drop across the core, lets say that both the inlet and outlet temperatures are lower by one degree, at 184dF and 166dF. That's still the right 18dF temperature drop, considering the faster flow rate, and now gives the same average temperature of 175dF across the core to expel the same original amount of heat. In fact the reduction of heat being expelled in a single pass of the fluid is made up for by having more passes of fluid in the same time interval

But then we have another problem. With radiator inlet and outlet temperatures being the same as engine outlet and inlet temperatures respectively, this means the thermostat would see a lower coolant temperature at 184dF, and would dearly want to regulate and control the minimum back up to the prior 185dF. So the thermostat would close down the aperture slightly to reduce the flow rate until the outlet temperature comes back up to 185dF. And where does that leave us? Right back where it all started with the original flow rate, 185dF at the top and 165dF at the bottom. The end result is that the belt drive to the water pump is working harder, but it is not moving any more coolant flow. So it appears that enough is enough sometimes, and more pumping may be a waste of energy.

But if you think about reducing the pumping to same energy, there is obviously some lower limit to that. The fluid obviously cannot stand still, or it would be boiling profusely in the cylinder head without moving any heat out to the radiator. So just where in between is the magic minimum flow rate? This is determined by the minimum fluid velocity required to prevent local boiling around the exhaust ports in the cylinder head. The critical point then is when the engine may be producing the most heat while the water pump may be producing the least flow. That could be immediately after a fast run on the highway when you drive into town and have to stop and wait at a traffic light. For a few moments the heat flow through the cylinder head remains high while it is dissipating the stored energy from the high internal temperature of the head casting near the internal surface of the combustion chamber and exhaust ports. But that condition is fleeting, and a little momentary percolating during cool down may not be particularly harmful. A more significant case might be when the engine is lugging under heavy load at near full throttle, but not turning very fast. This is all quite tricky stuff, and maybe best left to the engine designers, and don't fiddle with it unless you think you know what you're doing and have a pretty good reason to change something.

Here is another example, starting with the original static condition. Initially driving along on a level road you suddenly encounter a long steep upgrade. To maintain a constant road speed you step down on the throttle to increase engine power a bit. Now the engine is burning more fuel and generating more heat, so the cylinder head temperature starts to rise a bit. When the thermostat sees the higher temperature it will open the aperture a little larger to permit more coolant flow in the attempt to bring the coolant temperature back to the previously regulated minimum. This is that slightly tricky bit of logic, because the thermostat only controls minimum temperature regulation by restricting flow. But the end result is appropriate, as increased flow makes the radiator dump more heat (as in the prior example). Increased flow also allows the engine internally to expel more heat, as more flowing fluid can carry away more heat at a given temperature.

The problem we may finally run into happens when the engine load and fuel burning and heat production increase enough to exceed the capacity of the radiator to dispose of the excess heat. Then the thermostat is wide open doing all it can to cooperate with maximum coolant flow, and the coolant temperature continues to rise above the intended regulated minimum temperature anyway. This may not be a problem with the car at all, but a problem with our modern personal perception of the situation. That is when we SHOULD realize again that we are dealing with a vintage style cooling system, and we should maybe relax and smile understandingly at our vintage automobile with its vintage "design features", and don't worry too much. unless the vintage temperature needle gets suspiciously close to the vintage peg.

Now as a graduation award, please click and enjoy this vintage link.


Thank you for your comments -- Send e-mail to <Barney Gaylord>
© 2002 Barney Gaylord -- Copyright and reprint information