Feb 23, 2014

Saxophone Metals - Some Scientific Opinions

A while ago, I exchanged emails with my friend Jude Foulds regarding some aspects of saxophone construction, particularly metals. Jude’s opinions should carry some weight - he has an MS and PhD in Metallurgical Engineering, and has had an extensive career in failure analysis. He’s also a sax player (former student).

Our conversation was prompted by a Steve Goodson interview that another sax player friend had sent me, that included some assertions about the aging or “breaking in” of metal, cryogenic treatment, and silver/gold plating. I asked Jude about these things, and I had a few other questions as well.

Here are some excerpts from our email exchange, beginning with Jude’s response to the interview:

Jude:

[quoting the interview:] “Of course, the older horns sound different. I think that a large part of that difference is because the horn has been played for years and years and the metal has ‘broken in’ and responds more freely to a given range of frequencies. You can achieve the same effect of this breaking in by having the horn cryogenically treated to free up the stress in the metal.”

There is no aging effect for brass and most metals at ambient conditions. Also, using the horn would not have any significant effect on the response of the brass over time. There could be some, minor relaxation of stresses (induced from the original forming process) over time, but this would not have a significant effect on the metal's acoustic response. In short, the horn material should be no different than what it was the first day it was formed. Cryogenic exposure would have absolutely no relaxing effect on internal stresses. In fact, it could have the potentially opposite effect of inducing (residual) stresses simply from the uneven cooldown and subsequent heatup back to ambient. We typically heat up metals to remove stresses, not cool them down.

[quoting the interview:] “Around 1970, manufacturers began to use brass that contained more zinc than was used in the past. This gave the horns a brighter sound, and a little more projection, but at the expense of tonal complexity...I'm not sure this was a good idea. We've...found some brass recipes that give a nice mix of overtones. Our alloy is significantly higher in copper content than the alloy used by other makers. On our high end models, we hand hammer and hand burnish the bell, bow, and body so the metal doesn't become ‘work hardened’ and lose resonance.”

I'm having a terrible time believing the alloy effect. The alloy variation effect, I would estimate, to be a distant secondary one to the forming effects, geometric effects and the keys/pad design and construction...An important aspect of horn construction related to alloy is formability. I believe much of the tonal response (all other things being equal) is a function of the forming process, and I think Goodson may have something in his hand-forming of the bell. His higher Cu brass may make formability easier and this could be the "alloy" effect, rather than the chemistry itself.

Goodson is right-on [regarding] plating effects [that is, that plating material has little or no effect on tone].

Peter:

Thanks! ...I guess you wouldn't subscribe to the theory that pre-1960 French saxes sound better because they were made from recycled WWII shell casings. Then again, the best old horns do sound different, not all the same, often with a more complex tonal color. Is it all in the design and workmanship?

Is zinc harder than copper? I always thought that harder materials in mouthpieces reflected out high overtones more. Do you think that's correct?

Jude:

Here are a few facts and my thoughts on Zn in brass:

(1) Zn does increase the strength and hardness (resistance to plastic deformation and to nicks and scratches), while also making it more difficult to form.

(2) Zn also makes brass less stiff (elastically). This could, if anything, correlate more with tonal response than does hardness since elastic properties correlate with sound attenuation through the metal.

(3) Metals handbooks often refer to the 80-20 (Cu-Zn by weight percent) "low brass" as a musical instrument alloy. This is a far cry from the cartridge shell brass (70-30) that is much harder and more difficult to form. Did they make a lot of saxes from these shells? I can see where the significant Zn increase (from the "handbook" number) could affect tonal response, but they must have put a lot of work into making those saxes, particularly when much of the forming was done by hand. If you get a chance to, you may want to analyze a sliver of material from a discarded vintage and from a modern horn.

(4) An important effect on acoustic response comes from the metal grain structure and what's in the grains - greatly affected by the forming operation. This is a key factor that is much too in-depth an issue to have been seriously studied (good thesis topic).

(5) The industry may have increased Zn in the 70s when they became more automated in their forming; albeit more difficult to form, the harder materials probably gave a more reproducible formed product with fewer rejects. Also, they may have reduced the gage.

(6) I would also look into the gage thickness variability (vintage versus modern). I don't know how standard this is, but the first thing I noticed with the SML was the gage at the bell. I have nothing to compare it to, but it sure feels beefy.

Peter:

Thanks again...The idea that old French saxes sound better because the brass was from cartridges is a sort of urban myth among sax people. It certainly seems plausible that European brass was from scavenged sources for 10 or 15 years. But couldn't the metal suppliers have added copper if they wanted to?

Jude:

The remelt shops could have added Cu (I used to work in a brass foundry and process shop where, among other things, we made 60-40 leaded brass from a mix of scrap and raw Cu and Zn and Pb, for tubulars and rods, etc. that we extruded and drew at the plant). However, I guarantee you Cu was even more expensive back in the post-war years and not easy to come by. Any Cu additions would likely have ruined the economic advantage of remelting the cartridges.


To blog readers:

There it is, for your consideration. I was reminded of this exchange when one of my students mentioned a recent interview with Wayne Shorter where Wayne repeated the story about WWII shells.

I find it quite believable that late 40s - early 50s French saxes could have been made with brass from recycled shells, perhaps with a 70-30 “cartridge brass” ratio rather than an 80-20 “musical instrument” ratio. Whether this would have any significant effect on timbre, and by what mechanism that might work, is not clear to me. Oddly, it seems that more zinc increases sound attenuation, which one might think would not be helpful to timbre. I don't know. (Incidentally, I've also heard it suggested that 1930s European horns used brass from WWI shell casings. Maybe there is a 95-year-old French metallurgist out there somewhere who knows about this.)

And it's possible that they added copper anyway, in spite of the cost. No final answers here.

Jude implies that the hand-forming process might have affected the acoustic qualities of the brass, for the better.

I can agree that plating material has little effect, and it’s pretty much impossible for me to believe that lacquer (color, or lacquered/unlacquered) has any effect on tone. I don’t see any reason to believe the cryogenic stuff.

When I was getting this post ready, I asked Jude a couple more questions.

Peter:

1) You mentioned that a higher zinc ratio would result in a harder alloy, but would increase elasticity, which could affect tone. How would it affect tone? Would a more elastic alloy attenuate certain frequencies more/less?

2) You mentioned that "I believe much of the tonal response is a function of the forming process." Can you tell me how that might work?

Jude:

(1) We define elasticity as a modulus (stress divided by strain). A stiffer alloy (higher modulus) will typically transmit sound better (less attenuation). The addition of Zn makes the brass less stiff and therefore increases attenuation (less sound transmission efficiency). I don’t think hardness (not an elastic property) by itself affects attenuation...
(2) Tone for me is the quality of the desired frequency...What’s in the metal (grains, dislocations structure within grains, etc.) likely governs the purity of the emission. Forming methods that control grain structure and the material within grains (we call this microstructure) can affect tone. If you ask me what is optimal, I’d have to research the structure of the specific instrument, specifically the manner in which the sound travels through the metal and the relationship between dominant sound path(s) and the microstructure (e.g., orientation of grains).
Please recognize that I’m no expert on acoustics...as Clint Eastwood said in one of his Dirty Harry movies – Man’s got to know his limitations.

3 comments:

  1. 70/30 cartridge brass was used for most instruments until recent years. It was the material with the properties that all producers of sheet brass products (including musical instruments) based their designs and manufacturing processes on. There was little incentive, until the acoustic response spectra of other alloys came into consideration more recently, to go with anything other than what was standard and available, with known working properties and costs. Goodson may have been referring to Yamaha using higher zinc brass, as their brass is notoriously hard - resistant to dents, but also resistant to dent repair.

    Jude's last point is an important one. Within the constraints of standard brass, the acoustic response can be affected by the metalworking process - how much hammering, how much annealing, temperature and length of annealing, how much quenching,etc. - to achieve desired hardness and resilience. There is of course no available documentation of those parameters, which could be considered trade secrets and were likely passed on between craftsmen as a sort of oral tradition. That's more likely the key to what made certain Selmer saxophones so special, rather than the source of the brass. And a company that didn't even document design changes in the Mark VI would likely be completely dependent on the hands-on knowledge of its craftsmen for its metalworking processes. Once that knowledge base is out the door, it's gone. Did the process change as Selmer put more emphasis on production volume in the 1960s? Antigua claims to have reverse-engineered the metalworking on classic era Mark VIes by analyzing the microstructures of the brass, and to have incorporated that knowledge in the production of their Pro One saxophones.

    High copper alloys such as rose brass and phosphor bronze are intriguing for their potential sonic effects, but I suspect that it's better to have a standard brass horn with quality baked into the metalworking process than a a phosphor bronze imitation of a Yanagisawa slammed out in some Chinese factory.

    Stress cycling, either through strong vibration at subharmonic frequencies or deep thermal cycles, is an established practice for relieving residual stress in metal products. And, yes, stress does does deaden the sound of a saxophone.

    ReplyDelete
    Replies
    1. GFC - Thanks for the comment! Just out of curiosity, what is your technical/musical/industry background?

      Delete
  2. Serious amateur player and tech, with enough of an engineering background to get fascinated by the geeky aspect of saxophones.

    ReplyDelete