The methods at a glance
Scatter figures are approximate ranges commonly cited in bolting literature, including Bickford's work on bolted joints, for preload variation from bolt to bolt with good practice. Real results depend on tooling, procedure and crew.
| Method | What it controls | Typical preload scatter | Where it lives |
|---|---|---|---|
| Operator feel | Nothing measured | ±35% or worse | Where it shouldn't |
| Torque | Applied torque | ±25% or more | Most routine bolting everywhere |
| Turn-of-nut | Rotation past snug | About ±15% | Structural steel (RCSC) |
| Torque-and-angle | Snug torque, then measured angle | About ±10–15% | Machinery, some critical joints |
| Torque-to-yield | Onset of bolt yield | About ±5–8% | Automotive; single-use bolts |
| Hydraulic tensioning | Axial load on the stud | Under about ±10% | Large and critical studs, flanges, wind, subsea |
| Heat tightening | Thermal elongation of the stud | Depends on how elongation is controlled | Very large turbine and vessel studs |
| Direct tension indicators | Washer bump compression | About ±10% | Structural steel |
| Ultrasonic elongation | Measured stretch | About ±1–10% | Verification on critical joints |
| Load-monitoring fasteners | Measured load, continuously | A few percent | High-criticality and problem joints |
Torque
Torque is cheap, fast, and works on any size with the right tool, from a hand wrench up to a hydraulic torque wrench. Its weakness is that about 90% of the input goes to friction, so preload follows the nut factor, not the wrench. Torque dominates because most joints have the margin to absorb the scatter.
Turn-of-nut and torque-and-angle
Bring the joint snug, then rotate a defined amount. Past snug, rotation maps to stretch through the thread pitch and joint stiffness, and friction matters much less. Turn-of-nut is the structural steel standard. Torque-and-angle is the machinery version, and power tools with angle encoders make it practical in the field.
Torque-to-yield
Tighten until the torque-angle curve flattens at yield. Preload becomes as repeatable as the bolt's yield strength, at the cost of a bolt that is plastically stretched and generally not reused.
Hydraulic tensioning
Stretch the stud directly with an annular jack, run the nut down at zero torque, then release. Friction is out of the loop, every stud on a flange can be loaded at once, and nothing twists. It needs stud protrusion and clearance, and correct allowance for load transfer. Full comparison in torque vs. tensioning.
Heat tightening
For studs too big for practical tensioners, such as large turbine casings and vessel closures, a heater rod in a bore down the stud's center lengthens it thermally. The nut is run down while the stud is hot, and the stud shrinks back to clamp the joint. Accuracy depends on measuring elongation or controlling the nut rotation, not on the heat itself.
Ultrasonic measurement and load-monitoring fasteners
These measure the result rather than controlling the input. Ultrasonic gauges time a sound pulse down the bolt to read elongation, so they work as a check on any tightening method. Load-monitoring fasteners carry their own measurement and can report clamp load for the life of the joint.
Choosing a method
Work through these questions in order:
- What does a leak or failure cost? The higher the consequence, the further down the table you go.
- How much scatter can the joint absorb? That depends on gasket seating stress range, bolt yield margin and flange rotation limits. Your engineering basis or PCC-1's target bolt stress selection decides it.
- What does the geometry allow? Stud protrusion, radial clearance and access to both sides.
- Size and count. Large diameters and many studs favor tensioning, both for accuracy and for speed.
- Do you need proof? If QA, an insurer or the next failure investigation needs to know the load, measure it.
On a typical Gulf Coast turnaround, the answer is a mix: torque for most of the piping, tensioning for exchanger heads and large critical flanges, measurement on the handful of joints that have leaked before. Not sure which one your joint needs? Ask us. Telling you a $400 socket will do instead of a $12,000 tool is part of the job.
Frequently asked questions
What is the most accurate way to tighten a bolt?
Methods that measure the bolt's stretch or load directly, such as ultrasonic elongation, strain-gauged or load-monitoring fasteners, typically get within a few percent. Among methods that control an input, hydraulic tensioning is generally the most accurate for large studs, and torque alone the least.
What is turn-of-nut tightening?
A method used mainly for structural steel bolts under RCSC rules. The bolt is first brought snug tight, then the nut is turned a specified additional rotation, typically 1/3 to 1 turn depending on bolt length and diameter. Rotation past snug depends less on friction than torque does.
What is torque-to-yield?
Tightening until the bolt just begins to yield, detected by the change in slope of the torque-versus-angle curve. It gives very repeatable preload but plastically stretches the bolt, so the bolt is typically single-use. It is common in automotive engine assembly and rare in plant maintenance.