Fig. 1 Blistering under nickel-plating. (Photo courtesy of Dean Walters, Argonne National Labs)
The best “safety procedure” to use is to prevent those strong oxides from forming at all! This can be accomplished by nickel-plating the surfaces prior to brazing. Nickel-plating is compatible with most metals, and BFMs will readily bond to, and alloy with, the applied nickel-plating.
1. Two Types of Nickel-Plating
a. Electrolytic nickel-plating. This is pure nickel and melts at about 2600°F (1300°C). This is high enough in temp so that it will provide a tough, adherent surface to which to braze, and it will prevent any of the aluminum or titanium in the base metal from oxidizing; thus the plated surface should be quite brazeable!
b. Electroless nickel-plating. This is an alloy of nickel, not pure nickel. The primary added element is phosphorus (in varying percentages), but boron is sometimes used as well. In either case, electroless nickel-plating tends to melt at about 1600°F (850°C), which is often well below that needed for successful brazing. If electroless nickel plating is accidentally used instead of electrolytic nickel plating, then that electroless plating may begin to melt and flow away long before you have reached your desired brazing temp, thus allowing any of the aluminum and titanium in the base metal to become exposed to any available oxygen in the atmosphere on the surface of the part. This could then result in oxidation of the metal surface, which could then prevent brazing altogether.
2. Plating Houses Are Responsible for the Integrity of the Plating
Any good plating-company knows the importance of having clean surfaces on which to plate the nickel, and will normally take time to verify the cleanliness of the surfaces to be brazed, and will clean them as needed so that the applied nickel-plating will stick to ( i.e., electrolytically bond to) those metal surfaces.
However, not all plating-companies clean the base-metal surfaces as thoroughly as they perhaps should, and thus, there will be times when the plating does not “stick” to the surface of the parts as well as it should! This means that during a high temp brazing application, any oils/lubricants or surface-contamination on the surface of the part that was not completely removed prior to plating might actually volatilize during the brazing process, and this will cause blisters to form between the plating and the metal surface on which it has been plated, as shown in Fig. 1 for example. Then, during actual brazing, the molten BFM may bond very well to the top of the Ni-plating, but the plating itself may not be well bonded to the substrate metal below it. This happened to parts in my brazing shop on a couple of occasions. When my customer complained to me that the parts we brazed for them did not perform properly (the brazed assembly either leaked, or it failed prematurely from a strength perspective), we then examined the part when it was returned to us, and we found that our brazing was indeed perfect, but the plating had failed, i.e., the plating was not done properly, and had actually peeled off the underlying substrate to which it had been plated, thus causing the leak or tensile failure.
3. Code Words for “Quality” of Plating Done
So, based on those bad experiences, we developed some code-words for the types of jobs we got back from the nickel-plating companies with whom we worked. At that time we had our choice of many high-quality plating companies to choose from. But because each of those shops employed human-beings – and humans are prone to mistakes — we found that we could not find a “perfect plating shop”, and therefore, not all plating jobs would be perfect. We found that we needed to develop a method of checking the quality of the plating being sent to us by the plating shops because, unknown to us, some of the plated parts contained a number of small spots under the plating where the plating wasn’t completely bonded to the base metal. Then, when we brazed to those plated surfaces, we did not realize that those spots where the plating did not adhere to the base metal would cause problems for our customers. These customer complaints opened our eyes to the need for corrective action at the plating shops, and also introduced some new terms in our vocabulary when we talked about some of the plating we received. We determined that we did NOT want what we called “a Monday plating job” or “a Friday plating job”, code words for the idea that perhaps the plating shop’s equipment wasn’t quite warmed up correctly, or the people weren’t fully concentrating on their work, as if it was a Monday morning, and everyone and all the plating equipment was just warming up and having to change their focus back to the work environment. Or perhaps people were rushing too fast so as to get out on a Friday afternoon, etc. Of course, we didn’t know if that was actually the case or not, but if a plating job wasn’t perfect, we called it by those code-names (“Monday or Friday plating job”). Instead, we wanted what we called a “Wednesday plating job”, meaning that everything in the plating shop must be humming along well, the workers were being extremely attentive to doing everything absolutely well, the parts were expertly cleaned, etc. Again, these so-called “code words” had nothing to do with the actual day on which the plating was done. It merely represented our opinion regarding the quality of the plating work that we got back from those plating shops.
4. Blister Testing
Because of these problems that we experienced from time to time from the various plating shops we used for our parts, we instituted some QC-tests on the as-received plated parts in order to evaluate the quality of the plating. The so-called “Blister Test” was the most important and valuable one we developed.
a. Sample sent with parts. A sample part was sent along with all the work-pieces we sent our for plating, and the plating on the sample part was subjected to the blister-test. The sample was always of the same cross-sectional configuration and cross-sectional mass as the parts that were being plated for us. Thus, if the parts being plated were heavy and had deeply curved surfaces into which the plating had to go, then the sample we sent along with that load also was somewhat heavy and had a similar deeply curved surface. In other words, the sample that we would test had to very closely represent all the variables evidenced in the parts themselves, or else that plated sample would not be very representative of our plating needs. Thus, I wouldn’t send a small flat test coupon to be plated for subsequent blister-testing if the actual load of parts to be plated were heavy and curved, as just described above.
b. Test temps. We usually ran our plated samples in a hydrogen atmosphere at about 1000°F (500°C) to see if any blistering of the plating would occur. The time at temperature is NOT important at all, as long as you actually reach that temp. Thus, as soon as it is confirmed that you reached that test temp, you can bring it back down. A thermocouple was placed on the test piece, and as long as the sample reached about 1000°F (500°C), that was sufficient to reveal to us the quality of the plating.
IMPORTANT NOTE: The 1000°F (500°C) is not mandated, and you could perhaps use a lower temp (perhaps 700F/375C) as long as it is high enough in temp to be able to volatilize any surface contaminants on the part, so that if any contaminants are present, it will volatilize and cause a blister to show up on the surface of the plating. You may have to experiment to find the temp, and atmosphere, that works best for you in your shop.
Similarly, as just mentioned, the actual atmosphere used is not that critical, since these tests will also work in vacuum, or in an atmosphere of argon, nitrogen, etc., since the volatilization of any surface contaminant can cause a blister to form, irrespective of the type of atmosphere used. Bear in mind that in any of these atmospheres there is a sufficient amount of dewpoint or atoms of air/moisture so that any surface contaminant under the plating would volatilize and/or react with any given atmosphere’s constituents to form a blister.
c. Results. If the sample part did not show any blisters, we then accepted the plated lot and put those parts into our brazing production. But, if any blisters showed up on the plated surface of the test panel, we rejected the entire lot of plated parts, sent them back to the plating company, so that the plating could be stripped off, the surfaces re-cleaned properly, then re-plated. Again, a test-piece was sent along, and then blister-tested when the re-plated parts came back, and so on. NOTE: Never let the plating shop do the blister testing for you! That’s just like having the fox guard the chicken-coup. YOUR quality-control personnel should perform the blister testing.
Nickel-plating can be a reliable process for many companies to use in order to provide brazeable surfaces on metals that contain small additions of Ti, Al, etc. in their chemistries. But the quality of that plating should ALWAYS be verified by running a “blister-test” on a representative sample of those parts you plan to braze. Only after passing the blister-test should you allow the plated parts to be brazed.