Printable View
I live in canadian prairies, and in the winter here we get -30 to -40 C. I figure sticking a radiator outside might be an easy way to get a chilled loop. I have an old 85 Ford Ranger radiator, but it has an aluminum core. Is galvanic corrosion going to be a concern through 6'+ of tubing, because the resistance through that amount of water should be pretty high.
i had a long discussion about this with vapor.
Yes its true you need current for galavanic to start.
Or a presence of an electrolyte.
However i think its still too risky, since water can act as an electrolyte.
Can you goto the junk yard and look for an older chevy suburban radiator?
Those had copper cores.
the Ions will move with the flow, so yes, big problem. if you want to mitigate it, get nickel plated blocks, use LOTS of anti corrosive additive, and even then replace coolant every 6 months. i said mitigate, because it wont completely stop it. alternatively you could just buy a 480 GTX rad and put that outside. and sell the ranger on ebay.
Then you misunderstood what I said :p:Quote:
Originally Posted by NaeKuh
The water is the electrolyte but you need to complete the circuit electrically for galvanic corrosion to occur and water cannot be both the electrolyte and the 2nd electrical pathway. Basically, as long as the copper and alu aren't touching (or connected by other metal components, such as screws), you'll be fine.
OP, you'll be fine if you run plastic/rubber tubing. (just don't report back here that you're using alu in your loop--in that sense, the world is flat here)
There are other concerns with your setup though, like freezing and condensation due to the various PC-level components (CPU block, tubing) being much colder than ambient.
Thanks. I thought that would be the case. I think I may give it a shot, but I'm going to have to do a lot of research on insulation. This water could get cold! Well, actually I probably can't use water, but I'll address that in the Chilled Loop forum.
Yeah, do some research over in the Chilled section, because things like acrylic and plastic don't always play nicely in really cold temperatures. Anti-freeze/water should work for coolant. And if it manages to freeze, well, so did your car. :p:
thats because there running an ethanol based coolant.Quote:
Originally Posted by MpG
Oh was that it??? lol oops.. :PQuote:
Originally Posted by Vapor
Let's play with two examples of 'touching'
Sample A: Simple
You have alu and copper in the same block and it's assembled with metal bolts/screws. So long as the alu and the copper both touch the electrolyte (the water), the circuit is complete and galvanic corrosion will occur.
Sample B: Elaborate
Let's say you have a GTZ with the LGA1366 backplate and an aluminum radiator. If they're connected via rubber/plastic tubing, you're alright....until you mount it in a case. At that point, you have the alu radiator touching the case, the case touching its own motherboard tray, the LGA1366 backplate touching the motherboard tray, the thumbscrews screws touching the backplate, the mounting plate touching the thumbscrews, the assembly screws touching the mounting plate, and the copper base touching the assembly screws. In this situation, your alu rad and copper plate are electrically connected, there's an electrolyte, and galvanic corrosion will occur!
There are three ways to avoid galvanic corrosion:
1) don't mix metals (silver and copper counts as mixing, so does alu and copper, so does stainless and silver, so does stainless and copper, so does XXXXXXX and YYYYYY, even a small difference in nobility will lead to galvanic corrosion over time [nobility of metals such as alu and SS change a lot based on pH, so they're harder to look up, I'm not sure what they are--but silver and copper are surprisingly far apart]). This is practically impossible for us considering how important copper is to our thermal goals and the use of stainless steel in our favorite pumps (DDC and D5). Even using a killcoil 'breaks' this.
2) don't use an electrolyte. Also impossible...this is liquid cooling. There are basically no liquids that will not act as an electrolyte (and none that are worth our time, thermally). EDIT: there are actually many liquids that won't act as an electrolyte, distilled included, but they'll all eventually ionize over time (months to years), thus acting as an electrolyte.
3) don't complete the circuit, electrically. This is the only one we have some control over. We can use only plastic barbs, delrin tops, coated backplates, etc. It also largely invalidates our inherent mixed metals in #1 (the stainless steel in Laing pumps is not in circuit, electrically, neither is a kill coil in a reservoir or tubing).
Here's a list of blocks that I have in front of me that have a 'galvanic circuit' (multiple wetted metals in electrical contact) when used as intended (using included barbs counts):
Heatkiller 3.0 LT/Cu (LC does not, LT+Silicone mod does not either)
Apogee XT
Apogee GTZ SE
Sapphire Rev.A
Luna Rev.A
Aqua Computers Cuplex XT di
Aqua Computers Cuplex XT di2 (I don't actually have this one here, but it has the same fundamental construction as the original di)
Koolance CPU-350 (acrylic top)
Koolance CPU-345 (acrylic top)
EK Supreme
Alphacool Livingstone
Alphacool Yellowstone
I'm sure there are more on the market too...as well as many GPU blocks with dissimilar metals in direct contact with each other. Anything with a metal top will almost definitely fall into this category (using a metal barb instantly includes it). Of course, the less dissimilar the metal, the slower and less noticeable the g.c. will be, to the point where it's basically unnoticeable on the timescales we use our hardware.
All in all, the galvanic corrosion thing is way overblown (especially here at XS). Most of the times we've 'seen it' it's really something else at play that's gunking things up :yepp:
Also of note, oxidation (i.e., rust and the other metal's versions of it) is not the same thing as galvanic corrosion!
Totally agree Vapor. 99% of the sky is falling crap about Galvanic corrosion is blown WAY out of proportion. Its all about marketing a lot of the time. Fan boys beating on some other companies parts or mad about being out performed etc.
Copper and silver?! Will cause galvanic corrosion? Really? So using True Silver barbs and nickel plated copper blocks (in direct contact) will cause corrosion? Why hasn't anyone told me this before.
That's a large portion of my point!Quote:
Originally Posted by dreamaxx
Galvanic corrosion is so overblown (but somehow only between alu and copper) and misunderstood that it causes unnecessary fear. Don't worry, you're fine :yepp:
EDIT: there's tons of examples in the setups we use where galvanic corrosion should occur. Have we ever noticed it? Is it a big deal even if it is noticed? Not really. Its "doomsday" perception is totally overblown.
People like you change our views of watercooling, backed up by facts and research. Thank you. This non sense can now stop.
this is called red corrosion.Quote:
Originally Posted by dreamaxx
i addressed it on my silver rant.
But in practice, no one has seen it so far.
The key word is in practice. Not physics.
Because in 99.9999% of the case, your water will be an electrolyte. Its very difficult to separate this.
No the same arguement i had with eric.Quote:
Originally Posted by BoxGods
If it was that easily avoidable, then why do we see it on our the forum EVERY YEAR, year after year...
The prep is more hassle then its worth, that is why us "fanboys" rant about alu.
The moment your motherboard is touching the case, and it has standoff's, it will transfer current to your case.
The moment your rad is mounted to the case via screws, it now has completed its current.
Overblown? No... its called being precarious....
If it was truely overblown, we wouldn't have corrosion threads pop out each year now would we?
The part i bolded is why we rant about it.Quote:
Originally Posted by Vapor
Sure you wont start off with it, but over time you'll never see it? :P
Someone smarter than me was the one who clued me to the facts behind galvanic corrosion. I owe my thanks to him :)Quote:
Originally Posted by Boogerlad
(I'm just willing to bring forward what I've learned from him, lol)
Here's the links that convinced me:
Link number one
Link number two
Link number three
Plus there's info on wikipedia about the galvanic series and nobility. These are really guidelines as the environment we subject the metals to is slightly different than the standard environment they use for testing and standardizing.
We don't see instances of corrosion every year :shakes: Most of the time it's something else causing the "yuck factor."Quote:
Originally Posted by NaeKuh
In the 'old days' where alu was directly touching copper, it would happen on timescales users could see. In newer times, even my "Elaborate" contact scheme is a huge stretch--it requires that there's no paint in use and that the backplate makes contact with the metal of the case (both of which are far from given--I chose the GTZ in my example because, AFAIK, Swiftech uses the thickest backplate). The use of an alu radiator is basically safe. Alu reservoir? Safe. Stainless components in a loop? Safe. Why? Mostly because there's no electrical contact.
Galvanic corrosion really is overblown, we're safe from its effects these days because manufacturers have learned not to have the very dissimilar metals in direct contact :)
EDIT:
Even then you're still only satisfying two of the three requirements for galvanic corrosion. You still need the electrical circuit to be completed, which doesn't always happen (and is often easy to totally prevent).Quote:
Originally Posted by NaeKuh
did you mean base plate?Quote:
Originally Posted by Vapor
is the xt better thermally coz the base plate is sooo much thinner?
What?Quote:
Originally Posted by Hondacity
I just tried to think of the most round-about way for electrical contact to occur :p: The first thing that came to mind required a Swiftech LGA1366 backplate and a delrin-topped block. :lol:
EDIT: the timescales we'll see nickel/copper/brass/chrome-plating galvanic corrosion is an incredibly long time, probably not in our lifetimes.
Retardedly so:Quote:
Originally Posted by Vapor
- Make sure all your blocks are the same material - they're going to be grounded to the case. Electrically isolating them from the case without sacrificing performance is possible, but do you really want some electrically isolated piece of metal a fraction of a millimetre from your precious silicon? :)
- If the barbs are threading into metal, make sure they're the same metal, or plastic.
- If your radiator or whatever non-block component is a different metal from your blocks, don't mount it to your case directly with screws. Use rubber silent-blocks, tape, or something-or-another electrically insulating.
For 99% of people what this comes down to is using plastic barbs on your rad instead of stainless or nickel, and even then you wouldn't get measurable corrosion for oh, maybe 5-10 years? :)
BTW, doesn't galvanic corrosion go the lower your electrolyte's pH is? If so, you could paranoid-proof yourself by adding a tablespoon of baking soda ...
want me to dig them out? :yepp:Quote:
Originally Posted by Vapor
This year i remember reading 1.
Last year we had a few swiftech Stealth GPU block..
The year b4 that was the GTX
The years b4 that was the MCW6002.
The years during that we used a ton of water wetter.. :P
Nah eric.. we get at least 1 corrosion thread per year... its gaurentee'd.
Watch we'll have another one next year too. :P
Sorry gabe, dont mean to pick on your product alone... its just those threads were the most memorable.
Yeah, but those instances are old, surely it's less than 1/yr now (then again, my perception of time in WC is 'off,' I thought the GTZ and Koolance CPU-350 were nearly 2 years old) :p:
I'm not saying that galvanic corrosion doesn't exist or that it can be ignored, just that it's really easy to work around and still have alu/silver/copper/SS300/etc in your loop :) And that the galvanic corrosion issue is, frankly, way overblown here. Alu is not a bad material--it's pretty well priced, can be anodized to many colors, pretty easy to machine, not too dense (shipping costs), thermally viable, and has other redeeming qualities. I'm just saying that it can be used in our loops and galvanic corrosion can be totally avoided without additives :)
Is that the only reason? I thought differing thermal expansion rates came into play as well. Not that I have any experience myself in that arena, mind.Quote:
Originally Posted by NaeKuh
cold also cracks acrylic.
90%.Quote:
Originally Posted by MpG
Because glycol + water cant go down as low as ethanol based systems can.