|
My primary
Gravity Wheel experiment that was designed to
use swinging weights and 'holder rods' did not
produce excess power. The mechanical apparatus
worked great but the design concept was flawed
due to my ignorance.
It was supposed
to have the weights on the right side of the
wheel swing to an 'outer' radius and the weights
on the left side of the wheel swing to an
'inner' radius. The 'holder' rods were to
prevent the weights from swinging too far and
making the wrong chains tight (take
weight).
(Show drawing
and pictures)
So I'm calling
this version 1 and I've spent hours to figure
out WHY it didn't work. Here are some of the
things I discovered:
The holder rods
make no difference. The weights would 'balance'
and 'lever/torque' on the holder rods, so that
the 'wheel' still 'saw' the weight as where it
actually was, the same as if it was just hanging
from the chains.
The rising
weights (inner radius) are slower and tend to
accumulate an extra weight on that side of the
wheel, which balances the extra torque created
by the 'falling' weights that are farther from
the center of rotation.
VERY
IMPORTANT:
The wheel does
not care how 'high' or 'low' the weights are.
The wheel ONLY cares how far the weights are
(horizontally) from the center of rotation.
Equal weights will balance when they are equal
distance (vertically) from the center of
rotation NO MATTER what height they are.
(show 'balance'
position)
I think this is
the key to designing a working wheel. I am now
designing using a computer (vector based)program
where I can measure the total inches on the left
and compare with the total inches on the right.
If I can keep the total inches (of all weights)
on the left less that the right at every degree
of wheel rotation, I believe I will have a
'self-turning' wheel.
So the important
issue to solve is how to get the weights on the
left moved IN as quickly and as far as possible;
at the same time get the weights on the right
moved out as quickly and as far as possible. The
wheel will be self running when the total
distance of all of the actual weights positions
are consistently further out on the right than
on the left. This will keep the center of
rotation and the center of gravity offset, to
develop torque.
(show 'distance'
picture)
Moving weights
takes energy, lifting weights takes energy. We
want to move the weights horizontally (to take
the least energy) and we want to NOT raise them
any farther than needed. We also don't want them
to drop unless they are producing
power.
I discovered
that there is a 'power' position for the
weights, where they will provide power for
90° of the wheel's turn, to 'balance'
position.
(show 'power'
stroke)
We need the
wheel to turn 180° to 'reset' the wheel,
more degrees would then result in acceleration
of the wheel and usable power output.
Note: I did not
use low friction bearings for version 1. With no
weights, the disks would spin only 4 rotations
with a moderate hand pull. I figured it would
work with normal bearings or I wasn't
interested.
I think that a
'working' design will function with a
'balancing' set of only two weights (no need to
make eight weights). I think if it won't work
using two weights, then it isn't likely to be
produce enough power to be practical. The wheel
needs to 'reset' itself every 180 degrees.
In my first
version I'm about 25 degrees short of 'reset'. I
think I know how to redesign to get more than 25
degrees.
Examining the version 1 wheel with only 2
weights. I've discovered:
That the
momentum developed (from zero movement at start
of power position) would carry the wheel
55° past 'balance' 90°, for a total of
145°; need another 35° to
'reset.
If I removed the
'highest' weight on the center of the wheel, it
made no difference; it is a 'null' weight. Yet
it took 'work' to lift it there, so removing it
reduces work required to turn the wheel,
increasing the 'power' degrees (see
below).
If I move one
'high' weight horizontally from the left side of
the wheel to the right side of the wheel (with
the 'null' weight gone), that I gained 20°
as it leaves the left side (as per item 2 above)
and another 25° as it goes out onto the
right side of the wheel; for a total 'rotation
gain' of 45°.
I think, these
gains in degrees, coupled together with momentum
will make a working wheel possible.
Note:
It may be a
thought to entirely remove the upper left weight
from the wheel, having it travel from left to
right on a 'ramp' that keeps the weight off the
wheel until it arrives on the right side of the
wheel, where it will reenter the grooves.
Because of the rotation of the wheel, I expect
the weight to reenter the wheel 'advanced' one
position.
With all that in
mind, I am designing a version 2 wheel. I have
eliminated the chains. The weights will roll in
grooves cut into the disks. The grooves are
designed to handle 7 weights in 8 positions. The
weights on the left side move further in to the
center than they could with the chains. The
topmost weight on the left side transfers nearly
horizontally to an empty space on the right side
of the wheel.
Now that I know
what to look for, I am looking to put the
drawings on CAD so that I can check the total,
consistent distance (of the weights) from the
center of rotation. This should tell me (on
paper) if the design will work or
not.
Back
to Gravity Wheel home page
|
