Making Solderable Aluminium Baseplates.

[The_Doctor]

Best start a thread here with the others I made on similar ideas for making laser parts. This one's for using HTS-2000 brazing rods for aluminium in various ways, first mentioned in a thread in the Lounge sub-forum.








Ok, to explain:
First column is a basic impact test, was the first thing I did. It wets the surface well but I didn't try for anything thorough, I just turned it over onto a bit of HTS-2000 tinned 1/8" L section bar to fuse it there so I could hit it later. Then I hit it. The pictures show it well enough, but notice that the main bar was beaten out of line as well as the small bit being mangled well before it parted moorings. It shows that the bond wasn't good but it still took a sustained hammering.

Second column is far more interesting, it shows a staged process to get aluminium ready for LOW temperature soldering directly to it. Once the HTS-2000 (I used a few spilt bits from test 1) was melted on the block and scraped around gently with a stainless steel brush I'd let it cool to see if I could solder indium based solders to it. That wasn't successful but I then prepared a new surface with HTS-2000 and added a bit of standard lead/tin flux cored solder which merged easily, before scraping off the excess mix with a stainless steel edge. I then let it cool to solid and moved it to a hot plate at 125°C (hot enough to melt all three samples, HTS-2000 itself wants 400°C). The leftmost blob is 65% indium, 35% bismuth, melts at 74°C. The middle blob is 'Indalloy No. 8' which melts at 93°C (44% indium, 42% tin, 14% cadmium), and the rightmost blob is a eutectic mix of 52% indium and 48% tin melting at 111°C. They flow well with the help of a bit of resin flux, so heat conduction will be excellent, and cleaning will be easy.

The main problem (which I'm trying to get advice from the maker about) is that scored, etched looking finish, which is obviosly poorer than I started with. My guess is the HTS-2000 reduces the melting point of the aluminium substrate, which may have been too hot already. If I can't get rid of that problem, I can at least lay on a thin copper veneer that will spread heat and make a very fine finish. It should be possible to make up 6" wide slabs this way and then cut to smaller slabs on a bandsaw to get extremely nice base plates for small laser systems without any complex cleaning to do. It beats paying for copper ones.

[Dr Laser]

Weld it. Yes you need 600$+ welder with an argon. But it works great.

[carmangary]

I bet you could get as strong of a bond with JB Weld.

[Hemlock Mike]

The $600 might buy a spool gun for Alum wire. The argon and a mig welder are +++ !!! Also TIG with high frequency AC is nice -- So many options -- so little $$.

I've seen Alum solder but I don't know much about it.

Mike

[chad]

interesting progress. I assume you are trying to duplicate the coherent solder optics alignment method. It would be nice if you could figure a cheap and easy method to do this.

I am a little worried about the expansion problems with the alum in the heat and cool down phase might shift things too much.

Coherent gets around this using gold plated invar (cheap) and they still have to deal with a little moving on cool down and have to compensate for it.

Anyhow keep us updated!

Chad

[The_Doctor]

Well, Chad sees the point of this. If a welder was all you need, why would someone invent this stuff and why would someone want it?


JB Weld is NOT appropriate. I'm familiar with metal loaded epoxies, and rejected them a long time ago. I'm not after brute strength and heat tolerance, and it should be obvious from the kinds of stuff this forum is about, what properties I want this stuff for. Mainly, ease of getting stable optical/electronic structures with good heatsinking cheaply and without a lot of drilling and threading work.

To see the real point of this, don't look at what it isn't. Just see it for what it is, better yet, try some and see if you can adapt it. And the day someone manages LOW temperature soldering at finely differentiated temperatures for complex constructions like Chad mentioned in the Coherent lasers, and does so with a TIG welder, I hope they let me know exactly how they achieved this, in similar detail to my own report if not more. A terse post rejecting my idea isn't enough, especially if it isn't thought through.

Mike, aluminium solders vary, I tried one that was based on an acidic flux, it was truly vile stuff, it stinks and corrodes itself just sitting around, smelling like concentrated cats piss, and in use it spits and fizzes and is as likely to blow the join apart as to make it work. A welder is similarly ineffective on thin metal, but this new stuff might work better. My main problem is it seems to erode the substrate by lowering the melting point. I'll have to find a way to regulate the temperature closely to be sure whether this can be reduced, but if not, it won't be that good a process for small scale work. Good enough to lay a thin copper layer onto a thick Al heatsink though, so that small laser parts can be soldered directly to it.

Chad, that expansion thing bothers me too. With copper laid on aluminium it's even worse. (Bimetallic strip...) That's why the discovery that adding standard solder works to prepare a surface for indium based solders is so useful. Once you get to mount stuff at those low temperatures the problem is much less. For critical alignments I'd still provide adjustments with tiny screws. So long as the construction is for diode based lasers that don't need separated components to form a cavity, this method should be great. I'm not that good anyway, I won't be building anything really advanced, I'm just after a way to combine beams in as small and stable and cheap a space as I can get.

[MechEng3]

One company I have done some work for has started to employ ultrasonic fusion welders. They are very expensive, but the payoff comes in a high quality resulting process that saves them time and machining stages. I am not sure of it's capabilities of assembling dissimilar metals though. I will try to get some answers and specifics. You could always have the parts done for you....this equipment is rather expensive. I know that it can assemble 2mm stainless parts that are reasonably small without heat discoloring like a weld. May not be what you are seeking as to a solution.....may just be the ticket. What about the Coherent laser optics table....lots of glass? The waveplate assembly has a neat way of dealing with heat issues.

[chad]

I wonder if there is some easy plating solution you could use on the alum. Something that would take solder??

Any way just a thought.

That rough surface is a pain. If you machine it flat again then you loose all of your special coating. And a rough surface does not do you any good for thermal and makes things unstable on a plate...

Chad

[The_Doctor]

Ultrasonic welds are nice. Way out of reach for me though. I'll look into it though, maybe.

I tried another test. No pictures for that, it failed. DON'T bother trying to grind HTS-2000 to fine dust to get a level deposit on a surface before raising the heat. It won't work. I think it oxides too much or something, but all you get is hot useless powder, and keeping and applying the dust in a temporary binder like isopropanol doesn't help either.

Chad, that solder plate idea is a tad like my next idea. I've been looking at the erosion of HTS-2000 in various metals:


(Both sample sets have pieces of similar width. I chose GIF with dithering instead of JPG for all pictures as it preserves sharp detail and enlarges better if you want a closer look.)

The leftmost one was just above the melting point, so as I hoped, that minimises the erosion. The third sample was a deliberate attempt to pit the metal with a blob left sitting there well above melting point. The two leftmost samples have some added lead-tin solder. All but the leftmost sample were at the same higher temperature. Not sure exactly how much higher, estimated maybe 50°C.

The second image shows how much less copper erodes. Brass loses more, because the zinc behaves similarly to the copper, I think. The pits were too small to flush out so I just ground the surface back to clean metal to reveal what I could. If I ground out the bits in the brass with a dentists drill I might see copper, same as you do with brass if overheated in standard brazing when zinc and copper part company.

Anyway, the idea I have for the next test is to tin some copper, then slide a small block of aluminium on the surface to evenly run the HTS-2000 on the surface. By controlling both the amount available to the aluminium surface, and also the evenness of that amount, I hope to reduce the erosion to an even effect that changes the first few molecules depth to enable easy soldering at low temperatures.

---------------------------------------------------------------

Chad, that thermal coupling problem isn't that bad, one thing about HTS-2000 is it's extremely good, it's like aluminium. There are only about three or four known substances that conduct heat better at room temperature: copper, silver, and carbon (diamond, and graphite along plate planes). So a fill of HTS-2000 to bond a thin copper sheet is very close to as good as you can get, the main problem being the bimetal strip effect, so it's only good for thick, small assemblies. That's ok though, my diode laser designs are scaled about right for this, I just want to see how much better it can get before I settle on a single method.

MechEng3, your mention of glass has me thinking it might be useful, but probably not the same way you were thinking of. I suspect indium to be a major component of HTS-2000. Indium can wet glass. If I find that HTS-2000 can also wet glass, that might be even better (and cheaper) than copper as a means to evenly transfer HTS-2000 into the first few molecules depth of aluminium.

[p1t8ull]

This is all interesting stuff Doc.

A few years ago I used one of the older generation aluminium repair kits (one that needed plenty of abrading) to very good effect, and thought at the time that it was pretty cool.