history of screw threads

home > tools > fasteners

Modern machining was born in the industrial flowering in late 18th century Great Britain. The modern lathe, capable of cutting threads with great precision, was invented in 1797 by Henry Maudsley. Even today, for most purposes there is no need for any greater precision than that achieved by Maudslay. Creating threaded fasteners became much easier, but everyone made them to his own pattern. If you lost a nut from a machine, and the shop that made it was out of business, a new nut would have to be custom made to match the existing bolt. (Read an appreciation from a great contemporary.)

Maudslay took on an apprentice, Joseph Whitworth, who proved exceptionally talented. While he was with Maudslay, Whitworth invented the method for producing a true plane surface in steel, one of the fundamental operations in precision machining. He next worked at Joseph Clements, where they were trying to build Babbage's calculating engine, the first computer, and finally set up shop for himself as a toolmaker. By 1859 he had produced a machine capable of measurements to one two-millionth of an inch.

Whitworth set himself the task of devising a standard for threads. He had his own ideas about what would work best, but being a pragmatist he also collected bolts from all over England, noting which sizes experience had shown to be most useful, and the results of various thread forms. In 1841 he proposed as a standard a thread form with an included angle of 55°, and the tops and bottoms of the threads rounded with a radius equal to 0.1373 times the pitch. 

outline of form of Whitworth thread

In 1857 experience with the first proposal led Whitworth to greatly expand the original table.

Due in part to the immense prestige Whitworth gained from the display of his machines at the Crystal Palace Exhibition of 1851, Whitworth's system was in general use in Great Britain by 1860.   Later a second series with finer threads (BSF) was added. (For current values, see table.)

Sir Joseph Whitworth.
An uniform system of screw threads.
Minutes of Proceedings of the Institution of Civil Engineers, 1841, i, 157.

Engineering and Architecture Journal, 1857, page 262; 1858, page 48.

Americans experienced the same problems from lack of thread standardization that Britain did. The challenge was taken up by William Sellers, scion of an eminent family of American “mechanicians,” whose grandfather had made the plates with which the Continental Congress printed its currency. To William himself, among other things, we owe the color “machinery gray.” When others were decorating their machinery, he insisted on painting his a uniform light gray, in order not to obscure the functions of the parts. Sellers specified a thread form and a graded series of nuts and bolts that used it.

William Sellers.
A system of screw threads and nuts.
Journal of the Franklin Institute, volume 47, page 344 (May 1864).

See the same journal, volume 49, page 53 (1865), for the report of the committee, recommending the adoption of Seller's system.

In 1864, a committee of the Franklin Institute recommended the adoption of Seller’s system of screw threads. The thread form became known as the “Franklin thread,” or, more commonly “Seller's thread,” and later as the “United States Standard Thread.” It became the basis of the French standard thread, and then of the Système International thread. In May 1924 it was designated the “American Standard Thread.” 

outline of form of Seller's screw thread

The main difference between Seller's thread form and Whitworth's is that the tops and bottoms of the threads (the crests and roots) are flattened. The flattened roots was a bad choice. Such angular joins in metal concentrate stress, and the process of manufacture results in high stresses at the roots of threads anyway. The result is cracks and broken fasteners.

This problem was not so noticeable in Seller’s day for two reasons. One was that most machinery was stationary and the weight of a bolt rarely mattered. If a bolt broke it could be replaced with a larger one. The second reason was that thread roots tend to be rounded anyway as the tools that make the bolts become worn.

With airplanes, “just put in a bigger bolt” was not a satisfactory solution, and aerospace engineers finally introduced an American thread form (UNJ) with rounded roots. For example, by changing to this thread form an American car manufacturer finally solved a persistent problem with connecting rod breakage.

Manufacturers adopted Seller's thread form but rejected other parts of his system, such as the formulas for the size of square and hexagonal nuts and bolt heads, and they chose to use a different number of threads per inch for the 1116 inch and 1516 inch bolts.

National societies get into act

In 1907 the American Society of Mechanical Engineers (ASME) defined two series that used Seller's thread, numbering the sizes by gage numbers from 0 to 30. In the series the major diameter increased by 0.013 inch with each size. The obsolete ASME gages are described in this table. Yet another ASME “special” series used the same major diameters, but assigned different thread frequencies.

home | tools index | search |  to email Sizes drawing of envelope |  acknowledgements | 
help | privacy | terms of use