Plating corrosion on gold over nickel plated 6061-T6 aluminum 2

[THA88]

Scenario:
2 machined parts, both made of 6061-T6 aluminum, electroless nickel plated (MIL-C-26074, class 1, grade B - .002-.0025 inch thk), and gold plated (MIL-G-45204, type 2, class 1, grade D - .00005 to .0001 thk) were bolted together.

Problem:
A leak deposited water between the part surfaces and through cyclical heating and cooling (over a 4-week period) caused crystalline formation - ie possibly aluminum hydroxide pitting corrosion over the entire surface. This water may have some type of corrosion inhibitor since its from a closed loop cooling system with copper lines.

Question:
How can corrosion occur? Could this be a chemical and electrical process? I believe the plating is high quality but am surprised the gold and nickel could be breached so easily. Any help would be greatly appreciated!

 

[mshimko]

Galvanic corrosion is a possibility anytime you have different materials that are electrically connected and wetted.

I would think it very possible for the gold to cause caused galvanic corrosion of the nickel and/or aluminum.

 

[kenvlach]

Galvanic corrosion of aluminum (white hydroxide) from nickel plating is common if porous or cracked plating. Nickel is only a barrier coat, not sacrificial (like Zn on Fe). The corrosion can be quite rapid if the plating adhesion is poor. Cu (& Cl^-) ions in solution are also bad for aluminum.

Ask for a copy of the plater's monthly qualification results for salt spray (per ASTM B117) & adhesion testing (180^o bend). The salt spray required per MIL-C-26074 is 100 hours. Usually only a problem with low-phosphorus EN, mid-phos normally passes & high-phos passes with flying colors.
Heat treatment of the EN improves adhesion but worsens corrosion resistance. Although Class 1 specifies no HT, maybe done accidentally or via your heat cycle.

Are you sure about your EN Grade B thickness of 0.002-.0025 inch? Per MIL-C-26074E,
1.2.2 Grades. The nickel coatings shall be of the following grades:

Grade A - 0.0010-inch minimum thickness. {normal for Al alloys, see 3.3.2}

Grade B - 0.0005-inch minimum thickness. {normal for Co, Cu, Ni, Ti and Be}

Grade C - 0.0015-inch minimum thickness. {normal for Fe-based alloys}

I've added the info in italics.

What is the temperature cycle? Low-, medium- or high-phos EN?

Perhaps the temperature caused EN crystallization from the amorphous state, which cracked the thick EN. Can tell by microscopy & x-ray diffraction, or from salt spray test after thermal cycling.

Described the bolting. Any indication corrosion started at the edge of the unplated holes?

What is the cooling fluid? Maybe switch to auto antifreeze/coolant.

Was the gold added for corrosion resistance or for electrical purposes? Probably too thin to fix any EN problems. I don't think it played a significant role. Gold can cause galvanic corrosion of Ni (green hydroxide or blackish hydrated oxide), but in this case, the Al sacrificially protects the Ni.

 

[THA88]

Ken,
Thank you for such an excellent and extensive reply! This is a complicated issue because there are many things happening at once.

We are using one of the few platers in our area and unfortunately they do not conduct monthly salt-spray qualification tests. You may be correct that the plating is suspect. What’s funny is that the areas of least corrosion are around bolt-hole locations. It is as if the water could not permeate the localized area around the bolt hole. The plater is certain that the parts were not heat treated.

The plating spec. was actually .0005 to .001 and not the .002-.0025 I had specified (poor memory). However, as you point out it should be .001 min. and also, the gold should be .0001 min. Both the nickel and gold are used only for corrosion resistance. Interestingly, when some of the less corroded surfaces are wiped off, there is a green residue on the cotton.

The temperature cycle was from 25C - 70C so crystallization of the EN is not likely, however, there are other factors:

1) Internal pressure - the two parts are under bending stress when the system is operating.

2) There are large electrical fields that would create induced currents throughout the material during operation under a pulsed format. In addition, there would be leakage current while operating at continuous current.

3) Very corrosive cooling fluid - which is probably the main culprit. It is city tap water but there are copper ions because 90% of the piping is copper. There is also some type of anti-corrosion coolant added to protect the copper pipes. I do not yet know what this is but it could be ethylene glycol. One of the two aluminum parts has cavities for cooling and all the internal faces which are open to the liquid show pitting! This is the biggest indication that the liquid itself is causing corrosion. For your info. there is a great article at:

http://www.corrosionlab.com/papers/water-system/nace1998-water-system.htm

We have collected samples of the fluids and powder and will send them out for chemical and SEM analysis. I’ll let you know what we find out. Thanks again for the great information and insight.

Regards

Tim

 

[kenvlach]

Hi Tim,
Some brief comments while awaiting your results.

Bad plating can be worse than no plating due to localized galvanic & crevice effects. Anodized aluminum with an inhibited coolant might be better.

Also, if stuck using chlorinated tapwater, see if you can switch to a marine grade alloy, such as Al 5052 or 5083.

What size is the bending stress? SCC isn't usually a problem with Al 6061, & time doesn't seem long enough for fatigue. Maybe a factor if brittle coatings. Can you estimate the maximum surface strain?

Re bolts: Steel? Usually a sealant is used on through-bolts in water pumps and such. Also, under bolt heads for the purpose of minimizing galvanic contact. Info on sealant?

Good article; thanks. Note that glycols are used for thermal properties; corrosion inhibitors are separate additives.

Green corrosion product probably contains nickel. Copper-containing corrosion is usually blue-ish. The blue corrosion mentioned in the article is perhaps a combination of green & a bit of red, from Cu +2 ions reduced to +1 or 0 (metallic) by the Al. Other possibilities exist, could even be the ethylene glycol dye colorant (pp glycol is red).

For aerospace you need certified plating; plater must have passed a survey by the prime contractor. Guess this is internal plant equipment.

Ken