Sealing a brass tube - an adventure of soldering, brazing and welding (+failed attempts).

The hardest part in making a heat pipe is sealing it off and keeping it sealed. Soldering the end shut is not hard and it keeps the tube sealed but such heat pipe cannot be soldered to anything after it's sealed because the solder would melt and the seal would fail. Commercially made heat pipes are welded shut. Such a seal is the best you can get but it requires a TIG welder to achieve and TIG welders can be pricey. The next best thing is brazing. It's basically soldering but at a much higher temperature. This way you can solder your finished product without the risk of breaking the seal. 

The biggest problem I've come into is the step between pulling a vacuum and permanently sealing the heat pipe. That step being the crimp, because before you permanently seal the end of your heat pipe you need a way to keep it sealed while the permanent seal is being applied. Crimp being the most popular method. The crimp has to be really good because any leakage in or out will disrupt the brazing/welding process and will also put the integrity of your vacuum as well as the longevity and performance of your heat pipe at risk. I've tried many times to achieve a good crimp seal but had only partial success at best. The way I've tried to achieve the crimp seal was by first making my brass tube narrower by heating and hammering while rotating the tube. After some time the 8mm tube would reach about 50% of it's original diameter and the wall would also subsequently thicken.

One problem I've come to while doing this is that the inside wall of the narrow tube was quite rough and i think that might have contributed to some of the problems I was having with my crimp seal. Next time I will drill through the hole to make the inside wall smooth and clean. After this end was ready it was time to insert my mesh and seal the other end. The mesh was easy to put in, the braided shielding that I'm using is hard to put in if you're pushing but it's easy if you're pulling. For that I used some iron wire and put the wire through the heat pipe. A small hook was bent on the end of the wire and my mesh was hooked to it. The wire was pulled and so was the mesh, as soon as the mesh reached the end it would stop and the iron wire hook would bend itself back and pull out leaving the mesh inside. One thing I noticed was that my mesh had some springiness to it but tapping the tube before pulling the iron wire out would help the mesh move into place. After the mesh was in the other end had to be sealed off. For that I had the idea to try and weld it shut but without a TIG welder. My plan was to hook my arc welder to a carbon brush and my tube. By dragging the tube across the brush the end would melt and weld. I got this idea from people who DIY weld their thermocouples this way.

In reality it didn't work as planned. The tube did melt but it splashed a lot more than it melted and I ended up having to chop the end off and trying something else. That something else being zinc. Why zinc you may ask? Well when I was making one special tool, that will be at the end of this post, I had to heat and bend a piece of zinc plated steel and happened to notice that the zinc plating when heated melted nicely and was almost like water of the piece of steel. I then had the idea to braze the end of my tube shut with some zinc. Well you might call it soldering since zinc does melt at just 419.5C, which is exactly what I want actually. I got my zinc from some old zinc-carbon batteries. The zinc was old and dirty but with some cleaning was perfectly usable. Now I just need some flux and I'll be ready to get to work. Since the melting point of zinc is so low one might think that rosin can be used as flux when in reality it can't. At the kind of temperatures that need to be in order to braze with zinc rosin just starts burning and leaves nasty black residue behind, I think you can guess how I found that one out. The only other thing I had was borax. I actually got into some trouble with customs because I had to order the borax from another country and as you can imagine 500 grams of white powder in a bag did not slip through customs unnoticed. But that was a long time ago and now I have a heat pipe to braze shut so let's get going. My idea was to put some zinc in the tube as well as some borax to flux and to crimp the end before I heat it to melt the zinc brazing the end shut. Sound simple in theory but turned into a big fail in practice. The borax did not work as flux an by zinc did not wet the brass tube, which also tarnished quite a bit in the process.

The end was easily pried apart and after quite some cleaning was ready for a second try. This time I tried tinning the inner wall of the brass tube with regular solder to keep the brass from tarnishing. I then put some zinc in and pressed the end shut. When this was all done I put my butane torch in front of my heater that I had running because my workshop was quite chilly. When warmed up the torch put out a lot more heat and I managed to get the tube a lot hotter than before. After the tube had cooled down the first thing I noticed was that the borax did actually work as flux it's just that before I didn't get it hot enough. The zinc did melt very well this time and I was quite satisfied with the finished product.

You can see that where the borax flux touched the brass tube got a lot nicer. I noticed that a hole had formed at the spot I had crimped too hard so I tinned it and brazed some zinc on top.

 

Now all that was left to do was to fill this with my working fluid of choice, in this case acetone, and to pull a vacuum and seal it off. To pull the vacuum I used a refrigerator compressor which should easily get down to the vapor pressure of acetone. To crimp the end I used a pair of pliers that had cutters in them, the cutters happened to be very blunt and just about right for crimping. After I pulled the vacuum I pressed the tube in the cutters of mu pliers with all of my might, but as it turns out that was not enough. I quickly went for a second round with some long handle snips to get even more force on the crimp. 

After the second try while inspecting the tube I noticed that it did not seal and that the brazed end was also still leaking. At this point I was pretty bummed out, mostly because I was running out of brass tube. I was done with the heat pipe for today.

So let's talk about my working fluid a little, or rather the amount of it. I searched high and low for how much working fluid is optimal but couldn't find a satisfactory answer. Most places used the term "fill ratio" and from what I found 85% is optimal. I didn't know how fill ratio is defined and 85% can't possibly be the total volume of the heat pipe so I went looking. The only answer I found was "Fill ratio means the percentage of the evaporator section volume that is filled by the working fluids.". To me this is not a satisfactory answer because the evaporator section may stay constant but if the pipe is lengthened then more working fluid would need to be added to keep the wick wet. I might be getting stuff wrong of course. In my case I came up with a simple method to calculate the amount of working fluid. That amount being the volume taken by the wick by measuring the wick thickness and assuming that it's a hollow cylinder a volume of working fluid is calculated. Of course this is by no means the best way but it's a place to start at least.

Now for some fails. Hammering the tube to narrow it down takes time and effort and I though of making a tool to roll the tube and slowly make it narrower. This of course did not go as planned.

The tube would be put in between the rollers and while the tube was rotated the screw would be tightened thus the tube would be forced to narrow down. The tool itself turned out great but it didn't work as I had anticipated. What happened was that the tube at first would appear to be narrowing down but would quickly start turning into a square and the more I tried turning it down the more square it got.

Well, that's all for now, thanks for reading.