Tightening Methods Compared: Torque, Angle, and Yield-Controlled Tightening

Which tightening method you choose determines how much clamp force actually reaches the joint - and how much that force varies from bolt to bolt. The consequences of a wrong choice range from oversized fasteners and premature failures to full-scale recalls. This article explains the three most important methods with physical clarity, compares their key parameters, and shows when each method makes sense.

The Foundation: Why Torque Alone Is Not Enough

Torque is not a direct measure of clamp force. When tightening a bolt, the applied torque MA is split into three components: overcoming the thread pitch, friction in the thread (MG), and friction under the bolt head or nut (MK). Only a small fraction of the torque actually generates axial clamp force.

Roughly 90% of the tightening torque is consumed by friction - only about 10% generates clamp force. The friction coefficient is therefore the dominant parameter, and in practice it varies considerably: surface roughness, lubricant, coating, and temperature all influence it. Controlling only torque means controlling friction first and foremost - not clamp force.

This is precisely where advanced tightening methods come in.

The Three Methods in Detail

1. Torque-Controlled Tightening

Torque-controlled tightening is the most widely used method. The bolt is tightened until a specified target torque MA is reached - then the tool stops. Equipment is simple and cost-effective, and implementation is straightforward.

Physical limitation: Because friction variations feed directly into clamp force, the achievable accuracy is limited. With torque-controlled tightening, the tightening factor αA typically falls between 1.4 and 2.0 - meaning the maximum clamp force can be twice the minimum. In practice, this corresponds to clamp force scatter of ±50%.

To reliably achieve the minimum clamp force, the bolt must be designed for the worst case (maximum friction). This means the bolt is not optimally utilized under normal conditions, and the joint must be dimensioned conservatively.

Typical applications: Non-safety-critical joints, small production runs, rework, joints with well-controlled friction conditions.

2. Torque-and-Angle-Controlled Tightening

This method combines two control variables. First, a defined snug torque (seating torque) is applied to seat the joint and minimize embedment relaxation. From that point, the actual angle measurement begins: the bolt is turned a pre-calculated angle φ beyond the snug point.

Physical strength: Bolt elongation - and therefore clamp force - is proportional to the rotation angle in the elastic range. Because the influence of friction is largely eliminated after the snug torque is reached, clamp force scatter is significantly lower than with pure torque control. With angle-controlled tightening, the tightening factor αA drops to between 1.1 and 1.3.

The bolt can be tightened into the elastic or the plastic range. In the plastic range, angular errors have only a minor effect on torque scatter, which further improves the accuracy of the method. Torque-and-angle-controlled tightening is the established standard in the automotive industry.

Prerequisite: The rotation angle must be calculated or empirically determined for each joint configuration. This requires knowledge of bolt stiffness and grip length. A complete torque-angle curve from a joint analysis is the most reliable basis for this.

Typical applications: Safety-relevant joints in series production (automotive, mechanical engineering), joints with variable friction conditions.

3. Yield-Controlled Tightening

Yield-controlled tightening (YCT) is the most physically demanding method. The bolt is deliberately tightened into the plastic range - up to or just beyond the yield point. The goal: maximum bolt utilization with minimum clamp force scatter.

How the yield point is detected: During tightening, torque and angle are continuously recorded. In the elastic range, torque increases nearly linearly with increasing angle - the torque-angle curve has a constant slope (gradient). As the bolt approaches the yield point, the slope of the curve decreases: torque rises less steeply for the same angular increment. The tightening station identifies the inflection point of the torque-angle curve at the yield point through continuous gradient evaluation and stops the tightening process there.

This gradient-based shutoff makes the method largely independent of friction: the yield point is a material property of the bolt, not a function of friction.

A physical characteristic after tightening: During rotational tightening, the bolt develops not only tensile stress but also torsional stress from thread friction. The yield point is therefore reached through the combined equivalent stress of tension and torsion. Immediately after tightening, the torsional component springs back by approximately 50%, while the clamp force is retained - the joint thereby recovers an elastic reserve. This explains why yield-controlled joints do not fail in practice.

With yield-controlled tightening, the tightening factor αA drops to between 1.0 and 1.05 - clamp force scatter is below 5%.

Limitation: Because the bolt is plastically deformed, reuse is only possible to a limited extent. The method also requires bolts with a defined, reproducible yield point (property classes 8.8 to 12.9 per VDI 2230) and tools capable of simultaneously capturing torque and angle with sufficient resolution.

Typical applications: Highly loaded, safety-critical joints; joints where component weight or installation space demands maximum bolt utilization; wind turbines, automotive structural joints.

Comparison Table: The Three Methods at a Glance

Decision Logic: Which Method for Which Joint?

The choice of method is not a matter of convenience - it is a technical decision driven by safety class, component cost, production volume, and scatter requirements.

Guidance per VDI/VDE 2862

VDI/VDE 2862 defines three joint classes (A, B, C) and specifies the minimum requirements for tightening tools and process monitoring - differentiated by safety, availability, and environmental protection. Classification begins at the design stage: anyone who misjudges the risk relevance of a bolted joint may select a method that fails to meet the normative minimum requirements.

• Class C (no safety relevance): Torque-controlled tightening is generally sufficient.
• Class B (functional relevance, limited safety relevance): Torque-and-angle-controlled tightening is recommended; torque alone only with a tight tolerance band and documented friction coefficient.
• Class A (safety-critical): Both torque and angle must be monitored; yield-controlled tightening provides the highest level of process reliability.

The Torque-Angle Curve as the Basis for Method Selection

Regardless of which method you use: the torque-angle curve is the central analysis tool. It shows how tightening torque evolves over the rotation angle and makes visible what is actually happening in the joint - embedment relaxation, friction variations, the onset of yielding, or geometric errors in the joint.

For yield-controlled tightening, the curve is the control variable: the gradient is evaluated in real time, and the shutoff point lies at the inflection point of the curve. For angle-controlled tightening, the curve provides the basis for calculating the tightening angle. And for torque-controlled tightening, it shows whether friction conditions are stable enough for the method to be used reliably at all.

The QUANTEC MCS® analysis tool from GWK captures torque and angle simultaneously - with reference-point-free angle measurement that records the angle independently of a defined starting point. This is especially relevant when the tightening sequence does not begin at a precisely reproducible position. The complete torque-angle curve is immediately available for evaluation: as a basis for method selection in development, for process validation in quality assurance, and for parameterizing production tools.

The accuracy of the QUANTEC MCS® is ±1% between 10% and 100% of the nominal range - a prerequisite for reliable gradient evaluation in yield-controlled tightening. The recorded curves can be evaluated and archived directly in QuanLab Pro®.

Connection to VDI 2230: The Tightening Factor as a Design Parameter

VDI 2230 Part 1 is the authoritative guideline for the design of highly loaded bolted joints. The tightening factor αA plays a central role in its calculation procedure: it multiplies the required minimum clamp force and thereby determines the maximum load on the bolt.

A high αA value means the bolt must be designed for a significantly higher maximum force than the minimum clamp force alone would require. This leads to larger bolts, heavier components, and higher costs. Yield-controlled and angle-controlled tightening achieve a tightening factor approaching 1 - with torque-limited tightening, the factor can reach as high as 4.0 depending on the tool. The choice of method therefore has a direct impact on component design: a smaller bolt tightened with yield control can fulfill the same function as a larger bolt tightened with torque control.

Conclusion: Method and Tool Must Work Together

The three tightening methods are not interchangeable options - they have different physical foundations, different requirements for tooling and process, and different consequences for component design.

• Torque-controlled is simple and cost-effective, but friction-dependent and subject to high scatter. Sufficient for non-critical joints with stable friction conditions.
• Torque-and-angle-controlled significantly reduces the influence of friction, but requires careful parameterization based on the torque-angle curve. The standard for safety-relevant series joints.
• Yield-controlled delivers the lowest scatter and the highest bolt utilization, but demands precise measurement technology and bolts with a defined yield point. The single-use nature of the bolt must be accepted.

The torque-angle curve connects all three methods: it is analysis tool, parameterization basis, and process record all in one. Anyone who overlooks it is making method decisions based on assumptions - not data.

Accuracy by GWK.