ball bearings

[Christopher POVEY]

charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

The recent comments on ball bearings have been interesting. The comment fro=
m David that 'his' ball bearings had open sides reminded me of an inspectio=
n of a tower structure by a member of the Towers & Belfries Committee in (I=
think) a Derbyshire tower. It was pointed out to him that all the bells ha=
d been fitted with ball bearings. To his quite reasonable surprise he found=
that ordinary ball bearing races had been fitted onto the gudgeons and tha=
t the races themselves were held onto the frame by a flat-section inverted =
'U' straps. The balls themselves were fully exposed. The local ringer there=
was very proud of the conversion. This had clearly been done on a DIY basi=
s and clearly without the knowledge of the Diocesan authorities and a facul=
ty.

If David's bearings are anything like this, he needs to be worried. Firstly=
, ball bearings are manufactured to extremely fine tolerances and any grit =
entering the bearing (as the above arrangement) will very soon destroy the =
bearing. Secondly, the grease is there in part to allow the balls to slide =
against each other or against the cage, but more so to prevent corrosion in=
the surfaces; and this is the reason for using non-acidic grease. Should t=
his type of grease be in contact with air, it will soon oxidise, go hard an=
d be rendered useless. That is another reason why ball bearing housings are=
sealed. Thirdly, ball bearings used for bells are self-aligning, ie they c=
an accommodate a certain degree of angular movement of the gudgeon, which a=
rises from irregular movement of frame sides, for instance. An ordinary bea=
ring cannot do this and will soon wear out if used in such circumstances. F=
ourthly, ball bearings for bells are given end float in one of the housings=
, to accommodate longitudinal movement in the gudgeon (from frame sides tha=
t move towards and away from each other). If there is no end float, the bea=
rings will become axially-loaded and the bell can be harder to swing becaus=
e of this. Mears & Stainbank housings for single-row ball bearings initiall=
y failed to provide for end float. Lastly, bell bearings have to withstand =
4 times the deadweight when the bell is rung full circle. A 5cwt bell will =
exert 1 ton downwards as it revolves. The bearings need to be sized for tha=
t loading, otherwise (again) wear will occur. G&J's ball bearing were alway=
s on the small size. Alan Hughes told me they always change the bearings wh=
en they refurbish G&J-hung bells. Wooden bellframes tend to move around mor=
e than metal frames, so the self-aligning facility is very important. The e=
nd-float is similarly important with wooden frames, although temperature va=
riations in metal frames and fittings can produce similar results.

To get some historical aspects into this subject, it may be interesting to =
know that when ball bearings were metrically sized (generally Swedish: SKF)=
when initially used for bells. The British ball bearing industry then prod=
uced imperially sized bearings and these were used. The metric revolution o=
f the 70's encouraged the use of metric bearings world-wide and imperially-=
sized bearings were gradually phased out. The bell industry started to used=
metric bearings again. I understand that SKF no longer produce imperially-=
sized bearings. Imperially-sized bearings are now comparitively difficult t=
o obtain and expensive compared to their metric counterparts, which is bad =
news for those towers that have such bearings and where they have become wo=
rn (yes, ball bearings do wear out, even with the correct maintenance - whi=
ch is do nothing to them). However, those towers that had early ball bearin=
g use are still very well catered for, because metric bearings are cheap an=
d plentiful. It is a myth that the American ball bearing industry produces =
imperially-sized bearings in vast quantities. The Americans woke up to the =
fact many years ago that if they were producing goods to go all over the wo=
rld - particularly vehicles - the breakdown of an imperial bearing in, say,=
some Third World country would be catastrophic for future sales, as it wou=
ld mean the vehicle would be laid-up until an imperial bearing could be shi=
pped out. Far better to use metric bearings, which would be stocked at the =
local garage or bearing supplier. I understand the Americans just measure t=
heir bearings imperially.

The double-row self-aligning SKF bearings at Evesham did 43 years before mo=
st of them were either badly worn or becoming so. Single-row self-aligning =
bearings generally last much longer, because there is more metal-to-metal c=
ontact with the balls, but this type of bearing suffers from fretting betwe=
en the self-aligning collar and the outer bearing race. The double-row self=
-aligning type is (I think) used universally now. The interesting aspect of=
worn ball bearings is that the bell continues to ring well, unlike worn pl=
ain bearings, which affect go straightaway. Worn ball bearings emit rumblin=
g noises; the louder the noise, the greater the wear. A ball-bearinged bell=
only starts to go badly when the bearing is very badly worn, such that the=
wear has allowed the gudgeon to drop sufficiently to rub on the housing. I=
suspect there are a lot of worn ball bearings out there, but they are unde=
tected because the bells are still going well - and it's very difficult to =
hear the rumbling above the ringing.

There was a lot of correspondence in RW in the early/mid 1940s about the me=
rits of ball bearings and plain bearings. Some ringers thought ball bearing=
s were harder to ring on than properly-lubricated plain ones in good condit=
ion. Because the oil film in plain bearings reduces when the bell stops at =
mouth-up, it allows the bell to 'sit' at the balance, rather than having to=
hold the bell at the balance as with ball bearings. Most of us have either=
been taught on ball bearinged bells or become used to them, so the 'sittin=
g' of plain-bearinged bells has less significance. The greater difficulty o=
f ringing slowly with ball bearings has, of course, led to faster ringing, =
ie not up to the balance. The problem now appears to be getting ringers not=
used to plain bearings to ring slowly on plain bearings to take advantage =
of the 'sit'.

Chris Povey
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ringingworld.co.uk/pipermail/bell-historians/attachments/20030216/194c2ea5/attachment.html>