A. You are encountering what is termed erosion. In testing conducted years ago, we divided the erosion mechanism into three divisions:

    1) Solid-state diffusion, where the melting point depressant in the filler metal diffuses into the base metal. With the nickel-based filler metals, this was primarily the diffusion of boron into the base metal.

    2) Solution, when there is a large amount of nickel-based filler metal, such as a large fillet, enough boron would diffuse into the base metal to lower its melting point, and some of the base metal would become liquid and mix with the filler metal. After solidification, the only way to determine whether the base metal had dissolved into the filler metal is to section, polish, and etch. This will reveal the amount of solution that had occurred.

    3) Erosion is the final stage. This occurs when the fillet and melted base metal flow away from their original locations, leaving a thinned base-metal section, or a hole, in the base metal. This visible thinning, a hole, or streaks where filler metal ran down a part, or where the filler metal was sucked into the joint leaving a thinned area or hole, etc., is termed erosion.

    The furnace cycle is sometimes very important in preventing erosion.

    Consider an assembly consisting of thin tubing to be brazed onto a heavy fitting. A large quantity of filler metal must be placed against the tubing at the joint. When the assembly is heated rapidly to the brazing temperature, the tube will come to heat first and the filler metal will melt, while the heavy fitting lags in coming to brazing temperature. This condition, where the hot tube and molten filler metal are in contact for an extended time, promotes interaction between the molten filler metal and the tube. When the heavy fitting finally comes to the brazing temperature and capillary action sucks the filler metal into the joint, an eroded area will usually be seen on the tube above the joint.

    The solution to this problem is to hold the furnace temperature down to about 50°F below the solidus of the brazing filler metal. The hold time should be long enough to allow the heavy fitting to reach the same temperature as the tube. Then the furnace temperature can be raised to the brazing temperature, permitting the filler metal to melt and flow directly into the joint with only a small fillet remaining. This method minimizes the interaction between the filler metal and tube, thus preventing erosion.

    There is some concern that erosion will occur in the brazed joint. This will not occur, as it takes a large excess of brazing filler metal to dissolve the base metal and, if flowing away, to cause erosion. By holding the parts at the brazing temperature, we can produce diffusion in the brazed joint to strengthen the joint and increase the remelt temperature. This is a beneficial reaction.

Fig. 1 ‹ Puzzling even for the experts are the curious round-bottomed corrosion grooves radiating from the central brazed joint. The grooves were stained for clarity. This specimen is about 3/4 in. wide.

Solution and erosion are controllable, and should not occur with the proper application of filler metal and the proper furnace cycle.

    I liken this problem to my welding days. If the welder was continuing to undercut during welding, we blamed the welder because he or she did not know how to control the process. Similarly if we know how to control the brazing process, erosion should not occur.

    There is a very obscure type of erosion that we very seldom encounter. This is surface erosion caused by the filler metal running out of the joint and across the surface of the base metal, producing a spreading or tree-like effect ‹ Fig. 1. The branches get progressively smaller as they spread out. This example is a copper-tin-bronze base metal brazed with BAg-8 brazing filler metal at about 1500°F (815°C). A 1/8-in. (3.2-mm) diameter filler metal wire was placed at the step and the joint is 90 deg to the plane of the picture.  The branching ran both ways from the joint.
    We are still trying to find out what is the driving force that makes the branching continue to flow. One would think it would pick up enough base metal to freeze off and not continue to flow, but it continues. Micros indicate they are shallow round-bottom erosion grooves. We are still looking for the answer to this one. If you have some ideas on what causes this effect please share them with me.