The formed oxide surface is extremely hard to wet with the braze filler metal at the liquid temperature of the filler metal. To this end, it is a requirement of a successful braze that the metal surface be free of surface oxides. This is particularly true of aluminum, which has an affinity for oxygen.
This is the reason for the use of fluxes and atmospheres to reduce the oxide film on the metal surface. If the surface cannot be wetted, the filler-metal flow on the metal surface will not occur. This affects the joint metallurgy of what should be a successful joint. The result of the attempted braze will be voids within the filler-metal flow area, thus reducing the tensile strength of the so-called brazed joint.
Hydrogen is a reducing atmosphere. However, care must be taken when selecting hydrogen as an atmosphere for reducing the metal surface of oxides because of the potential for hydrogen embrittlement. The hydrogen gas must be of a very low dewpoint to prevent it being contaminated with moisture. The moisture will have the potential to accelerate the formation of metallic surface oxides that would be contaminated with moisture/oxygen.
Because of its small size, hydrogen is the most diffusible gas, and it will readily diffuse into the surface grain boundaries of the metal being brazed. If oxygen is present at the surface of the metal (and if the temperature is high enough), the hydrogen can react with the oxygen to form water vapor molecules at the grain boundaries. The depth of diffusion is dependent on how much hydrogen and oxygen is present and how long the metal takes to get to the braze temperature and its complete cycle (heat-up and cooldown).
In steel brazing, the oxygen will not diffuse but only react with the surface. The hydrogen, however, will diffuse and will tend to form hydrogen molecules around the surface grain boundaries.
Careful selection of the flux is critical. Some fluxes, such as the fluoride fluxes, can reduce some oxides and, in conjunction with hydrogen, reduce the oxides further (dependent on the braze temperature selection). The chemistry formation of the oxides and the rate at which an oxide reaction forms will be dependent on the base metal itself and the oxidizing agent.
The metallurgy of the brazed joint can only be seen with X-ray or perhaps with ultrasonic methods. This means that the joint integrity can only be seen at the surface of the braze filler metal. One would then have to rely on mechanical testing or destructive metallurgical examinations such as tensile testing. Therefore, joint preparation, as far as oxide contamination is concerned, is of a critical nature. This leads the way to vacuum, or low-pressure, brazing techniques, which will keep the metal surface free and clear of oxide formation.
By: Dan Kay
Posted: March 30, 2010 4:41 PM
Certainly you start out well by reminding the readers that surfaces must be free of oxidation in order to be brazed. A number of other points concern me, and need to be clarified:
1. All metals, as they are heated, have a greater and greater affinity for oxygenthe hotter they get, and thus all metals need to be appropriately protected during heating.
2. To prevent confusion, be very careful about the way fluxes are mentioned as far as their ability to "reduce the oxide film" on surfaces to be brazed. I encounter too many situations in brazing all the time where folks actually use flux as a method of cleaning the surface, thinking that the flux will react with whatever is on the surface "to clean it up and make the surface ready to braze". That's totally the wrong way to use flux. A flux is supposed to be placed onto carefully cleaned surfaces in order to keep them clean, and to prevent further surface oxidation during the brazing cycle.
3. You mention the presence of voids in the joint if the joint is not properly cleaned. Although that may be true, it is also quite possible that there may be no brazing filler metal (BFM) flow at all into some of these joints because of surface contamination. BFM does not want to bond to, or flow over, any oils, dirt, greases, or oxides on the faying surfaces to be joined. Readers need to know that POORLY cleaned joints may not braze at all, let alone have voids in them.
4. Your point about hydrogen embrittlement is important, and needs an article on its own, since there are so many variables to that phenomenon. But readers need to know when the presence of hydrogen atoms will be a problem, and when it will NOT be a problem. Certainly hydrogen should not be used when brazing certain base metals in which irreversible "embrittlement" will occur, such as with tough pitch copper, or any base metal containing significant amounts of oxygen in its matrix. But with many other metals the hydrogen, when held at intermediate temps for a while, migrates out of the metal, and no further problem exists. This is true in many steels, and even in many titanium alloys, as just a couple of examples.
5. Be careful about implying that brazing fluxes can be used in hydrogen furnace atmospheres to get the good benefits of each in surface oxide reduction. That's very poor brazing practice. Flux should be used when brazing out in air (torch or induction, etc.) with most kinds of BFMs, and also when furnace brazing in an air-furnace. But, it is NOT recommended that folks combine paste flux with any kind of non-oxidizing furnace atmosphere (such as hydrogen, nitrogen, argon, etc.), since that defeats the purpose of using the non-oxidizing atmosphere to begin with. Plus, it necessitates additional clean-up steps afterwards to remove the flux.
6. Be VERY careful about ever recommending X-ray as a valid inspection technique for brazements! To even be able to "see" the inside of a brazed joint via x-ray, the 2-percent rule should be used, which states that the brazed-joint thickness must be at least 2-percent (or greater) of the cross-section thickness through which the x-rays are being sent, or the joint cannot be readily seen by x-ray, let alone any imperfections in the joint. Thus, if a brazed joint is about 0.002" thick, then the total thickness of the brazed assembly should not be greater than 0.100" (that's less than an eight of an inch thick!). I see too many people trying to x-ray large parts, and such attempts are not meaningful. Ultrasonics is the preferred method for brazements, and even then, there are many limitations.
Yes, we must all do what we can to get the word out that surfaces to be brazed have to be kept clean, and free from oils, dirt, greases, or oxides, if anyone expects to have success in brazing. That's always VERY important for good brazing success!
Dan Kay
Kay & Associates Brazing
860-651-5595