Re: [CAD_CAM_EDM_DRO] Sorting Steel Balls

On 5/1/02 9:33 AM, "Jon Elson" <elson@...> wrote:

> Ethan wrote:
>
>> Good morning.
>>
>> I'm looking for some ideas on how to sort some steel balls. I need to do
>> about 4,000 per day. Sizes range from 19/32 to 7/8 with some sizes being
>> within .001 from each other.
>>
>> I'd like to use a PC or PLC based machine that will measure the balls and
>> them sort them according to size.
>
> The most obvious method is a tilted track with graded holes in it, the
> smaller ones first. You could have a mechanism that holds a track with
> 20 holes in it, and tubes below that go to 20 different bins. The balls that
> are
> too large for any hole wind up in a reject bin, where all balls are larger
> than
> any hole in your track. You then run these through again after switching
> to the next larger track. It will take a little fooling around to find the
> right
> slope, spacing of holes, etc. One thing that comes to mind is to have
> wire rails that don't allow the balls to drop into the holes unless they are
> very close to being able to fit through the holes. You might need a 'wiper'
> to bump the balls out of the holes that they almost can fit through. A
> wheel with spokes, kind of like a water wheel could be used to control
> the flow of balls into the sorter.
>
> Another method could use bars made of hardened and ground shaft, with
> the balls rolling down an incline, and the shafts slightly farther apart at
> the downstream end. You might be able to sort a fairly large range of ball
> sizes with a few feet of this. Maybe a 1" square funnel leading to a
> container
> for each size, and the rails spreading apart at the rate of .001" per inch.
> I see you want 281 different sizes. That would be 23 feet long to do it all
> in one pass. But, it would only be 6 feet long to do in 4 passes.
>
> A commercial method I've heard of used a resilient board of some sort, and
> the balls were dropped from a calibrated height. They bounced to a different
> height depending on size, and fell into troughs based on size. I think this
> was
> in a line where balls were all within a few thousandths of the same size.
>
> I can't see how a PLC or PC would do anything but increase cost (VASTLY)
> and slow the process down.
>
> Jon
>
>

Hi all,
Jon's mechanical system using the ground rods sounds good to me. You could
cascade these if necessary.

Any sort by weighing system with a little sensitivity will work. On the
small end those balls are still better than a 100 mg different in weight and
closer to 300 mg difference on the large end of the range.

Dave