When you torque a fastener, you are applying a rotational force and relying on the torque-tension relationship to translate that force into clamp load.

The relationship works like this: clamp load equals applied torque divided by the product of the nut factor and the nominal bolt diameter. Nut factor, also called the K-factor, captures the friction characteristics of the joint: friction under the bolt head, at the bearing surface, and at the thread interface.

In a typical unlubricated carbon steel joint, the nut factor runs roughly 0.20. With a lubricant or coating applied, it can drop to 0.15 or lower. That difference matters. The same applied torque produces meaningfully different clamp loads depending on which surface condition the spec was written for.

Approximately 85% to 90% of the torque you apply overcomes friction. Only 10% to 15% translates into the clamp load that holds the joint together. That ratio is manageable at room temperature, where friction characteristics are stable and predictable. At elevated temperatures, both variables in the torque-tension equation start to move.

Key Takeaways:

• Applying torque does not directly control clamp load. Roughly 85-90% of applied torque overcomes friction, and any change in friction at temperature changes the clamp load the joint actually retains.

• Stress relaxation causes permanent, invisible clamp load loss at elevated temperatures. A joint can lose 30%-40% of its designed clamp load with no nut rotation, no visible loosening, and no external indication of failure.
• The original assembly torque value is not always correct for retorquing. If lubrication has burned off or threads have oxidized, the nut factor has changed and the torque value must be recalculated for the new surface condition.
• Selecting the right alloy is not separate from the torque specification. The torque value sets the initial clamp load; the alloy determines how much of that clamp load the joint retains at operating temperature over time.

How Temperature Degrades the Torque-Tension Relationship

High-temperature service introduces two compounding problems that standard torque specifications do not account for. Lubrication degrades or burns off above its rated temperature ceiling, raising the friction coefficient and changing the nut factor the original torque value was written for.

And beyond the first heat cycle, oxidation begins modifying the thread interface with each successive cycle. Both mechanisms shift the torque-tension relationship away from the conditions assumed at assembly, and both do so without any visible external indication that the joint has changed.

Lubrication Burns Off

Most lubricants and thread-interface compounds have a service temperature ceiling well below exhaust operating conditions. When a joint reaches temperatures above the lubricant's rated limit, the lubricant degrades or burns off. The friction coefficient rises. The nut factor increases.

The result: a joint assembled to specification at room temperature now has a different torque-tension relationship in service. The original torque value that produced adequate clamp load at installation no longer produces the same clamp load at temperature.

A shift in nut factor from 0.15 to 0.22 at the same torque value translates directly to a proportional drop in clamp load. The fastener has not failed. The joint has not loosened visibly. The clamp load has fallen because the friction characteristics of the interface changed.

Oxidation Changes Thread Surfaces With Every Cycle

Beyond the first heat cycle, oxidation begins to modify the thread interface. Each thermal cycle changes the surface characteristics of the mating threads. The effective friction coefficient shifts with each cycle, and not always in the same direction.

This makes retorquing a more complex operation than reapplying the original torque value. The surface condition of the thread interface after 50 thermal cycles is different from the surface condition at initial assembly. If the retorque specification does not account for that change, the resulting clamp load may differ from what the spec assumes.

Stress Relaxation and Rotation-Free Clamp Load Loss

Clamp load loss from changed friction conditions is one mechanism. Stress relaxation is a separate and often more significant one.

Stress relaxation is the permanent, time-dependent loss of clamp load that occurs in a fastener under sustained stress at elevated temperature. The fastener elongates slowly under load. That elongation bleeds off clamping force. The nut has not moved. The fastener has not stripped or fractured. But the joint has lost preload through accumulated strain.

This type of clamp load loss leaves no external evidence. A joint that passes room-temperature inspection after assembly, with all fasteners torqued to specification, can lose significant clamp load in service through stress relaxation alone. Visual inspection will not reveal it. Nut rotation will not occur. The only way to confirm clamp load retention is through preload measurement or an audit torque study.

Thermal cycling exacerbates the problem. The fastener and the joint flange expand and contract at different rates with each temperature change. Every cycle works the assembly and accumulates strain beyond what the original torque-tension calculation assumed. In joints with dissimilar metals, the differential expansion effect is larger and may require engineered compensation in the joint geometry.

Verifying Clamp Load in High-Temperature Joints

Standard visual inspection is not sufficient to verify clamp load retention in a high-temperature joint. A joint that looks correct from the outside may have lost 30% to 40% of its designed clamp load through stress relaxation, with no visible indication.

Two verification methods are appropriate. Preload measurement with ultrasound gives direct clamp load data without requiring disassembly. An audit torque study applies a known torque and measures the breakaway torque required to produce rotation. The gap between audit torque and original assembly torque indicates how much preload the joint has retained.

An audit torque study is an indirect measurement. Where direct clamp load data is required, ultrasound preload measurement is the appropriate tool. MacLean-Fogg's in-house test capability includes torque-tension analysis with ultrasound for exactly this purpose.

Not sure whether your joint is retaining clamp load after heat cycling?

MacLean-Fogg's engineering team runs torque-tension analysis, including ultrasound measurement, for specific applications. Contact us to request a joint review.

Retorquing Best Practices for High-Temperature Applications

High-temperature joints do not behave the same way at the second assembly as they did at the first. Lubricant burnoff, oxidation, and stress relaxation through the initial heat cycles all change the conditions under which the joint needs to be retorqued. Understanding what has changed at the thread interface is the prerequisite to retorquing correctly.

When Retorquing Is Appropriate

For joints that see significant clamp load loss through the initial heat cycles, a planned retorque after the first thermal exposure is standard practice. This applies to exhaust joints, turbocharger mounting hardware, and other applications where the fastener reaches operating temperature early in service life.

The retorque should occur after the joint has reached operating temperature and cooled back to ambient. The goal is to restore clamp load that the initial thermal cycle bled off through stress relaxation and, where applicable, lubricant burnoff.

Accounting for Changed Surface Conditions

The original assembly torque value is not always correct for reassembly. If the thread interface has changed through lubricant burnoff or oxidation, the nut factor has changed. Applying the original torque to a joint with a higher nut factor produces less clamp load than the torque spec assumes.

Where the application uses a dry or oxidized thread interface at the time of retorque, the torque value should be recalculated for that surface condition. Applying a lubricated-surface torque value to a dry, oxidized interface underdelivers on clamp load.

Fastener Torque Chart Considerations: Alloy Selection and Clamp Load Retention

A fastener torque chart gives you target values for achieving a specific clamp load at room temperature in defined surface conditions. It does not tell you how much of that clamp load the fastener will retain at 500 or 600 degrees C.

For carbon alloy fasteners, the answer is: significantly less. At 500 degrees C, a Grade 10.9 bolt retains approximately 56% of its room-temperature yield strength. At 600 degrees C, that drops to around 24%. A bolt softened to 24% of its design strength cannot hold the clamp load the torque specification assumed, regardless of the torque value.