OK - I spoke with Bill Davidson about his recollections of this test and as usual, he remembers a lot more detail than I do. I had most of it right but I hate to say Brian is correct and I remembered wrong on one point. None of the silver joints pulled away from the tubes in this particular test. More about that in a second.
The device we used was borrowed from Charlie Cunningham. Just as I said, it held the front triangle rigidly while a lever was used much as a fork would be in a front end crash, to wrinkle the front end, using a hydraulic jack as the power. A strain gauge was hooked up to this to measure the force required, which would tell you how resistant the frame was against an impact. The absolute amount of force required is irrelevant due to the leverage involved, only the relative differences are significant. Here are the results that Bill remembers: With Silver, the frame wrinkled right near the points of the lugs. With brass, it wrinkled about 1 cm down from the lugs. This didn't change with the different tubing or lugs, just with silver vs brass. This is as you might expect, since brass requires a higher temperature than silver and so the heat affected zone is further out. Bill says this is significant because it means that the weakest point of the tube, the margin of the heat-affected zone, is right near those nasty stress risers - the points of the lugs - in a silver brazed frame. In a lifetime of flex cycles, if it is going to break from fatigue, that is where it will go first. We have seen a lot of frames go like that. The brass-brazed frame which wrinkles farther down, would at least be stronger or at least more uniform where the points are and less likely to be taken out by fatigue cycles.
Another significant thing Bill mentioned: Remember I told you about brazing with a big rosebud torch to preheat the joints so you can do it faster? We did some that way and some the usual way with only a small torch. Bill said the preheated joints withstood a lot more force according to the strain gauge, although ultimately they failed in the same manner. That means the frame should be able to survive a somewhat harder impact. Since we only braze this way with brass (too much heat for silver) that is a point in brass's favor.
Now back to those silver joints coming apart. I didn't make this up. It just didn't happen during this test. We get a lot of crashed bikes in for repairs, and Bill says he has seen many frames brazed with silver pull apart during a crash. Not just ours, but from other very well known builders too. I should point out that we have used Silver to braze also for many years on very light tubing. I have seen a number of these too, and somehow remembered them as being in the test. Some particular lugs are more prone to this than others. The Otsuya lugs that we used back the early 80s had very small surface area behind the head tube, and these were more prone to coming apart, as you might expect. I crashed my Davidson with these lugs in a high speed front ender in 1988, and while it did wrinkle the frame pretty good, it didn't come apart. Luck of the draw I think. I straightened and rode it a bit more, and it eventually the .7/.4 Prestige downtube cracked so I rebuilt it. Bill has seen a number of times where when the joint pulled apart, it left a thin layer of steel on the brazing material. He suggests that this has something to do with how the metal cooled in the investment casting process, leaving a weak surface layer. He started shining up the inside of lugs when he noticed this to remove that shear layer.
Conclusions? Draw your own. It still seems to me that a frame can be stronger when brazed with brass and using a preheat torch in the hands of a good brazer who can do it quickly, but there is nothing wrong with silver as long as you don't plan to crash it. And if you crash, well, the frame is probably the least of your worries.
Bob Freeman Elliott Bay Bicycles 2116 Western Ave Seattle, WA 98121 206-441-8144 Home of Davidson Handbuilt Bicycles
In a message dated 9/24/2005 11:31:13 P.M. Pacific Standard Time, firstname.lastname@example.org writes:
I would have to see the testing device, the way you bulit the "test triangles", and a whole lot more before I would believe your story. No properly brazed silver joint will fail before the tube in an impact. Seen lots of them. Seen more broken joints (where tube did not break but braze joint did, which is just as wrong with brass as with silver) on brass brazed frames than with silver by quite a large margin. I've seen a good number of relavatively poorly brazed silver joints not fail after impact. It sounds to me like you tested a bunch of improperly brazed silver joints. Probably better to use brass in that situation, but properly silver brazed joints have at least equal strength to those of brass.
For those of you out there who do not or have not built a good number of frames; there is no way to really decide which information is closeset to the truth. Since I happen to know what I've seen and experienced, I know for sure I don't buy this particular story.
It is obvious that good bikes and excellent bikes can be made useing either method and that the primary factor in that is the ability of the builder and their intentions towards producing a quality product. But there is such a thing as reality. 700 frames a year is not the place for silver. Silver is for those individually producing handmade frames with traits and features that aren't part of low production work. It's a different world, different circumstances. Like I said, good that you did some testing and concluded that for your circumstances brass worked better for you. Silver works better for the type of work I do.
La Mesa, CA
Show me, talk is cheap.
I came into this thread late but I thought I would share a couple of tidbits. In the mid-80s, when we (Davidson Cycles) were building around 700 frames per year, we wanted some idea of what was going to happen to them in a crash, so borrowed a destructive testing device (I forget who had it. Bicycling Mag, maybe?) that simulated a front end crash. Made a bunch of front ends to test in it, with all different materials and methods but otherwise the same, and found, almost without exception, that the joint failed (came apart) when done with silver, and the tube failed behind the intact joint when brazed with brass. What does that tell you? Well, given a hard enough impact you will destroy a steel bike no matter, but the brass did make a stronger joint. Yes you can braze at a lower temperature with silver, but if you are burning the snot out of it getting it to move around then you will do a lot more harm. I think an experienced builder using brass will build a much superior bike to a rookie with silver.
It is not only the absolute temperature that is reached that degrades the steel, it is the length of time at that temperature. The longer it stays hot, the farther away from the joint the heat affected zone travels. It is that margin of the heat affected zone that is where a frame will usually fail in a crash. Keeping it closer to the lug, in the butted section, makes it more likely to withstand a crash. So brazing it quickly is the answer to making the strongest frame. We developed a way to braze them very quickly. We don't do a lot of lugged steel frames any more so I don't mine sharing the tricks. We made little rings of brass that were the shape of the inside of the joint, and assembled the frame with them in there. The joint is heated quickly and evenly with a large, broad flame until red hot. You can see the brass ring in there as a shadow. Once it is gone you know the brass has melted, we switch to a small torch and work the heat around until brass comes out all the edges evenly, and voila, you are done. Almost no cleanup involved afterwards. A few of our brazers would use the big torch in one hand and the small torch in the other. We had fork crowns made with a shelf that the steerer butted against, and stamped out washers of brass that fit right in there and did the same thing for the crown to steerer braze, and had rings inside the blade sockets too. Two guys could braze up 30 steerer, crown, blade assemblies an hour that way. Again with almost no cleanup. And it leaves a little fillet of brass on the inside of the joint too to further strengthen it. You would never get penetration that good by brazing from the outside only, and not nearly as fast.
Another tidbit, and all framebuilders know this, but 853 Reynolds is always
brazed with brass. It doesn't get hot enough for the air-hardening to