Steel cord conveyor belts are specified with a factor of safety of 6.67 against maximum working tension, whereas fabric (textile) belts are specified with a factor of safety of 10, per manufacturer guidance published 2026-05-28 [S1].
The split exists because a steel cord splice retains more than 90% of parent belt strength on average, while a multi-ply fabric splice typically retains 70% to 80%, so a lower numerical margin still delivers equal or better joint reliability [S2]. DIN 22101, the German standard that originated the convention in 1982, sets a default S₀ ≈ 8 for fabric belts and 6.7 for steel cord belts, with documented splice quality permitting further reductions to roughly 6.7 (fabric) or 5.4 (steel cord) [S2][S4].
Definition and Scope of the Belt Safety Factor
The belt safety factor SF (also written S₀) is a dimensionless ratio: SF = (breaking strength × belt width) / T1_max, where breaking strength is the belt rating in N/mm and T1_max is the maximum tight-side tension in newtons [S2]. It is not a generic material safety margin; it is a splice-driven multiplier applied to the calculated maximum force that the weakest link, which is almost always the splice, will see during steady operation and transient peaks [S4]. Because splices fail before the parent belt, raising the splice retention percentage is mechanically equivalent to lowering the required safety factor on the same duty. The 0.3% elongation of steel cord belts versus 4% for NN fabric belts also reduces dynamic tension excursions, which is the second physical reason that the steel-cord margin can be smaller [S1].
Standard Values Across DIN 22101, CEMA, and Manufacturer Data
DIN 22101 sets S₀ = 8.0 for fabric belts in steady-state operation and S₀ = 6.7 for steel cord belts, with splice quality verification allowing reductions toward 5.4 for steel cord and 6.7 for fabric [S2]. Forech lists 6.67 (steel cord) and 10 (textile) as the operating targets when redesigning a fabric line as steel cord [S1], while Sungda documents conventional practice in the 6.7 to 10 range tied to splice strength [S4]. Metso's Flexocord steelcord brochure uses a 6.7 to 8 band for steelcord belts and publishes a kN-rating table keyed to local belt tension T in kN [S7]. CEMA uses a belt service factor formulation that produces broadly similar rating steps for the same duty, though the underlying algebra differs from the European convention [S2].
A 2021 peer-reviewed study of textile belt failures in transition sections reports that the total safety factor in steady motion for textile belts typically remains within 9 to 12, an empirical range consistent with the upper end of the DIN 22101 envelope and the historical 10:1 nominal practice [S3].
Selection Criteria: When 6.67 Beats 10, and When It Does Not
The decision turns on splice retention, belt length, dynamic loading, and acceptable pulley diameter, not on raw safety-factor arithmetic. Steel cord at SF 6.67 is appropriate for long overland conveyors, high-tonnage duty, and installations where 90%+ splice efficiency is documented by the supplier; fabric at SF 8 to 10 is appropriate for shorter runs, more variable loading, and operations without a documented splice-fatigue program [S1][S2][S4]. The Forech conversion table quantifies the cost: a 2000/4 fabric belt at 20.8 mm and 25.3 kg/m² drops to a ST1400 steel cord belt at 14.5 mm and 23.4 kg/m², a 30% thickness reduction and 8% weight reduction at the same duty, because the lower SF permits a lower N/mm rating [S1].
The 0.3% elongation of steel cord versus 4% for NN fabric means smaller take-up travel, lower slip losses on the drive pulley, and tighter control of T1_max, which is what makes the lower numerical margin safe in practice rather than merely defensible on paper [S1]. On the flip side, fabric belts remain the default for shorter lines, applications with sharp impact loading at transfer points, and sites that prefer the mature multi-ply splicing workflow over vulcanized cord splicing crews.
Side-by-Side Comparison on Decision Criteria
On four criteria, the two constructions diverge sharply. Safety factor: 10 (textile nominal) vs 6.67 (steel cord) per Forech [S1]; 8 vs 6.7 per DIN 22101 [S2]. Splice retention is approximately 70–80% for fabric belts versus 90%+ for steel cord belts, per BisonConvey [S2]. Elongation at operating tension: 4% (NN fabric) vs 0.3% (steel cord) per Forech [S1]. Troughability improvement at 30° side roll angle: 33% per Forech troughability table, meaning the same belt width moves more material when running steel cord [S1]. Joint efficiency ratio between the two constructions is published as 3.3:1 in steel cord's favor by the same Forech source, and flex-fatigue resistance is 4:1 in steel cord's favor, while impact-energy absorption runs 3:1 against steel cord, the one criterion where fabric's stretchier carcass absorbs shock better [S1]. A useful working rule: if the conveyor is longer than roughly 1 km, the duty exceeds several thousand tonnes per hour, or the operator can document splice fatigue data, the steel cord 6.67 figure is the defensible pick; if impact energy at transfer points dominates the design, the 10:1 fabric figure should stay.
Real Use Cases and Failure Modes
Long overland coal and iron ore conveyors, port loading lines, and high-tonnage aggregate feeds are the canonical steel cord applications, and the 6.67 figure was developed for exactly these duty cycles [S1][S7]. Within those installations, the documented failure mode is not belt rupture at the parent carcass but cord damage accumulation: with safety factor at 6.67 or higher, it is not uncommon to see up to 10% of cords damaged or broken before reduced tonnage protocols are triggered, which is why condition monitoring rather than higher SF is the controlling practice [S9]. Fabric belts fail differently, with ply separation, splice peel under impact, and edge cuts dominating the downtime ledger, and the higher 10:1 SF exists to absorb the larger variability in retained splice strength [S1][S4].
The 2021 MDPI study adds a non-uniformity warning: replacing one belt construction with another whose elastic properties differ, even at the same nominal rating, can increase transition-section load non-uniformity by up to 100%, a load-peaking effect that can offset some of the static SF margin in real service [S3].
Limits, Caveats, and Sourcing
The 6.67 and 10 figures are manufacturer guidance, not universal constants; DIN 22101's 8 and 6.7 are the European default, and the actual selected number must be defended against the splice-fatigue test data for the specific belt being purchased [S1][S2][S4]. Lower factors (down to 4.7) are documented in the Sungda steel-cord test case where splice efficiency reaches 50%, but those values are not transferable to a belt whose splice retention has not been independently fatigue-tested [S4]. Woźniak's 9 to 12 textile envelope is the empirical upper bound; designs consistently below 9 should be justified by dynamic analysis rather than accepted on nominal rating alone [S3]. For comparison shopping on a given duty, the EP630/4 belt rating decoder walks through how a fabric belt's N/mm number interacts with the same SF logic.
Trackable next signals: a revision to DIN 22101's S₀ defaults, more peer-reviewed splice-fatigue datasets on ST-rated steel cord above ST3500, and any CEMA alignment of its service-factor formulation with the European S₀ convention. The belt conveyor kWh per tonne 2026 breakdown is the natural follow-on read for the energy side of the steel cord versus fabric choice.