Brazing Heat- and Corrosion-Resistant Alloys for Robust Service (part 1)
Brazing is a highly versatile and cost-effective high-temperature method to permanently join a wide range of heat- and corrosion-resistant alloys for service in such diverse industries as automotive, aerospace, medical, electrical, hand-tools, cutlery and food-handling. Brazing methods used range from hand-held torches for brazing one part at a time, to highly sophisticated, automated equipment that can braze hundreds or thousands of parts on a continuous basis or all at one time in large batch furnaces.
Brazing technology and research is steadily moving forward, and many end-use applications of heat- and corrosion-resistant alloys (referred to as HCR alloys in this article) are directly benefiting from these advances. HCR alloys today are not only being brazed to a variety of different metals but also to ceramics for use in conditions of extreme temperatures and/or corrosion. One proprietary recent example of this requires operations exceeding 2700°F (1500°C) in steady service. Brazing indeed has a bright future.
What are HCR alloys?
When referring to HCR alloys, what kind of metals are we talking about? The field is large and includes standard stainless steels, Inconels, Hastelloys, Rene and Haynes alloys, and titanium alloys (among others). The key consideration here is that each of these alloys is used in end-use applications where they must be tolerant of and strongly resistant to heat and/or chemical corrosion over extended periods of time.
To achieve this, the chemistries of these HCR alloys usually contain significant amounts of nickel and/or chromium. They often contain additions of molybdenum, tungsten, niobium, titanium and/or aluminum, and others. A few typical HCR-alloy chemistries are shown in Table 1.
It must be noted that any brazing filler metal (BFM) chosen to join an HCR alloy must not only be chemically compatible with that HCR alloy, but it must also be able to bond to and diffuse into that alloy and then be capable of handling the same end-use conditions to which the HCR alloy is exposed. It also needs to be understood that properly made brazed joints will be as strong as, or stronger than, the HCR alloys themselves. That means that if a component were to fail in service for any reason, the failure should occur in the HCR-metal itself and not in the brazed joint. Is this a tall order? Perhaps not.
For the successful brazing of HCR alloys, the principles of brazing must be thoroughly understood and conscientiously applied. If not, disaster awaits. I have seen too many examples, in too many industries, of half-hearted attempts at following the required steps for proper brazing that resulted in poor-quality brazements and premature failure of brazed assemblies out in the field.
Achieving High-Quality Brazing
Successful brazing is not difficult to achieve. There are a number of steps needed to achieve high-quality brazing of HCR alloys. The two most important steps are proper joint design and cleanliness of joint surfaces.
Proper Joint Design
Don’t brush this one off. This, in my experience, is the most important, and it is too often overlooked by designers. It is often assumed that if a joint can be welded, then that same joint design could probably be brazed. That’s totally wrong. Good joint design for brazing is very different from a good welding design. As shown by the criteria in Figure 1, a good brazing designer will be looking inside the joint where the BFM is supposed to fill the inside surfaces (faying surfaces) rather than looking on the outside (fillets).
Good brazing design does not look for external fillets but instead seeks to optimize the gap clearance and total amount of joint overlap of the faying surfaces.
As shown in Fig. 2, sharp corners can cause a problem in service. Good brazing design provides for smooth, contoured shoulders at the edge of a braze joint rather than sharp corners. Sharp corners, being points of high stress concentration, are often the place where cracks are initiated in service. And since those sharp corners are often right at the end of a brazed joint, any cracks seen in the sharp corners are often immediately blamed on the braze.
Notice that the joint design in Fig. 2a resulted in a crack during service due to high stress concentrations in the sharp corner, which then progressed through the base metal due to the bending stresses the joint faced in service.
The cross-sectional drawing shows that the brazed joint itself was fine, did not fail and had nothing to do with the field failure. But, because the crack was observed to start at the external exposed edge of the braze joint, the brazement was incorrectly blamed for the failure. Instead, the designer of the joint should have been blamed for that service failure, because he/she did not properly contour the joint edge so as to spread the service-stresses (Fig. 2b).
For good brazing results, the surface roughness of the HCR alloys being brazed should be in the as-rolled, as-drawn or as-machined condition, typically in the range of 16-64 RMS (root mean square) surface roughness. This is fine for brazing. The “peaks and valleys” of that surface roughness can actually help the wetting (spreading and alloying) by the BFM on the HCR-alloy surface and help to maintain a reasonable gap-clearance between the faying surfaces inside the joint. This allows capillary flow of the BFM through the joint, even with so-called “metal-to-metal” contact of the parts being brazed (Fig. 3).
Cleanliness of Joint Surfaces
This, too, is critical to the success of the braze. I have heard too many people say something like, “Oh, don’t worry about cleaning. The furnace will take care of that.”
BFMs do NOT like to bond to or flow over oils, dirt or oxides. Each of these contaminants will prevent the BFM from not only flowing into the joint, but also from diffusing into and alloying with the HCR metal. If diffusion/alloying is prevented or degraded, then the joint strength will be marginal at best and may fail in service. A “catch-22” scenario arises in which some people then say, “See, I told you brazing isn’t really good for this part. You should have welded it.”
If designers and production personnel will take these first two important brazing criteria seriously, great brazing results can (and will) be achieved.
BFM Alloys for HCR Brazing
The BFMs that are used to join HCR alloys should be selected so that, when brazed, they are able to meet the end-use service conditions in which the HCR alloys will operate. Proper selection of the BFM is important so as to prevent excessive base-metal erosion during brazing or re-melting of the BFM in service. This would obviously be very undesirable.
A few of the HCR alloys commonly used in industry are shown in Table 1. Please note that there are many different commercially available HCR alloys that are not shown in this table. The few that are shown in Table 1 are merely there for brief illustrative use in this article. Similarly, the BFM alloys shown in Table 2
are only a small fraction of the many excellent commercially available alloys for joining HCR base metals.
Many of the metallic elements in traditional HCR alloys are easily brazeable. Nickel, chromium, tungsten, molybdenum, niobium and copper are all easily wetted by most of the readily available BFMs, such as the standard nickel-based, cobalt-based, gold-based and even the copper- and silver-based BFMs shown in Table 2.
As metals are heated, they have a greater and greater desire to react with oxygen and form an oxide. This is true for ALL metals, including gold. Of course, not all metal oxides are stable at room temperature under ambient conditions. Many oxides, such as those of nickel, gold, etc., are not stable (and will not form) under such conditions and, thus, are not of any real concern to us.
Other metals form oxides that are quite stable under ordinary room-temperature conditions. They will be readily observable and will have a negative effect on brazing unless dealt with appropriately. Many of these oxides are not stable when heated, however, and can easily be dissociated into pure metal with the release of the oxygen. Such is the case with iron, copper and silver at moderately low temperatures. Even the oxides of tungsten and molybdenum can be readily dissociated on the way up to brazing temperature and should present no real brazing issues.
The oxides of chromium, manganese, niobium, titanium and aluminum are quite stable, however, and require careful elevated-temperature controls in order to appropriately and effectively deal with and remove those oxides prior to the actual melting and flowing of the BFM. This is very important because BFMs do not like to bond to or flow over any oxides on the faying surfaces of the HCR alloys. It is very important, therefore, that the faying surfaces of the HCR alloys be thoroughly cleaned prior to assembling for brazing, thereby removing all oil, dirt, grease, fingerprints, oxides, etc. so that the molten BFM can flow over each of the faying surfaces and diffuse into those surfaces to form a strong metallurgical bond. Such BFM flow and diffusion cannot occur on uncleaned surfaces.
Because many of the HCR alloys that are brazed contain chromium – and some of them also contain small amounts of niobium, manganese, titanium and/or aluminum – the tenacious oxides of these metals may interfere strongly with brazing and, thus, require tight controls on brazing procedures to ensure high-quality brazements.
Each of the curved lines in Figure 4 represents the threshold for oxide reduction for a given metallic oxide, based on the brazing temperature (shown along the horizontal axis of the chart), and either the dew point of the furnace atmosphere (the left-side vertical axis) or the level of vacuum used in a furnace (the right-side vertical axis).
For instance, look at the line for chromium-oxide (Cr2O3) in the center of the chart. 304-stainless contains about 18% chromium in its chemistry. So, as that HCR alloy is heated, the chromium in that alloy will progressively oxidize more and more until it reaches the Cr2O3 line. At temperatures beyond the Cr2O3 line, the oxides will tend to be reduced (dissociated) more and more.
At all temperatures and dew points/vacuum levels to the right of that curved line, chromium oxides should effectively be removed, thereby allowing molten BFM to wet and braze the faying surfaces of that alloy.
For example, look at the chart and determine if you should be able to effectively braze 304-stainless if you are operating a brazing furnace at -30?F dew point and 1800?F (950?C) brazing temperature. The point where the horizontal line of dew point crossed the vertical line of the chosen temperature will still be to the left of the Cr2O3 line, and you would expect to have real difficulties in effectively brazing 304-stainless with those furnace parameters.
Instead, you can readily see that you must use a much drier furnace atmosphere and a much higher brazing temperature in order to ensure a quality braze in which the chromium-oxide has been removed. Experience indicates that you should be a distance equivalent to at least the diagonal of one full “block” to the right of the Cr2O3 line on the chart if you expect to be able to reduce (dissociate) Cr2O3 from the faying surfaces to be brazed. Brazing at 1950°F (1000°C) and a dew point of -60°F (-55°C) or drier will achieve that in this example.
Thus, if you are brazing Inconel 625 or Hastelloy X, both of which contain chromium, similar precautions would be needed for furnace brazing, and a reasonable combination of temperature and furnace dew point (or vacuum level) can be determined from the chart to ensure high-quality brazements.
Problems come up, however, when you are attempting to use the chart to help you specify a temperature and vacuum-level combination that will allow you to braze HCR alloys such as Inconel 738 (see chemistry in Table 1). Notice that this alloy contains some titanium and aluminum, which was added to achieve certain metallurgical benefits in end-use service (hardenability, corrosion resistance, etc.). In order to be brazeable, we have to remove (dissociate/”reduce”) any oxides of titanium and aluminum that will normally form as we heat Inconel 738 (Inc738) to brazing temperature in a vacuum furnace.(Yes, there will still be enough oxygen in a vacuum furnace at brazing temp to cause significant oxidation of Inconel 738.)
Shouldn’t it then be merely a matter of brazing the Inc738 at a temperature/vacuum-level combination that is at least one block to the right of the titanium-oxide (TiO) curve in order to dissociate the TiO? Well, that’s a problem because that would require a brazing temperature well above the melting point of the Inconel 738 itself. This becomes the reason why you would electrolytically nickel plate the Inc738 faying surfaces prior to assembly for brazing. The plating will shield the surface of the Inc738, preventing any oxygen from being able to reach any of the titanium or aluminum atoms in the Inc738 to form oxides.
The nickel plating, as was mentioned earlier, will not oxidize at the brazing temperatures involved. Therefore, it can be a highly effective way to allow normal brazing of Inc738 components for aerospace applications since the molten BFM can readily wet and spread over the nickel-plated surface throughout the braze joint.
Always remember, however, that the electrolytic nickel plating will slowly diffuse into the base metal and the BFM, so you must complete your brazing process before the nickel plating has fully diffused away, once again exposing Ti and Al in the HCR alloy to any oxygen in the furnace atmosphere.
In part 2 in September, we continue our discussion looking at topics such as carbide precipitation, honeycomb brazing and new developments in joining HCR alloys to ceramics. IH
Dan Kay and Robert Peaslee, “Furnace brazing: How to Boost Quality and Productivity,” Heat Treating Magazine, April 1981. Author’s note: Although developed for hydrogen furnace atmospheres, the chart can also be used for nitrogen, argon and helium atmospheres. It is recommended that you operate at a point at least one diagonal (of the boxes in the grid) to the right of the given oxide you wish to dissociate. Always use the most sensitive metal (farthest to the right) in the HCR or BFM compositions to set your brazing parameters.