Figure 1. Cap dimensions do not match tube top.

I published a four-part article called "Voids in Brazed Joints" back in early 2009. Recently, a reader, Damian, noticed that more information about the fourth major reason for voids in brazed joints – poor joint fit-up – was needed. Let me address it now more completely.

It is also one of the topics covered extensively in each of my three-day brazing seminars, two of which are coming up this fall (October in South Carolina, and November in Hartford, Conn). Clickherefor more information.

Figure 1 shows a classic example of poor joint fit-up leading to voids in a brazed joint. This was encountered a few years ago at one of my client companies when someone copper-brazed a steel cap onto a specialized steel tube. The cap was supposed to fit nicely onto the formed top surface of the tube, but (as is readily apparent in the drawing) the cap's dimensions were such that it did not fit the curvature of the formed top of the tube.

Fig. 2. Left side of assembly showing very nice braze.

It is an unfortunate fact of brazing that parts seldom "self-center" during a brazing process. Extreme caution can be exercised to ensure self-centering. This can be done under lab conditions, if the brazer is carefully monitoring the brazing process visually or if furnace brazing is used where the parts that are mating to each other have been very carefully machined for tight fit-up. However, in the real world of normal brazing, in which a lot of commercial tolerances are used (or worse) and fit-ups are often less than desirable, problems with voids in brazed joints often occur because gaps are too wide.

Look again at Fig. 1. Notice that the cap is too big for the curved surface at the top of the tube. It can easily drift to one side or the other during the furnace brazing process. Since there is a layer of brazing filler metal (BFM) between the cap and the tube, the cap can "float" when the BFM has melted and becomes liquid. Unless the tube remains perfectly vertical during the furnace process, the cap can "skate" to one side or the other until solidification takes place. When parts are heated in a furnace brazing cycle, temperature differentials occur between each part of the load and within each component, since hotter sections will grow more than cooler sections. Thus, the part is likely to move, tilt, etc., as these thermal differences continue throughout the cycle.

If the cap is poorly matched dimensionally, as shown in Fig. 1, then there is a high probability (as occurred in fact) that the cap will shift and the braze will solidify in a position as shown in Fig. 1. Let's look further at what happened.

As metals are heated, they expand. Of course, as they cool, they contract. This is true for liquids as well as for solids. Another fact to hold onto is that very thin brazed joints will solidify quicker than thick joints, simply due to the huge difference in liquid mass present in the joint. BFMs in a joint usually solidify from the outside to the center of the joint. And, since diffusion is also happening in any braze-joint, a very thin joint will usually solidify much faster than a thick joint with a lot more BFM in it.

Thus, the BFM in the thin section of the joint on the left side of Fig. 1 will solidify first, locking the cap in its position on top of the tube. The solidified joint of the left side is shown in Fig. 2. Note the smooth consistency of the tight braze joint, and the lack of voids.

Fig. 3. Wide gap on the right side of the joint with lots of shrinkage voids.

Now see photo of the right side of same joint (Fig. 3). Note all the voids in the joint, which actually gave rise to a leak-path through that section of the joint (much like a spiraling, uneven worm hole).

Remember that all metals shrink when they cool down, even liquids. On the right side of the joint the liquid BFM filled the gap clearance, but upon cooling it wanted to shrink. However, the gap was "locked" because the joint had already solidified on the left side with the tight gap. Thus, the liquid BFM on the right side of the joint could not force the cap to move closer together as the BFM liquid on that thicker, right side shrank. As the liquid shrank to occupy a smaller volume and the walls of the joint could not move closer together, voids were opened up in the solidifying liquid.

This is a very real issue in a lot of brazing shops I visit. That is why I continue to "preach" about the need for good fit-up at brazing temps for brazements. For this to happen it may be necessary to tighten up on the tolerances of parts made by a shop or purchased from outside vendors. Yes, this may increase costs a bit. This cost is more than recovered in significantly reduced scrap and rework, greater productivity, and in a better name for your company in the eyes of your customers.

Dan Kay operates Kay & Associates Brazing, a brazing consulting firm specializing in problem solving and brazing-training. His company is sponsoring two brazing seminars this fall: Oct. 5-7 in Greenville, SC, and again in Hartford, CT from Nov. 16-18. Contact Dan for more information at 860-651-5595 or visit hiswebsite.