Fastener Technology International — August/September 2015
Change Language:
Bolt Fatigue Failure Due To Insufficient Tightening
Bill Eccles PhD

Ensuring fasteners are tightened sufficiently is one of the key factors to ensuring product reliability with bolted joints.

Not all fastener failures are due to a manufacturing or material defect, or even a design flaw. With bolts and threaded fasteners in general, failure to tighten them sufficiently is a common cause of failure. It is by no means obvious why fasteners should be tightened to the extent that they are, especially those of larger diameter. Hence, on occasion without the right controls in place, they can easily be under tightened.

The plates comprising a joint are clamped together by tightening the bolts. In doing so, any subsequent loading that is applied to the joint unclamps it rather than additionally loading the bolt to any significant amount. If the preload is insufficient, the joint can be completely unloaded, that is, no clamp force present between the plates. The terms frequently used for this condition are joint decompression or joint separation. Any subsequent additional loading, once the joint decompression point is reached, is completely sustained by the bolt. If the load that is applied to the joint fluctuates, it results in the bolt sustaining a high alternating stress. Fatigue failure, normally in the threaded part of the bolt, is often the consequence. If joint decompression can be prevented by adequate tightening, in most but not all instances, fatigue failure will not occur.

A Case of Fatigue Failure

Fatigue is a common, possibly the most common, cause of failure for fasteners in service. The root cause in many such situations is however a lack of bolt preload. A case of fatigue failures of M27 bolts used in a baling press joint illustrates this point.

The joint originally comprised eight M27 property class 8.8 bolts. Due to the repeated failure of these bolts, it was decided to increase the bolt strength. As a result of local lack of availability of property class 12.9 bolts, property class 12. 9 socket head cap screws were used. The joint sustains the reaction force of the hydraulic cylinder that is used to compress bales of paper. Prevailing torque lock nuts were used with the bolts with a washer being present under the bolt head and nut face. The property class 12.9 bolts failed after they had been in use for approximately six months. The machine was in use on a typical day for 12 out of 24 hours.

The nature of the failed surfaces indicated that the failures were due to fatigue. The failures typically occurred at the nut/joint interface. A failed bolt is shown in Figure 1 and a cross section view is shown in Figure 2. The nut/joint interface is often the region where the bolt fails, since usually the stress concentration at this location is greater than in the thread runout and under the bolt head.

Based upon the machine operational rate, the bolts were failing after approximately 100,000 load cycles. There was some nonlinear loading occurring, but an approximation was that the alternating load cycles were between 0 to 100 kN. This is illustrated in Figure 3.

There is a difference between the loading that is being applied to the joint and that which is being sustained by the bolt. Until joint separation occurs, the majority of the applied load will reduce the clamp force on the interface. Calculations indicate that up to the point of joint separation, for this specific case, the joint will sustain only about 7.5% of the applied loading, 92.5% of the loading reducing the clamp force on the interface.

The nuts were tightened without a torque wrench, essentially the same torque was being used with the 12.9 bolts that had been used with the 8.8 bolts. Essentially, reliance was placed upon the individual doing the tightening, knowing and deciding that the bolts were sufficiently tight. An estimate of the torque that was being applied, based upon the size of the wrench being used, was anticipated to be approximately 500 Nm.

FTI EMPHASIS: Fastener Failures

Allowing for the likely scatter in the preload in a joint, an analysis can be achieved, and controlled tightening processes (torque tightening, torque-angle, tensioning) can be accounted for by use of a tightening factor. Essentially, this factor is the ratio of the maximum anticipated preload to the minimum anticipated preload for a particular tightening process. For torque tightening, this ratio is in the order of 1. 6, that is, the maximum preload is likely to be 1.6 times the minimum preload. When a person’s judgement is used rather than a torque wrench, it is likely that a significant amount of scatter would occur. For this analysis, a tightening factor of two was used for the manual tightening, but it may well have been greater.

The results of a joint analysis calculation are presented in the preload chart in Figure 4. As can be seen, the expected preload requirement exceeds that which is anticipated to be provided by a bolt by a significant margin. The consequence of this is that joint separation would occur under the applied loading. Every time the hydraulic cylinder exerts maximum load, a small gap would occur in the joint. When this happens, the bolt sustains the full magnitude of the loading. The load would fluctuate from whatever the preload was provided by the bolt to the maximum applied loading. The consequence is that a higher alternating force would be sustained by the bolt increasing the alternating stress being sustained by the threaded region of the bolt. For preloaded bolts, it is the alternating stress rather than the magnitude of the preload stress that is crucial from a fatigue standpoint.

Repeated joint separation will usually result in the alternating stress exceeding the fatigue endurance limit for the thread. The fatigue endurance limit is relatively low for threaded fasteners. In this case, for a M27 thread, it is on the order of 45 Mpa, less than a twentieth of the bolt’s tensile strength. When joint separation occurs, especially considering the number of cycles likely in this application, fretting will occur that would result in a further loss of preload over time steadily making the situation worse. This would also be coupled with a loss of preload due to crack propagation. As the crack propagates through the threaded section, the stiffness of the bolt reduces, which in turn reduces the preload.

To prevent joint separation, a higher and more consistent bolt preload is required. To achieve a higher preload, an appropriate torque value needs to be specified that would fully utilize the strength of the bolt. Specifically:

• Based on the friction characteristics in the thread and under the bolt head, an analysis indicates a 1200 Nm torque is appropriate for this size and strength of bolt.

• To deliver this level of torque either a manual torque wrench with a torque multiplier is required or a hydraulic wrench or similar controlled tightening method.

Based upon a tightening torque of 1200 Nm delivered by a hydraulic torque wrench, the Preload Requirement Chart is shown in Figure 5. When tightened to this torque level, the preload is significantly greater than what was being achieved previously. Although the preload stress will be greater, since the alternating stress will be less and below the thread’s endurance limit, fatigue failure is not anticipated. Lessons that can be learned from such failures are:

• Full benefit of using a higher-strength bolt will not be realized unless you tighten it to fully utilize its strength.

Fatigue endurance strength of a tightened property class 12. 9 bolt is similar to that of a 8.8, rather surprisingly.

• Use only property class 12.9 bolts if there is a need and it is intended to fully utilize the bolt strength. Additional risk is being sustained from hydrogen embrittlement and stress corrosion cracking if 12.9 bolts are tightened to the same level as 8.8 bolts for little or no benefit.

• Guessing whether bolts are sufficiently tight is not a good approach if long/trouble-free product life is required.

Conclusion

Fatigue failure of bolts as a result of inadequate tightening is relatively common. Ensuring that fasteners are tightened sufficiently is one of the key factors to ensuring product reliability with bolted joints. Www.boltscience.com

Company Profile:

Bolt Science is a UK-based company specializing in consultancy on the technical aspects of bolting. Besides consultancy, the company provides training including on-line training on bolting topics. Bolt Science’s website provides details about other case studies and general information on bolting technology. Contact Bill Eccles at bill.eccles@boltscience.com. www.indfast.com
VIEW ALL ARTICLES
Message
SEND