Ok, I think I have the story on the melting fuel pump fuse problem. You know the one- where the fuse doesn't blow but melts itself and the surrounding plastic on the fuse panel? Here is the rundown:
Start with corrosion at the fuse contacts. It gets worse with time, increasing the resistance of the circuit that feeds the pump. The voltage drops at the pump, and it draws more current up to a point. Since power dissipation goes up with current and resistance, this means more heat is produced as well. In this example, power = heat. This is not a prob if you have a heatsink that takes care of the heat like on your cpu, but there ain't much dissipation area at the blades of the fuse. The fuse gets warm. Some dude named Georg Ohm figgered out the numbers. Typical contact resistance should be less than 1/10 ohm but we'll use 0.1 for this example. Using 8 amps for the pump and 0.1 ohm, we get 0.64 watts - no prob. Now jack up the resistance due to corrosion to 1/2 ohm and poof - the watts shoot up to 32 W. That's a 32W light bulb dumping heat into a small space much less than that of a bulb. The pump randomly stalls and you eventually get melted plastic:
And an unblown melted fuse that is stuck in the panel:
Solution? Replace with a spanking new fuse holder:
I'm figuring the fuel pump is the most common example of this problem since it draws sizable current and is always running with the car. Hope this helps somebody.