Re: CROWNING ROLLERS
- --- In VanDeGraaffGenerator@yahoogroups.com, "Dick Linder"
> Hal & all:OK, Dick...here's how it's done, and it works for me!
> Am I ready? Here I sit scratching my head, looking at a bunch of
> assorted blanks ready to be crowned.
> Am I ready he asks.....I think so.
> Graaffer Dick
First, forget about roller diameter. The belt is concerned only
with the surface it rides on, whether it's the size of a toilet paper
roll or a 55-gallon drum.
The formula for generating a convex crown is 1/8-inch for every 12-
inches of roller width; that is, the center of the roller will be
1/8" higher than the edges, as observed with a straightedge placed
across the width of the roller. In my case, my 4.5" diameter rollers
are 6.5" wide, so my crown was calculated at a bit over 1/16", using
the formula 1/8" x (6.5/12).
Our lathes can cut only linear tapers using our compound feeds, so
I had to use files and sandpaper to generate my crowns, checking
periodically with the straightedge. While running the spindle in
reverse, (with the top of the roller spinning away from me), I used
progressively finer files on my aluminum roller, followed with
progressively finer sandpaper, and finally Scotchbrite for a nice
smooth crowned surface. On non-metallic rollers, the same technique
is used, but with the cutting tools chosen for the material being
crowned. Hold the files backward, with the roller pulling AWAY from
you for safety and better control. The crown should be a smooth,
convex transition across the entire width of the roller.
Do not use a belt speed high enough where centrifugal force begins
to lift the belt from the rollers! Even a perfect crown won't help in
this case! If the charging rate must be increased, this can be
accomplished with a wider belt, which has the same effect as a faster
belt, but with more stabilty. Don't forget...centrifugal force
increases as the SQUARE of the belt speed! Increase your belt speed
by a factor of 1.414, and you DOUBLE the centrifugal force around the
- Hal & Graaffers all:
> The formula for generating a convex crown is 1/8-inch for every12-
> inches of roller width;OK..... got it so far
> Our lathes can cut only linear tapers using our compound feeds,so
> I had to use files and sandpaper to generate my crowns, checkingThat's what I was afraid you were going to say :-(
> periodically with the straightedge.
> Do not use a belt speed high enough where centrifugal forcebegins
> to lift the belt from the rollers!"Been there, done that" Another approach is to use a thinner lighter
belt. I can run my .003" thick mylar belt at 100 fps without any sign
I tried to cut a one degree taper with my lathe, first starting at
one end of the roller and ending at the middle and then starting at
the other end and ending in the middle. Even though the procedure
should work, my old Craftsman lathe just doesn't have the precision
any more to do that approach.
In order to check the uniformity of your manual approach I might make
a template with the proper arc radius to hold against the piece in
order to monitor the filing process.
Thanks for the info Hal.
By the way, I have not calculated the energy released by the 15'
diameter sphere at the MOS but I did have the opportunity to be in
the operator's cage when it took a strike. I would guess the distance
from the sphere to the cage is at least 25'. Sounded like someone
fired a .45 right next to me!
- DICK:A template is a good idea, using a small arc from a large circle (how large would that circle have to be?), but since we're talking about only 1/8-inch out of one foot, that's only 1/96 or just over one per cent, so this slight curve can just as easily be discerned by eye as by referencing it to a template, which would be difficult to produce accurately to such a slight curve anyway.I would suggest using my method, since the same amount of material must be removed from the roller whether you're using a template or not. Just be prepared to sit at your lathe for a long time, using files and sandpaper! You'll produce a fine crown! The eye can be quite accurate in geometric evaluations!Cheerio...HALSWORTH