Fig. 1 Cross-section drawing of an actual 6061-aluminum tube-in-fitting joint.

Figure 1 shows a cross-sectional drawing of an actual tubular joint made from 6061-aluminum, furnace brazed using 4047-aluminum brazing filler metal (BFM). Notice how the brazing filler metal (BFM) has formed a nice concave-shaped fillet on the outside of the joint (right side of the drawing) but also displays a nice fillet at the far left (bottom) of the joint. But there are also a lot of trapped-air voids in the middle of the joint. How then did the BFM get all the way down to the bottom of that long braze-joint? Is that even possible?

We were told that a ring of 4047-aluminum was used on the outside of the joint, but the rest of the joint was a bit of a mystery that required investigation.

Question for Readers: How did the BFM get all the way down to the bottom of that long braze-joint? Is that even possible for that externally applied 4047-ring to do that? Aren’t we supposed to be operating with the Rule of Brazing that states: “Feed the BFM from one end of the joint, and inspect the other end of the joint to verify that the molten BFM has pulled all the way through”?

First of all, notice that the joint depth is quite extensive. Using the wall thickness of the inner tube as “T” (thickness), it appears that the tube was pushed down into the fitting to approximately eleven (11) times the wall thickness (“T”) of that inner tube. That’s a long way!

Extensive testing of brazed joints over the years has shown that a full-strength brazed joint is achieved when the amount of overlap is about 3T-to-6T for most metals (EXCEPT aluminum), where “T” is the thickness of the thinner of the two members being joined, as illustrated in Fig. 2. Extensive testing has also shown that when such overlaps are used, any failures to that brazed joint occur in the base metal outside of the joint, and not in the joint itself! Yes, part of the reason is due to the double-thickness of the joint as compared to the single-thickness of the metal on either side of the joint area, but it is also due to the amount of overlap used, since overlaps of only 1-to-2 “T” invariably have failed in the joint itself. Only when the amount of joint overlap exceeded approximately 2.3-to-2.5 “T” (for most metals) was the joint immune from failure in the joint itself. Very interesting results.

Typical amount of overlap of two pieces of metal being brazed together, where “T” is the thickness of the thinner of the two members being joined.

Fig. 2 Typical amount of overlap of two pieces of metal being brazed together, where “T” is the thickness of the thinner of the two members being joined.

Aluminum is the overlap-exception. When similar testing was done for aluminum, the results were uniquely different. Because the melting temperature of the aluminum BFM is so close to that of the parent base-metals being joined, the molten BFM reacts with, and aggressively alloys with, the base-metal (and vice-versa), rapidly forming a strong metallurgical bond and alloy-compositional changes in the joint area — so rapidly in fact, that the molten BFM does not (and really cannot) flow extensive distances through the joint by capillary action. Because of this extensive diffusion and its rapidity, aluminum overlap distances are typically 1T-to-3T maximum!

Aluminum overlaps should be 1T-to-3T maximum!

Overlap for aluminum brazed joints should typically be only 1T-to-3T

Fig. 3 Overlap for aluminum brazed joints should typically be only 1T-to-3T

So, back to question at the start of this article — how did that aluminum-BFM get all the way down to the bottom of that aluminum brazed joint shown in Fig. 1? The obvious, and only answer is that the BFM was not only placed as a BFM-ring on the outside of the joint but also HAD to be placed at the bottom of the joint as well, so as to form nice “fillets” at both ends of the joint! And our investigation showed that this was actually the case. Early tests of that braze joint revealed to the brazing-shop that the aluminum-BFM could not reach the end of the joint, and left the bottom edges open so that contaminants could be trapped in the open space at the end of the joint. So, they opted to put BFM at each end of the joint to seal the joint ends adequately.

Difficulties: By placing the BFM at each end of the joint, the brazing shop caused problems:

1. Trapped air inside the joint because of the molten BFM barriers on each end of the joint preventing the expanding air inside the joint from finding an escape path to the outside. Thus, a large trapped-air “bubble” remained inside the joint. This might have been alleviated by putting some circumferential holes in the inner tube, allowing this expanding air to escape, as shown in Fig. 4. As the air escapes, the molten BFM can follow and should be seen filling those vent holes. The holes could have been drilled into the tube at about 120-degrees apart, and perhaps on two or three different points along the tube-length, as shown.

2. Excessive length of aluminum tube. Sometimes designers make tubular joints very long (deep) in order to provide dimensional stability during assembly so that the inner tube does not “wobble” when inserted all the way into the fitting, thus allowing a more “rigid” assembly for stability during manufacture. Although this might be nice from a design and assembly perspective, it can – and does – have a potentially negative effect on the ability to effectively braze the entire length of the tubing (if required)!

Place vent holes in the inner-tube wall to allow expanding air to escape from inside the joint, so there will be less trapped-air in joint.

Fig. 4 Place vent holes in the inner-tube wall to allow expanding air to escape from inside the joint, so there will be less trapped-air in joint.

Conclusion

When dealing with aluminum tubular assemblies, remember that the molten aluminum-BFM can travel only a very limited distance through that capillary space between the tube and the fitting due to the extensive reactivity between the molten BFM and the base-metals it is joining. Thus, only a short overlap of about 1T-to-3T is needed, where “T” is the wall-thickness — or sheet thickness — of the thinner of the two members being joined. IF the designer needs to exceed this amount of overlap for whatever reason, and it is determined that BFM must be applied from both ends (for sealing purposes, or for vibration resistance, etc.) — although feeding molten BFM from both ends of a joint is NOT recommended — then the designer MUST provide vent holes to allow expanding air trapped in the joint to escape, as shown in Fig. 4, if any trapped voids are to minimized.

Problem solve and improve with our technical experts.