I ran across a nice article about the different ball bearing metal and plastic alloys they use and where they use them.
http://www.precisionballs.com/ball_m..._selection.htm
Now then in this article they state that steel corrosion will increase in contact with a more anodic metal such as copper. So the effect may still happen with the ball bearings.Also the ball bearings are made from a different steel alloy than the races are and then there is the bearing cage alloys to consider as well. So we wind up with several different metals in contact with each other. Throw in some combustion byproducts, raw fuel residue and some moisture and we have a recipe for corrosion to occur. ref
http://www.dieselduck.ca/machine/04%.../corrosion.htm
The example they mentioned was having a steel bolt fastening a copper plate to the hull of a boat. The steel bolt would corrode very rapidly in that environment. You would normally think that the copper plate would corrode faster but instead it was the steel bolt that rusted away.
But then it may all hinge on that the company spec'd the wrong bearings to be used in their engines, or they wound up getting a bad batch of bearings that they should not have used in the first place.
<font face="Trebuchet MS">
Designers who want to prevent corrosion usually like to make structures and devices out of corrosion resistant materials. However, they may not consider the interaction between the different materials that they choose. For example, some aluminum alloys do not corrode very fast in seawater, and are used for boat hulls. Some bronze alloys also do not corrode very fast in seawater, and are used for propellers. As long as the propeller does not come in electrical contact with the hull, everything works well. But if the two come in contact through a bearing, gearing, or the boat engine itself, the galvanic series tells us what will happen. The aluminum is very negative compared to the bronze, so the electrical contact will cause the aluminum hull to be an anode and its corrosion rate to increase, causing heavy pitting and eventual failure of the aluminum hull.</p>
The galvanic series tells us that the more negative metal will corrode more quickly when electrically coupled in seawater, but not how fast. Two metals far apart in the series will not necessarily experience more corrosion than two metals close together. Finding the rate of corrosion in a galvanic couple requires knowledge of polarization, the ability of a metal to change voltage while accepting or giving up a certain amount of electrons. A metal that polarizes easily, that changes voltage quickly with a small amount of current, will not cause much corrosion of metals coupled to it. It also will not have much increase in corrosion when it is the anode in a couple. An example of a metal that polarizes easily in seawater is titanium. Metals that are harder to polarize, such that it is hard to change their voltage when current is applied, will cause or experience a lot of galvanic corrosion, depending on the other metal in the couple. Examples of metals that are hard to polarize include copper alloys and some aluminum alloys. So, a piece of aluminum will corrode faster if it is coupled to hard-to-polarize copper than it will if coupled to easy-to-polarize titanium in seawater, even though the voltage of the titanium is farther away from aluminum than the voltage for copper.</p>
The larger the wetted surface area of the cathode, the worse will be the corrosion on the anode. For example, steel corrosion will be increased by contact with copper, according to the galvanic series. A steel fastener used to hold a copper plate will corrode quickly, because there is a large area of copper and a small area of steel. However, a copper fastener will not cause much increase in corrosion of a steel plate because its area is so small compared to the steel. This effect was first discovered by Sir Humphry Davy when he was exploring attaching copper plates to ship bottoms to prevent barnacle growth. This leads to an interesting rule of thumb: always paint the cathode. To slow down galvanic corrosion on the anode, you can paint the cathode (which is not corroding) to decrease its wetted surface area. Painting the anode will only increase its corrosion rate at defects in the paint.</p>
Recognizing galvanic corrosion is not always easy. If a metal normally corrodes by pitting, it will just pit faster when it’s the anode in a galvanic couple. If it normally corrodes uniformly, it will do so more quickly when coupled. So galvanic corrosion can’t be recognized by the form the corrosion attack takes. Sometimes galvanic corrosion can be recognized because it is usually worse close to the cathode that is causing it. In the copper fastener case above, the steel will corrode more quickly close to the fastener than far from it. Galvanic corrosion will usually be worse near joints between dissimilar metals. But the best way to recognize galvanic corrosion is to know the order of metals in the galvanic series and look for the more positive metals in the vicinity of the corrosion failure. If they are there, they likely contributed to the problem.</p>
Now one other thought is where they use a iron anode to protect copper. Hummm. So if the aluminum alloy has copper as one of alloying metals then that could be the problem.
ref
http://en.wikipedia.org/wiki/Galvanic_anode
Since the operation of a galvanic anode relies on the difference in electropotential between the anode and the cathode, practically any metal can be used to protect any other, providing there is a sufficient difference in potential. For example, iron anodes can be used to protect copper.<sup id="cite_ref-10" class="reference">
<span>[</span>11<span>]</span></sup>
I would not normally think of this as being a problem unless the steel used in the bearings is a pretty low quality steel. But maybe it is the bearing cages that are lower quality or something since usually people report the bearing cage has come apart on them. </p></font><font face="Trebuchet MS">
</p></font>