Joseph Gwinn wrote:
In article ,
Bob Engelhardt wrote:


Joseph Gwinn wrote:

Bob Engelhardt wrote:

I still don't understand them [hinges].

The patent isn't awfully clear. But the idea cannot be all that
complex. It's probably a beefy variation on the invisible hinges used
on kitchen cabinets.

That was my thought, then I realized that with cabinet hinges the axis
of rotation isn't important. The axis for the MagnaBend hinges has to
be the intersection of the clamp plane and the bending bar plane.


Building them would be totally out the question for me.

That isn't at all obvious just yet. Some of the later hinge designs look
perfectly practical for a HSM, being two or three orthogonal
pin-in-sleeve hinge joints in mechanical series.

Maybe. One thing that troubled me was the patent's description of one
axis of rotation intersecting another axis at yet a third axis.


I think I figured the fancy hinge out. They don't come out and say it
(in 4,513,475), but they are more-or-less implementing a virtual ball
joint: The center of rotation is the intersection point of the three
hinge axes. In some variants, one of the hinge axes is a goinometer
mechanism (the cylinder-segment bearings). In all cases, the point is to
make the center of rotation be outside of the actual hinge mechanism.

It takes a minimum of two such hinges to define the axis line about
which rotation occurs, just like with ball joints in automobile steering
gear.

The three-pin hinge (figure 20) isn't stiff against side-to-side motion,
so the hinges are provided in pairs, one hinge right-hand the other
hinge left-hand, just like gloves.



The thread on the Practical Machinist forum was started by a guy who was
going to make a brake and had made a prototype or mock up of the hinge
that he claimed worked. I could try one in pine, just to get the idea.


Sounds like a real good idea.



Although there has been a thread on using discarded microwave oven power
transformers as the magnet, it isn't obvious that this is necessary.
Given that the excitation current will be DC, laminated steel is not
needed, so one could cobble a magnetic circuit from ordinary mild steel.

Yeah, that's a sub-thread in this thread G.

I don't think that I want to do that much coil winding. You'd need a
couple of hundred turns to keep the current at reasonable levels, and
I'd want one at least 24" long.


The MagnaBend patent (4,111,027) gives some coil data in Column 5 Lines
5-12:

"A specific construction of the above described tool had a length of 600
mm, a weight of 20 kg. (not including keepers), a coil formed from 22
guage copper wire and weighing 2.4 kg., operated on a 240 volt, single
phase, 50 cycles per second AC supply and consumed, intermittently, 4
amps. That specific construction was able to exert a holding force on
sheet metal of about 4 tonnes. "

Apparently, the Australians used AWG (American Wire Gauge) sizes for
copper back then, and probably have gone over to IEC metric wire sizes.
In any event, #22 AWG wire with single build (thickness) insulation is
1.972 pounds per 1000 feet, and 2.4 Kg is (2.4)(2.2)= 5.28 pounds of
wire, which would be 2,677 feet of #22 wire. The brake is 600mm wide,
which is 600/25.4= 23.62" wide, call it 24" or 2 feet. A turn is
therefore 4 feet, so 2677/4= 669.4 turns, call it 670 turns.

Can you be certain the Australians were using AWG and not SWG, it makes
a difference. Their video mentions it bending "16 gauge" and their
specifications mention 16g/1.6mm which would indicate SWG is in use at
least for the metal specs, US metal gauges are thinner for the same number.

This sounds like a lot, but it is certainly doable by hand, especially
if one cobbles together a simple winding machine out of wood and powered
by hand. One would wind on a wooden form, not on the iron, just as is
done when winding a motor.

From a cross-section drawing in the manual, the winding space is 20 by
28 mm (0.787" by 1.102", 0.868 square inches), which will accommodate
1191 turns of single build, so there is space. In practice, one would
most likely use double build (to better handle the voltage in a single
winding), allowing 1099 turns. There will also be heavy insulation
between the coil and the iron; this will reduce the area available for
winding. But it looks like we have a viable solution. This, for 220
volt systems. Fewer turns of heavier wire will yield the same magnetic
flux in a 120 volt system. Roughly, 335 turns of #19 AWG wire, pulling
8 amps.


The ampere-turns product is (4)(670)= 2,680 amp-turns. This yields a 4
metric tones clamping force in a length of 0.6 meters, or 4000/0.6 =
6,667 kilograms per meter, which is 372.5 pounds per inch.


Joe Gwinn