Of course the coolant is a conductor, but just as you can't light a
light bulb with one copper wire going to it, to cant have
electrolysis without a return path.

Now before David B. chimes in with stray current problems in marinas,
I understand their can be funny things going on with electrolytes,
dissimilar metals, relative sizes of dissimilar metals etc. Indeed
there are probably all kinds of cool electrolytic things going on in
the stock set up.

but in the car, one seemingly can isolate metal components "on the
return path". And yes, there are dissimilar metals in the stock
vanagon set up, metals of differing "nobility" and of differing
surface area (exposed to coolant).

Think of the brass in radiators, think of the cast iron, brass in
sensors, aluminium in heads. I'd say that refreshing the coolant once
a year would do much to quench runaway corrosion as any worrying
about "isolated" copper in the system. i could be proved wrong of
course :)

Anyhoo, how many of us actually test our coolant , not for glycol
conc, but for pH (which is an indication of corrosion inhibitor
levels), and how many of us DO change and flush coolant every year?

Alistair

On 29-Jun-06, at 6:54 PM, Jake de Villiers wrote:

Well, don't argue with me, I'm not a professor! The guy who laid it
out on the Subaruvanagon list made a convincing case for the coolant
being an electrical conduit. Electrons are pulled from the aluminium
by the copper through the coolant.

I'm pretty sure  the low coolant level sensor uses the water as a
conductor.

On 6/29/06, Alistair Bell <albell@uvic.ca > wrote:
Jake,

in circuits and electrolysis one needs a return path.

isolate dissimilar metals with non conductor, its not a new idea.

Alistair