Fail Of The Week: Learning How Not To Silver Solder

The interior of a failed boiler.

Sure, there are subtleties, but by and large it’s pretty easy to pick up soldering skills with a little practice. But wait! Not all soldering is created equal, and as [Quinn Dunki] learned, silver soldering is far harder to get right.

Granted, the job [Quinn] is working on is much more demanding than tacking some components to a PCB. She has been building a model steam engine, a task fit to put anyone’s machining skills to the test. And a steam engine needs a boiler, which is where the silver soldering comes in. As she explains in the video below, silver soldering, or “hard” soldering, uses solder that melts at a much higher temperature than “soft” solders like we’re used to in electronics. That’s a big advantage in the heat and pressure of a boiler, but it does pose some problems, many of which [Quinn] managed to discover as she tried to assemble her copper beast.

It turns out that heating a big hunk of copper evenly without burning off the flux actually isn’t that easy, though you can’t say she didn’t give it the old college try. In the process, she managed to share a number of tidbits that were really interesting, like the fact that drawing acetylene from a tank too fast can be dangerous, or that model steam boilers have to be certified by qualified inspectors. In the end, her boiler couldn’t be salvaged, and was put to the saw to determine the problem, which seems to be her initial choice of heating with oxyacetylene; after that initial failure, there was little she could do to save the boiler.

As [Quinn] says, “Failure is only failure if you don’t learn from it.” And so it may be a bit unfair to hang “Fail of the Week” on this one, but still — she has to go back to the beginning on the boiler. And we already know that model steam engines aren’t easy.

26 thoughts on “Fail Of The Week: Learning How Not To Silver Solder

  1. Put the whole assembly on a charcoal grill or hot plate and then try heating specific points with a torch? That looks like *really complicated* and picky soldering compared to anything I know; which was usually in circumstances where pressure was specifically something to avoid.

  2. She is darn good at this stuff and very meticulous. In this case flux is the thing. Or heat in a little oven with argon or hydrogen. Hydrogen is the best because it removes the oxidation as temperature goes up and is the flux. It isn’t hard to do the hydrogen over, keep some positive flow and plug air gaps as much as possible. The results are spectacular. Use around 90% nitrogen or argon and 10% hydrogen is the easiest and not explody. The proper amount of SS is important.

    I wonder if the heat on one end allowed the tubes to suck heat shrink too quickly lengthwise and diameter when heat is removed and thus the cracks. I think one could do it in a small electric kiln for ceramics. Since flux is not needed beyond the hydrogen, the rate of heating will not matter.

    1. Flux helps a LOT with copper because copper oxides are pretty frustrating. The silver solder will kinda crawl through but you get a better fillet if you have enough flux.
      We used boric acid in alcohol, like, a LOT of it, to the point where the whole surface of the item is covered in a melted liquid coat of it, and introduce the silver solder.
      But for something this big I’d be tempted to do it the way bike frame manufacturers used to, and wrap the solder wire around all the joints, then cover everything in borax, and then reflow it. Bike manufacturers would stick a silver ring in the space between where tubes intersected, inside the lug, and it would run out into the narrow space via capillary action.
      (Silver solder is easier than brazing coz it follows heat so you can often pull the silver along with the torch tip.)

      1. That is where the hydrogen shines. Under heat it will reduce the oxides and form water vapor. You need to start with a clean surface, but no new oxides will form and at least when I watched this, everything came out sparkling clean.

      2. Brazing of steel (especially bike frames) is a long used technique.

        Idea is to keep the temperature as low as possible. This reduces bending, avoids phase changes and all the other problems welding has. Part tolerances can be smaller and post-joining reworks is lesser. Needed tools are simpler, energy consumption is lower and it is easier to do in manual labor.

        Main drawback is less structural stability. With process optimization for mass production most disadvantages can be avoided and thus it is not used any more nowadays.

      3. The stress concentration on bike frames is at the joints. You can deal with this in steel frames by butted tubing, where you use a hydraulic ram to compress the tubing enough so that it’s thicker at the ends (which was prohibitively expensive for most of bike history) or by adding metal on the outside. One way was with brazing, which adds comparatively large amounts of metal with very smooth fillets to minimize stress concentrations. Another was with lugs, formed steel tubing joints very much like copper plumbing fittings, that used silver solder wicked into the capillary clearance space. The advantage of silver solder was that if you were a talented welder you could flow the silver solder at temperatures below the threshold to remove the work hardening from the steel tubing, so it would retain its hardness; brazing and welding require post-treatment heat treating to get back to a similar hardness because they fully temper the metal.
        Now most everyone uses hydroformed and double butted tubing for commercial steel tubing.
        Aside, the lugged frames became an art item in their own right.
        is an example of an art silver-soldered lugged bike frame.

    2. Question: How does one get hydrogen for this? Is this a gas mixture I could purchase?

      I’m interested in doing some experiments with small amounts of hydrogen, but had thought that you can’t just purchase it in tanks. I’m thinking of making it on the fly (that’s easy), but would rather purchase a tank and have less hassle.

      (A small bernz-o-matic-sized tank or pony bottle would be enough for my purposes.)

  3. liquid solder has surface tension..

    does anyone continuously rotate the work piece while applying heat, so that gravity continually redistributes the solder around joint, with surface tension pulling it back from dripping?

    just a random thought.

    the process notes in the video are very interesting and helpful for learning.

    never soldered anything bigger than a circuit board before personally.

    when i saw the huge silver solder waterfall, and wondered if continually rotating the boiler while heating it would have caused a meniscus to bulge out where the waterfall went through. i am just randomly guessing, lol. melting metal is like magic to me. thanks for the project documentation.

    1. The “proper” way involves carefully measuring the gaps and using bits of silver solder to get just the right volume. It will wet the surfaces and only bulge or drip if there is too much.

      Quoting an industrial spec sheet, “In Controlled Atmosphere Brazing, oxygen is removed from the atmosphere of the brazing oven and replaced with a mixture of hydrogen and nitrogen that does not contain oxygen molecules.”

      Then there is the idea of using an induction heater in a box with nitrogen and hydrogen.

  4. you can have my lead solder when you pry it out of my cold dead hands.

    honestly im still not out of the ginormous spool of 60-40 i pilfered from my dead grandpa’s closet 2 decades ago. rumors of solder expiring are grossly exaggerated.

    1. Silver solder for mechanical stuff like this isn’t lead-free to be lead-free; it’s a different material entirely, with deliberately different physical properties. Substituting 60-40 here would be the same kind of wrong as using electronics solder when the design calls for brazing with brass.

  5. The tube plates looked nice and bright externally, suggesting a good reducing atmosphere externally while brazing, but the internal copper looked oxidised suggesting perhaps an inert atmosphere within the boiler such as argon or nitrogen may have helped to reduce scale formation. Equally, heating on a charcoal brazier, as suggested by others, may have achieved the same reducing atmosphere.

    The higher the silver percentage of the brazing rod, the runnier the solder is. It may be that using a lower silver percentage brazing rod would allow the joins to be built up more easily without the internal silver solder waterfall noted on cutting open.

    Hard soldering, silver soldering, and brazing all refer to soldering where the filler melts at over 420C…. Not sure what soddering is… :-)

    If all else fails, a monotube flash boiler might be easier to build!

  6. I work in HVAC/R and braze tubing almost daily. The filler rod we use is Harris staysilv 10 or 15. The number represents the silver percentage. I had a friend bring me a boiler for repair. The boiler was for a model steam engine and the fire tubes at the bottom had cracked due to water freezing in them. I closed the cracks as much as possible and brazed them shut. For safety the unit was filled with water and pressurized usind a hydraulic pump to twice maximum working pressure. This is referred to as hydro testing. If the vessel were to fail it would not explode but rather pop and leak. One other note is the torches that I use are oxy/acetylene and the torches themselves are Smith little torch. This a torch designed for working with up to 1-1/4 inch copper tubing and comes with various tips for brazing tubing. One other note is that in HVAC/R we use dry nitrogen(obtainable from any welding source) to purge the inside of the tubing to prevent oxidation.

  7. I thought she was too quick to put the work in the picking solution, rather than letting it cool down and anneal. If I’m right, it would have hardened the material, while introducing thermal strain.

    1. IIRC brass and copper will work harden but do not change hardness from cooling. There are hardening schemes for silver and copper but I think they involve soaking at some temperature for 20 minutes to several hours.

      1. Ah, you’re right about annealing copper, of course.
        Still, quenching too quickly is bound to cause stress due to complicated geometry, nonuniform cooling, etc. That’s the best explanation of the hairline cracks of the solder joints that I could come up with.

  8. The biggest error was not making practice parts to quickly test the skill/result of the process. Never put 100% of the work effort into a part you are not sure you know how to finish. It’s better to do smaller work-fraction test pieces. A length of pipe with just one perforated support and a couple of tubes, not every tube.

    Live and learn.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.