Help needed: laser welding niobium
Greetings:
I have a problem: I’ve got a commission that’s due early next week
that requires several pieces of niobium to be laser welded together.
No problem: I’ve got a buddy with a Rofin Starweld that I can use.
Unfortunately, he’s a chip-engineer, and uses it for micro work on
computer prototypes. Thus, he knows nothing about using it for
reactive metals. (Didn’t even know what they were until yesterday.)
We did some samples, and they worked reasonably well, enough that I
pulled out the real pieces and started tacking them. Suddenly, the
god Murphy reared his ugly head, and the welds went to hell. We’d
been having trouble with getting brittle welds, but we increased the
argon flow, and that seemed to help. As soon as I pulled out the real
pieces, everything got glass brittle. The samples were done in the
off-cuts from the real pieces, so they were exactly the same pieces
of metal.
Here’s the setup: The critical welds are on 12Ga Nb wire straight
through a sheet of 20Ga Nb sheet, so as to get a ‘through’ joint. I
wasn’t too trusting of the welds, so I designed it so that most of
the stress will really go to the pin going through. They’re coming in
at an angle, so there’s some unavoidable slop in the joint initially.
The main problem is that the joints are just brittle as hell, even
with the argon blasting on them from about 1/4" away. In reading the
archives (yes, of course I read the archives before posting. Doesn’t
everybody?) I note a post that makes me think that high-velocity
argon close up may be part of the problem, and I should slow it down
and back it away a bit. I’ll try that tomorrow. This particular Rofin
has the ability to tailor the profile of the laser pulse. Currently,
we’re using 350-400 volts, 10Ms, and the beam diameter set to about
.5mm, with the beam at full power for the entire 10Ms pulse. We’ve
both had the thought that perhaps a high power spike initially
followed by a ‘cool down’ phase might be useful. Or perhaps a slow
ramp up to a big hit. We don’t know. In one of Peter Rowe’s old post
he says something about messing about with the nature of the pulse
being useful for reactives, but unfortunately, he doesn’t say what
changes in pulse profile to use. Anybody got any clues? This is a
"cannot fail" sort of job.
One other issue: I’m getting intermittent coatings of black crud
that looks like carbon flashing…except there’s no carbon
anywhere to be burning. What is that, and what causes it? I’m also
getting a little dusting of some sort of light blue vapor deposits
coming out of some of the joints. (Looks like the zinc that burns
out of overheated brass. But it isn’t, of course. So what is
it?)
I’m thinking that I may have the power up too high as well? The
initial welds I’m doing with it jacked up high enough to burn about
half way through a piece of 12Ga wire. The idea being to burn in
half way, then fill a little with some 24 Ga wire, and finish up with
the beam down low and wide to polish the surface. They looked
great…and snapped like glass. Pretty clearly it’s gas embrittlement
of some sort, but we’re out of ideas on what or how. Rather than
burning irreplaceable parts all night, I decided to head home and
look for answers.
Any help will be most appreciated.
Brian Meek.
PS–> anybody else with Rofin experience: the only manuals we’ve got
are in German, and mine isn’t that good any more. There are two flex
hoses for gas delivery in the weld area. The right side one is the
argon. Looking at the manual, I suspect the left one is an air jet
for air-cooling of parts? It’s not another argon jet is it?