Aluminum brazing is a growing industry around the world today, and I teach a lot about this topic in each of my brazing training seminars. A question that is often asked during those seminars relates to the ever-present aluminum-oxide coating on the surface of the aluminum and how that can be dealt with during brazing. Because brazing filler metals (BFMs) cannot bond to oxides effectively at all, how then does one need to deal with this ever-present aluminum-oxide so as to be able to braze aluminum components (Fig. 1) effectively?

Please look at the chart in Table 1, showing the relative expansion characteristics of some metals compared to some ceramics. Especially notice the expansion rate of aluminum metals near the top of the chart and the aluminum-oxide “ceramic” expansion rate nearer to the bottom of the chart. This difference is very important.

For many years it was common for many people trying to braze aluminum to spend a considerable amount of time trying to remove as much of the aluminum oxide layer as possible from the surface of the aluminum components to be brazed, often by complex acid-cleaning processes. Then attempts would be made to try to keep those surfaces as oxide-free as possible by holding them in tanks of argon or nitrogen, often at low temperatures, until brazing took place.

Unfortunately, due to the highly reactive nature of aluminum with oxygen, the aluminum-oxide layer on the parts would almost instantly re-form, and all the fancy oxide-removal procedures were not that effective at all. Additionally, since each of the argon- or nitrogen-containing boxes in which the parts were stored contained some moisture (the dew point of the gases being used), that moisture (and the oxygen in that moisture) recontaminated the surface of the aluminum as well.

Apparently, a number of years ago, some engineer somewhere noted the very significant difference in the expansion rates between the aluminum base metals being brazed and the aluminum-oxide coating on its surface and asked, “Can’t we use that expansion difference to our advantage instead of doing all that acid cleaning?” And, of course, the answer to that question is a resounding “YES!”

We will finish our discussion next time.