Thank you for actually responding to my point this time, it was helpful in pointing out a flaw in my thinking.

However, nothing prevents the addition of a current limiting resistor, as i said before, like so:

http://swc2.hccs.edu/evan/battery3a.gif
Now, If we make Radd some high ohm value, like 1000, then we end up with
2 volts / 1000.001 ohms = .002amps
2 volts * .002amps = .004 watts.

Unless i am missing something.

You haven't calculated all the currents and voltages correctly.

I've redrawn the picture with the added resistor, because it makes the math a bit easier without changing anything of significance.

http://uffish-thought.net/wc-gifs/battery4.gif

The voltage drop across Rs is 2 microvolts.

So the only difference between this setup and my Figure 2 above is that the voltage differential across Rbat is:

2V - 0.000002V = 1.999998 volts (instead of 2.0 volts)

The current through Rbat will be 1.999998 microamps instead of 2.0 microamps. The rate of Galvanic corrosion will only be slowed by 0.0002%.

You simply can't have a significant effect on the current through Rbat without either; dissipating a prohibitive amount of power in Rs, or increasing the value of Rs substantially. If you can increase the value of Rs substantially, then you can as easily make it effectively infinite.

ah thats right.
Damn, i took circuit analysis 4 years ago, i've already forgotten alot.
thanks.

a much easier and cheaper sollution would be to just use redline water wetter.

i have had a cup with a chunk of copper and aluminum sitting next to eachother in a mixture of water wetter for several years now. The aluminum gained an anodized like coating in the first few days, and there has been NO corrosion.

but the same cups containing only distilled or tap water broke down pretty fast.