​​Bill,

The peak in temperature and pressure at around 25 minutes is due to the climb.  The peaks at about 65 minutes are due to slow flight, the first with no flaps and the second (highest) peak with full flaps.

The lower density of the air at high altitude does have a negative effect on cooling since it is the mass of air passing through the heat exchangers per unit time that affects the cooling more so than just the velocity.  This is compensated by the lower amount of power that can be produced at high altitude with the normally aspirated engine.  Also, as you note, the OAT normally decreases with altitude which also helps.  My experience is that if the cooling system is adequate at low altitude, it will also be adequate high altitude because of the offsetting factors.

With my system, one pint of airspace is not sufficient to prevent coolant from exiting the overflow.  If the airspace is too small, the first time the system is taken to operating conditions coolant will be expelled and the entire system will be filled with coolant.  Without an overflow bottle, when the system cools off, air is drawn into the system through the overflow, establishing an air volume in the expansion tank just sufficient to accommodate that operating condition.  Subsequent operation at that same or less stressful operating condition would not expel coolant from the overflow.  In other words, the air volume in the expansion tank will reflect the most stressful operating condition encountered.  One would expect to see evidence of overflow of coolant if it occurred.

 As far as whether to provide additional air space in the expansion tank, it depends on how you want the system to operate.  If you want the system to operate at the filler cap pressure whenever the overall coolant temperature is increasing, fill the expansion tank all the way.  Provide an overflow bottle with sufficient volume to accommodate the expansion of the coolant and set things up to draw the expelled coolant back into the expansion tank when the system cools down.  With this setup, the pressure observed when the overall coolant temperature starts to drop will be more a function of the vapor pressure of the coolant than a function of the filler cap relief pressure.  Pressure changes will be more pronounced with this type of system.

If you want the pressure in the expansion tank to change more gradually with overall coolant temperature changes, then provide an air volume in the expansion tank greater than the volume change of the coolant due to heating.  An overflow bottle would not be required then.

As far as I know, either system works.  A possible advantage to the first setup is that no air is introduced into the system as a result of thermal cycling.  Keeping oxygen out of the system may reduce the chance for corrosion.  My system has an overflow bottle even though the air volume in the expansion tank should be sufficient to accommodate the expansion of the coolant.  This provides a record of anomalous coolant expansion and is probably just an indication of my paranoia.  I remove the filler cap before each flight to check the coolant level.  As a result, keeping oxygen from entering the system is not an option. 

   

Steve