Galvanic corrosion: theoretical prevention idea.
 

When placed in an electrolyte, different metals/alloys assume different corrosion potentials. It is this potential difference that is the driving force for galvanic current flow. The less noble material in the galvanic couple will become the anode and will undergo accelerated corrosion. There are several things done to reduce this effect, such as the addition of various corrosion inhibitors (Dex-cool, Zerex, Water Wetter, ect.) to commonly used distilled water used by most of us in our water-cooling systems.

http://www.sc0rian.cjb.net/procoolin.../galvanic1.jpg

As we all know, pure distilled water is the best medium for heat transfer. The addition of said additives will lower the performance of pure water. How much of a performance hit is dependant on the water/additive ratio. I personally would like to find a way to prevent galvanic corrosion without taking a performance hit.

What I am proposing is to counter the potential difference with an equal and opposite potential via a DC power source and potentiometer.

Measure the electrical potential with a multimeter, then tweak the potentiometer until the meter reads zero.

http://www.sc0rian.cjb.net/procoolin.../galvanic2.jpg

I have a feeling our systems are electrically dynamic due to variations in the metal, oxidation, and water chemistry. If this were so, it would require constant tweaking of the potentiometer to maintain a zero potential, and could lead to huge problems if left unattended for long periods of time. These problems could include increased rates of corrosion, electrolysis of the water leading to Hydrogen-build up, and leaks due to added pressure associated with Hydrogen build-up.

To remedy this, an active feedback loop consisting of a voltage control unit and a voltmeter could be utilized to constantly and automatically measure, and correct for, any deviations from zero potential.

http://www.sc0rian.cjb.net/procoolin...galvanic_3.jpg


Of course all of this is mostly speculation, and not very practical. The reason I posted this is I'm looking for either confirmation or disproval of my idea, based on factual information, not speculation, and the little details that I have overlooked that could make this completely pointless.

I realize it will be practically impossible to completely eliminate galvanic corrosion with the above method. I would be happy if it could limit galvanic corrosion to the same levels as corrosion inhibitor additives.

i argued this same point awhile back, and finally saw the error of my ways.

Here

Good thread, but not exactly what I'm trying to achieve. From my comprehension of that thread, it seems that ludicrous voltage and current is needed to prevent cathodic corrosion between dissimilar metals with an external connection outside of the electrolyte.

What my idea is more geared toward the poor saps with an all aluminum block and a brass/copper radiator without any outside electrical connection. True galvanic corrosion won’t happen nearly as fast, but it still happens. If I read right, it doesn’t take much power to counter that at all.

What I really want is a way to get away with not having to use any additives at all, and have piece of mind that my radiators aren't rotting away.

Spank me if I’m wrong. I’m here to learn.

my understanding of galvanic corrosion requires there to be an external connection outside the electrolyte. If there is not one, then a galvanic cell cannot form....

Right. If the electrical resistance between the two metals is infinite, then there is no flow of current for the external power source to counteract.

Much easier to provide electrical insulation between the metal components. (in the rad vs wb case)

This may not be correct but the way I understand it is that if you have a copper/brass radiator and an aluminum WB, the WB will 'rot away' not the radiator. The radiator will get a coating of aluminum oxide.

May not be a dead short between the different metals. What about high impedance secondary paths? WB> thermal compound> CPU die> socket> motherboard traces> power supply> case> radiator. I don’t have my multimeter handy or I’d measure, but I bet at least an M Ohm resistance. Heck, even at 1 K Ohm, it still wouldn’t take much power to counter cathodic corrosion

If it’s not high impedance secondary paths, then what is the source of corrosion in our water-cooling systems? Oxidation? Shouldn't be that big of a problem, in a sealed system any oxygen would be quickly scavenged and chemically locked away.

If it’s not cathodic corrosion, are we needlessly dumping additives into our water?


True, but in my case my water block will be more noble than my copper/brass radiators. I confused my personal application with the theoretical example I threw out there.

http://uffish-thought.net/wc-gifs/battery1and2.gif
Note, i am stealing Since87's drawings :)

Examine figure 2:

We previously postulated that the value for Rbat would be roughly 1,000,000 ohms (thats through the electrolyte) and Rs is, as you say, also 1,000,000 ohms, then you have:
The current flow in the system will be:

2 Volts / (1,000,000 ohms + 1,000,00 ohms) = 1 microamp

The power dissipated in the system will be:

2 Volts * 1 microamp = 2 microwatts

Is this still going to be a problem? I guess. But it relies on the value for secondary paths to be 1,000,000 ohms, which im not sure of. Could be much higher, or lower.

Still:http://uffish-thought.net/wc-gifs/battery3.gif
Figure 3 represents a system that attempts to prevent the Galvanic corrosion using an external power source. By connecting an external power source so that the voltage differential across Rbat is zero, the Galvanic corrosion can be stopped.

BUT, the voltage applied by the external power source is applied across Rs. Therefore the current through Rs is:
2 Volts / 1,000,000 Ohms = 2 microamps

And, the power dissipated in Rs is:

2 Volts * 2 microamps = 4 microWatts
Which is hardly any current, so you should be able to do this.

Either way, wouldnt it simply be easier to add some of Swiftech's anti-corrosion additive, which BillA says will prtect even bare aluminum in a copper system?

Note: Since87, i stole much of your pics and test, hope you dont mind :) (and i hope i did the circuit analysis correctly this time also)

Can you,please, provide a link to the original work by Since87.

No problem. It looks like the circuit analysis is right.

Les,

Althornin did post a link in his first post in the thread. Here it is.

Damn. I had written a somewhat lengthy post last night and probably got distracted and closed the explorer window without actually posting.

Anyway, check this thread.

There's a lot of good info there, though I don't necessarily agree with the conclusion.

Ta.
I should have read all the thread.
Apologies Althornin.

So let me get this straight… Chloride ions will break down A12O3 (anodization), and bare aluminum will decompose in the manner outlined above?

It looks like to me aluminum is going to corrode away eventually no matter what you do.

For those of us without aluminum in our systems, what are the methods of corrosion? Are they similar to those of aluminum just not as severe? That would make sense to me seeing they are in the same column on the periodic table.

What do the various ions floating around in our water do to copper/brass/silver?


Thanks guys, I’m eager to learn.