On 13/12/2016 19:05, Chris Hogg wrote:
On Tue, 13 Dec 2016 16:37:54 +0000, Chris B wrote:

The torsion spring on my garage door transmits load to the door through
a large threaded "collar" end fitting similar to those shown on this
google image search.

Snip


Unless of course this is the way garage door mechanisms wear out and it
is time for a new door. (Parts other than cables for this door seem to
be unavailable).


I don't have experience of torsion springs (the two garage doors I'm
familiar with either have tension springs or concrete counterweights,
respectively), but looking at the pictures you linked to, and reading
some of the descriptions as to how to fit and torsion up such springs,
I don't understand quite why your spring works its way off the
fitting. AIUI once torsioned up, the spring should hold itself onto
the 'thread' on the end fitting. The only way I can see it working its
way off is if the torsion gradually weakens over the years. Also AIUI,
the end fitting that gets turned in order to torsion the spring has a
couple of locking bolts that are tightened up onto the torsion shaft
when the spring is torsioned by the correct amount. If over time those
bolts gradually slip on the shaft, it will allow the torsion in the
spring to relax, meaning that it won't be screwed onto the 'thread'
quite so strongly, and may eventually work its way off.

Does that make any sense at all? :-)


Well I have been trying to get my head around it as well, and
struggling, but the way I think it works is this.

I should just say that (in case it makes any difference) this is a
Cardale compact door which has the spring on the back of the door rather
than on the door frame above the door. I believe these are a bit rare
with only about 900 of them being made in the 1980s. Most advice on the
net seems to be for torsion springs on the frame above the door.

All fitting is done with the door closed

First of all the spring can be threaded over the end fitting by hand to
the end of the thread it is now effectively "fixed" to one side of the
door, the other end goes over (but is not yet clamped to) a shaft that
connects the cable pulleys on both sides of the door.

To fit the spring I wind up (with 2 spanners) the free end over the
shaft many times. As the spring is rotated it increases in length (to
my initial surprise but I think it makes sense now) and when it has been
wound up enough I hold the spring/spanner with one hand whilst turning 2
allan keys to lock the other end of the spring to the shaft. I can now
let go of the spanner, the spring doesn't move but the cables at each
end of the door become tensioned. This part of the fitting has not moved.

As the door rises the end of the spring now fixed to the shaft rotates
to help lift the door, but as it rotates both ends are now a fixed
linear distance apart so it must be coming under linear tension (even
though the rotational tension will be less) since as we were fitting it
with one free end the length increases as it is wound up.

When lowering the door the spring is now "wound up again" by the weight
of the door to its original position increasing the torsional tension
but reducing the linear tension.

I think it is this repeated increase/reduction in linear tension which
is making the spring slowly walk off the thread.

Just my thoughts on how it must work anyway - happy to be corrected.

--
Chris B (News)