View Single Post
  #110   Report Post  
Posted to rec.woodworking
Prometheus Prometheus is offline
external usenet poster
 
Posts: 833
Default TS_Alignment_Reinventing_the_Wheel? -- Apology to Ed

On 12 Nov 2006 12:27:14 -0800, wrote:

I assume that most of your work with the laser is cutting all the way
through. So, yes, you don't have to worry much about the Z axis. And,
if you are just going to cut shapes from large sheets then there's
really no precise setup needed. Just set your working origin to the
machine origin.


Yes and no- there is a lot of etching scales and part numbers, but
that's more a function of cut conditions than Z-height. Leaving it
higher to etch would not only cause terribly ugly etching, but also
scatter laser radiation all over the shop. For through cutting.
there's a very small sweet spot with the focus that changes with
material composition and thickness.

And while you're correct that cutting shapes laser-complete does not
require precise setup, about 60% of what I do comes off a Vipros punch
first (that thing is a whole other ball of wax, and may be more than
most people think as well), and the setup requires careful adjustment
to get the cutouts to properly locate punched holes and formed
features. It's a little different than a mill, of course, because the
machine's zero is defined by a set of three stationary guage blocks
rather than points determined by an edge finder.

The tricky stuff really happens when the Vipros guy is loading
manually, and the sheets are cockeyed by an unspecified amount. I
need to compensate for that later by shimming the blocks. In effect,
the sheet itself has to be treated like a sine bar- where the guage
blocks on the X axis are like the pins, and feeler guages between the
sheet and the blocks are used to define the angle. It's a PITA,
because rather than using a set hypotenuse, it needs to be calculated
for each sheet by using the adjacent leg and a measured hypotenuse
(from the center of a punched hole or inside corner of a square punch
to a guage block) to find the angle of error, and then compensated
for. If that gets skipped or is calculated incorrectly, the
cumulative error on a sheet with 50 parts (or even 6 parts) can easily
throw the parts in the bottom corner out of spec, even if the first
part is apparently good.

The setup comes in with parts that have been to the brake as well-
sometimes, they need to be cut after they are broke to insure that the
holes are not deformed, or to remove alignment tabs that the brakes
use for their back gauges . In that case, the same technique applies,
as well as adjustment of of the Zht, often at multiple levels for
different features. Especially in the case of cutting off alignment
tabs, the setup has to be dead-on, or else the laser will make a very
nasty burr or divot on the lead-in or coming out of the cut- our
customers will generally not accept a ground or repaired edge (we just
cut off the tabs with a notcher for those that will), so it has to be
right the first time.

Add in the macro handling and nesting, and it's fairly involved.

I was avoiding the use of "sheet metal" when talking about it for an
obvious (to me) reason- as soon as those words get used, people get a
picture of making duct work out of gavanized steel with a shear and a
die grinder in thier heads right away, and immediately dismiss the
whole concept as contemptable- that exact tactic was even used by
another poster as an insult to your product in another thread.

What we do is about as similar to duct work as a pipe organ is to a
child's xylophone. Sure, the material comes in sheet form rather than
bar stock, but it fills a need in machining that would be wasteful to
do on a mill, and impossible to do on a lathe. The next time you open
your PC, try and imagine making those parts on a mill or machining
center, and I imagine you'll see what I mean- it can be done, but it'd
be a whole lot of senseless work and waste to use cutters, when
compared to the thin kerf of a laser. We make similar parts for
supercomputers that take up an entire room (yes, they still make them,
even with really good PCs available) where thousands of parts need to
fit together perfectly with both each other and the circuit board
features or the cumlative error is huge- and you can bet the end-user
is bound to get a little edgy if their multi-million dollar toy isn't
up to snuff.

It's the setup of a milling machine (manual or CNC) that requires the
skill. You can get unskilled labor to run a Bridgeport once it's set
up. Have you ever seen a handgun factory? Literally hundreds of
manual milling machines all set up with customized fixturing. Each one
set up to do one specific operation (drill a hole, mill a slot, face a
surface, etc.). The operator mounts the part and pulls a lever (the
quill feed or the table feed). All other adjustments are locked down.


Nope, but I see what you're saying- I have set up guys to do single
operations like countersinking or reaming holes on it when I'm too
busy to stand there and do it.

Again, the problem arose when you described doing this sort of work
without using dial indicators. Fact is that you do use dial indicators
- not for the sheet metal fab but for the real machinist work. To
bring it full circle, I would classify a table saw more like the
woodworker's version of a Bridgeport manual mill. It has a bunch of
alignments and adjustments that greatly benefit from the use of a dial
indicator. Strongman's jig addresses one of those adjustments.

It is possible to align a vise on a milling machine without an
indicator. But, what would you think of a machiniist who spends an
hour checking test cuts with a square (or by flipping parts over
looking for a gap)?


All right, point taken- I wouldn't think much of him.

(As this has been way off-topic for the group, I'll note that the
laser cutter can and occasionally does cut plywood as well.)