☙

ball bearings

Because the best ball bearings in the early 20ᵗʰ century were made by German manufacturers, ball bearings have usually been sized in millimeters, even those made in the United States – but the balls themselves were frequently sized in inches. The Anti-Friction Bearing Manufacturers Assn. adopted a classification of ball bearings according to their construction, with each type identified by a symbol (such as “BA”). Symbols that include the letter “I” identify types that are nominally sized in inches; of ten types, only two have an “I.”

Embedded in the manufacturer's model number for most ball bearings is a three-digit code that gives its dimensions in millimeters. The first digit refers to the series. There are four principal series:

100, extra light
200, light
300, medium
400, heavy

The series designation indicated the outer diameter of the bearing. It is followed by two digits which indicate the size of the bore, but beginning with 04 the bore was 5 times the designation number.

Designation	Bore, mm	Series 100 (x = 1)		Series 200 (x = 2)		Series 300 (x = 3)		Series 400 (x = 4)
Designation	Bore, mm	OD, mm	Width, mm	OD, mm	Width, mm	OD, mm	Width, mm	OD, mm	Width, mm
x00	10	26	8	30	9	35	11	—	—
x01	12	28	8	32	10	37	12	—	—
x02	15	32	9	35	11	42	13	—	—
x03	17	35	10	40	12	47	14	62	17
x04	20	42	12	47	14	52	15	72	19
x05	25	47	12	52	15	62	17	80	21
x06	30	55	13	62	16	72	19	90	23
x07	35	62	14	72	17	80	21	100	25
x08	40	68	15	80	18	90	23	110	27
x09	45	75	16	85	19	100	25	120	29
x10	50	80	16	90	20	110	27	130	31

The table shows the sizes up through a bore of 50 mm, but much bigger sizes were defined. In the medium series, for example, a 356 bearing takes a shaft with a diameter of 280 mm — more than eleven inches.

Notice that this system specified only the size of the bearing, not the size of the balls inside the bearing.

resources

Bearing manufacturers have developed some outstanding websites with detailed information. In particular, see

www.skf.com

historical sources

Up to the date of the advent of the modern bicycle, ball bearings had no practical application. They were scientific toys, mechanical curiosities, of admitted excellence, truly, but far too complicated and delicate for ordinary use. But when man became his own horse, the first task for his ingenuity was the devising of means for lightening his labor. And the ball bearing proved to be foremost among such means.

W. H. Hale.
The Theory and Construction of Ball Bearings.
Scientific American, vol. 81, no. 1 (July 1, 1899).
Page 8.

The first practical use of ball bearings was in a crane in 1845 at the Sayner Hütte Foundry in Sayn, near Endorf, in Germany. By 1871 the Krupp Works was also using them in cranes, and by 1885 in naval gun carriages. But the author correctly emphasizes the importance of the bicycle; its role in the history of ball bearings can hardly be overestimated.

Without ball bearings the bicycle would have been a commercial failure. Those whose memories carry them back to the year 1890 may have recollections of trying to push a bicycle equipped with plain bearings; which was a few degrees worse than having no bicycle.

Ball bearings were such a wonderful success, commercially, in bicycles, that attempts were made to apply the same design to heavy service. The loads and speeds of bicycle work are trivial. Attempts to use such construction in serious engineering work were usually dismal failures. Sometimes they worked fairly well, sometimes they failed utterly.

A few thinking men began to search for the reasons. It was clear to them that the failures were not to be attributed to any mysterious causes, but just to plain ignorance of facts. Therefore, they sought the facts.

One very clear reason was defective balls; out of round and not true to size.

Otto Hoffman, an American, a resident it is believed of the State of Ohio, conceived a machine wherein the balls would generate themselves as true spheres and to remarkably close limits of size. Also, it was a quantity process. The Hoffman process is still the best and apparently always will be, as it is fundamentally correct since it is based on the laws of rolling motion, later explained.

Hoffman was unable to raise sufficient capital in the United States, so he went to England, where he started a company which became very successful, and the prophet without honor in his own country saw many million American dollars pass to England to buy his product.

About 1896, the German Small Arms and Ammunition Company, a subsidiary of the German Government, started an investigation of ball bearings, the report on which has become classical, and is now most commonly known as “Stribeck's Treatise” (pronounced Streebeck). As a result of the investigation, the German Small Arms and Ammunition Company started the manufacture of Ball Bearings according to the first design which can claim any scientific, though defective basis.

The sale of their bearings was carried on in this country.

A public report of Stribeck's treatise may now be found in Vol. 29, May 1907, transactions of the American Society of Mechanical Engineers.

Louis Langhaar.
Principles of Bearings.
Aurora, Indiana: Langhaar Ball Bearing Co., 1922.
Pages 31-32.

The history of the anti-friction bearing is closely identified with the development of the bicycle and the automobile.…Crude ball bearings were first considered about 1862 to 1870, the Patent Office files showing several patent applications at about this date. These early ball bearings, however, were not practical and are merely of theoretical interest. The first practical applications came in the early nineties with the use of the cup and cone bearings on bicycles. The tremendous development of this industry furnished the stimulus for a very intensive effort toward the production of a bearing which could be more general in its application. The first type of the annular or ring ball bearing was developed about 1901, based upon investigation made by Professor Strybeck. The annular bearing consists of two entirely independent concentric rings of practically rectangular section and with groove shaped raceways in the outer surface of the inner ring, and the inner surface of the outer ring. In this groove a row of balls is placed and the whole is a self-contained, non-adjusting bearing. It was recognized that theoretically the annular bearing should contain as many balls as the space permitted.

…At about this time a large number of ball and roller bearings were designed and built, some of which proved to be very satisfactory, and some quite the contrary. Little was known as to the proper selection of the size of balls, and many misapplications were made.

In 1905 the first commercial application of annular ball bearings was made in America, although these bearings had been previously used in some of the foreign countries. The application was on a marble polishing machine for the Empire City Iron Works in Long Island, and the application proved quite satisfactory. At about the same time the first application of ball bearings was made on electric motors. It is at least interesting to note that the two electrical manufacturing companies who were the pioneers in the use of hall bearings for motors are still using such bearings, and consider them entirely satisfactory. In 1905, ball bearings were being favorably considered by the automobile industry, and during that year they were applied to the transmission of a 50 Hp Thomas Flyer. In 1906, an Apperson built for the Vanderbilt Cup Race had its transmission so equipped. In 1910, one of the machine tool companies exhibited at Atlantic City and elsewhere a vertical spindle drill press mounted on ball bearings, which drove an inch and a half drill through 31 inches of cast iron per minute, and a three and a half inch drill through 11½ inches of metal per minute. This of course meant a very heavy end thrust on the bearings. Also in 1910 ball bearings were used to take the thrust on belted vertical milling machines. It is interesting to know that ball bearings have been used on power punch presses where there is undoubtedly a very heavy shock condition. The claim is made that on such a press the use of anti-friction bearings decreased the idle stroke power 54 per cent and the working stroke power 20 per cent, and that the power saving for the complete cycle was 40 per cent. 1910 marked another interesting application of anti-friction bearings to a cold saw which ran at 4100 rpm. The use of plain bearings had required renewal of the bearing about once in three weeks with the consequent shutting down of the machine. Twenty hp was required to drive the saw when equipped with oil type bearings, and water cooling was employed. After the change in the bearings a 7½ hp motor was substituted for the 20 hp, and there is no evidence that further bearing trouble occurred on the machine.…

In no other line of manufacture has there been a greater attempt at standardization and many of the bearings of different manufacturers are absolutely interchangeable.

A. M. MacCutcheon.
Anti-Friction Bearings in the Steel Mill.
The Blast Furnace and Steel Plant, vol. 9, no. 10 (October 1921), page 600 ff.

That a very small amount of wear takes place when ball bearings are properly managed is well known, but this subject has only been actually investigated a short time. Mr. C. V. Boys has recently given the results of some measurements made by him upon the actual wear due to friction taking place in such ball bearings. At the end of every 200 miles run he cleaned the balls, and then weighed them with all the possible care and accuracy that the resources of a physical laboratory would permit. The set of twelve when new weighed 25.80400 grammes, and after running 1,000 miles they weighed 25.80088 grammes, the loss being 3.12 milligrammes, or less than 1/20th of a grain—that is to say, each ball in running 1,000 miles lost only 1/250th of a grain. This corresponds to a wear of only 1/158000 of an inch on the surface. Thus, at this rate of wear the balls would lose less than 1/34 of their weight when in use in travelling as far as from the earth to the moon. Further experiments have been since made by Mr. Boys, in which the balls were not taken out for the whole distance of 1,000 miles. The balls actually showed only 1/5 of the wear above mentioned, from which it would appear that the wear due to such grit and dirt as cannot be entirely excluded at first, but which does not get in after the bearing is properly screwed up, does not continue after the dirt is ground up, after which the balls literally do not wear at all.

These astonishing results point to the fact that ball bearings have a high mechanical efficiency, and it is extremely probable that as the cost of their production becomes less, and the mode of accuracy and quality of their manufacture improves, they will be employed not merely as now for the occasional use in such cases as for the foot of crane posts, but for many other purposes on a large scale, such, for instance, as in the thrust bearing of screw steamers, and possibly even to carry the shafting itself.*

*Since delivering this lecture I have seen, for the first time, an account of the actual application of ball bearings to this purpose (Engineering, Aug. 6th, 1886. p. 132). Mr. Wells, of Northwich, Cheshire, more than 18 months since, applied a ball thrust bearing to the steam launch Delaware, the speed of the engines being in consequence increased from 177 to 187 revolutions, with the same boiler pressure and grade of expansion. A ball bearing was afterwards fitted to the steam tug Volunteer with very satisfactory results. With a 5-inch shaft there are two concentric circles of balls employed, one pair for going ahead and one pair for going astern, and it is stated that the bearing requires little oil, that the wear is very slight, that there is in consequence no end play on the shaft, and that the first cost is moderate.

H. S. Hele Shaw.
The Methods of Reducing the Resistance of Friction.
Journal of the Society of Arts, vol. 34, (Nov 12, 1886), page 1289.

Standardization of ball bearings in America received its first impetus in 1911 when the Society [of Automotive Engineers] adopted the three series of metric annular ball bearings known as the light, medium and heavy series. These bearings followed the German standard sizes in their main dimensions as at that time Germany was the principal manufacturer of ball bearings and many of them had been imported into the United States. It was recognized that standardization would have to follow closely the practice which had become well established in the American industries.

R. S. Burnett
International Ball Bearing Standardization.
Journal of the Society of Automotive Engineers, vol. 8, no. 6 (June 1921).

The dimensions of ball bearings are well standardized. Oddly enough, ball diameters are universally expressed in inches and fractions thereof, while the dimensions of the races are given in either millimeters or English units. German builders use millimeters with a single exception where a firm has developed a series in English units, adapting them for the British trade, though that firm also uses chiefly ball bearings in millimeters. Even in England most ball bearings are made to millimeters as is also the more general practice of American manufacturers who have followed the German example. This general adoption of the millimeter dimensions is due to the fact that early German makers rehabilitated the ball bearing by the development of the principles and construction data of the modern type and secured a wide vogue for their products that made the sizes standard. As a rule each manufacturer makes a wide and narrow type of radial bearing, with three series for each; namely, light, medium, and heavy. In some cases a fourth series has been standardized, known as extra heavy.

F. A. Halsey.
Handbook for Machine Designers and Draftsmen.
New York: McGraw-Hill, 1913.