[Prev][Next][Index][Thread]

Re: torque conv./ inner tubes





On Sat, 11 Jan 1997, Tesla List wrote:

> Subscriber: music-at-triumf.ca Sat Jan 11 20:44:07 1997
> Date: Sat, 11 Jan 1997 17:48:14 PST
> From: "Fred W. Bach, TRIUMF Operations" <music-at-triumf.ca>
> To: tesla-at-pupman-dot-com
> Cc: music-at-triumf.ca
> Subject: Re: torque conv./ inner tubes
> 
>    As do I.  Plating is a very good idea.  Starting out small is an
>    excellent idea.  One thing that we should remember is that plating
>    any object with pure smooth DC will likely result in uneven
>    thicknesses depending on the distance from the plating point or
>    plating surface to the opposite electrode.  Unless you saturate the
>    solution electrically then the points closest to the other
>    electrode will plate thicker.

Fred, are you sure about this? Commercial plating has used DC since 
plating's been done, and none of the standard plating texts (at least 
those in my personal library)
mention anything about using pulses. Getting an even plate is a classic
problem, but there's no mention I've seen of solving it with pulses. It's
odd, because it would be such a simple fix, if it were workable. Might
you have any references to share?

>    So, either very good dimensions of the plating tank are needed
>    (i.e., to plate a toroid the plating tank should be exactly
>    toroidal as well, with the graphite-coated inner tube suspended

Actually, any convenient shape is acceptable for a plating tank; what you
need be concerned about is the shape of the anode, which should really
not be the plating tank, as the tank's integrity would slowly be eaten
away, resulting in a disastrous break-through at an unexpected moment.
The standard fix is often to use multiple anodes, with different spacing
to the work piece, to get a more uniform current density at the work piece
surface. Also, the further the anode(s) is (are) from the work piece, the
more uniform the current density becomes at the surface.

>    inside the plating chamber it would be like a doughnut within a
>    doughnut), or else deliver the plating current in high-peak-current
>    low-average short duration pulses which saturate the
>    current-carrying capacity of the plating solution. This "saturation
>    effect" will make for a more even plating of copper all around.  I
>    am not sure of the frequency of the pulses, I would imagine that
>    anything between 10 Hz and 1KHz would work.  60Hz sounds likely. 
>    One drawback with this pulsing technique is that the resistance of
>    the graphite will be a big problem to high-current pulses at the
>    beginning of the plating job.  So it would be best to start the
>    plating job with smooth DC until you get a thin copper plating all

Actually, they may want to start with a large resistance in series. The
resistance should be much larger than the resistance of the toroid with its
graphite-loaded binder. (Per my previous note on this subject, a loose 
graphite coating will work very poorly. It may be fine to demonstrate 
that plating can happen on a small object, but it will yield poor results
if you want to produce a usable plate.) The large resistance will swamp
out the varying resistance of the graphite coat as it gets further and
further from the point where the wire is attatched. If this isn't done,
the current density will start out highest at the point where the wire
attatches. (It will anyway, of course, but the variation will be reduced.)
This will be the first spot to get any significant copper. After
it gets the copper, that small spot will have such a small resistance 
compared to the rest of the toroid that almost all the current will flow to
that spot, and it will be the only spot to get any significant plate.

>    over.  To do this, you might have to change the position of the
>    other electrode often.  Once you have an even thin coat of copper
>    all over the tube, then I think you can go over to pulsed DC for
>    more even thicker plating.
> 
>    This experience is from back in my chem lab days.  It is why
>    lead-acid cells charge better on rectified but UNfiltered DC,
>    rather than smoothed DC.  The charge is distributed better over
>    the surface of the plates.  
> 
>    In some plating cells if you use pure DC in a very quiet
>    environment, the plating will even grow in long spiky filaments! 
>    Why?  Because this process displays positive-feedback upon itself. 

While such dendritic growths are possible, we shouldn't see them under these
circumstances. Not that you're suggesting that we should, but this should be
made clear for the other readers.

>    The plating rate depends on the current, and as the filaments grow
>    from the plated surface to the other electrode, the resistance of
>    the liquid between the plating spike and the other electrode falls. 
>    The current goes up in that part of the solution and
>    correspondingly DOWN in other parts of the solution (the spike hogs
>    the plating current),  Thus the plating rate goes selectively up on

While this description may not be particularly descriptive of what they're
likely to see while plating a toroid, it's perfectly applicable to describe
why the portions of the toroid closer to the anode will receive a heavier 
plate than those portions farther away. The closer a point on the cathode
is to the anode, the higher the current density will be at that point on
the cathode, and the heavier the plate will be. 

>    the spiky filament, and so the plating on the spiky filament grows
>    faster than the surrounding area.  You can modify or ward-off this
>    annoying effect with with external magnetic fields to stir up the
>    ions in the solution, or by just by using a low-duty-cycle pulsed
>    DC source for the plating current.  Instantaneously forcing the
>    solution between the electrode and the end of the spike into
>    saturation forces the current to move out and flow to other more
>    far-away places on the plating surface.
> 
>    Does anyone else have any plating experience in this area?  I know
>    the effects but I don't know the currents or duty cycles used with
>    the different strengths of CuSO4 solution.  I suppose I could

Nah. No need to calculate it when we have libraries. The literature's
filled with data on this often-solved problem. While not directed at you,
there's an unusual tendency on this list to forget that a lot of the
problems we seek to solve have already been heavily researched, or at least,
many problems similar enough to be very useful. There's mountains of data
out there, just waiting to fascinate us...

A plain CuSO4 solution is only useful to demonstrate that plating "works" 
in a beaker. In the real world, it will produce a spongy plate, with 
abysmally poor adherence to the work peice. Unless they discover 
something new, this is likely to produce only disappointment. To get useful
results, they'll need a more complex plating bath.

These come in three main types (safety glasses recommended!):

1) Cyanide/copper. Can't be plated fast, has very good throwing power 
(related to its ability to fill in cracks, and evenly plate irregular
surfaces) and is quite toxic. This isn't recommended for beginners, although
anyone who's accepted the hazards of coiling shouldn't get too bent out of
shape over this level of toxicity. Still, it's more applicable to plating
metals that are too active to take plating any other way, like iron or
zinc/die cast metals. Those interested in formulations may consult the
literature, in which they're widely published; I don't think they're
appropriate here.

2) Acid/Copper. This is the bath I referred to in my previous post. A mix of
CuSO4/H2SO4, in a fairly standard range of proportions. Normally, you use:

CuSO4.5H2O - 200-250 g/l
H2SO4      - 45-75 g/l

The plating rate on this stuff is almost unlimited. It has much poorer 
throwing power than the cyanide/copper. Slight toxicity due to the copper 
sulfate. Main problem with the acid is that any spatters will make swiss 
cheese out of your clothes. This stuff's easy to make, easy to maintain and
plates a nice, solid layer of metal. Lastly, the materials for this bath are
available most anywhere. Wear old clothes when using this stuff.

3) Phosphate/Copper. Only mentioned for completeness, since I've never used
it, due to its higher complexity. This stuff involves fairly complex mixes
of phosphates, pyrophosphates, nitrates, and aqueous ammonia. It has good
throwing power (enough so that it's what's used for plated-through holes in
printed circuit boards). This is also the bath of choice for plating over 
conductive-coated plastics, which is what I'm doing, and is most like what's
also been suggested with the graphite-coated inner tubes. That would seem to
make this stuff worthy of investigation, and I'll probably at least give 
it a try someday.

Just to make it interesting, various plating houses will sometimes add their
own "secret ingredients" to the bath. Often, various organics will cause the
copper to be deposited in smaller crystals, making for a stronger plate. 
That's just mentioned as a trivia item, though. I haven't gotten that 
deeply into it yet, and don't expect to have the time to do so for some time.

Wes B.