Underdrive pulley recommendations?

[blue2000s]

Originally Posted by JFP in PA

While it may seem counter intuitive, slowing down the water pump can actually produce lower coolant temperatures (within limits). In any cooling system, there is an “ideal” coolant flow rate, at which the heat source is exposed to the coolant for the optimum amount of time for heat transfer to be accomplished. The radiators also have an “ideal” flow rate to get rid of the heat as well. In an Ideal application, the pump would run at a constant speed to provide the best possible cooling (this is one of the several reasons why a lot of race cars use electrically driven water pumps). Unfortunately, because most street automotive cooling systems are mechanically driven (by the belt), once the thermostat is open, the coolant is typically either moving too fast or too slow for optimum transfer due to the range of RPM’s them pump is seeing, resulting in suboptimum heat transfer.

In years gone by, a lot of racers used to pull the thermostats out of engines using belt driven water pumps, assuming “the more flow, the better”, only to find the engine actually ran hotter. But when they replaced the thermostat with a restrictor plate, the engine ran cooler. It had too much flow without the restrictor or the thermostat………..

Faster flow rates keep temperatures more uniform. There's less temperature difference between the inlet and outlet of the engine if coolant is moving faster. This will translate into component temperatures directly. Flow a fluid over a surface and if you don't change the fluid temperature or the surface heat output, the surface will be cooler with faster fluid flow. But the fluid temperature rise decreases as it's speed increases.

An ideal flow rate will keep the most sensitive component below it's critical temperature for any given load/speed/environment condition. This may be faster or slower flow depending on where the component is or even if it's always the same component.

The important point here is that the temperature of the coolant at the radiator outlet isn't telling what the component temperatures are inside the engine. The temperatures have to be directly measured or at least modeled with some very expensive software.

Is there a reliability impact from a smaller diameter pulley? I suspect there is, but I don't know. All I know is what's happening when the smaller pulley is installed and if I personally am comfortable with the risk.