On the kart: the fins cast into the LO206's cylinder and head, the blower fan on the flywheel under the rewind shroud, the sheet-metal shroud that ducts the fan's air over the fins (all stock, all tech items, and "no taping, covering, or restricting of air to the rewind shroud" is in the rules), and the thermocouple under the plug that reads the result.
The principle in one sentence
Heat flows from hot to cold at a rate proportional to how much surface is touching, how big the temperature gap is, and how fast the air is moving past.
The one equation
heat out per second goes as area x (T_head - T_air) x airflowT_head is the metal temperature, T_air the air hitting it. The engine puts heat IN at a rate set by how hard it's working (mechanism 7: a share of every power stroke ends up in the head). The head settles at whatever temperature makes heat out equal heat in. Raise heat in (full throttle) and the head climbs until the bigger gap pushes enough heat out; drop heat in (coasting) and it falls.
Math: proportional to a product. When a quantity goes as A x B x C, changing any one factor by some percentage changes the whole thing by that same percentage, as long as the others stay put: double the area, double the heat out; 4% more temperature gap, 4% more heat out. On a calculator: turn each change into a factor (4% more = 1.04, 20% less = 0.80) and multiply the factors together; the result is the factor for the whole thing (1.04 x 1.026 = 1.067, a 6.7% rise). To see what one factor does, hold the others fixed and change it alone. Physically, here: fins change the area, the day and the draft change the gap, rpm changes the airflow, and each one moves the heat out by its own percentage. That's the same "one variable at a time" rule the pressure tests used, in equation form.
Why it works
Heat is atoms jiggling. Where hot metal touches cooler air, the jiggling gets handed across the surface into the air molecules, which then have to be carried away or they warm up and stop taking heat. So: more surface (fins), a bigger difference to push across (temperature gap), and moving air to keep the surface supplied with cool molecules (airflow). A bare cylinder would melt; fins multiply the surface twenty times over without making the engine bigger.
On our kart, in numbers
What the fins buy. A bare cylinder about 3 in across and 2.5 in tall has roughly 24 in^2 of side surface. Twenty fins each an inch deep (typical proportions) add about 500 in^2. Twenty times the area, twenty times the heat out at the same gap and airflow.
The fan, not the wind. The LO206 is fan-cooled: the flywheel carries blades that pull air in through the rewind shroud and blow it across the fins. Airflow goes with engine rpm, not kart speed. So tucked in behind another kart the fins still get the same volume of air; what changes is the air's TEMPERATURE, because you're breathing the kart ahead's exhaust and radiator wash. A smaller T_head - T_air gap, less heat out, a hotter head. That's the mechanism behind "it runs hot in a pack".
Day versus night. On 9/19 the air went from 84 F to about 70 F. With the head around 400 F (typical), the gap went from 316 to 330 degrees, 4.4% more. Night air is also 2.6% denser (lesson 11), so each liter the fan moves carries more molecules. Both help; the engine runs cooler at night for the same work, which is part of why night laps can be quicker.
Why CHT lags. The head is a lump of aluminum that takes time to heat and cool. Lesson 11's model uses a 25 s time constant (typical): after 25 s at a new throttle setting the head has made about 63% of its move; after 75 s, 95%. So the CHT trace is a smoothed, delayed picture of throttle. A 10 s straight doesn't show up as a 10 s spike; it shows up as a slow climb that the next braking zone barely dents.
Two things to notice
- The shroud is a duct, and a duct with a hole is not a duct. A missing screw or a bent shroud lets the fan's air escape before it reaches the far fins. The rule requiring stock, properly installed shrouds is a cooling rule, and it's also why "shroud bolts tight" belongs on the fastener wiggle-pass.
- Overheating and a mechanical scrape look different on the trace. Overheating: CHT climbs past its normal plateau lap after lap and the engine goes soft everywhere. A scrape (the 9/19 main): CHT looks normal and the kart gets slower anyway. The thermocouple isn't only an engine gauge; it's a way of ruling the engine out.
Problems
1. One factor at a time. Heat out goes as area x gap x airflow. Tucked in a draft, the fan turns the same but the air arriving is 20 F hotter (typical). With the head at 400 F and open air at 84 F, by what percentage does the temperature gap shrink? If heat in is unchanged, does the head warm up or cool down, and by roughly how many degrees does it have to move to restore the balance?
2. Night gain. Air 70 F instead of 84 F, head at 400 F: by what percentage does the gap grow? Air also 2.6% denser. Combine the two (multiply the factors) for the total change in heat out at the same rpm. If the engine wants to run at the same heat-out as in the day, how much cooler can the head settle, in degrees?
3. The lag. The head's time constant is 25 s (typical), meaning after 25 s at a new throttle setting it has made 63% of its move. After a 10 s full-throttle straight, what fraction of the way from its coasting temperature to its full-throttle temperature has the head moved? (Fraction = 1 - e^(-t/25), where e^ is the "e^x" key on a scientific calculator: enter -10/25 = -0.4, press e^x, subtract the result from 1. Or use the shortcut 1 - e^(-0.4) = 0.33.) If the full-throttle plateau is 60 F above the coasting temperature, how many degrees does the head climb during that straight?
Go look: with the engine off and cool, find the fan blades by looking in through the rewind shroud's grille, then trace where the shroud sends the air: over the cylinder fins on one side, under the head on the other. Feel a fin edge. Then find the cylinder side the shroud doesn't cover, which is why the engine's hot side is always the same side.