It's beneficial to understand how the head and barrel are mated. The head is heated to expand and the barrel is cooled to contract. The head is then screwed on to the proper alignment. As the two metals reach temperature equilibrium, the interference fit is attained. The design is to allow for adequate strength absent the threads (which will shear at a much lower pressure) by the interference fit up to the design redline temp. The shear strength of aluminum is fairly low (can't recall the exact value). BTW, the interference fit is NOT at the same place as the threads. It is below the threads. The threads do not hold the head on during operation.
As Yr Right as alluded to but not exactly explained accurately, the cyclical fatigue on the aluminum is the issue that results in failure over time--NOT the time in service. If the cylinder is operated below about 380dF, the time in service will be MUCH longer than if the metal gets hotter. In addition, if the internal cylinder pressures are controlled (and by our best estimate based on a LOT of test data) and kept under about 800 psi peak pressures, the longevity is vastly improved.
Remember, the enemies of metal are heat and pressure. Control those and the design parameters are adequate. Operating at high CHTs (above about 380dF) and with high ICPs and the metal will fatigue and fail much earlier.
ROUGH, back of the napkin calculations:
Diameter of the cylinder approx. 5"
Area of cylinder head approx. 20 sq. in.
ICPs during a various mixtures range from 600psi LOP to 1100 psi at 50dF ROP.
20 x 600 = 12,000 pounds of force down on piston and up on head.
20 x 1100 = 22,000 pounds of force down on the piston and up on the head.
NOTE: These are peak pressures (stress) not mean pressures (HP).
Both of the above stress scenarios can actually produce the same HP!!
The former will have CHTs lower by about 50dF.
In which scenario is the cylinder going to last longer?