Copper/Aluminum rig. Are you an anti-corrosion expert?

[Jayzilla]

Hi. I’m having a very diverse liquid-cooled case mod project. The system will be running with some copper and aluminum parts in a single loop that I can’t and will not avoid. Here are the cooling parts:

Copper:
1X 240MM rad.
2X 120MM rads.
1X20MM~rad.

Aluminum parts:
1X40MM rad.
4X aluminum tunnel OCZ Flex RAM.

So here my questions:
What is the chance for corrosion to occur when running under Anti-corrosion coolant, a piece of sliver coil, changing coolant every three months? I’m talking about long term for at least 5 years and 2hrs of usage every day.

If the corrosion is unavoidable what else I could do to slow down or stop the corrosion process?

Where the corrosion will takes place first? On aluminum or copper or both?

How can I tell if there is corrosion inside of the radiators? What are the effective ways to clean them?

These might be hard to answer but I want to hear your real experience and thoughts. Some of the parts in my project are really hard to find/replace if something went wrong. Thanks!

[zanzabar]

how do you have ddr2 flex that is micron and not degraded to the point of unusable or a promos so you do not need the liquid. and just do not mix them, there is no way to stop the corrosion but your best bet is to run half anti freeze (the blue euro kind) and half water but you are better off just not doing it. if you have ddr3 buy a new 8GB kit and you will be much better off.

just make 2 loops and use a cheap cpx for the ram if you have to, or you could even go passive.

[musicfan]

Hi Jayzilla. As Zanzabar says, try not to mix them but if you have to he recommends 50% blue anti-freeze and 50% water.

I am not an expert in corrosion. And there are many types of corrosion - only some are relevant to water-cooling. I have read many items I will share with you and hope it will help.

In your questions, you mention "Anti-corrosion coolant" so is that the manufacturer's proprietary coolant? Those have anti-corrosives and usually a biocide, which would make the silver unnecessary.

A "wild-card" issue is that some of the problems with blocks and fittings are "plating failure". Gary, the owner of Sidewindercomputers feels there is a certainly % of plating failures, which will not be prevented by anti-corrosives. So you need to make sure you don't lump those under "corrosion".

In your situation, it sounds like you have discrete Aluminum parts that do not touch copper or create a galvanic cell. The corrosion of oxidation-reduction reactions where Aluminum donates electrons to Copper do not seem to proceed at the same speed (in water-cooling anyway) as situations which create a Galvanic cell.

My sources say anti-corrosives work by creating a barrier so the metal is less likely to donate electrons either through redox or Galvanic means. Now you can see why it won't protect from bad plating.

Most avoid wetted Aluminum so do not use anti-corrosives in our water/biocide mix. But if you need anti-corrosives, you either use the proprietary coolants, anti-freeze in various concentrations (because it has anti-corrosives not because of its thermal properties), or a special anti-corrosive additive like Fesers. No one can guarantee success but this is what others do and have done in the past when mixing metals was more common.

A great source is to search this forum on "corrosion" to get this: http://www.xtremesystems.org/forums/...archid=1910374

Other Helpful Sources on Corrosion are:
http://en.wikipedia.org/wiki/Corrosion
http://www.chem1.com/acad/webtext/elchem/index.html
http://corrosion-doctors.org/Contents.htm
http://www.engineersedge.com/galvanic_capatability.htm
http://en.wikipedia.org/wiki/Standar...28data_page%29
http://www.engineersedge.com/corrosi..._corrosion.htm
http://en.wikipedia.org/wiki/Galvanic_corrosion
http://en.wikipedia.org/wiki/Redox
http://www.wisegeek.com/what-is-oxidation.htm

I really like the Chemistry book as it is easy to understand. I am going to speculate on your questions. The electrons should leave the Aluminum and since you may lose mass, this can result in a hole. Some will stay in solution - some will deposit, usually onto "more noble" metals lower on redox tables if there is no Galvanic cell. Oxidation is losing electrons - reduction is gaining.

The Aluminum radiators should in theory, leak. If the copper ones are seeing Aluminum deposit, they may plug and lessen flow. I don't know how or whether to revive old Aluminum radiators. The copper ones should be able to be flushed. Some use scale remover such as sodium citrate. That's my best guess from the sources I found helpful. Good luck.

[El Mano]

Cool post dude
What do you think about installing a piece of any metal more prone to be corroded? Does this make sense at all?

[dengyong]

Cool post dude
What do you think about installing a piece of any metal more prone to be corroded? Does this make sense at all?

That's called a sacrificial anode. It would leave your loop full of junk as it decays.

[musicfan]

Cool post dude
What do you think about installing a piece of any metal more prone to be corroded? Does this make sense at all?

It works for a zinc anode in the steel water heater in our homes. But I do not know whether it works in water-cooling.

BoxGods made Bucky Plugs. At one point in the thread, BG does not recommend it be the ONLY product used if one mixes metals. And if one doesn't mix metals, does one need anything other than regular maintenance? That makes me think the best way is always prevention or risk management.

This is just my understanding in the tiny world of water-cooling. Others please comment.

[Agent11]

http://en.wikipedia.org/wiki/Galvanic_corrosion

'In most HVAC systems, the concept of anodes and cathodes are not an option as they would need to be applied within the plumbing of the system and, over time, would corrode and cause potential mechanical damage to circulating pumps, heat exchangers, etc.'

The best policy is to not mix at all in my opinion.

[MattC]

Just insulate the Aluminum parts from everything else. Check continuity with a multi meter before you fill the loop. Use rubber washers to help isolate. Galvanic cell between the Al and anything will be impossible at that point.

BUT!....should an insulator fail at some point.....

Also keep in mind that Aluminum will create a Galvanic corrosion cell with itself. This self consumption rate is pretty low, so I wouldnt really worry about it.

[CrazyNutz]

^^^ This 100%.

Keep the aluminum parts electrically isolated from the copper parts, and you will greatly reduce the possibility of corrosion. Also keep your coolant as non-conductive (try to keep it from becoming an electrolyte) as possible.

Corrosion of dissimilar metals happens due to the "galvanic cell effect".

[jomama22]

Take a look at the galvanic corrosion charts and the placement of metals. You will see copper, nickel and brass ( most common metals in a loop) are within tolerance of .15 of each other which is considered safe. Aluminum is about .2 or .3 away from these and is over the limit of what is considered safe. As others said, keep them away from each other and use corrosion inhibitor to help. The biggest problem stems from any of the electrons from the aluminum or copper away from each other. It is possible for these to flow through your loop and create corrosion on one or both of the respective metals, usually copper will corrode from the aluminum rather than the other way around

[MattC]

Take a look at the galvanic corrosion charts and the placement of metals. You will see copper, nickel and brass ( most common metals in a loop) are within tolerance of .15 of each other which is considered safe. Aluminum is about .2 or .3 away from these and is over the limit of what is considered safe. As others said, keep them away from each other and use corrosion inhibitor to help. The biggest problem stems from any of the electrons from the aluminum or copper away from each other. It is possible for these to flow through your loop and create corrosion on one or both of the respective metals, usually copper will corrode from the aluminum rather than the other way around

Aluminum will corrode 100% of the time when connected to copper. And in the water cooling world, you won't have an "electrolytic short" through the liquid. The driving voltage is too low for that too occur, and the insulated path (tubing) is way too long. This does happen to ships with huge impressed current systems, and in oil and gas pipelines ALLLLLL the time.

[Church]

MattC: actual experience shows that those weak currents via liquid coolant are sufficient for corrosion to occur soon in mixed aluminium/copper loop and start "easting" components. As distilled water is insulator only at the very beginning but soon in actual loop takes in metallic ions from components or from dissolved air, then soon it becomes conductor. +You have to remember, that most LC fittings are made from copper or brass (haven't heard of such made from aluminium), so here you go, direct metal-to-metal contact. One can use plastic fittings of course, but by imho ruining looks of build though.

Preferred way to stop corrosion: to NOT use aluminium components (those are mostly blocks from stone age anyway, or other components from ancient LC kits, mostly by Zalman, so not much to loose).
Most common way to stop or reduce corrosion if one has to use aluminium components: coolant with corrosion inhibitors, commonly propylene glycol antifreezes (due to toxicity of ethylene glycol based ones), or some of the premixes for LC. Less preferred then not using aluminium, as most efficient thermal-properties wise/cheap/commonly available coolant in LC is distilled water. Dilluting it with other liquids lowers efficiency (acceptably though) / rises price / are less commonly available.

In OP case - to heck with that mini alu rad or LCing the ram with lame vendor alu ram blocks. +O/C of ram won't increase that much from LC, +those 100-200MHz+ almost won't net any extra performance in actual apps/games, only in very few and little synthetic benchmarks. If one must absolutely LC ram (for no practical purpose, as 5$ fan might be more then sufficient, only for bling), best bet would be using something like corsair dominator (or mushkin copperhead) series ram and use united single block for 4-6 modules of such ram. They are made by EK/MIPS/BP. And to heck aluminium crap designed and sold by few ram vendors with little to no clue in LC (especially when those blocks often are one per single ram module = very little cross section / often extra two 90deg turns / fixed unchangeable barbs for small ID tubing = need to spend extra effort by parallelizing flow there for this unneeded LC to not overly hurt flow in main loop, where it counts more).

[CrazyNutz]

There is one thing about this galvanic corrosion that is misunderstood by many, and it's really difficult to get this point through for some reason. I'll try to explain:

For rapid galvanic corrosion to occur there needs to be a low resistance electrical path between the components. It's the case!
1). The rad is mounted to the metal case.
2). the Block is mounted to the metal CPU heat spreader.
3). the heat spreader is connected to the retention bracket.
4). the retention bracket is connected to the motherboard ground plane.
5). the motherboard ground plane is connected to the motherboard mounting holes.
6). the motherboard mounting holes are connected to the metal case.

See the the connection. This %100 completes a near idea Galvanic Cell.



Look at this:


Now in your mind replace the wire connecting the anode and cathode with the case,
and replace the salt bridge with the tubing.

Snipping that wire would stop the reaction.


There are two good solutions:
1). Do not use mixed non-compatible metals.
2). Or just keep them electrically isolated.

If the rad was the only part that is aluminum, then it is as simple as using rubber washers between the rad and the case, and use plastic screws.

[Church]

washers? no wire in between? it's simply not enough. OK, it may slow down corrosion, but to me corrosion after 1 year would still be bad enough to bother even while it might be longer then vs eg. 1 month. I'll rather trust actual experience of those with alu components @loop (for me those are several users having reserator xt at one other forum). If such simple measures would help with no noticeable corrosion signs for .. let's say 3 years, that might be more acceptable. But they aren't efficient enough.

[zanzabar]

thinking of cells, if you got a high surface area cathode with more electronegativity than copper like led or tin to pull the aluminum particles out of the water so less attack the copper. but it is still a bad idea to mix.

[CrazyNutz]

washers? no wire in between? it's simply not enough. OK, it may slow down corrosion, but to me corrosion after 1 year would still be bad enough to bother even while it might be longer then vs eg. 1 month. I'll rather trust actual experience of those with alu components @loop (for me those are several users having reserator xt at one other forum). If such simple measures would help with no noticeable corrosion signs for .. let's say 3 years, that might be more acceptable. But they aren't efficient enough.

Don't get me wrong I don't advocate doing the mixed metal thing, just some people have spare parts they want to use, or are just hell bent on doing it I agree with what you're saying. Keeping parts completely isolated may not guaranty immunity to corrosion, however it will greatly reduce the risk. Also anyone using aluminum with copper must avoid additives that raise the conductivity of the coolant, as in do not use copper or silver based biocides. Also flushing the loop, and filling with fresh coolant would be a good idea biyearly or yearly, as the coolant conductivity can raise over time.

BTW I have done mixed metal loops with 0 corrosion after years of use. You just have to know the do's & don'ts to make it work

[Patrick[PC-Max]]

So here my questions:
What is the chance for corrosion to occur when running under Anti-corrosion coolant, a piece of sliver coil, changing coolant every three months? I’m talking about long term for at least 5 years and 2hrs of usage every day.

If you use the right coolant, you don't have to fear about corrosion.
A few years ago I used some anodised alu parts in a typical loop with some copper parts. With Innovatek Protect there was no corrosion at all and I didn't change the coolant for more than a year.

[jomama22]

Distilled water + nuke-phn + feser base corrosion blocker = coolant for life for me. Never corrosion or algae and the water, tubes and blocks look crystal clear.

[xivlia]

main thing id worry about is corrosion inside the blocks, the microfins should not get any corrosion on them, if they do.. then might aswell buy a new block.

[bmaverick]

Back in the early DIY WCing days of the 00's most of the WCing loops were mixed metals of copper/aluminum. That is the reason why coolants like FluidXP and others were on the market unlike today.

With that said, Corsair, Swiftech, Danger Den and others shipped a glycol mix with their kits. The content of the coolant was simply 33% Propylene Glycol (Sierra PG brand) + 66% Distilled Water. The PG is pet safe, really aided in lubricating the pumps, and only required flushing one a year in a mixed metal loop. The 50/50 mixture is overkill and limits cooling performance.

If your loop runs with any DDC pump (stock), D4, D5, the plastic involute and chamber is electrically isolated within the plastic body. The impeller has a stainless steel base.

If you use nylon plastic fasteners of washers, nuts and bolt/screws, you can really keep the galvanic loop more isolated.

Hope this helps

[nateman_doo]

STICKY THIS POST. There is actual data on this subject that people ask all day every day.

[bmaverick]

STICKY THIS POST. There is actual data on this subject that people ask all day every day.

Good general questions lead very well into detailed information. Hope to see others provide good input as well.