Jim Wilkins wrote:
"N_Cook" wrote in message
news
On 23/12/2017 13:32, Ed Huntress wrote:
On Sat, 23 Dec 2017 09:31:23 +0000, N_Cook
wrote:

Metal tape with round or square holes in, engaging with balls or
pegs
set in pulleys. Firstly any generic name for this?
I'm after design criteria, for steel tape, 1.25 or 1.5 inches
wide,
flexible enough to wrap around 1 foot diameter pulleys.
So the trade-off between tape thickness , size and spacing of
holes/sprockets for maximum tension/torque transfer and tape
integrity.

That sounds like a prescription for fatigue failure, if I
understand
you correctly. A lot depends on how big the pulleys are and how
long
they're expected to last. Bandsaw blades are made to take it, but
the
wheel size has to be large enough to keep the flexing within
bounds.

Have you worked that into your design?


Its actually very very slow action.
Trying to get some insight into a failure mode of this type and
model of tide gauge, this pic and the original in a 1920s
publication is other wise the only info.
http://trigtools.co.uk/data/2GL_AutoTG.JPG
The relevant pair of pulleys are the wide ones , one under the
mainframe and one with the brass cogs.
But resolution in that pic , or the original book , not up to seeing
the sprocket spacing.
The plotter drum records 24x1inch paper for 1 days record, so
circumference of 24 inches, for overall sizing.
The band may well have been bronze as a marine context.
There is surviving plot of when this machine "broke" in service as
well as another one. Both subjected to an extreme tide surge.
The traverse pen mechanism hit the designed-in endstop .
The perforated tape jumped some sprockets and resettled , neatly
still drawing , but displaced downwards. What was the surge level
that in a sense , it recorded. One insight would be the spacing of
the teeth, if only we knew, as it could only be an integre number of
the spacing , for the jump slippage. It was of the order 2 inches to
2.5 inches from meteorological data etc.


I suspect the strength of the engagement between the tape and sprocket
wasn't a primary consideration, as the sprocket drove only the
plotting mechanism. Does an easy-to-index factor of 360 such as 12, 24
or 36 make sense as the number of sprocket pins?

A sprocket of 11.459" diameter would revolve once per 3' of band
travel. It might have 24 pins spaced 1.500" apart, or 36 at 1.000". A
7.639" diameter sprocket would correspond to 2'.


He's in England, so it may be metric with a one meter circumference
which would make the diameter closer to 12.5"


The fusee and the frame suggest that a tower (turret) clock maker
built it, so a 30 tooth sprocket is possible.
http://www.model-engineer.co.uk/foru...s.asp?th=75826

This shows a clock built on a similar frame:
https://en.wikipedia.org/wiki/Turret...k_movement.png
-jsw