All torque values green. SPC chart looks fine. Tools freshly calibrated. And yet the complaint rate is rising.
This is exactly how we often encounter it with our customers: a problem that does not show up in the statistics - because the wrong parameter is being monitored. This article describes a typical real-world scenario from automotive production. A scenario that shows what happens when you stop looking only at torque monitoring and start analysing the complete torque-angle curve.
Initial situation: Everything green - but something is wrong
Imagine an assembly line for safety-critical chassis components. Control arm bolted joints, A-class according to VDI/VDE 2862. An experienced team, modern screwdriving technology, established processes.
Torque values are cleanly within tolerance. The SPC chart shows green, batch after batch. No alarms, no outliers.
Then, during final inspection, something is noticed: individual joints show lower-than-expected breakaway torques in random re-checks. Not catastrophic, but reproducible. The complaint rate rises - slowly, but measurably.
The classic reaction: tools are recalibrated. Torque tolerances are tightened. Both are sensible measures in torque analysis - but the problem remains. Because its root cause lies somewhere else.
The blind spot: Torque ≠ preload force
Torque is only a reliable indicator of clamp force (bolt preload) if the friction coefficient of the bolt joint is constant. In reality, that is rarely the case in bolt joint analysis.
Friction coefficients of screw joints typically scatter between 0.08 and 0.30 in series assembly - these variations are the rule for many bolted joints. Physically, the applied torque is divided into three portions: thread friction, underhead friction and the portion that is actually converted into preload force. The preload force achieved at a given torque therefore depends directly on the friction coefficient.
The consequence: fluctuating friction means that even with very repeatable torque values, variations in the resulting preload force of 50% and more can occur.
50% variation - at the same torque value. This is not a theoretical edge case. This is everyday reality in series screw assembly quality.
The key insight: Torque values can be fully within tolerance—and yet the actual preload in the bolted connection is too low. Why? Because torque only provides a reliable statement about the clamping force if the coefficient of friction is constant. Fluctuations in the bolt coating make this relationship unreliable. The torque-angle curve shows what is really happening in the connection.
In our scenario, exactly this problem has been running invisibly through the line for weeks - hidden behind green torque values. To make it visible, you need a different kind of torque monitoring tool.
The investigation: QUANTEC MCS® reveals the complete tightening curve
GWK is brought in and deploys QUANTEC MCS® analysis tools on the affected station. The key difference compared with the previous torque monitoring approach: individual torque values at the end of tightening are no longer the only parameter being captured.
For the first time, complete torque-angle curves are recorded for every single screw assembly - directly under real production conditions, without a test bench and without a laboratory set-up. This delivers a true torque angle analysis of the process.
The reference-free angle measurement of QUANTEC MCS® is crucial here: it does not require a fixed reaction surface and delivers precise tightening curves with a measurement accuracy of ±1% between 10 and 100% of the nominal range - even under the conditions of a running assembly line. All data is captured and archived in real time via the QuanLabPro software for robust assembly quality assurance.
The result after a few hundred measurements: a clear picture - and a surprising discovery.
The discovery: The problem was hidden in the angle all along
The torque values are - as expected - all OK. That was already known from conventional torque monitoring.
But the angle values show significant scatter. The torque-angle curves reveal a clear pattern: around 15% of the bolted joints have a noticeably flatter gradient in the elastic range than the other connections.
What does that mean physically? In the elastic range of the tightening curve, the gradient is a direct measure of the bolt stiffness - and thus of the preload force that has been built up. A flatter gradient means that the target torque is reached at a larger torque angle. More friction consumes more torque before it is converted into elongation - the result: lower preload force.
The root cause is quickly identified: batch-to-batch differences in the surface coating of the bolts.
If the friction coefficient is too high, tightening stops too early - because torque, as the only directly measured variable during tightening, is reached or exceeded first. However, the desired bolt preload is not achieved. The result: the parts are only insufficiently clamped together.
This is exactly what the curves show: batches with higher surface friction reach the target torque earlier - but at that point the bolt has not elongated sufficiently. Clamp force is too low. Torque says "OK". Angle tells the truth.
Without torque angle analysis, this problem could not be detected. The line would have continued to run - with structurally under-preloaded chassis joints.
The solution: Three measures that eliminate the problem permanently
After the diagnosis, a structured response follows - immediate and medium- to long-term.
Immediate action: Friction value incoming inspection
The affected bolt batch is quarantined. For future deliveries, a tighter specification for the surface coating is agreed with the supplier - including documented proof of the friction coefficient.
In parallel, the Q-CHECK® QS and audit tool is deployed in incoming inspection. With a measuring range of 3-1000 Nm and an accuracy of ±1% between 10 and 100% of the nominal range, it enables fast checking of friction values on samples from every bolt delivery - before they enter production. This targeted bolt testing significantly improves production process reliability.
Process improvement: Angle-controlled tightening
The affected joints are converted to angle-monitored tightening. This means: torque is no longer the only stop criterion. A defined torque angle window in the elastic range is added. Joints that fall outside this window are automatically classified as NOK.
This method is robust against friction variations - because the angle is directly correlated with bolt elongation, regardless of how much torque is lost to friction. You can read more about the fundamentals of this torque angle method in our article on torque-angle analysis and tightening curves.
Monitoring: Regular PFU with QUANTEC MCS®
QUANTEC MCS® is permanently integrated into the quality assurance system. At regular intervals, process capability studies (PFU) according to VDI/VDE 2645-3 are carried out. These detect friction variations and process drifts at an early stage - long before they lead to customer complaints in automotive quality management.
How Cpk and Cmk values are correctly calculated and interpreted is explained in detail in a separate article.
The results: Numbers that speak for themselves
The effect of the measures is measurable and unambiguous.
| Metric | Before | After | Change |
|---|---|---|---|
| Complaint rate (Station) | Baseline value 100% | ~20% of the starting value | ↓ >80 % |
| Preload variability | High (visible in rotation angle values) | Significantly reduced | ↓ significantly |
| Cpk value (screw location) | Borderline (< 1.33) | Comfortable > 1.67 | ↑ process-capable |
| Proportion of noticeable fastenings | ~15% with a flat curvature gradient | < 1% after measures | ↓ drastic |
| Breakaway torque at final inspection | Irregular, downward outliers | Stable in the target range | ↑ consistent |
The most important number: The complaint rate at this station drops by more than 80% - not through more expensive bolts, not through tighter torque tolerances, but through the shift from one-dimensional torque control to two-dimensional torque-angle analysis with Quantec MCS.
The Cpk value of the affected joint rises from a borderline level to comfortably above 1.67 - the minimum value required in the automotive industry for safety-critical processes. What MFU and PFU mean and how both studies interact is explained in our dedicated guide to production process reliability in screwdriving technology.
Valuable side effect: similar analyses are conducted at other stations on the same line. This reveals additional hidden process issues that had previously remained under the radar of pure torque monitoring and torque analysis.
The lesson: What this scenario means for your production
This typical scenario from automotive practice has a clear message for bolt joint analysis and screw assembly quality:
- Torque alone creates a false sense of security. Green SPC charts and calibrated tools do not rule out quality issues if the friction coefficient is fluctuating.
- The torque-angle curve is the "fingerprint" of every bolted joint. It shows what is happening inside the joint - not just how much torque was applied.
- Friction variations are not an exception. The relationship between torque and preload force depends on bolt geometry and friction. From torque alone you cannot reliably derive the preload force achieved in screw assembly.
- Regular analyses with QUANTEC MCS® are an investment in process reliability - and in many cases the most direct route to a stable, auditable tightening process for safety-critical A-class joints.
The question every production engineer should ask after this scenario: Would I notice if the same problem were currently occurring on my line?
Frequently asked questions
Why is torque control alone not sufficient to ensure secure screw connections?
Torque is only a reliable indicator of preload if the coefficient of friction is constant. In practice, friction coefficients for screws vary considerably - depending on coating, lubricant, and surface condition. Even when torque values are kept exactly within bounds, the actual clamping force can deviate significantly from the target value. The torque-angle analysis makes these deviations visible.
What does the rotation angle reveal about the bolt connection that the torque does not show?
The rotation angle in the elastic range is a direct measure of the screw elongation - and thus of the built-up preload. With above-average friction, the screw reaches the target torque earlier, i.e., at a smaller rotation angle. That means less screw elongation, less clamping force - even though the torque is OK.
What are typical causes of friction fluctuations in screw assembly?
Batch-to-batch differences in the surface coating of the fasteners are one of the most common causes. In addition, lubricant fluctuations, temperature fluctuations, corrosion at contact surfaces, and varying component tolerances also play a role. In the automotive industry, therefore friction windows (e.g., µges = 0,08-0,14 according to the VDA standard) are defined, whose compliance must be regularly checked.
How often should a torque-angle analysis be performed with QUANTEC MCS®?
For safety-critical A-class bolted joints, we recommend regular process capability investigations (PFU) according to VDI/VDE 2645-3. The frequency depends on process stability, supplier changes, and the bolt class. As a rule of thumb: after every supplier change, after coating changes, and at least annually for A-class bolted joints.
Can QUANTEC MCS® be used directly in the production line?
Yes. QUANTEC MCS® is designed to be used under real production conditions. The reference-free rotation-angle measurement does not require a fixed reference surface, which enables use directly on the assembly line. The recorded data can be archived and analyzed using the QuanLabPro, Ceus, or QS-Torque software.
Ready to uncover hidden issues on your line?
QUANTEC MCS® with reference-free angle measurement is the only analysis tool that provides the full torque-angle curve under real production conditions - with a measurement accuracy of ±1% between 10 and 100% of the nominal range, in a robust aluminium-titanium design, compatible with QuanLabPro, Ceus and QS-Torque. It is a powerful basis for advanced torque monitoring, angle analysis and bolt testing in modern assembly quality assurance.
You would like to test QUANTEC MCS® without investment risk first? With GWK ToolRent® you can access calibrated analysis tools on demand - weekly, monthly or annually, with worldwide shipping.
Accuracy by GWK.
