USA 02The Age of Precision and Specialist Makers

Brown & Sharpe and precision tools (1833–1863)

Brown & Sharpe began as a small clock and instrument shop opened in 1833 by David Brown and his son Joseph R. Brown in Providence, Rhode Island. For nearly 20 years the shop had no power, and even in 1853, when Lucian Sharpe joined and it became J. R. Brown & Sharpe, it employed only 14 people^[1].

The graduating engine and the vernier caliper

The turning point came in 1850, when J. R. Brown built a dividing engine that graduated straight scales automatically. It had devices to compensate for errors and wear in the machine itself, and similar machines were added in 1854 and 1859. With scales graduated on this machine, he brought out the vernier caliper in 1851, which could be read to 1/1000 inch (about 0.025 mm)^[1]. Roe writes that it was the first tool an ordinary machinist could afford that measured with precision. Sales were slow, and only four were made in the first year, but within a few years it was selling well^[1].

Sharpe also noticed how much wire and sheet gauges varied, and devised a gauge whose sizes changed by a regular progression. In February 1857 eight leading brass makers adopted it, and it became what is now the American Wire Gauge (AWG)^[1]. In 1855 they also built a precise gear dividing engine for making clock gears and index plates. The copper ring carrying its graduations was graduated for Brown at the U.S. Coast Survey in Washington, where he went in person^[1]. Owning their own tools for measuring, graduating and dividing became the foundation for the company's later machine tool building.

The universal milling machine (1862)

When the Civil War began, the Providence Tool Company took a contract for Springfield-pattern muskets. The drills used to bore the musket's vent were made from tool-steel wire, with the spiral grooves cut by hand with a round file. Superintendent F. W. Howe (the Robbins & Lawrence designer of Chapter 1) told J. R. Brown about this, and Brown, whose company also used the same drills for parts of the Willcox & Gibbs sewing machines it made, set out to design a new machine^[2][1]. The result was the universal milling machine, and the first one was sold to the Providence Tool Company in March 1862. Ten were sold within the year, and in December Scientific American featured it^[1][3].

What made it "universal"? The workpiece was held in a dividing head (a device that turns the work through set angles), the table feed screw was geared to the dividing head so that the work turned slowly as the table advanced, and the table could also be swiveled in the horizontal plane. This made possible helical milling, as for the spiral flutes of a twist drill. The same machine could also cut gear teeth and irregular shapes^[2]. The base was a single hollow casting, with the table carried on a knee that moved up and down. Rosenberg calls it the best example of a machine built to solve one narrow problem that spread to unexpectedly broad uses^[2]. The structure of the universal milling machine and how it led to the knee-type milling machine are covered in Chapter 7.

Setting standards: threads, gauges and length (1864–1879)

The Sellers thread (1864)

Screw threads used to differ in pitch and form from one shop to the next. Nuts had to be recut, and replacement parts did not fit. On April 21, 1864, at a meeting of the Franklin Institute in Philadelphia, its president William Sellers presented "A System of Screw Threads and Nuts"^[4]. He proposed a thread angle of 60 degrees, with the crest and root cut off flat to a width of 1/8 of the pitch, and the widths of bolt heads and nuts fixed by formula^[1].

Britain at the time had the Whitworth thread (55 degrees, with rounded crests and roots). Sellers rejected it for three reasons: rounded crests and roots are hard to fit exactly, a 55-degree angle is hard to verify, and cutters of that exact form are expensive to make. In Sellers's view, 60 degrees was easy to make and already widely used in the U.S., and flat crests were simple both to make and to check^[1]. It was a standard that put what shops could make and inspect ahead of theory. Recommended by a Franklin Institute committee, it was adopted by the U.S. government in 1868, the Pennsylvania Railroad in 1869 and the railway car builders' association in 1872, and in 1898 an international congress in Zurich adopted it as the standard thread form^[1]. The first issue of American Machinist in 1877 explained the U.S. standard and Britain's Whitworth side by side^[5]. Thread types and today's standards are summarized in the Threads pages.

Formed milling cutters and the micrometer

In November 1864, J. R. Brown's formed milling cutter that keeps its shape when only the face is ground won him a patent^[6][3]. The back of each tooth is relieved along a curve that follows the cutting profile, so the shape of the cutting edge stays the same when the face is ground. Before this, formed cutters that grew dull had to be annealed, recut and hardened again. The 1880 census report says that with Brown & Sharpe's formed cutters, any number of identical cutters could be made, and since grinding did not change their shape, they were important for interchangeable work^[7]. Their first purpose was cutting gear teeth, and from here grew the company's system for cutting involute gears (see Chapter 9)^[1].

Reading length by means of a screw, the micrometer also became a shop tool in this era. The principle was old: in 1848 Palmer of Paris had patented a "screw caliper." In 1867 Bridgeport Brass, troubled that three sheet gauges that should have been identical all disagreed, brought in a prototype. In August that year Brown and Sharpe saw Palmer's caliper at the Paris Exposition, adopted Palmer's method of dividing the scale for easy reading, added a clamping knob and wear adjustment, and in 1868 put it on sale as the "pocket sheet-metal gauge"^[1]. The name "micrometer caliper" came into use with the 1-inch model listed in the 1877 catalog^[1].

Pratt & Whitney and the standard of length

F. A. Pratt and Amos Whitney, who had honed their skills at Colt's factory, rented a small room in Hartford in the summer of 1860 and began building thread-winding machines on the side while still employed. The Civil War brought more work than they could handle; in 1865 they built a factory on the present site, and in 1869 the business became The Pratt & Whitney Company, with capital of $350,000^[1]. Besides machines for making guns and sewing machines, it made thread gauges and tools, and its pipe thread gauges were long the national standard^[1].

To exchange interchangeable parts between factories, every company's gauges had to mean the same "inch." But when Pratt & Whitney compared national and foreign standards with its own, the same bar came out longer or shorter depending on how it was measured, and the results were unusable. So W. A. Rogers of Harvard and G. M. Bond of the company went back to reliable copies of the London yard and the Paris meter and, with the help of the U.S. Coast Survey, built the Rogers-Bond comparator to compare them anew. It was based not on end measurement but on reading graduation lines under a microscope, and the results, three years and several thousand dollars in the making, were endorsed by a committee of ASME (the American Society of Mechanical Engineers)^[1]. Brown & Sharpe, too, found discrepancies among the sets of Whitworth plug and ring gauges it had long relied on, and in 1878 built a measuring machine to serve as its own standard^[1].

The universal grinder and the automatic screw machine

Hardened parts cannot be cut, so they are finished with a grinding wheel. From the early 1860s, to finish parts such as sewing-machine needle bars, Brown & Sharpe used a converted lathe, the grinding lathe, and from around 1865 also sold it to other companies^[1][2]. Over several years from 1868 it developed a machine designed from scratch for grinding, the universal grinder, and exhibited it at the 1876 Centennial Exhibition in Philadelphia. Rosenberg regards it as the direct ancestor of today's heavy production grinders^[1][2]. The patent was granted to Brown's heirs in 1877, after his death^[12][8].

C. M. Spencer, who came from Colt, took his idea from a machine for turning sewing-thread spools and built an automatic turret lathe that cut screws automatically. A cam drum with interchangeable plates set the sequence and feed of the tools, and Rosenberg regards it as the origin of all modern automatic lathes. The most important part, the cam drum, was overlooked by the patent attorney and left out of the patent^[1][2]. The patent granted in September 1873 (U.S. Patent 143,306) covers a mechanism for cutting threads with the chuck kept turning, and automatic stock feed^[9].

Such machines also went out to the arsenals of other countries. In the 1870s Pratt & Whitney supplied machines for making Mauser rifles to the Prussian arsenals at Spandau, Erfurt and Danzig, and an 1875 letter of thanks from the Prussian government said they had largely freed production from dependence on craftsmen's skill and halved wages^[7][1]. The spread of the American system as seen from Germany is covered in the German Machine Tool Archive.

1877: a trade journal is born. In November of that year, American Machinist, a monthly for machinists, engineers, founders, boilermakers, patternmakers and blacksmiths, was launched in New York. The first article of its first issue was on a universal chuck made by Pratt & Whitney, and page 3 explained thread standards^[5]. New machines and ideas became known nationwide through its pages, and Roe's 1916 book was itself partly based on articles first published in the magazine^[1].

Specialist makers and Midwestern machine tools (1880–1899)

According to Roe, before 1880 American machine tool building was almost entirely east of the Allegheny Mountains. Machines built in Ohio or Indiana were copies of eastern designs, and not of high quality^[1]. In the 1880s machine tool makers sprang up in Cleveland, Cincinnati and Hamilton, and within ten years they were competing with the eastern makers. Rosenberg sums up the 30 years from 1880 as a period with few new machines on the scale of the milling machine or the turret lathe, in which instead machines narrowed to specific purposes multiplied explosively^[2].

Warner & Swasey and Cleveland

Worcester R. Warner and Ambrose Swasey, who had been foremen at Pratt & Whitney, set up on their own around 1880, starting in Chicago, but moved to Cleveland the next year because they could not find skilled workers. Their first order was for 12 turret lathes, and they built turret lathes from then on^[1][11]. The company also built the mounting for the Lick Observatory's 36-inch telescope and made its name as a builder of precision machinery. Roe writes that Warner & Swasey was the first in the Midwest to do work comparable with the best in the East^[1]. Cleveland also gave birth to the multi-spindle automatic. Hakewessel and Henn, both from Pratt & Whitney, built a five-spindle automatic in a Hartford attic in 1895 to make bicycle parts, and in 1902 moved to Cleveland, becoming National-Acme^[1].

Cincinnati: a town of specialists

In Cincinnati, William Lodge, born in Leeds, England, started a shop with William Barker on January 1, 1880, with $1,000. Within ten years its assets had reached $400,000, and one skilled worker after another in the city set up on his own. When Charles Davis replaced his partner in 1886 and the firm became Lodge & Davis, Lodge concentrated on lathes and ordered shapers, tool rests and the like from shops just started by former colleagues. This fostered small specialist makers that did not compete with one another^[1]. Lodge parted from Davis in 1892, and a year later formed Lodge & Shipley with Murray Shipley^[1]. In 1889 he had won his first export order, from Alfred Herbert of Coventry, England^[1].

Around the same time, Frederick Holz began making screws and taps in his kitchen, and built a milling machine on a wooden base to mill the flutes of his taps. A customer who saw it ordered a milling machine, and in 1889 The Cincinnati Milling Machine Company was born^[1]. In the 1900 census, Cincinnati had 30 machine tool plants with output of just over $3.37 million. Seven built only lathes, two only planers, two only milling machines, five only drilling machines and three only shapers^[2]. Concentrating on one type of machine and building it in volume was itself an application of the interchangeable system.

Jones & Lamson and Hartness; Landis

Jones & Lamson, descended from the Robbins & Lawrence of Chapter 1, moved to Springfield, Vermont, in 1889 and brought in James Hartness. Hartness narrowed the sprawling company down to turret lathes and devised the flat turret lathe, whose turret was low and flat^[1][2]. The patent was filed in January 1891 and granted in August (U.S. Patent 457,967)^[14]. Out of the company came Fellows with the gear shaper (1896), Gridley with the multi-spindle automatic and Bryant with the chucking grinder, and small Springfield became a machine tool town^[1]. In grinding, the Landis brothers of Pennsylvania founded a company in 1890 and, while using a small Brown & Sharpe grinder, designed grinders of their own^[1]. In 1897, after the plant burned down, the company was reorganized as the Landis Tool Company with capital of $100,000^[18]. As for the vertical lathe (vertical boring mill), E. P. Bullard designed the first machine in 1883^[1] (Chapter 10).

The bicycle changed machine tools. Rosenberg writes that the bicycle, and later the automobile, played a major part in the improvement of machines from 1880 on. The bicycle demanded lightness, precise hardened parts, efficient power transmission and low friction. In meeting those demands, techniques such as ball bearing grinding, thin-wall tube work and pressing were refined and became usable in other industries^[2]. The Acme multi-spindle automatic, too, was originally a machine for making bicycle parts^[1].

High-speed steel, heavy grinding and electric motors (1892–1914)

Taylor and high-speed steel

Before he became known for time study, F. W. Taylor was a researcher of cutting tools. In 1880 he began metal-cutting experiments at the Midvale Steel Company in Philadelphia (whose president was William Sellers), spending 26 years and more than $200,000^[1][11]. From 1898, working with Maunsel White at Bethlehem Steel, he found a method of increasing the tungsten and chromium and hardening at a far higher temperature than before. This was high-speed steel (the Taylor-White process). At the 1900 Paris Exposition he hurriedly brought in a lathe, its driving motor and mild steel stock, and showed it cutting on even with the tool tip glowing red, at a depth of cut of 3/16 inch (about 4.8 mm), a feed of 1/16 inch (about 1.6 mm) and 120 feet (about 37 m) per minute^[11].

The trouble was on the machine side. Rosenberg writes that although high-speed steel made faster cutting possible, machines designed for carbon-steel tools could withstand neither the forces nor the speeds, and their structure, drives and controls all had to be redesigned. In the first decade of the 20th century, lathe beds and carriages grew heavier, feed mechanisms stronger, and cone pulleys were made for wider belts. The legs of large lathes grew shorter until they disappeared and the bed came down to the floor^[2]. Roe likewise sums up that high-speed steel raised cutting speeds 2 to 2.5 times and that machine tools were largely redesigned over those 15 years^[1]. Cincinnati Milling Machine's "High Power" milling machine (1907–08), built for heavy cutting, is a leading example^[3]. In 1906 Taylor presented "On the Art of Cutting Metals" at ASME, setting out the relationships among cutting speed, feed, tool shape, cooling and tool life^[2].

Driving each machine with its own motor

Until then, factories turned long overhead line shafts with steam engines or water wheels and distributed power to each machine by belts. Driving machine tools with electric motors began around 1892, and Taylor used motors in his experiments from 1894. In 1895 the Baldwin Locomotive Works in Philadelphia drove its wheel-turning lathes with individual motors, removed the overhead shafting and installed cranes for moving heavy work^[11]. Being able to arrange machines by the flow of work rather than by the belts paved the way for the flow-production factories of the next era (Chapter 3).

Norton and heavy grinding

For a long time grinding was the "finish after the finish," a light polish on a surface already finished on the lathe. Charles H. Norton, in charge of grinders at Brown & Sharpe, reasoned that by fitting a wide, large wheel to a heavy, rigid machine, plunging the wheel in without traversing the work (plunge grinding), and letting the machine's accuracy set the size, the finishing cut on the lathe could be skipped and the part finished in one go by grinding. When he proposed this at a company meeting in 1897 he was laughed at, and Norton left the company^[8][11]. In 1900 the Norton Grinding Company was formed with investment from the wheel maker Norton Emery Wheel, and its first machine was an 18 × 96-inch cylindrical grinder with a wheel 24 inches in diameter and 2 inches wide^[1][8]. In 1903 it finished in 15 minutes an automobile crankshaft that had taken 5 hours of turning, filing and polishing, and in 1905 Norton and Landis began supplying grinders to automobile companies^[11][8].

Heavy grinding was made possible by artificial abrasives. In 1891 E. G. Acheson discovered silicon carbide (Carborundum), and in 1897 C. B. Jacobs developed artificial alumina. Norton produced the alumina as "Alundum" at Niagara Falls from 1901, and by around 1906 it had replaced natural emery and corundum^[8]. Note that F. B. Norton, who founded the grinding wheel maker Norton Company, and Charles H. Norton of the grinding machines were not related^[1]. How grinding machines evolved is covered in Chapter 8.

The picture in 1914

In the 1914 census the U.S. had 409 machine tool plants with output of just over $31.44 million. The same year American Machinist published a map of 570 plants, including tool makers: 117 in Ohio, 98 in Massachusetts and 66 in Connecticut. Thirty years earlier there had been almost none west of Buffalo; now more than half lay to the west^[2][1]. That year Roe judged that the basic forms of machine tools were now settled, and that the coming era would be one of higher speeds and heavier cuts, clever fixtures and automatic machines that ran without skilled workers. And orders from the war in Europe were pushing American machine tool makers to first place in the world^[1].

What it means on the shop floor. What was settled in this era is the foundation of today's shops: the 60-degree thread, calipers and micrometers that read to a thousandth, formed cutters that keep their shape when sharpened, the dividing head and helical milling, cam-driven automatics, plunge grinding, artificial abrasives, and the rule that "when the tool material changes, the machine's rigidity and horsepower must be redesigned." The same thing happened again later when carbide and CBN arrived.

⛩Links with Japan

From the late Meiji era into the Taisho era (roughly 1900–1926), the American machine tools of this period were the models for Japanese machine tool building. Below are events confirmed in the Japanese Archive chapters, set alongside the American companies involved.

• 1897: gear cutting on a Cincinnati universal milling machine: The first machine-cut gear teeth in Japan are said to have been cut one space at a time on a No. 2 universal milling machine from Cincinnati (U.S.) with a formed cutter. The universal milling machine and the formed cutter were both born in this chapter's era (see Japanese Archive, Chapter 4)^[23].
• 1903–1905: Francis and Ikegai's American-style lathes: In 1903 the Tokyo Higher Technical School invited the American Charles A. Francis as a practical instructor, and in 1905 Ikegai began volume production of American-style lathes built to his specifications. Until then Japanese lathes had mostly been English-style (see Japanese Archive, Chapter 1)^[21].
• 1906: the American lathes of Kumamoto Higher Technical School: The American-built lathes and other machine tools preserved at Kumamoto University are still in working order and are designated an Important Cultural Property of Japan and Mechanical Engineering Heritage^[22].
• 1916–1923: Okuma and Brown & Sharpe, Pratt & Whitney and Norton: Okuma used a lead screw bought from Brown & Sharpe in its 1916 KE lathe, and in 1919 built the OP lathe from drawings of a Pratt & Whitney lathe supplied by an army arsenal. In grinders, it built the GA-1 in 1918, modeled on Norton's universal tool grinder, and the GA-2 around 1923, following Brown & Sharpe's No.13 (see Japanese Archive, Chapter 3)^[24].

For Japanese shops, America's precision tools and the machines of its specialist makers were things "to buy and use" and, at the same time, "to copy and learn from." How Japanese milling machines of the Taisho and early Showa eras took the Cincinnati type as their model is covered in Japanese Archive, Chapter 2.

FAQ

What is the difference between the Sellers thread and the Whitworth thread?

The Sellers thread (1864, later the U.S. Standard thread) has a 60-degree thread angle, with the crests and roots cut off flat. Britain's Whitworth thread has a 55-degree angle, with rounded crests and roots. Sellers chose 60 degrees because flat crests and a 60-degree angle make cutters easier to make and threads easier to inspect. The 60-degree form was carried on into the later Unified and metric threads.

What made Brown & Sharpe's universal milling machine "universal"?

Gearing the dividing head to the table feed and making the table swivel allowed "helical milling," as for the spiral flutes of twist drills. The same machine could also cut gear teeth and irregular shapes. The first machine was sold in March 1862, and its design, with the table carried on a knee that moves up and down, became the basis of later knee-type milling machines.

What changed in machine tools when high-speed steel arrived?

Tools could now cut even at temperatures that made them glow red, and cutting speeds rose 2 to 2.5 times. Machines therefore needed much more force and speed: beds and carriages grew heavier, feed mechanisms stronger and belts wider, and driving each machine with its own motor spread. The rule that a new tool material means redesigning the machine began here.

Are Norton Company and Norton Grinding Company the same company?

No, they are different companies. Norton Company (originally the Norton Emery Wheel Company) was the grinding wheel company started by F. B. Norton, and the Norton Grinding Company was founded in 1900 by Charles H. Norton to build grinding machines. Roe writes that the two Nortons were not related.