Simple QA for wind-turbine bolts

It is common knowledge that torque control is a risky way to achieve consistent pretension in structural bolts. It all comes down to friction or what’s known as the k-factor. The k-factor is the relationship between the torque applied to a fastener assembly and the actual tension achieved in the bolt.

The short form of the relationship is:

Tension = Torque/kd

where:
k = k-factor, and d = diameter.

K-factors vary from lot-to-lot, from bolt-to-bolt within a lot, and as a consequence of handling, storage, and lubri-cation. K-factors can range from 0.10 to over 0.20. Add to this the variability of the torque tool, and you can bet that when using torque control, actual bolt tensions will vary ±40% or more.

A better tension indicator

The Squirter Direct Tension Indicator is a compressible washer that tells when a target tension has been reached in a bolt, regardless of the applied torque. When the bumps on the washer compress sufficiently, orange silicone squirts out of cavities under the bumps. The bump compression and squirt appearance are calibrated to tension in the manu-facturer’s ISO/IEC 17025 laboratory. Millions have been used around the world. They conform to ASTM and EN standards, and Germanischer Lloyd certifies that the devices for M36, M42, and M48 bolts will reach target tension ±10%.

Wind turbine owners and erectors are looking for a better, quicker, and cheaper way to assemble and maintain their towers. Most turbine OEMs specify bolt torques, and some require retightening tower bolts after commissioning. The direct tension indicator or DTI can be used as a simple quality assurance system to guarantee target tension during initial installation and when retightened.

Flange tests

The results from a recent test with flange sections demonstrate the reliability of the device. We ultrasonically instrumented 20, M42x280 10.9 HV bolt assemblies. In a temperature controlled environment, we lubricated the threads and nut faces and installed them in actual flange sections. Each bolt was tensioned using squirt appearance from the DTI alone. The bolt elongation and calculated tension was recorded, along with the torque indicated by a Tone Tower Master electric wrench.

The minimum preload for a M42 10.9 is 710 kN. The purpose of the DTI is to ensure a minimum preload. So, in the calibration exercise using a Skidmore Wilhelm, we shoot for 780 kN, roughly 10% higher than the minimum pretenstion. We then match that squirt appearance when tightening the assemblies in the flange.

The graph First squirt shows results along with a photograph of a typical first squirt. The average tension achieved using only the squirt appearance was

761 kN with a standard deviation of 20kN. All values were above the minimum preload of 710kN and well below the

930 kN proof load. Notice that tension varies less than ±10% while the torque varies more than ±20%. K-factors ranged from 0.09 to 0.16 (±28%), even when using ample lubrication. While this degree of variation is not uncommon, one quickly understands why experienced owners and erectors want something better than torque to indicate the tension in flange bolts.

After cleaning off the original squirt and letting the flange assembly relax for a day, we retorqued the nut until the second squirt appeared. The graph Second squirt shows results along with a photograph of a typical second squirt. The second squirt is not as voluminous as the first, but it is distinct. The average tension achieved at second squirt was 810 kN with a standard deviation of 40 kN. This value is 49 kN or 6% higher than the average original tension. There is more variation in the data, yet the tensions are still well below the proof loads. In three cases, our wrench didn’t have enough torque to achieve the second squirt, so the sample size is slightly smaller (17 versus 20).

Trouble with torque

A recent experience with an OEM underscores the unreliability of torque as an indicator of tension. After tightening its M42 10.9 HV fastener assembly to its specified torque in a Skidmore Wilhelm tension calibrator, the indicated tension was only 500 kN, 30% below the 710 kN minimum. Lubricating the threads and nut face with moly-disulphide increased the tension to 700 kN, but the minimum pretension was still not achieved.

We then performed a Squirter DTI calibration exercise. On his first try using squirt appearance alone, a technician was able to hit an initial pretension of 815kN, or 4% above the target. We backed the nut off a bit, and a different technician retightened until the second squirt just started to emerge. The tension was 835kN, 2.5% above the initial value.

The DTI described here lets erectors achieve initial target tensions ±10% in tower flange bolts as certified by Germanischer Lloyd. What’s more, the second squirt provides a slightly higher tension when retensioning is required at a later date. Given the variation in k-factors allowed by specification, the Squirter is a simple structural quality assurance system for wind turbine towers. A patent is pending for Applied Bolting’s second-squirt capability. WPE