Torque Control, Torque-Angle and Yield-Point Control: A Comparison of Three Tightening Strategies
Executive Summary
There is only one tightening-torque knob, but three tightening strategies: torque control, torque-angle, and yield-point control. Their prices differ several-fold, yet the preload scatter they deliver differs three-fold. This article uses concrete parameters to explain the mechanism, equipment requirements, precision limits and selection criteria of the three strategies, and gives an engineering approach for assigning strategy by joint safety class.
Table of Contents
Same Torque Setting, Two Different Fates for the Joint
On one gearbox-housing line, two stations tightened the same M10×1.25 grade 10.9 bolt, both set to 55 N·m. Three months into production, torque-clamp-force tests on samples showed line A preload at 26–34 kN; line B, with the same bolt lot, at 20–42 kN, with the loosest already below the slip-resistant requirement. The equipment, torque and parts were identical; the difference came from the tightening strategy: line A used a servo spindle with torque-angle control, line B used a pneumatic tool in pure torque control.
Fastener tightening strategies have only three basic forms: torque control, torque-angle, and yield-point control. They differ not in level of precision but in what variable is controlled—one controls torque, one controls angle, one controls the slope-change point.
Mechanism of the Three Strategies Broken Down
Torque control: convenience at the cost of friction
Stop when the set torque is reached. The bolt axial force is back-calculated from T = K × F × d. The problem is that K itself scatters: within one bolt lot, K may vary between 0.11 and 0.16 due to fluctuations in coating thickness, lubrication and surface roughness. K is a multiplicative factor, so K fluctuation is amplified directly into preload fluctuation.
Torque-angle: bypassing the friction coefficient
First apply a small snug torque (usually 20%–30% of target torque) to eliminate clearance, then turn a fixed angle from that point. The bolt elongation is determined precisely by pitch and angle, largely independent of the friction coefficient. The core parameter is the angle:
Δl = (P / 360°) × θ, F ≈ Δl × KS
P: pitch; θ: turned angle; KS: bolt stiffness
For an M10×1.25 bolt, turning 90° corresponds to about 0.31 mm of axial elongation. As long as the bolt has not entered the plastic zone, this elongation is unrelated to friction.
Yield-point control: running close to the material limit
The torque-angle curve is captured continuously and its slope dT/dθ computed in real time. The point where the curve slope begins to fall is the yield onset; the tool stops a fixed angle after that point. It pushes preload to 90%–100% of the material yield strength and is the only strategy that automatically compensates for friction differences.
Parameter Comparison of the Three Strategies
| Comparison | Torque control | Torque-angle | Yield-point control |
|---|---|---|---|
| Controlled variable | Torque | Snug torque + angle | Torque-angle curve slope |
| Preload utilization factor ν | 0.60–0.70 | 0.80–0.90 | 0.90–1.00 |
| Tightening factor αA | 1.3–2.0 | 1.1–1.25 | 1.05–1.15 |
| Preload scatter | ±25%–±35% | ±10%–±15% | ±8%–±12% |
| Friction sensitivity | High | Low | Very low |
| Equipment requirement | Torque wrench / pneumatic tool | Servo spindle with angle encoder | High-end spindle with curve capture and algorithm |
| Thread plastic zone | Must not enter | Near but not entering | Allowed and exploited |
| Bolt reusability | Reusable | Case by case; not recommended | Not recommended for reuse |
Failure Limits of Each Strategy
Three typical failures of torque control
- Abnormal friction causes insufficient preload: missed coating, lubricant migration or surface rust raises K while torque stays unchanged, dropping preload by as much as 30%.
- Over-tightening: when dry unlubricated threads drop K to 0.10, the same torque produces much higher preload and may enter yield directly.
- Invisible false pass: the torque reading is normal, actual preload is insufficient, line data is all green, yet a hazard is already hidden in the joint.
Limits of torque-angle
- Over-tightening when bolt strength is low: the angle is fixed; if bolt strength is low or friction abnormally high, the bolt enters plasticity or even snaps before the angle is reached. Therefore this method requires a stated lower limit of material yield strength and an upper limit of friction coefficient.
- Crushing of clamped parts: soft clamped parts (aluminium, magnesium) dent under large preload; the angle is reached but effective preload is insufficient. Bearing-surface pressure must be checked and hardened washers added.
- Misjudged snug point: burrs, paint layers or incomplete seating on the assembly shift the snug point backward, leaving the actual angle too small.
Limits of yield-point control
- The yield-detection algorithm depends on curve quality; insufficient sampling or noisy signal causes misjudgement, possibly reading the elastic region as yield and yielding insufficient preload.
- Risk rises when material scatter is large: when the curve slope changes subtly (e.g. low-strength steel, softened material), the algorithm struggles to catch it.
- After tightening the bolt is near yield, so it cannot be reused; it must be replaced after disassembly.
Assigning Strategy by Safety Class
The selection criterion is not "more expensive is better" but the consequence of joint failure. A practical grading approach:
| Joint safety class | Typical locations | Recommended strategy | Additional requirements |
|---|---|---|---|
| Class S (failure endangers life) | Brakes, steering, airbags, load-bearing lifting points, battery-pack fastening | Yield-point or torque-angle control | 100% curve recording; threads not reusable |
| Class A (failure causes line-down / major overhaul) | Engine, gearbox, e-drive, shock absorbers | Torque-angle control | Spot-check torque-clamp-force; specify friction-coefficient upper limit |
| Class B (failure repairable) | General housings, brackets, covers | Torque control + lubricating coating + servo tool | Staged tightening; spot-check preload |
| Class C (non-structural / cosmetic) | Interior, guards, shrouds | Torque control | Consistency control only |
Putting the Strategy into Process Documents
Implementing a tightening strategy requires three things in place at once: equipment capability, friction state and process parameters. It is recommended to specify the following in the process document:
- Strategy and parameters: e.g. "snug torque 12 N·m, angle 90°±5°" or "yield-point control, start torque 20 N·m".
- Friction-state definition: coating type, application method, acceptable K range (e.g. 0.11–0.15), and incoming-inspection frequency.
- Tightening sequence and staging: multi-bolt joints require a diagonal-cross sequence and staged load ratios.
- Process-data requirements: whether torque-angle curves are retained, retention period, and fault criteria (e.g. angle out of tolerance, abnormally high torque).
- Tool calibration interval: servo spindles monthly or per 10,000 cycles; torque wrenches daily spot-check.
The difference among the three strategies ultimately comes down to one number: how wide the usable preload window is. The more refined the strategy, the smaller αA, the wider the usable window, and the stronger the slip, separation and fatigue resistance the same bolt can deliver.
Conclusion
Torque control is cheap, universal and reusable, suiting most joints. Torque-angle bypasses the friction coefficient with one snug torque plus a fixed angle and is the workhorse for critical batch joints. Yield-point control runs close to the material limit and is used where failure consequences are most severe.
Before choosing a strategy, first determine the consequence class of joint failure, then the acceptable cost, and finally whether the friction state can be controlled. Once all three are clear, strategy choice is a simple question; conversely, if the friction state is out of control, even the most expensive tightening equipment merely packages the uncertainty more beautifully.
