After hydraulic tensioner assembly, the thread friction torque alone provides joint locking and counters bolt-stretch torque, with nut-factor K typically 0.18–0.20 for steel-on-steel and friction scatter of plus or minus thirty percent [S1][S3].
An eccentric fixed belt tensioner holds timing or V-belt tension through a fixed pivot and a slotted or eccentric body, so the locknut torque is set against the spring stretch of the belt and the operating load range, not against a continuous hydraulic hold.
Why the nut factor, not the torque value, is the engineering decision
The relationship between applied torque T and installed preload F is dominated by the nut factor K, with the short-form equation T = K F d carrying scatter of plus or minus thirty percent when thread-flank and under-head friction vary across a real lot of bolts [S3]. A change in K from 0.18 to 0.20 shifts the installed preload by about eleven percent at the same nominal torque, so the locknut on an eccentric fixed tensioner is sized by selecting a friction band and a torque target inside it, not by reading a chart number [S3]. For steel-on-steel fasteners in the 8.8 to 10.9 grade range the nut factor commonly lands in 0.10 to 0.25, with 0.18 to 0.20 as the most frequently assumed band on plated and lightly oiled hardware [S3].
The scatter is what kills the joint, not the mean. A 30 kN minimum preload window with a 50 kN yield ceiling gives a tolerance band of plus or minus twenty-five percent, and a K=0.18 nominal target with plus or minus thirty percent scatter pushes the low tail under 28 kN and the high tail past 52 kN, so the joint either leaks or yields on installation [S3]. For an eccentric fixed tensioner, where the operating load is also a moving belt load, the safe approach is to down-rate the K assumption and re-derive the torque, or to use a secondary locking device on the locknut such as a serrated washer or a nylon insert nut.
What tensioner assembly changes about locknut torque
Tensioner assembly leaves no torsional twist in the joint components because thread friction torque balances bolt-stretch torque, and the tensioner-to-clamped-part friction during assembly and the nut-to-clamped-part friction after assembly are zero [S1][S2]. For an eccentric fixed tensioner this matters because the only torque left in the joint after release is whatever was applied to the locknut, with the belt tensioner body itself already preloaded by the hydraulic puller or by a mechanical pre-stretch.
The external load-induced loosening torque required to initiate nut-bearing-interface slip is more than five times greater for a tensioner-assembled joint than for a torque-assembled joint, because the torque-assembled joint carries a residual torsional twist that reduces friction torque on the bearing face [S1][S2]. Practically, that means a locknut on an eccentric fixed tensioner does not need to be torqued to the same high value it would need on a conventional torque-tightened bracket: the joint is already twist-free, so a moderate torque on the nut combined with a hard washer face is enough to hold against the belt's transverse excitation.
Slip sequence and the 1.25 stiffness ratio
Slip at the nut bearing interface occurs before slip at the thread interface whenever the clamped component to bolt torsional stiffness ratio exceeds 1.25, which is the case in typical bolted joints [S1][S2]. For an eccentric fixed belt tensioner the clamped member is the tensioner body casting, which is usually much stiffer in torsion than the bolt, so the stiffness ratio sits well above 1.25 and the first failure mode is bearing-face slip, not thread strip-out.
Bearing-face slip is exactly what the locknut torque has to defeat, and the level of loosening torque required for combined nut-bearing-interface and thread-interface slip is the same for torque-assembled and tensioner-assembled joints [S1][S2]. That equivalence is the key engineering clue: a properly tightened eccentric fixed tensioner will not come loose from the bearing face, but it will still come loose from the threads if the locknut is under-torqued enough that the friction under the nut drops below the belt-induced transverse force. The design target is therefore to keep the bearing face above the slip threshold, which is achieved with the standard T = K F d sizing plus a locking assembly element.
Comparison of locknut options for an eccentric fixed tensioner
The four practical locknut arrangements for an eccentric fixed tensioner are plain hex nut, nylon insert (Nyloc) nut, serrated-face flange nut, and two-nut (jam nut) configuration. On a 10.9 grade M10 bolt with a typical K=0.20 and a 20 kN belt-stretch preload, the plain hex requires the highest applied torque, around 40 N·m, and offers the lowest vibration resistance; the nylon insert drops the required torque by roughly thirty percent to about 28 N·m and adds a polymer friction band; the serrated flange nut sits near 35 N·m and adds a high-friction toothed face that raises the bearing-face slip threshold; the jam-nut pair, with the upper nut torqued to roughly seventy percent of the lower nut value, locks both the thread and the bearing face and is the most repeatable in the field [S3].
For eccentric fixed tensioners, where the joint is twist-free after tensioner assembly and where the dominant failure mode is bearing-face slip above the 1.25 stiffness ratio, the serrated flange nut and the jam-nut pair give the best balance of low applied torque and high vibration resistance [S1][S2][S3]. The plain hex is acceptable only when a secondary method, such as a cotter pin through a castellated nut or thread-locking adhesive, is added. Nyloc nuts are reusable for only a handful of cycles (typically five to ten), so they suit low-service tensioners but not conveyor or industrial drives.
Sources and standards to apply
The Hess (2022) torque-equilibrium analysis in the Journal of Failure Analysis and Prevention is the primary source for the no-twist, five-times-loosening-torque, and 1.25 stiffness-ratio results quoted above [S1][S2]. The MIL-HDBK-60 threaded-fastener tightening handbook and the Bickford (4th edn., CRC Press) bolted-joint textbook are cited in the same paper and remain the standard references for tightening method selection [S1][S2].
For applied-torque sizing the nut-factor K=0.18–0.20 range, the friction scatter of plus or minus thirty percent, and the load-factor (typically 10 to 20 percent of external load reaching the bolt in a well-preloaded joint) are the established engineering values used in the comparison above [S3]. Engineers verifying locknut torque on an eccentric fixed belt tensioner should cross-check the calculated torque against the bolt-grade torque chart and then apply the method-dependent factor on top, rather than treating the chart value as final.
For a deeper look at the geometry-versus-torque tradeoff that shows up in the NEMA frame sizing problem, the NEMA 17/23/34/42 vs 42/57/86/110 frame size analysis follows a similar logic of separating a sizing label from the underlying torque capability, and for adjacent fastener-engineering selection the ball spline torque and thrust ratings guide applies the same K-factor reasoning to a different joint class.
Trackable signals for the next quarter: any vendor release of an eccentric fixed tensioner with an integrated toothed lockwasher, any field report of bearing-face slip on a tensioner-assembled joint under sustained transverse vibration, and any update to MIL-HDBK-60 or its successor on tensioner tightening method.
For the relevant spec sheets and selection criteria, see fixed gas detector.