Was it the head gasket? And what was the chemical that made the white powder?
Sorry, I’m not a chemistry person so no idea which of the substances make a white powder (assuming the chemicals suggested actually exist! 
). Just curious.
I saw that once, in a Toyota when I replaced the radiator with an all metal one from a third world country. It looked like an all brass/copper radiator but there is no telling what the actual allow was, but it set of a galvanic action that was dissolving all the aluminum it came in contact with and forming aluminum salts of some kind. I took the radiator back to the parts store where I got it from and had them replace it with a plastic/aluminum OEM type radiator.
aluminum rusts white. bad head gaskets have blown my heater cores. Id start with the additives. and pressure test.
also a bad hose to the reservoir will bled out the radiator and over heat. White smoke could be steam that is not showing up in the water temperature gauge or is and high pressure steam is blown the a weaker core. maybe the radiator cap is crappy or left open or wont release pressure.
I am convinced that the first answer (It’s the ground!) was actually the correct one. In this case (like in most) the lack of. This is the case when sometimes loose ground causes the coolant to be charged from the action of the alternator because it somewhat serves as a “secondary” ground. Causing the coolant to become very corrosive. First attempts of fixing it did not include complete replacement of the coolant with only weeks between the reappearance of the problem and when coolant was replaced it lasted three month. There is still possibility of the head gasket been bad because of the speed of the heater core “exploding” with such frequency, but I like the loose ground theory still.
Yep, agreed. I would be checking all factory grounding points.
I don’t know much about Subarus but Volkswagen had a similar problem with the VR6 in 1992. I replaced my heater core twice before getting a recall notice. The dealer installed a relay to extend the run time after engine shutdown. The relay ran the electric water pump after the engine was shut off because pressure would build up and blow up the heater core. I never had any problem after this repair. The white powder thing is another issue. No clue about that.
I’m thinking along the lines of many others here, that this is some kind of electrolytic current problem. I’d be suspicious of anything that takes a lot of current not having the designed current path – usually it would be a ground, but not always – due to a bad or missing electrical connection somewhere. Esp att’n to the starter motor and coolant fan current paths.
I, too, am leaning toward an electrolysis problem. I am thinking that the heater core is somehow positive with respect to the engine block because of a bad ground for the blower motor(?). The white powder is probably aluminum hydroxide and it is being produced from the heater core. That is the reason the heater core starts to leak soon after the replacement. I think the two month core life after the full flush was because there was no salts or ions to carry current and facilitate the oxidation of the aluminum.
My researcher did not yield any reference to the existence of aluminum carbonate. So a leaking head gasket would be unlikely to cause the corrosion of the heater core.
I think Tom’s suggestion of a bad ground was the closest correct answer.
The fellow with the Nissan Pickup hesitation problem might try this fix. I drove various carburated Datsuns and had this issue several times. Remove the cover from the air cleaner and locate the float chamber vent. Sometimes it is not obvious, usually it is a tube poking up near the top of the air intake and cut off at an angel. Start the car and using a good carburetor cleaner with a straw on it spray the cleaner into this tube. The engine should stall. Good, you found it. If not keep spraying into every orifice until you kill the engine. Best to use a face shield. Start the engine and slowly spray into the vent while you work the throttle through all three circuits: idle, cut over and main metering jets. Keep at it until you have used about half the can then go test for the problem. I think it is the cut over circuit. This is supposed to keep the engine running when under gentle acceleration when the idle circuit is no longer working but the main metering jets are not fully on line yet. I found this works better than a cleaner added to the tank because the cut over circuit does not get much use during normal driving. My two cents! Al
Another idea to consider at least. I’ve notice when there’s a small exhaust leak in the exhaust manifold or early in the tailpipe where the exhaust gas is still really hot, there’s often a noticeable deposit of a white power-like substance on the outside surface of the pipe, near the hole where the exhaust gas is escaping. I’ve always thought this was some kind of carbonate salt, but no certain exactly what it is. But it might be a clue anway.
I ran into a problem with my sons 1976 MustangII 302. It overheated on him on night and when I opened the Radiator it was full of white gunk. Come to find out, he had a leake between his engine cooling and transmission cooling tanks in the radiator. After replacing the radiator and installing an external transmission cooler, and extensive flushing of the cooling system. I was able to have a safe running car for him. I wonder if they don’t have a pinhole leak between the transmission colling tank and the cooling system tank. Just saying!