Fig. 1 Walls and doors of aluminum-brazing vacuum-furnace coated with a special high-temp paint to prevent Mg from sticking to them. (Photo courtesy of Steelcraft Inc., Stratford, Ontario).

As mentioned in a previous blog-article, magnesium (Mg), often referred to simply as “mag”, is a highly effective “getter” that is used when vacuum-brazing aluminum. Because Mg is very effective at gettering (reacting with and removing) both oxygen and moisture that may be present in a vacuum-furnace atmosphere during aluminum-brazing operations, it can effectively prevent (or minimize) the reaction of these elements with aluminum, thus allowing aluminum-brazing to occur. However, magnesium is a highly combustible metal, and when it condenses on the walls of a vacuum-furnace during aluminum brazing operations, extreme caution must be exercised in removing the condensed mag from the furnace walls during subsequent furnace clean-up, so that no sparks are generated which could cause rapid ignition of the condensed magnesium, resulting in explosive combustion, and even death. To prevent this, coating the walls of the vacuum furnace with a “non-stick” surface, such as shown in Fig. 1 may be highly effective.

If Mg Is Dangerous, Why Is It Used for Furnace “Gettering”?

Mg is one of the few metals that reacts with oxygen more aggressively than does aluminum. It can also react with aluminum-oxide and take the oxygen away from the aluminum, resulting in a clean, oxide-free aluminum-alloy surface that can be readily brazed. The typical reactions that occur with magnesium (Mg) during the aluminum vacuum-brazing process are:

Mg + H2O → MgO +H2
Mg + O2 → 2 MgO
Mg+ Al2O3 → MgO + Al

From these equations it can be seen that when Mg reacts with oxygen and water-vapor it does so by taking the oxygen and “bonding” the oxygen to itself to form MgO. Yes, magnesium is indeed a strong and effective gettering material for use when vacuum brazing aluminum components.

When the rapidly expanding aluminum base metal causes the lower-expanding oxide layer to break apart, any free oxygen in the atmosphere wants to quickly try to re-form new Al-oxide to “heal” the breach.

Fig. 2 When the rapidly expanding aluminum base metal causes the lower-expanding oxide layer to break apart, any free oxygen in the atmosphere wants to quickly try to re-form new Al-oxide to “heal” the breach.

It is interesting to recall that Al-oxide is actually a very stable “ceramic” type of material, and as such, expands/contracts at much lower rates than most metals, and certainly at a far lower rate than aluminum itself. This big difference in expansion rates between aluminum and the oxide layer on its surface can actually become a real “plus” when vacuum-brazing aluminum components because as the rapidly expanding aluminum base metal heats up to brazing temperature, the lower-expanding oxide layer will crack and break apart (similar to ice breaking apart on rivers in early spring), exposing clean oxide-free aluminum base-metal to the BFM when the BFM melts and flows at brazing-temperature.

And because no vacuum brazing “environment” is actually a “perfect vacuum”, there will always be (relatively speaking) a lot of oxygen atoms and water-molecules (which also represents oxygen) present in that furnace atmosphere at brazing temp, and that oxygen will want to quickly move into those cracks/openings in the Al-oxide layer to form new aluminum-oxides that can “heal” the cracks in order to make the Al-oxide layer continuous once again, as shown in Fig. 2.

Therefore, to prevent this “healing” process from occurring, magnesium (Mg) is added to the aluminum base-metal, or to the brazing filler metal (BFM), or it may be added as a separate entity (as chips or powder) in a small crucible placed near the parts being brazed, so that the Mg can react with the oxygen and water-vapor before those elements can react with the clean aluminum metal at the bottom of the cracks in the Al-oxide layer.

For commercial brazing, pure magnesium (in the form of chips, powder, etc.) may be expected to volatilize and become most active at temperatures above approximately 1000F (525C). When the Mg is alloyed into either the aluminum base metal or the BFM, then the temperature needed for volatilization may be in the range of about 1050-1060F (570C), since the Mg is alloyed with the aluminum and silicon in the base materials.

Mg-condensation as a white/grey coating on component.

Fig. 3. Mg-condensation as a white/grey coating on component.

Then, when Mg volatilizes, it forms a gaseous Mg-cloud that, as mentioned earlier, will grab onto any available oxygen (to form MgO) before that oxygen can react with the clean aluminum base metal surfaces revealed at the bottom of the cracked Al-oxide layer.

Because the vacuum-furnaces that are used for aluminum-brazing are what are known as “cold-wall” vacuum furnaces, the Mg gas will condense on those cooler furnace walls, building up a layer of pure magnesium over time, which must be removed. As mentioned at the beginning of this article, Mg-buildup can be dangerous and highly pyrophoric, presenting a danger to furnace operators if that buildup is not removed. Fig. 3 shows an example of condensed Mg on a vacuum furnace component.

Removing Mg-Buildup from Furnace Walls

There are a number of methods available for handling the Mg condensation on furnace walls:

1. Special removable inserts along the furnace wall, as shown in Fig. 4, allow the Mg to condense on their surfaces. These inserts, when removed from the furnace, can easily be cleaned outside the furnace, and then re-inserted into place in the furnace.

2. Coat the inside walls of furnace — There are a number of high-temperature coatings that can be applied to the inside walls of vacuum-furnaces to keep the Mg from sticking to the walls:

Mg-condensation board inserted into furnace at left can be removed to scrape off Mg and then re-inserted into furnace. Notice the Mg coating on the board does not stick, and is already wanting to peel off the surface of the insert. (Photo courtesy of PV/T.)

Fig. 4 Mg-condensation board inserted into furnace at left can be removed to scrape off Mg and then re-inserted into furnace. Notice the Mg coating on the board does not stick, and is already wanting to peel off the surface of the insert. (Photo courtesy of PV/T.)

A. Coat with a high-temp paint (see. Fig. 1 at the beginning of this article). These high-temp paints can handle the temperatures used for aluminum brazing, and present a non-bonding surface for the Mg vapors when they cool down and condense on those coated surfaces. One such manufacturer, Glyptol Inc., in Massachusetts has been providing such coatings for a number of years.

When the furnace hot-zone is removed from the vacuum-furnace chamber, the coated walls can be scraped with a non-metallic scraper (or a bronze tool), and the condensed Mg will come out without too much effort. The painted surface may also be scrubbed with abrasive cloth materials to remove the Mg, but re-painting of damaged areas may be required.

B. Coat with a high-temp brazing stop-off. Some brazing shops have used some of the commercially available brazing stop-offs to coat their aluminum-brazing vacuum-furnace walls to prevent the Mg from sticking to the furnace walls, and have reported much success with it. Because of the proprietary nature of all brazing stop-off materials, it is strongly recommended to evaluate different commercial stop-off materials to find out which one works for you.

3. High-temp furnace burn-out. Heat up empty contaminated furnace to high enough temp to volatilize the Mg coating on the walls, thus “burning” the Mg off those surfaces. Be careful with this procedure, since the walls of the furnace must become hot enough to cause the Mg to volatilize. Experimentation is required to find out what kind of “burn-out” cycle will work in your furnace.

4. “Flaming Newspaper” technique. NOT RECOMMENDED! Believe it or not, some people have actually rolled up a newspaper tightly, ignited the end of it (like a torch), thrown it into the furnace chamber and quickly locked the furnace door so that the flame would ignite the Mg and quickly burn it. This kind of foolish action is EXTREMELY dangerous, and I have heard that at least one death has resulted from it.

Problem solve and improve with our technical experts.