Mixing metals in the Watercooling loop

F

fx9

[H]ard|Gawd
Joined
Apr 2, 2008
Messages
1,148
Mixing copper + aluminum = very bad.

As for everything else, I don't think you'll have any trouble.
 
Tsumi said:
Mixing copper + aluminum = very bad.

As for everything else, I don't think you'll have any trouble.
Click to expand...

Little over board with the "very"

Its not the greatest thing to do, but if you take the right measures like others have said to do, you will be just fine.
 
It's too easy in this day and age not to mix metals, and still have high-quality blocks/parts.

@OP: If you're jammed for cash, by all means use what you have for now plus a designed-for-water cooling corrosion inhibiter. However, when you can, swap out any aluminum in your loop.
 
I personally think that if you cant do it right don't do it at all for water cooling, it is just to big of a risk to take(if you cant afford to do it right you cant afford a new computer if one of your cheap parts fails and waters down your system). and alum. anything in water cooling just isn't worth it, the whole point to water cooling is high performance.

But back to the main topic, mixing really isn't that big of a deal, as long as you take the right precautions.
 
The OP was concerned with Nickel and Acetal. Both of which will not cause galvanic corrosion.

I just threw aluminum in there to keep him informed.
 
@Nick: I understand what you're saying, but not everyone is going to view water cooling this light.

As for your view of mixed metals, I think I'm missing something. On one hand your saying that water cooling is worth doing right. While on the other it's not important to avoid low density metals, and all the complications that they bring. The truth is that it's difficult to find AL parts, primarily because OEMs have gotten the message that knowledgeable folks won't buy/suggest them.

My loop is comprised of cooper and nickel. With this simple choice I can use distilled (still the best coolant) and one kill coil. I just don't see the downside of one less complication.
 
people around here blow galvanic corrosion way out of proportion. yes its a real thing, no its not going to destroy your loop anytime this decade even if you run metals as dissimilar as silver and aluminum. after a year you might notice some tarnish/discoloration on the copper parts, but we are talking about maybe a few micrometres of buildup over the course of years. not nearly fast enough to damage anything before the parts reach end of life.

just so you know- almost all radiators are brass, with some being aluminum. almost all water blocks are copper. everyone is running with dissimilar metals. i have never once in my life heard of someone who had actual problems with galvanic corrosion in a water cooling system, yet people talk about it like the plague.


see this thread, and specifically post #15
http://www.hardforum.com/showthread.php?t=1571679
 
As long as the metals being mixed are of the noble variety there's no problem that can be measured in the duty cycle of the average WCing FRU.
 
there are no common water cooling components made from noble metals (noble metals being gold, silver, iridium, platinum, paladium and a few more exotic ones).

as long as you dont use a combo like magnesium and copper, there is nothing to worry about. take a look at this chart
http://www.corrosionist.com/Corros1.gif

the farther two metals are from each other on the chart, the greater the potential for corrosion. metals at the top will corrode preferentially to all metals below them.

for instance, if you had a loop with lead and 70/30 copper/nickel, there would be very little corrosion. if you had lead and aluminum, the aluminum would corrode and deposit on the lead much more readily.

the point is moot though, because there needs to be an electrical potential difference in each piece of metal (electrons flowing). almost all water cooling components are grounded out, you might be able to measure a few single micro-amps between two parts, but that is not enough to do much of anything.
 
danger den had a TDX silver edition on the market, but it was not cost effective for the performance.
 
So, cooper isn't considered in the range of noble metals now?

This is insane, and why folks are so afraid of water cooling. Look, it's really easy and you do NOT have to be a meatball materials scientist to play THIS game:

1. No water cooling OEM worth looking up is going to offer AL parts even if they're plated. This means that you'll have to actively seek out AL parts. Why expose yourself to a situation that WILL fail you VERY quickly?

2. Corrosion inhibiters are are all well and good, but do they really make sense? The shearing force created by a pump will rapidly change the nature of any chemical solution. The simple truth is that distilled and a bit of silver will keep your loop clean and safe, and sacrifice nothing in terms of performance. If you don't see the need in creating an uncontrolled science experiment in your cooling system, keep it simple.

Do I have a dog in this hunt? Not really. Use what you want, but go into it with your eyes open, and with Google as your best bud. :) My experience? Ten years of making every possible mistake with water cooling, and learning from each one. Hence, this is not theory for me. If I offer advice you can bet that I won't use you as a crash dummy, or as a canvas for my brilliance.
 
I cant believe how much some people are down playing copper and aluminum corrosion.

I've had 3 machines in my shop the last month using cheap coolermaster and koolance kits that literally have chunks of metal just flowing around (2 of which were cyberpowers. ewwwww). And not just a few, probably an entire tablespoon worth. All of them were running the manufactures recommended fluids which include additives to prevent corrosion.

As far as the OPs concerned you are fine.

I don't know any companies that still make aluminum components. You would have to look really hard, or buy old cheap stock for it to be a problem.
 
Yeah, it really depends on whats in the water that determines how bad the corrosion will be. If you use lab-grade RO water (not the shit that you buy from the crap machines at the grocery store) then you dont have much to worry about. I bet most here have some salts in their water, in particular anyone from the midwest where salts are in the water. Then you can start to see some rapid electron exchange.
 
Swiftech Apogee with corrosion. 6 months old.

swiftech_apogee.jpg


BeCooling SlitEdge.

IMG_3225.jpg


Neither of these would have been doing any good for the pump. Galvanic corrosion is a big problem. Bad anodizing in both these cases was to blame.
 
Whoa that is crazy. I've never seen corrosion like that before. Don't even think it can be salvaged by a simple cleaning either.
 
The GTX needed a replacement copper top which I had to buy. The other one is a museum piece and paper weight. The lesson here is never trust anodizing.
 
Yes, galvanic corrosion is not as huge a concern as some make it out to be - but it is a VERY real problem and has been discussed here (and elsewhere) at length. In short, there's no reason to mix metals, ever. If you absolutely must mix metals (say, for the inclusion of a very foreign part, perhaps for aesthetics in a professional quality case mod) you can do it "safelier" by isolating both metals from ground. Galvanic corrosion is markedly reduced when the electrical path between the dissimilar metals (external to the electrolyte solution) is removed (as a battery's chemical components stabilize when it's poles are disconnected). A 10% solution of any anitfreeze is enough to slow corrosion at this point to a non-issue.

But there's no reason to mix metals, ever, and yes, it can cause some serious damage. Don't ever trust plating. I don't even use brass in my loops because of some Koolance fittings I used recently. Even the brass side had obvious corrosion through the nickle plating, which ruined the seal on the barb.

Don't do it, but if you must, stop at nothing to do it right. Use nylon screws and washers if you must, but break ground.
 
You must log in or register to reply here.
Back
Top