Fig. 1 Tools some shops use to roughen surface of parts for brazing (coarse and fine-grit emery paper, steel wool, and a course file).


In recent months we have been examining the essential criteria for good brazing, many of which are, unfortunately, overlooked by designers and brazing personnel, thus causing problems in production. We are currently looking at critical issues related to proper joint fit-up, and one of these issues concerns the finish (roughness) of the faying surfaces inside a braze-joint. Surface roughness can range from highly-polished to very rough, and can, in fact, have a significant effect on the ability of molten brazing filler metal (BFM) to flow into and through a braze-joint.

As shown in Fig. 1, some brazing personnel have resorted to a variety of ways to roughen surfaces for brazing so that, in their opinion, the molten brazing filler metal (BFM) can “flow better”. This can be quite misleading, since good brazing depends primarily on proper joint design, proper cleanliness of the parts, and joint-gaps that are quite thin, three criteria that have been discussed in my prior articles in this series.

Over the years it has been demonstrated (in my own experience as well) that rather than taking a lot of extra time and effort to roughen surfaces by hand prior to brazing, the “as-received” (as-rolled, as-drawn, as-machined, etc.) surface roughness of the material coming into a brazing shop is quite adequate to achieve excellent brazing results, with no additional surface roughening or polishing needed.

An illustration of what this surface roughness might look like, under high magnification, is shown in Fig. 2.

The as-rolled, as-machined, as-drawn surface finish will add the appropriate amount of surface “roughness” to the part to aid in capillary action of BFM through the brazed joint.

Fig. 2 An assembly diagram showing how the carbide-facing was brazed to the front of the extrusion-die. The carbide had to be made in tightly fitting pie-shapes, and the BFM was BNi-2 brazing foil.

Surface roughness obviously increases the total surface area of each faying surface inside the joint, when compared to a flat, polished surface. And, due to this “roughness”, it can be seen that there are many capillary paths for brazing filler metal (BFM) to follow through between all the valleys and “peaks” on that roughened surface.

I have personally roughened test-pieces in a laboratory, using the coarsest file I could find. The resulting surfaces were rough enough to lacerate skin if a hand were quickly rubbed along the surface! When those two roughened surfaces were brazed together, the joint was very, very strong (failing in the base metal when pulled apart on a tensile-testing machine). But, due to the severity of the surface-roughness, the gap-clearance tolerance was way outside of normal guideline parameters for brazing. It was interesting to see in that lab-test how the very rough surfaces nested together, the peaks and valleys of one surface fitting into the peaks and valleys of the other surface well enough to allow BFM flow through the joint, and a good “mechanical grip” to exist between the faying surfaces in addition to the metallurgical bonding due to BFM diffusion into each of the two components being joined.

Surface Roughness Measurements

Surface roughness refers to the “texture” of a surface, the measurement of which is often expressed in several different ways, including Ra (Roughness average), RMS (Root Mean Square), AA (Arithmetic Average), and CLA (Center Line Average). I’ve found RMS to be a frequently used measure in my brazing experience. Others may perhaps have found otherwise.

Is there a conversion-chart that people can use to compare these different surface-roughness values? Yes! The company, L.J. Star Inc., has published the following excellent chart, reproduced in Table 1, which provides comparisons between several different surface-roughness measures. What I like about this chart is that it also adds in a column for “grit-polishing” for approximate comparison to the other columns. “Ra” is in microns (metric), whereas the remaining three columns are all measured in micro-inches. Please note that this chart represents just one possible approach to this topic of roughness-comparisons. But I have found the chart to be helpful, and pass it along to readers here.

Conversion chart for equivalent expressions of roughness.

Table 1. Conversion chart for equivalent expressions of roughness.

For the balance of this article I will be using RMS for my discussion of surface-roughness. In my experience, RMS surface-roughness may range from about 16 RMS, to 32 RMS, 64 RMS, or even 125 RMS or greater when it comes to the as-received, as-machined, or as-drawn surface roughness of the part to be brazed. Please remember once again that RMS means “root mean square”, and is a mathematical average between the peaks and valleys of the surface roughness.

Do I Need Spacers to Keep Faying Surfaces Apart?

The as-received, as-machined, as-drawn surface-roughness of the part being brazed should allow parts to be assembled without the need for adding any “spacers” in between the surfaces to be brazed such as many people do in order to provide a specific amount of gap clearance between the two faying surfaces of the joint. By allowing so-called “metal-to-metal contact” between the parts being brazed, the surface roughness of the faying surfaces should therefore be quite sufficient to allow BFM to flow through the gaps created by the surface roughness of each component, even when those surfaces are touching each other (so-called “zero clearance” between the surfaces).

Should Faying Surfaces Be Polished Prior to Brazing?

Generally speaking, no, it is NOT RECOMMENDED to polish the metal surface prior to brazing. Some people have felt that polishing of the surfaces was needed in order to make them clean and smooth enough to allow brazing to occur between those surfaces. That’s not really true at all. Polishing the faying surfaces will remove the surface-roughness peaks on those faying surfaces, and, with metal-to-metal contact fit-up (zero-clearance) prior to brazing, could actually shut off the possibility of any capillary flow into the joint by molten BFM applied to the outside of brazing joints (such as when paste is applied at the outside edge of a joint). Such a situation would then, in fact, require the addition of some kind of shims between the faying surfaces to allow molten BFM to flow into those joints from the outside.

IMPORTANT NOTE: Having said that, please note that if the BFM is pre-placed as a preform inside the joint (such as with a BFM foil), or if the BFM has already been roll-bonded (clad) to the faying surfaces prior to assembly of the parts, then, upon heating to brazing temp, the BFM will merely melt in place inside the joint, and will be able to alloy with, and bond to, the base metals surfaces without any problems, irrespective of whether the faying surfaces were rough or polished ahead of time.

Thus, the surface-roughness of the faying surfaces is really only a concern when the BFM is going to be applied outside the joint and needs a joint clearance in order to be able to flow into the joint from the outside by capillary action.

Increased Joint Strength via Surface Roughness?

Questions have also been raised about the relationship of joint strength to surface roughness. Will a rougher surface, with its larger peaks and valleys (and thus greater surface area inside the joint) give the joint enhanced joint strength when compared to a similar part with polished faying surfaces in the joint? Although I will not deny that possibility, the nature of the question seems to assume a higher importance to the “mechanical gripping” action to prevent sliding (in shear) of a brazed joint, than to the metallurgical bonding that occurs inside a brazed joint due to the diffusion of the BFM into the base metals that invariably occurs in any good brazed joint. As I discussed earlier in this article, I’ve very successfully brazed rough surfaces together, and although mechanical-gripping of the surface roughness may have occurred in addition to the metallurgical bonding, the parts still failed in the base metal, away from the brazed joint, just as it should have done with surfaces that were not nearly so rough. A negative effect of the roughened surfaces I worked with was that I had lost overall dimensional control of the desired tight joint gap-clearance joint fit-up that I would have been preferred to have.

Can the extra work of surface-roughening be beneficial? Much depends on the nature of the service conditions that the parts will encounter after brazing. If the components are to be used in a way that the brazed joints will see heavy shear forces, then increased surface roughness may be helpful. For that matter, perhaps even further support to the joint could be made by designing it with a shoulder in the joint to prevent slippage/sliding via shear loading. The use of a pin or threaded screw might also be desirable.

But if the part is to be used in only tensile or compressive loading (perpendicular to the brazed joint) such as stamping dies, etc., then added surface roughness may not be sufficiently advantageous to warrant the extra labor and process procedures to further roughen the faying surfaces prior to braze.


Surface roughness of the faying surfaces inside a joint to be brazed should always be taken into account prior to assembly of the parts for brazing. Normally, the as-received, as-drawn, as-machined finish is perfectly adequate for brazing, with no further work required to roughen or smoothen those surfaces.

Problem solve and improve with our technical experts.