UK 01The Birth of the Machine Tool: Maudslay and His Pupils

The steam engine and the boring machine (1769–1800)

The cylinder that troubled Watt

When James Watt began improving the steam engine, there were hardly any tools that could cut metal accurately. Machine frames were made of wood, and metal parts were castings or forgings fitted with a file. Lathes, too, were commonly ones in which a cord worked by a treadle turned the work back and forth while the tool was held by hand. According to Roe, Watt complained that the long and short diameters of an 18-inch (about 46 cm) cylinder differed by as much as 3/8 inch (about 9.5 mm), and tried to fill the gap around the piston with cork, oiled rags and old hats^[1].

The boring machines of the day had a cutter head on the end of a slender bar that was pushed into the cast bore, supported at one end only. The machine John Smeaton built at the Carron ironworks in 1769 was of this type, and the Science Museum has a model of it^[1][2]. Because the bar was slender and the cutter was not guided, the cutter simply followed the curve of the cast bore. However carefully the work was done, it could do no more than "tidy up the shape of the original bore a little."

Wilkinson's boring bar: supported at both ends

The ironmaster John Wilkinson found the answer while making cannon. His patent of January 27, 1774 (No.1063) was a method of casting cannon solid, turning them by waterwheel or steam engine, and boring the hole with a cutter on a boring bar^[3]. For cylinders, he passed a thick boring bar through the cylinder and supported it in bearings at both ends, and fed the cutter head along the bar. Since the bar itself acted as a straightedge, however crooked the casting, as long as there was enough stock the bore came out straight and truly round^[1][2].

In 1775 Wilkinson bored the first cylinder for Boulton & Watt at his Bersham ironworks, and for many years afterward he supplied the cylinders for Watt's engines^[1][2]. In 1776 Boulton wrote that a 50-inch (about 1.27 m) cylinder bored by Wilkinson was nowhere out of true by more than the thickness of an old shilling^[1]. The Science Museum catalog calls it the first truly large power-driven machine tool, the machine that made possible Watt's rotative engine and, through it, the power-driven factory^[2]. An actual boring bar about 4.5 m long, said to have been installed at Bersham, is also in the museum's collection^[2].

How it connects to machining today. "Let the path of the cutter be set by an accurate element of the machine (here, the boring bar), not by the workpiece." This is the most important point of this machine. When boring long holes on today's horizontal boring machines, supporting the boring bar at both ends or guiding it with a guide bushing follows the same logic. Conversely, when a long overhung boring bar chatters or the hole comes out crooked, it is because it has the same weakness as Smeaton's machine.

Factories build their own machines

As demand for steam engines grew, the works that built them began equipping themselves with their own machine tools. In 1798, before Watt's patent expired, a horizontal cylinder boring machine of Murdock's design at Boulton & Watt's Soho Foundry bored a 64-inch cylinder, and a list of the machinery in 1801 shows 10 lathes, 3 drilling machines and more^[2]. The machine tools of this period were not for sale; they were first of all made by a works for its own work. They began to be sold to others in the era of Maudslay's pupils, which we look at next.

Bramah, Maudslay and the Portsmouth block mills (1784–1831)

Machine tools for locks

Joseph Bramah of London patented a lock in 1784. It was said for a long time that no one could open it, but it worked only if its precision parts were made to consistent dimensions. Even skilled craftsmen could not make many of the same quality by hand^[4][2]. Bramah decided to install special machines for making the locks in his workshop.

The man he brought in was Henry Maudslay, 18 years old, who had been working in the smithy of the Woolwich Arsenal. He joined Bramah's workshop in 1789, was accepted among the journeymen after restoring a worn vise to as-new condition in a single day, and became foreman within about a year^[4][1]. According to contemporary testimony quoted in the Science Museum catalog, the machines for making lock parts in Bramah's secret workshop were made by Maudslay's own hands^[2]. Bramah's hydraulic press, patented in 1795, also would not work because the packing around the ram leaked, until Maudslay devised a leather cup that seals tighter the higher the pressure and made it practical^[4][1].

The screw-cutting lathe: lead screw and change gears

In 1797, refused a raise on his wage of 30 shillings a week, Maudslay left Bramah's workshop and opened a small workshop of his own with one workman^[1][2]. What he worked on there was a lathe for cutting accurate screws.

The idea of fixing the tool not in the hand but on a rest moved by screws (the toolholder, or slide rest) already existed in the world of clockmakers and ornamental lathes. The idea of linking the spindle and the lead screw by gears also appears in old French books and in Bentham's 1793 patent^[1][2]. What Maudslay did was to build it all in metal, solidly, and put it into a form usable for everyday work. His first lathe, of about 1797, had a bed of two triangular iron bars, and gears linking the spindle and the lead screw fed the toolholder. At first he changed the lead screw for each pitch^[1][2].

His second machine, of about 1800, had a single lead screw of 30 threads per inch on a cast-iron bed, and by rearranging 28 change gears of 15 to 50 teeth it cut threads from 16 to 100 per inch. An intermediate gear was mounted on a swinging arm so that it could mesh with gears of different sizes^[1][2]. Cutting many pitches with one accurate lead screw and change gears. The basic form of thread cutting on today's lathes is here. To make the first accurate lead screw, Maudslay is said to have used the fact that a tilted blade held against a soft metal cylinder advances by itself as the cylinder turns, making hundreds of screws and comparing them^[1]. The slide rest was at first mocked as "Maudslay's go-cart," but in time it found its way into every works^[4].

What changed on the shop floor. Once the toolholder held the tool and a screw fed it, cutting accuracy shifted from "the steadiness of the craftsman's hand" to "the accuracy of the machine's guideways and feed screw." The order of things, in which good machines are needed to make good machines, also begins here. When you set the gearbox levers to cut a thread on a lathe, you are using this rearranging of change gears, now enclosed in a box.

Standardizing screws

Maudslay also improved taps and dies so that they "cut" the thread instead of squeezing it. He standardized diameters in steps of 1/8 and 1/16 inch, and fixed the pitch for each diameter within his works^[1][2]. According to his pupil Nasmyth, before then every bolt and nut was a matched pair, and they would not fit unless each pair was marked^[1]. The dimensions fixed within this works spread to other works through his pupils and became the foundation of Whitworth's later threads.

The Portsmouth block mills

A sailing ship uses hundreds or thousands of pulley blocks. According to Roe, the navy bought about 100,000 blocks a year^[1]. Samuel Bentham of the navy, troubled by the shortage of skilled workers, had been planning since the 1790s a factory at Portsmouth Dockyard to make blocks with woodworking machines^[1]. Then Marc Isambard Brunel, an engineer who had fled the French Revolution to America and come to Britain in 1799, brought in a plan for making blocks by machine. Bentham acknowledged that Brunel's plan was better than his own and adopted it^[1].

The search for a craftsman who could give shape to Brunel's drawings led to Maudslay, who had only just set up in business. Model-making began in 1800, and the navy approved it in 1801^[4]. Sawing machines, boring machines, mortising machines, machines for shaping the outside, machines for cutting the rope groove, lathes for the sheaves and more: 44 machines in all. The frames were cast iron and the fast-moving parts hardened steel, and Smiles writes that Maudslay found his own slide rest indispensable for machining the parts^[4][1]. The Science Museum catalog dates their construction to 1802–1807 and calls the mills the first large-scale mass-production factory using machine tools^[2].

In 1808, when the mills were in full operation, 10 unskilled men did the work previously done by 110 skilled men, producing more than 130,000 blocks a year (Smiles says 160,000)^[1][4]. The work was divided into operations, a special machine was placed at each, and the material was gripped at fixed reference positions. Roe points out that although these mills had already achieved the "interchangeable system," they had almost no influence on other British industries for two generations, and writes of the irony that in 1855 Britain bought the Enfield machines from America and adopted them under the name of the "American System"^[1]. That story continues in Chapter 2.

The Lambeth works and the "Lord Chancellor"

Having made his name with the block machinery, Maudslay moved his works in 1810 to Lambeth, south of the Thames^[4]. Joshua Field, who had been a draftsman at Portsmouth, joined him, and Maudslay & Field (later Maudslay, Sons & Field) became a works employing hundreds of people, building machine tools, flour mills, minting machinery, and land and marine steam engines^[1][2].

Maudslay was a stickler for accuracy. Beside his workbench was a bench measuring instrument that he called the "Lord Chancellor." The object was held between two hardened, flat steel faces, one face was advanced by a fine screw, and it was stopped at the lightest touch. The graduations were 1/10 inch on the scale and 100 divisions on the screw head. Nasmyth writes that it could detect a difference of 1/1,000 inch^[5][1]. The Science Museum catalog says it was accurate to 0.0001 inch^[2]. It is one of the ancestors of today's micrometer.

The same went for flat surfaces. According to Nasmyth, Maudslay's works made surface plates in sets of three, rubbed them against one another to remove the high spots, finally took off the small bearing spots with a hard steel scraper, and placed them on the workmen's benches. With only two plates, one convex and one concave can still match, but if all three match one another, all of them are flat^[5][2]. Checking their own work on these plates taught the workmen "what truly flat means." After Maudslay died in 1831 his sons and Field carried on the works, and his pupils took Maudslay's standards all over Britain^[1].

The pupils and the planer (1813–1844)

Roe sets out the early British machine tool builders in a single family tree. Maudslay and Clement came out of Bramah's workshop, and Roberts, Nasmyth, Whitworth, Muir and others came out of Maudslay's works: a chain of skilled men who worked in a good works, then set up on their own and founded the next one^[1][2]. Within this chain was born the planer, the most important machine after the lathe.

Who invented the planer?

Until the beginning of the 19th century, making a flat surface was hard work: roughing with a chisel and finishing with a file. According to Whitworth's 1856 address, 30 years earlier the labor cost of finishing a cast-iron surface true by hand had been 12 shillings per square foot, and with the planer it fell to less than a penny^[2][4].

No single inventor of the planer can be named. Workmen later testified that Matthew Murray of Leeds was using a machine to plane the flat faces of steam engine slide valves by 1814. He did not patent it, and kept it hidden in a locked room^[1][4]. James Fox of Derby is also said to have built a planer with automatic reversing around 1814. Fox issued a price list of planers in five sizes and lathes in seven sizes in 1829, and in 1826 supplied a complete set of machine tools to a Polish works. This was the business of selling machine tools, in one of its early forms^[1][2].

The oldest surviving planer is that of Richard Roberts, who after working in Maudslay's works set up on his own in Manchester in 1817. The table, 52 inches long and 11 inches wide, was moved back and forth by turning a chain by hand; the tool could be fed vertically and crosswise and was hinged so that it lifted clear on the return. The guideways of the machine itself still show the marks of chisel and file. In an age without planers, he made a planer by hand^[1][2].

Joseph Clement of London built a planer in 1820 to machine the triangular bed bars of lathes, and in 1825 completed a large planer that could take work up to 6 feet square. The table ran on rollers, and it had tools that cut on both the forward and return strokes. Clement claimed such accuracy that "a single sheet of paper slipped under one roller would stop the whole," no machine of the same size existed for about ten years, and contract planing at 18 shillings per square foot became Clement's main income^[4][2][1].

What the planer changed. The accuracy of a machine tool is determined by the flatness and straightness of its guideways. With the planer, the guideways themselves could be machined. The Science Museum catalog mentions the view that beam engines stayed in use so long, in the days when there were no tools for making flat surfaces, because their form had few flat surfaces^[2]. Even the form of machines was bound by the ability to make flat surfaces.

Roberts' back gear and the slotting machine

In the same year, 1817, Roberts also built a lathe with back gear. This mechanism reduced the speed of the belt cone pulley through gears, greatly widening the range of spindle speeds, so that thick or hard work could be cut slowly and powerfully. The long lead screw was used not for threading but for the ordinary feed of the toolholder, and the feed disengaged automatically at a set position. This lathe was used in Roberts' works until 1854 and is still at the Science Museum^[2][1]. Roberts also devised a slotting machine for cutting keyways in gears, and Holtzapffel wrote that no one could match Roberts in improvements to drilling machines either^[1][2]. In 1825, at the request of cotton spinners, he built the self-acting mule, and Sharp, Roberts & Co. went on to build locomotives. Those locomotives were built with parts standardized by templates and gauges, making it one of the first works in Britain to adopt the interchangeable system^[1].

Clement's taps and Babbage's calculating engine

In 1817 Clement opened a workshop at Newington Butts, London, with £500 of capital. He employed at most about 30 men, but all of them were highly skilled^[4][1]. From 1828 he mass-produced taps and dies based on Maudslay's screw dimensions, cutting the flutes of the taps with rotary milling cutters and making the shanks slender so that they would pass down through the threaded hole^[1]. It was also Clement's workshop that made the precision parts of Charles Babbage's calculating machine (the Difference Engine), and among the craftsmen who worked there was Joseph Whitworth^[1][4].

To Manchester

When the Liverpool & Manchester Railway opened in 1830, orders for locomotives and the machine tools to build them gathered in the north. William Fairbairn rented a shed in Manchester in 1817 and replaced thick wooden shafts with slender wrought-iron ones, raising the speed of mill gearing two or threefold. His brother Peter Fairbairn took over the Wellington Foundry in Leeds in 1828 and built large machine tools for boring, planing and slotting^[1][4]. "Diametral pitch," which expresses the size of gear teeth in terms of diameter, also began at the Manchester works of the Swiss-born Bodmer and was called "Manchester pitch." In 1837 Willis of Cambridge gave it its name and set out the idea that gears of the same diametral pitch would all mesh with one another^[1] (gears are covered in Chapter 8).

Manchester and Whitworth: standards of flatness and threads (1833–1857)

Nasmyth's shaper and steam hammer

In 1829 the 20-year-old James Nasmyth of Edinburgh called on Maudslay with a model engine and drawings he had made, and was taken on as assistant in Maudslay's private workshop. While there he devised a milling machine for indexing and cutting the flats of hexagon nuts^[1][2]. After Maudslay's death Nasmyth built his own machine tools in Edinburgh, and according to his own chronology he started in business in Manchester in 1834, built the Bridgewater Foundry at Patricroft on the outskirts in 1836, and set it fully to work in August^[5]. It was a site between the new railway and the Bridgewater Canal, convenient for bringing in materials and shipping out products^[1]. Of the 54 machine tools illustrated in Buchanan's 1841 book, 21 were Nasmyth's^[2].

Devised by Nasmyth in 1836, the shaper moved a tool slide back and forth by a crank to cut small flat surfaces and keyways, and was called "Nasmyth's Steam Arm." Small flat surfaces on work not worth putting on a large planer could now be machined easily. Whitworth later added a quick-return mechanism to speed up the return stroke^[2][1].

On November 24, 1839, asked how to forge the 30-inch (about 76 cm) diameter paddle shaft for the steamship Great Britain, Nasmyth sketched a steam hammer in his notebook in about 30 minutes: a heavy iron block lifted straight up by a steam cylinder and dropped along guides. Earlier lever-type hammers could not be raised high enough when a large piece was placed under them^[5][1]. In the end the ship was changed to screw propulsion and the paddle shaft was never made; then Nasmyth, seeing at the Le Creusot works in France a steam hammer built from the sketch in his own notebook, hurriedly took out a patent in June 1842^[1][5]. In 1843 his partner Wilson devised a self-acting valve gear, so that the hammer could strike continuously with uniform force and frequency^[2]. The catalog of the 1851 exhibition marveled that its force could be adjusted down to just enough to crack an eggshell^[2].

Whitworth: flat surfaces

Joseph Whitworth, born in Stockport, joined Maudslay's works in 1825 and worked next to the skilled John Hampson. There he worked on finishing three plane surfaces against one another; then, after working at Holtzapffel's and Clement's workshops, in 1833 he put up a sign in Manchester reading "tool maker from London"^[1][2].

In the words of his 1856 address, Whitworth's thinking rested on true plane surfaces and the power of measurement alike^[1]. In an 1840 paper in Glasgow, Whitworth rejected finishing plane surfaces by grinding them together with abrasive. Abrasive collects at the edges and rounds them, and the workman cannot control where and how much is removed. Scraping point by point while checking the bearing marks was, he argued, the only way to produce an accurate plane with bearing points spread fine and even^[6]. He also described experiments showing that if planes matched well, an upper plate would glide as if floating on a layer of air, and dry surfaces would cling together under atmospheric pressure and could not be pulled apart^[6]. Roe considers that matching three plates was itself an old technique already found in Maudslay's works, and that what Whitworth added was finishing entirely by scraping^[1].

The same logic as today's scraping. Scraping is still used to finish the guideways of machine tools and surface plates for the same reasons as in this 1840 argument. Spreading the bearing points fine and even creates oil pockets for smooth movement and avoids wear from local contact. Some works deliberately finish ground surfaces by scraping, which extends the same logic.

Whitworth: threads

In 1841, at the Institution of Civil Engineers, Whitworth presented "a uniform system of screw threads." It was a proposal to end, by standardizing threads, the waste caused by every works having its own threads, so that thread-cutting tools had to be gathered anew for every repair^[6]. His method was practical. He collected bolts from the main works across Britain, averaged the number of threads for each diameter, fixed some diameters as references to determine the sizes in between, and extended the system to 6 inches. For the thread angle, the average on 1-inch screws was about 55 degrees and much the same on other diameters, so he made them all 55 degrees. The crests and roots were rounded by one-sixth of the depth each, so that screws, taps and dies would be less likely to chip^[6][1].

Not seeking a theoretical optimum, but taking the average of what everyone was using and having everyone conform to it. Whitworth writes that since the form of a thread cannot be settled from principles, it can only be settled by compromise^[6]. The paper says that the system had already been adopted by many railways and large works and introduced at Woolwich Arsenal, and a note in the 1858 collected papers says that "since 1841 this system has been widely adopted"^[6]. Roe says that by 1860 it was in use all over Britain^[1]. This became the Whitworth thread. Its later history as a standard is covered in Chapter 2 (the British Standard Whitworth of 1905).

1851: the peak

Whitworth's machine tools stood out less for novelty than for accuracy and quality of construction. A hollow box-section bed was far stiffer for the same weight than a bed of triangular bars, protected the lead screw from dirt, and engaged it with a split nut^[2]. His planer, patented in 1835, was power-driven with automatic feed and cut on both the forward and return strokes^[2]. At the 1851 Great Exhibition in London he exhibited 23 items, including lathes, planers, shapers, slotting machines, drilling machines, screw-cutting and gear-cutting machines, a measuring machine and gauges, while the other 20 firms showed one to three each^[2]. Roe calls Whitworth in 1850 the world's leading machine tool builder^[1].

Roe goes on, however, to suggest that the very dominance of Whitworth's methods over British machine tools for generations may have been one reason Britain lost the lead, noting that most of the new developments of the next 50 years took place in the US^[1]. At the same 1851 exhibition were America's interchangeable rifles, and the American Hobbs, who opened Bramah's lock in 16 days^[1][4]. The age in which Britain gave birth to the machine tool reached its peak here, and moved into the "age of being overtaken" of the next chapter. In 1857 Nasmyth gave a working model of the steam hammer to the new museum at South Kensington, and this was the beginning of the Science Museum's machine tool collection^[2].

⛩Links with Japan

In the period of this chapter, Japan had no machine tools yet. But what the factories of Meiji Japan took as their first model were the lathes that had taken shape in Britain, from Maudslay to Whitworth. The Ikegai works, opened in Tokyo by Ikegai Shotaro in 1889, had "English-type" lathes, and the English-type 9-foot lathe modeled on them is the oldest surviving powered lathe made in Japan (Mechanical Engineering Heritage No. 53, National Museum of Nature and Science)^[11].

According to the Mechanical Engineering Heritage description, this lathe has six speeds from a three-step belt pulley and back gear, and a gap in the bed so that it can take large work^[11]. Back gear is the mechanism Roberts fitted to his lathe in 1817, as we saw in this chapter, and 70 years later in Tokyo it was handed down as a standard feature of the British-style lathe^[2][11]. The switch in Meiji Japan from British to American styles, and the story of Henry Dyer, the British teacher at the Imperial College of Engineering, are covered in Chapter 2, and the development of Japanese lathes in Japanese Archive, Chapter 1, "The Lineage of Japanese Lathes".

The same goes for threads. Whitworth's 55-degree thread is known in Japan as the "Whit thread." For today's thread standards and basics, see the Threads page.

FAQ

Did Maudslay invent the screw-cutting lathe?

The slide rest and the idea of linking the spindle and lead screw by gears already appeared in earlier French books, clockmakers' machines and Bentham's 1793 patent, and David Wilkinson in the US also took out a patent in 1798. Maudslay's achievement was to build them solidly, all in metal, in a form usable for everyday work that could cut many pitches with one lead screw and change gears, and to spread it. Two machines, of about 1797 and about 1800, survive at the Science Museum.

Who built the first planer?

No single person can be named. Murray of Leeds and Fox of Derby are said to have been using planers around 1814, and Clement built them in 1820 and 1825. The oldest surviving one is the planer built by Roberts of Manchester in 1817, in the Science Museum's collection.

How was the 55 degrees of the Whitworth thread decided?

In 1841 Whitworth measured the thread angles of bolts collected from works all over Britain; the average on 1-inch screws was about 55 degrees and much the same on other diameters, so he made them all 55 degrees. The pitches, too, were set from the average of the threads actually used in works. Rather than seeking a theoretical optimum, it was a standard that chose values everyone could conform to.

Did the Portsmouth block mills come before the American System?

As a mass-production factory that divided the work into operations, lined up special machines and replaced the handwork of skilled men, the Portsmouth block mills of 1802–08 are an early example. But as Roe points out, their methods hardly spread to other British industries, and in 1855 Britain ended up buying the machines for the Enfield small arms factory from the US and adopting the "American System."