Aluminium togehter with Copper 1

[lasselur]

I have a cooling circuit where some aluminium is used, this have forced me to use Stainless steel and not copper everywhere else du to elecromagnetic corrosion. The problem i now have is that certain components I need does not exist in SS. Is it ok to use Crome plated Carbon steel or Brass togeter with aluminium. The liquid running inside the pipes is water. Could normal corbon steel be used with Aluminium ?

 

[Metalguy]

"Electromagnetic corrosion?" Try electrochemical.

Anyway, if you can mix SS with Al you should be able to use Cu too. Both of them are quite cathodic to Al.

 

[lasselur]

Well the thing is that Cu and Al does not go together due to the fact that Cu is much more Noble.

 

[kenvlach]

Cu + Al is a very bad combination due to galvanic corrosion. Even alloying Cu in Al, as in Al 2024, creates such corrosion-prone galvanic behavior between the CuAl₂ intermetallic and the Al matrix (with an emf potential of 0.73 V) such that Alclad 2024 was invented some 70 years ago (as Alclad 24S).

To answer your question: Different metals can be used with enough corrosion inhibitor in the water (e.g., coat the entire wetted surface with a silicate film) and/or with electrical isolation between them (think of dielectric pipe unions or radiator hoses or lots of Teflon(R) tape on threads).

 

[unclesyd]

As kenvlach stated, this system will require a lot of isolation to stand any chance of not corroding without an inhibitor. I like rubber or if the pressure or temperature is too high they make some Teflon lined S/S flexible hoses that do an excellant job of isolation. If this is a closed loop there are inhibitor systems, as stated silicates and others,that can be used to maintain the film which will be required to protect all the components. I once used a lead lined pipe in a loop to help with MIC as the makeup water was loaded with SRB's and his gang.
A once thru raw water system it is going to be extremely hard to protect the components without inhibitors. It can be done, we once used a continual injection of sodium silicate (desired 7-9 ppm) to a extremely pure raw water system to increase the hardness which prevented dissolution of the Cu pipe.
I hope this is not a potable system, an entirely different game.

 

[lasselur]

The cooling system is going to be used in the new CMS particle detector at CERN. The muon chambers in this detecor was designed with aluminum pipes inside. Now this chambers have to be connected and we desided to use Stainless steel in order to avoid galvanic corrosion. since the geometry of the cooling system is so complex I need Air seperators, the problem is that i can not get this component in stainless steel. However it is possible to get this components in normal carbon steel or brass with a Crome coating. Does anybody know if these crome plated componets could be used in the Aluminium circuit. The liquid used in the cooling circuit is normal tap water that have been demieralized.

http://isaksen.home.cern.ch/isaksen/

 

[Metalguy]

Is this a "closed-loop" system? If it is, and you have very low conductivity water, then you won't have a galvanic corrosion problem. All you need to do is use an inhibitor for Cu as per Kenvlach. SS and Al can also cause gal. corr. in high-cond. water, so I don't understand why you selected that material.

You have to keep the flow velocity low with Cu, so if you can't you might see if Cu-Ni is avail. for what you need. Cr-plated brass might also work OK, but Cr is difficult to plate without lots of fine cracks.

 

[unclesyd]

Lasslur,
As I understand it an inhibitor would be prohibitive in your system. Demineralized water will corrode Cu. The mechanism is the simple dissolving of the Cu to achieve an equilibrium with the water. I checked some piping codes/practices that I’ve used in the past on extremely low conductivity pure and demineralized water. All the piping and components in these systems were Al or S/S. No plating allowed. One reason is that Cr plating a complex object will probably leave porosity, though minute, a pore can become extremely active.

What type of separator are you trying to use?

Do you mind losing very small amounts of your demineralized water?
You could put a S/S separator, similar to a steam purifier at the high point and take the air off the top and dump through a air eliminator. This will dump a very small amount of Demineralized water. We have done this.

Can you use a separator prior to the demineralizer?

 

[AlanD]

Quote [Cu + Al is a very bad combination due to galvanic corrosion. Even alloying Cu in Al, as in Al 2024, creates such corrosion-prone galvanic behavior between the CuAl2 intermetallic and the Al matrix (with an emf potential of 0.73 V) such that Alclad 2024 was invented some 70 years ago (as Alclad 24S).]

Just a pedantic pull up

Aluminum bronzes are corrosion resistant to both in atmospheric and water conditions. They have a nominal composition of 90%Cu and 10%Al. Although I fully agree at the aluminum rich end of the spectrum they have huge corrosion problems, at the copper end, they have good corrosion resistance.

 

[IFRs]

The nuclear power industry uses de-mineralized water, I have seen aluminum storage tanks. I realize it's of little help, but perhaps you could contact a few nuke plants to see what they use for their water piping systems??

 

[rnd2]

Is there any reason why you could not use glycol?

 

[lasselur]

yes beacuse I the feed water system is already installed

 

[odle1128]

No comments on the area of the two metals?

Copper is noble to aluminum and will be the cathode. Aluminum is the anode and will dissolve. If the aluminum area is large and the copper area is small, and/or the water is not very corrosive/conductive, and/or they are not in electric contact, you might get away with the two metals in the same system.

In short, the answer is in the details.

Robert

 

[moltenmetal]

I've always wondered why copper is listed as incompatible with deionized water. I've seen it listed as incompatible in so many different places that I believe it to be true- but I don't understand why.

Is copper similarly corroded by steam condensate? (i.e. very hot de-mineralized water?) If so, what about all the miles of copper steam tracing out there in industry? Is there some volatile additive in everyone's boiler feed water which renders the copper exposed to it passive against attack by condensate? Or is condensate different somehow?

Copper is widely used in domestic hot/cold water and hydronic heating tubing/piping and lasts a very long time indeed in those services.

If the mechanism of corrosion in deionized water is, as a previous poster put it, "...the simple dissolving of the Cu to achieve an equilibrium with the water...", then surely the same would be true in more conductive but essentially copper-free water? Surely some copper may dissolve and result in an increase of the conductivity of the de-min water (possibly rendering it unfit for its intended use), but does this happen fast enough to lead to failure of the copper components?

Other posters have indicated that silicate passivates copper. Surely there's not that much silicate in potable water? Is there something else which is present in most potable water which renders copper passive? I've read that "hardness" (carbonate) precipitation can result in an enhanced corrosion of copper.

 

[unclesyd]

molten metal,
This was our problem in real life. I don’t remember the conductivity of the water but it was very low. At the time there was no detectable minerals in the H2O. No one was concerned until we started having leaks in the potable water system, very large, the plant at that time covered 45 acres. Metallurgical analysis revealed small areas of island type corrosion within an area of general thinning. We begin to pick up Cu in the water at the distant stations from the source and not as many leaks. The source water had no Cu. After several phone discussions my boss and several big names in the corrosion business concluded that the cause was there was no Cu ions in the water so there was an equilibrium being established. As I stated the end result was to add sodium silicate at ppm range to establish a silicate film on the pipe. At the time the plant was constructed the piping code for pure water and deionized water was Aluminum or S/S.

 

[unclesyd]

Our BFW treatment was very simple run it through a DA the adjust the ph and add a little sulfite. I don’t know what the complete treatment system was at the time. We made our own. We kept superheat in the HP (650) steam so there was very little condensate. We maintained superheat until we got to 175#. The condensate wasn’t very corrosive in part due to the iron in the condensate. We had the usual erosion/corrosion in the condensate system.
At that time the CTW was treated with chromates, so no corrosion.
I just call and talked to the water chemist that came in about the time the leaks started and he said that the leaking started when the piping was about 7 years of age.

In this area most houses are built on a slab with embedded Cu pipe. There are a couple of plumbing companies that make a living rerouting the water piping overhead in the house. Most of the first to go is the exit line at the hot water heater. The water from their wells is also very pure, the only thing is that they have 1-2 ppm CO2. There is sometimes a CU carbonate film, but usually this is scoured off. This weekend I was shown a section of 3/8 Cu tubing from the exit of a water heater with a hole just below the tee. The recommendation was to repipe the hot water system and bing me a sample of the cold water. The house is 6 years old.

 

[Metalguy]

Sure sounds like a very simple problem of excessive water velocity caused by the use of undersize pipe/tubing, as does your previous incident. Contractors stay awake nights trying to think of ways to save every nickel.

 

[unclesyd]

I’m not going to take offense at your remarks. One has to always consider the source. My initial statements were made to assist “lasselur” with his problem. I assumed that he had very high purity water and this may be problem even if he could isolate the systems. If you have very low conductivity there one problem that can arise. I was told it was always hard to get an electrical couple with an nonconductor. One insulating washer or two should make it alright. I don’t think so.
There a three Princeton Potentiostats sitting idle at the old place, maybe I should try to get some time on one.
The statement about design was made totally with out facts, a big problem at times. This plant happened to be designed by engineers from two very large chemical companies along with a competent core of engineers for the new plant. Even though they might not have had the computational power available today they were at least slide rule accurate and if warranted there were logarithms. The degree of accuracy was always to the positive side because we always rounded up. The system was designed to supply 10,000 people and never over 7700 availed themselves of the potable water.


I can assure you this is not a case of poor design. My example was happened to a 3/8" pipe off a local heater for an upstairs bath room. I can also state that I don't think the owner was shorted on anything in the plumbing line. He owns the supply house
How is water no solid impurities at this pressure probably, 50 psig at the most, and velocity the shower heads are of the 1 ½ gal type, toilets are tankless, don't know about the tub. I'll let you know next week how much pipe is affected.

Just for my own edification how much pure or low conductivity water piping have you been involved with. Though not directly involved with the water system I saw all the failures of any kind on this site for 42 years. Every failure regardless of origin from day one on this site passed through the Metallurgical laboratory. We were doing RCFA a long time ago.

 

[Metalguy]

No offense was intended (believe me you'd know it if it was). Guess I just have a thick skin.

Anyway, excessive water velocity has caused many (probably millions) of Cu pipe/tubing leaks, usually just downstream of elbows, tees, etc. It frequently leaves "islands" of Cu just like you describe. The rest of the system is unaffected-just the turbulent zones get attacked.

One case I did the FA on involved a very large engr. co.-probably the largest in the US for nuclear-related construction. Doesn't mean that even the giants don't sometimes make mistakes.

I should have used the word "builders" instead of contractors, esp. home builders. They just love the slightly cheaper thin type M Cu tubing instead of type L. I lived in a devel. in AZ where most of the houses, mine included, had leaking pipes buried in the slabs. Mine had the pitting as described above. Also performed a FA for a huge VA hospital with leaks all over their CU potable system. They couldn't understand why no leaks for the 1st 20 years or so, then leaks all over. I collected many samples-all had the leaks right after fittings. No straight tubing problems. I asked about increased water usage. They answered with "problem started after they added a new hosp. wing". I told them that made sense, because of higher/longer flow. They saw the light and agreed-thought I was a god for coming up with their answer. I love it when people think like that!

 

[moltenmetal]

Thanks for the anecdotal information, but I still don't know if this is an issue for copper in demin water, or for copper with water generally. I've got copper piping in potable water service in my house that's at least 50 years old, so copper clearly can last a very long time in water service. It's treated Great Lakes water, so there's relatively low alkalinity relative to groundwater. The carbonate problem is documented in Corrosion (Denny Jones, Prentice-Hall), but no mention is made of the problems with demin water etc.

Though I'd never use copper for demin process feed water, I still don't know if I should be concerned about spec'ing copper for BFW, condensate handling or trace service, or not- clearly, you can get away with it sometimes and not others. Copper corrosion has got to be worth megabucks- surely this has been studied on more than just an anecdotal basis?