Why this story matters
The bolted joint is the most common engineered connection on Earth. Refineries, offshore platforms, pipelines, reactor heads, ships and power plants are all held together by threaded fasteners under controlled clamp load. It's one of the unsung technologies of the modern world.
The story of how we got here spans twenty-five centuries. It runs through Mesopotamian engineers, a Greek mathematician in Sicily, an Italian polymath, a London machinist, a Philadelphia toolmaker, a New York City Water Department employee tired of leaky mains, and a German engineer who finally explained why bolts come loose under vibration.
Every era of bolting history has been about closing the gap between what we wanted on the joint and what we actually delivered.
The first screws (c. 700 BC)
The screw was one of the last of the six classical simple machines to be invented. The lever, the wheel and axle, the pulley, the inclined plane and the wedge all came earlier. Once it arrived, it changed everything.
The earliest evidence for screw mechanisms comes from the Neo-Assyrian period in Mesopotamia, roughly 911 to 609 BC. From there the idea spread to Egypt and Greece, where it was first described by the mathematician Archytas of Tarentum (c. 428–350 BC). By the 1st century BC, wooden screws were common across the Mediterranean in presses for olive oil and wine.
Two caveats for anyone telling this story honestly. First, the screws of antiquity were not fasteners. They were lifting devices, presses and water-raising machines that used the helix to do work, not to hold things together. Threaded fasteners in meaningful numbers were nearly two thousand years away. Second, these screws were almost all wood. Metal threading needed a precision that wouldn't exist until the Renaissance.
Archimedes and the water screw (c. 250 BC)
The most famous screw of antiquity bears the name of a man who probably didn't invent it. Archimedes of Syracuse (287–212 BC) is traditionally credited with the water screw: a helix inside an inclined cylinder that lifts water when turned.
Modern scholarship is more cautious. Historians including John Peter Oleson argue that Archimedes likely encountered an existing device in Egypt, already used for irrigation in the Nile delta, and some evidence points to Assyrian use centuries earlier. His contribution may have been the mathematics of the helix and the popularity of the machine rather than its invention.
What isn't disputed is that the Archimedes screw became a foundational technology of the ancient world. The Romans used it to drain mines, sometimes cascading several screws, to irrigate and drain farmland, and to bail ships. It's still in use in wastewater plants and drainage stations. For bolting, the Archimedes era established the helix as a reliable engineering primitive: rotation turned into linear motion by a helix wrapped around a cylinder. Every bolt since is built on that idea.
The Wolfegg Housebook (c. 1480)
The screw largely fell out of use in Europe after the 3rd century AD. For a thousand years it hibernated, surviving in occasional presses and manuscripts but not developing.
It resurfaces in the late medieval period. One of the earliest visual records of screws used for fastening is the Medieval Housebook of Wolfegg Castle, compiled around 1480 and illustrated by an artist known only as the Master of the Housebook. Among its drawings of military equipment, mining, hunting and household life are what is often cited as the first known illustration of a screwdriver, and an early screw-cutting lathe with slide rests and a leadscrew.
In the same era, Johannes Gutenberg used screws in his printing presses. From presses, screws spread to clocks, armor and, slowly, general construction.
Leonardo's notebooks (c. 1500)
Around 1500, Leonardo da Vinci sketched a series of screw-cutting machines. One shows a lathe with two leadscrews and interchangeable gears, a single machine able to cut different pitches. Others show taps and dies. There's no evidence any were built. The sketches sat in his notebooks for centuries.
The idea in them was enormous all the same: that making screws could be mechanized, and the work of hand-filing threads handed to a purpose-built machine. The world wouldn't build that in steel for three hundred years.
The Industrial Revolution and Maudslay's lathe (1760–1800)
Through the 17th and most of the 18th centuries, threaded fasteners were hand-made, slow to produce and inconsistent from shop to shop. A nut from one workshop wouldn't fit a bolt from another.
Machines began to change that. In 1760 the English firm of J. and W. Wyatt introduced a factory process for making wood screws. In 1770 the instrument maker Jesse Ramsden built what is generally considered the first satisfactory screw-cutting lathe.
The figure who made screw-cutting the foundation of modern manufacturing was Henry Maudslay. Born in 1771 at the Royal Arsenal in Woolwich, he started work at twelve and by his twenties was building precision locks and hydraulic presses for Joseph Bramah. Around 1800, Maudslay built a screw-cutting lathe whose leadscrew was geared to the spindle, so the cutting tool advanced at a constant, mechanically fixed rate. The threads were accurate, repeatable and fast to produce.
Maudslay's Lambeth workshop trained a generation of British engineers: Joseph Whitworth, James Nasmyth, Joseph Clement and Richard Roberts among them. His lathe made the standardized, mass-produced threaded fastener possible. Steam engines, locomotives, ships and machine tools had their fastener.
Whitworth standardizes threads (1841)
Maudslay's lathe could make precise threads. It couldn't make matching threads across shops. Every manufacturer picked its own angle, pitch and diameters, and a Birmingham bolt wouldn't take a Glasgow nut.
Joseph Whitworth, one of Maudslay's apprentices, solved it. Obsessed with precision, he collected sample screws from shops across Britain to study their variation. In 1841 he presented a uniform system to the Institution of Civil Engineers: a 55° thread angle, rounded crests and roots, and a defined number of threads per inch for each diameter. It became British Standard Whitworth (BSW), the world's first national thread standard. The railways, which needed interchangeable parts, pushed adoption hard.
Whitworth's lasting insight was that interchangeability needed agreement, not just precision.
Sellers and the U.S. standard (1864)
The United States had the same chaos two decades later. William Sellers of Philadelphia, a machine tool builder and president of the Franklin Institute, admired Whitworth's achievement but not its details. The 55° angle was hard for ordinary mechanics to gauge, and the rounded forms were harder to make.
In April 1864, Sellers presented A System of Screw Threads and Nuts to the Franklin Institute: a 60° thread angle, flat crests and roots, and a graded series of sizes. The Institute endorsed it that December. The U.S. Navy adopted it in 1868, and the Master Car Builders' Association recommended it for railroad work in 1871. It became the United States Standard thread.
Two nations now had thread standards close enough to be confused and different enough to fail when mixed. Fixing that would take a world war.
The pre-industrial toolkit (1850–1920)
Thread standards solved manufacturing. They did nothing for assembly. A flange could be made to fit, but nobody could yet tighten it to a known, repeatable, leak-free load. Flange bolting was a craft, done largely by boilermakers with a brutal kit:
- Hammer (slugging) wrenches: a box wrench with a striking face, driven with a sledge. They're still used on wellheads and large flanges where powered tools don't fit, and still the original uncalibrated tool. The load depends on the swing.
- Podger spanners and drift pins: a wrench on one end and a tapered spike on the other, used to drive misaligned holes into line. The podger is the direct ancestor of every modern flange aligning pin.
- Cheater bars: pipe over a wrench handle for leverage. Controlled bolting was invented to replace them.
- Rivets instead of bolts: for the most critical pressure boundaries, such as boilers, hulls and vessels, hot rivets were considered more reliable than the bolts of the day. The move to bolted and welded construction came gradually through the 20th century.
The problems didn't change. The solutions got engineered. The history of bolting is the history of replacing feel with measurement, strength with leverage, and luck with engineering.
The torque wrench (1918–1938)
The first big step in controlled bolting came from a frustrated municipal employee. Conrad Bahr worked for the New York City Water Department, where cast-iron water mains were joined with bolted flanges, and the flanges leaked. Not from bad gaskets or bad bolts, but from inconsistent tightening. One crew cracked a casting, and the next left a slow weep. Nothing on the jobsite measured what a wrench was doing.
Around 1918, Bahr devised a wrench that indicated the torque being applied, so a mechanic could tighten to a value rather than a feeling. The patent record is messier than the popular story. John H. Sharp filed for a beam-type torque measuring wrench in 1931 (granted 1935). Bahr and George Pfefferle patented an adjustable ratcheting wrench with an audible signal at the set torque in 1935. Paul Sturtevant patented and sold beam-type wrenches from 1938.
Bahr is still the right place to anchor the era, because the problem he was solving, consistent clamp load on a flange to stop leaks, is the same one bolting crews work on today. The first torque wrench was invented to seal a flange. So is nearly every one since.
The Unified Thread Standard (1948)
In the Second World War, American and British forces shared equipment on an enormous scale, and their fasteners didn't interchange. A mechanic in North Africa could be sitting on a crate of bolts that didn't fit the truck in front of him.
On November 18, 1948, representatives of the United States, the United Kingdom and Canada signed the agreement establishing the Unified Thread Standard (UTS) for inch fasteners. UTS kept the 60° angle of the Sellers system, added refined tolerance classes and gauging, and defined the UNC and UNF series. It's still the dominant inch thread standard in North America, maintained as ASME B1.1. The 8-thread series on most large flange studs belongs to it.
The hydraulic torque wrench (1960s–1970s)
By the 1960s, petrochemical plants were building reactors, exchangers and vessels at scales that outran the tools. Hand torque wrenches topped out at a few hundred foot-pounds. Impact guns were loud and inaccurate. Mechanical multipliers were slow. Big joints meant several workers on a long bar.
Hydraulic power was the answer. Hydraulically powered torque wrenches first appeared in basic form in the early 1960s. Through the late 1960s and early 1970s, ratcheting designs driven by a remote pump made them practical tools. Several companies founded in that period are still in the business. The principle has never changed: a piston, fed at high pressure, drives a ratchet, and output torque is directly proportional to pump pressure, readable on a gauge.
For the first time, one operator could deliver controlled, repeatable, documented torque on the largest fasteners in a plant. Refinements since have been large: lighter bodies, slimmer nose radii for tight access, low-profile hex cassette tools, swivel couplers, and several tools run from one pump. The core idea is the same.
Junker and self-loosening (1969)
A joint has to be tight at installation and has to stay tight through vibration, thermal cycles and years of service. By the 1960s everyone knew bolts loosened under vibration. Nobody had a rigorous explanation.
The German engineer Gerhard Junker supplied one. In New Criteria for Self-Loosening of Fasteners Under Vibration (SAE paper 690055, 1969), he showed that the most damaging loading isn't vibration along the bolt axis, as intuition suggested. It's transverse loading, perpendicular to the axis, which causes slip in the threads and under the nut and lets the fastener unwind a little with each cycle.
Junker's test rig, which vibrates a joint sideways while monitoring clamp load, became known as the Junker test. It's codified in standards including DIN 65151 and ISO 16130 and is still the benchmark for qualifying locking devices such as wedge-locking washers and prevailing-torque nuts. If you've specified a Nord-Lock washer on rotating equipment, you've relied on Junker's work.
Hydraulic bolt tensioning (1970s)
Even with controlled hydraulic torque, the highest-stakes joints had a problem no wrench could fix: friction. About half the torque goes into friction under the nut, about 40% into thread friction, and roughly 10% into stretching the bolt. Preload from torque scatters widely from bolt to bolt. More in nut factor explained.
Hydraulic tensioning bypasses the nut. An annular jack pulls the stud axially, the nut is run down at zero torque, and the pressure is released. What you set on the gauge, allowing for load transfer, is what the stud keeps, and scatter drops from ±25% or more to under about ±10%. Tensioners can be linked on one hydraulic circuit to load every stud on a flange at once. Through the 1970s to the 1990s, tensioning spread across offshore, pipeline, power and petrochemical work, including subsea tools for diver and ROV use. See torque vs. tensioning.
Bickford writes the textbook (1981)
By the 1980s, bolted joint engineering had the tools, the vibration science and the materials. What it lacked was a single reference text. John H. Bickford, of Raymond Engineering's Power-Dyne division in Middletown, Connecticut, wrote it: An Introduction to the Design and Behavior of Bolted Joints, first published in 1981 and revised through several editions since. It remains the foundational text on joint diagrams, torque-tension scatter and preload selection.
With Sayed Nassar, Bickford also edited the Handbook of Bolts and Bolted Joints, covering design, assembly, in-service behavior and failure analysis across industries. He was a leading figure in professionalizing bolting as a discipline of its own. If Junker gave bolting its science, Bickford gave it its textbook.
Load-monitoring fasteners (1990s)
Every method so far controls the input and assumes the load. Load-monitoring fasteners measure the output. A common design anchors a reference pin in an axial hole down the bolt. As the bolt stretches, the head moves relative to the pin, and that movement is a direct measure of clamp load. Through the 1990s these became commercial products: direct-reading bolts, probe-read systems, and later wireless versions that report to plant monitoring systems.
They're concentrated where being wrong is expensive, such as reactor closures, inaccessible offshore joints and vibrating slew rings. They also make excellent auditors. Put four on a troublesome flange, run the usual procedure, and see what it really delivers. See load-monitoring fasteners.
ASME PCC-1 (2000)
Through most of the 20th century, even after the torque wrench, tensioning and Bickford's book, flange assembly was taught person to person, and every contractor had its own sequence and targets. Leaks were treated as a fact of life.
ASME's Post-Construction Committee changed that. In 2000 it published the first edition of ASME PCC-1, Guidelines for Pressure Boundary Bolted Flange Joint Assembly, one document covering preparation, alignment, lubrication, target torque, tightening patterns and troubleshooting. Later editions became an ANSI American National Standard, refined the alignment and washer guidance, and added methods for selecting target bolt stress and alternative assembly patterns. The 2013 edition added Appendix A, Training and Qualification of Bolted Joint Assembly Personnel, which made bolted joint assembly a qualified skill with documented training and demonstrated competence. The current edition is PCC-1-2022.
The cross pattern, the staged passes, the alignment limits and the qualification of the people doing the work all come from PCC-1. It's why a turnaround in Louisiana and one in Singapore can be specified and audited to the same documented process. See flange bolt tightening sequence.
Today: the methods toolkit
The remarkable thing about modern bolting is that every era's method is still in use, because each earns its place on a particular class of joint. The hammer wrench still tightens wellhead nuts where nothing else fits. The torque wrench still does most routine bolt-up in every plant in the world. The tensioner owns the biggest, most critical studs. The smart bolt watches the joints nobody can afford to be wrong about.
- Torque: cheap, fast, friction-sensitive.
- Turn-of-nut: the structural steel method.
- Torque-and-angle: much of torque's speed with far less friction sensitivity.
- Torque-to-yield: very repeatable, single-use bolts.
- Hydraulic tensioning: friction out of the loop, every stud at once.
- Heat tightening: for studs too large to tension.
- Ultrasonic measurement: checks the stretch, whatever method made it.
- Load-monitoring fasteners: measure the load, for the life of the joint.
The detailed comparison, with typical scatter and a selection framework, is in bolt tightening methods compared.
Why history matters at the joint
Every milestone here solved a problem the previous generation accepted as the cost of doing business. Archimedes' helix, Maudslay's leadscrew, Whitworth and Sellers ending shop-by-shop chaos, Bahr ending "tighten until it feels right", Junker explaining why bolts come loose, hydraulics bringing control to the largest joints, Bickford's textbook, PCC-1's qualified assemblers, and fasteners that finally tell us what's happening inside the joint.
The same is happening now with wireless bolts, connected tools that log every pass, and better coatings for joints that live in salt air. Most of it will quietly become the new baseline, the way a hydraulic torque wrench is now a boring everyday tool.
The tools change. The problem stays the same: get the joint right.
Sources and further reading
- ASME PCC-1, Pressure Boundary Bolted Flange Joint Assembly (current edition, 2022).
- Bickford, J. H., An Introduction to the Design and Behavior of Bolted Joints.
- Bickford, J. H. and Nassar, S. (eds.), Handbook of Bolts and Bolted Joints.
- Junker, G. H., "New Criteria for Self-Loosening of Fasteners Under Vibration," SAE Technical Paper 690055, 1969.
- ASME B1.1, Unified Inch Screw Threads.
- Sellers, W., "On a System of Screw Threads and Nuts," Journal of the Franklin Institute, 1864.
- Records of the Institution of Civil Engineers on Whitworth's 1841 thread proposal; ASME engineering history landmarks.
- The Medieval Housebook of Wolfegg Castle (c. 1480).
This paper first appeared on processbolting.com and was revised and corrected for this edition.