UNCONSIDERED DETAILS.

Otago Witness, Issue 2410, 10 May 1900, Page 45

UNCONSIDERED DETAILS.

<a It is not possible for the average rider to examine intelligently every part of his ma- ( , chine (says "I. W. B." in the Field), and •] these hints are not given with the idea of in- , chicing him lo dismantle his bicycle .and pry , curiously into it? hidden mysteries. As a t j'lile, it i* best to " let well enough alone," and t ,bo content if the machine seems to run well. c Still, it falls to the lot of every owner to take c his machine . apart at tome time or other, ]; ] either personally or by deputy, and when r , , this is done it is well examine those parts in i] which even a non-mechanical person can r easily learn to distinguish the good from the tiad. The steel balls, which form so impor- p tant a part of a modern bicycle, rea<Uly lend tl I themselves to such examination, and a few 1< , remarks about these may be of interest. H I Including the pedals, some 10 to 12 dozeii g balk are used in the construction of an ordir A nary safety, the number differing according to gj the size of balls used, some makers using Jin, S where others use 5-16 in. It is of the utmost fi importance that all the balls iised should be tl perfectly round and well hardened, and that •« all those in any particular bearing— c. g., the w

bracket or hub— should be exactly uniform' in size. It is not by any means an easy matter fo.? any manufacturer to ensure that these three conditions are fulfilled, and I venture to say that, if carefully tested, most cheap machines, and not a few expensive ones, would be found to be defective in one or other respect. The most usual 'sizes are 3-16 in, and 5-16 in, and it seems a simple matter for a manufacturer to order a quantity of these from a good maker of steel balls, and to get what he orders. They will come to him packed in neat little boxes, and marked with the decimal size and " guaranteed," but in the course of a long experience I have never been able to rely implicitly on the guarantee. To ensure accuracy every ball should be gauged in several positions by an intelligent and reliable person, and with a micrometer gauge divided into thousandths or half-thou-sandths. Notwithstanding the wonderful accuracy of production, it is not commerciallypossible lo turn out balls which shall all be oi the " dead " size. For example, a iin ball should, of course, correspond to .250 on the micrometer ; but in the finishing only a proportion will rome out this size, while some will oaliper .251 or .252/ and others .249 or .248. The best ball manufacturers sort them out accordingly, and the slight variation does not matter much if all the balls used by aparticular bicycle manufacturer are. of the same decimal size, though it would be preferable, of course, to keep to the exect size. In gauging a' box of balls, all nominally iin or 5-16 in, I have often come across one which ' varied in one of its dimensions several thou- - sandths from the rest, and eueh a ball would be' quite enough to interfere with the free running and to increase the wear of a bear- • ing. Such a ball as I refer "to might "be of the pr-oper size in , one direction, "but confeiderably "out of round," and this can pnly be detected by careful - individual gauging, Of course, pome infinitesimal amount of varia-* tion must be', allowed; but I do not think that any ball in a bearing should vary more than half a thousandth from the standard decimal ' or one-lhoitsandth from any other I'al 1 . Greater accuracy than this\is desirable. In many cheap machines a Aariation of four otC.five thousandths would probably be found, and it is quite obvious that the larger balls will have to do nearly all the work, a'ld that the,Jbearing will suffer from the strain coming upon a reduced surface. \ \ It- is necessary to be very^ careful in replacing' balls after any repairs have been done. A rgpairei should place each set of balls in ( a -separate receptacle, so thai they are used again'^h "the same ploce. If he is a careful man -'h^s .will do this, but if not he may mix them upi with a miscellaneous lot, „and your .nvachroVtoay come bank wi*h quite a different set and pf"-unequal size, and you will wonder : why it .does not run so easily as it did. If you -have the misfortune to break or losa & ball, and -want one to replace, do not be content to ask for or p 5-16 in 1-all. but go to a good-place with a .sample, and a4c foi one gauged "the same size. In legard to hardness there is' little to fear. Even the commoner qualities are generally well-hardened, and apart from, irregularity in size the only danger would probably be in their being made of common steel and being liable to break up in use, which r -of course, plays havoc with' the bearing. ■ Many readers of the Field orobably remem» ber- the introduction-,of ball bearings some 18 "or -20 years ago.. Balls were then very ex,pensive and yery jiriaecurate. As the " demand increased new methods of making and . automatic gauging were invented, and at the present time good balls can be bought at very moderate prices, so that 'there is no excuse for using rubbish. ' In- 1 >the early days few people bothered' themselves about accurate calibration. I think I was one of the first to give this matter attention. I got Whitworth's to make me a sleel plate with three holes for each size of ; ball. One hole was the ' "dead" size — say, 5-16 iu — one was a thousandth smaller, and the third was a thousandth larger. If the balls would pass through the large hole and. would not pass • the small one, we considered them _ near enough. Subsequently, several Midland firms hal similar plates made. But shortly afterwards I ■ possessed myself of a' Brown and Sharpe's micrometer, which I have used for tho last 15 years. 1 may add that the balls in bracket and hubs are the most important, the uniformity of those in the head atld pedals being of less consequence.