ISO 6336 and AGMA 2001 both rate spur and helical involute gears by comparing tooth-bending stress and pitting (contact) stress against allowable limits, yet a side-by-side rating on the same manufactured industrial gear pair routinely produces different numbers because each standard defines geometry factors, load-distribution factors, and safety treatment differently [S1][S4].
The current ISO edition, ISO 6336:2019 (parts 1, 2, 3, 5 and 6), superseded the 2006 issue and superseded part 5 from 2003 [S5]. The corresponding U.S. standard is ANSI/AGMA 2001-D04, which covers pitting resistance and bending strength of spur and helical involute gear pairs [S7].
Calculation Framework and Common Ground
Both standards reduce the problem to a stress-versus-allowable comparison: tooth-root bending stress σF (or SF) must stay below an allowable σFP (or σFP), and Hertzian contact stress σH (or SH) must stay below σHP (or scP), with geometry factors, dynamic factor, load-distribution factor, and application/service factors applied multiplicatively [S4][S5].
ISO 6336 splits this across six parts: part 1 principles and general influence factors, part 2 surface durability (pitting), part 3 tooth-root bending, part 5 strength and quality of materials, and part 6 service-life calculation under variable load [S2][S5]. The basic bending formula in part 3 is the familiar σF = (Ft / (b·mn)) · (YF · YS · Yβ · YB · K) where the K-cluster covers application, dynamic, and load-distribution factors [S2].
AGMA 2001-D04 keeps the same physical input (tangential load Ft, face width b, transverse module mn) and the same Y–J (geometry) factor family, but assembles the load-side multipliers into the overloading-distribution factor Km and the dynamic factor Kv, plus the application factor Ko for driven/ driver uniformity [S7].
Where the Two Methods Diverge
The first divergence is in how contact-ratio effects, tip-relief, and edge contact feed into YJ; ISO 6336 applies separate YZ (zone factor) and Yδ (tooth-form factor) terms within a stricter tooth-root-bending equation, while AGMA 2001 uses a single combined geometry factor J which already bundles profile-shift, pressure-angle, and operating-stresses effects for bending [S2][S7].
The second divergence sits in load distribution. ISO 6336-1 breaks the load-distribution factor KHβ (or KFβ) into KHB = KHβ·KHα·Kv with explicit face-load and edge-contact components, and includes a Method C for equivalent mesh misalignment that has been shown to over- or underestimate the face load factor relative to finite-element analysis on helical gears [S3]. A 2026 paper (Mechanism and Machine Theory, vol. 227, article 106502, October 2026) demonstrates that an improved analytical model for elastic deformation, including shaft bending and gear-body torsion, matches finite-element predictions across helix-hand, power-path, gear-location, and mounting variations, outperforming Method C in several study cases [S3].
AGMA 2001 instead uses the elastic-deformation-based Kh (formerly Cmc) and the size, load-sharing, and thermal-distortion contributions under one combined factor mN, then converts to surface durability through the I-factor [S7]. On the same industrial gear, the resulting KHβ versus Kh spread alone is enough to shift the rated capacity by 10–20 percent, before material ratings are even considered [S1][S4].
Safety Factors and Material Allowables
ISO 6336 uses partial safety factors on load and on material (S·F·K_A versus σ-lim with material factor YST), and part 5 distinguishes through-hardened, case-hardened (nitrided, carburised, induction-hardened), and tempered steels with σ-H-lim and σ-F-lim endurance values tabulated against hardness, alloy content, and surface condition [S2][S5].
AGMA 2001-D04 keeps the AGMA-style bending-allowable stress sat and contact-allowable sc (or σc,all), set against material grades, hardness ranges, and reliability targets with AGMA quality numbers from Q6 to Q15, and an application factor Ko that varies from 1.00 (uniform power) to 1.75 (heavy shock) depending on the driven machine [S7]. The practical effect is that for a 58 HRC case-hardened pinion rated under both systems, AGMA's contact allowable at 10^9 cycles sits measurably above ISO's σ-H-lim for the same steel grade, while ISO's bending-side partial safety system tightens up at low cycle counts [S1][S4].
Decision Matrix: When to Use Which
Choose ISO 6336 (with part 6 for variable duty) when the gearbox ships to Europe, the OEM documentation must quote EN/ISO-compliant load capacity, the duty cycle is irregular with frequent overloads, and the input data includes quality grade ISO 1328, helix modification, and surface roughness Rz, because the standard's factor stack rewards that granularity [S2][S5].
Choose AGMA 2001-D04 when the gearbox is built to U.S. specifications, the customer requires AGMA quality numbers on the drawing, the application is steady-state (fans, conveyors, mixers) where Ko = 1.00–1.25 covers the duty, and the input data is conventional (no detailed contact-ratio or relief data), because the AGMA factor set is forgiving and the allowable tables are widely published [S7].
For a 10-module, 25-tooth, 20-degree pressure-angle spur pinion run at 1500 rpm against a 75-tooth gear, the rated tangential load differs by roughly 8–18 percent between the two standards, with AGMA higher on contact-pitting capacity and ISO higher on tooth-bending capacity in most reported comparisons [S1][S4]. This is the order of magnitude engineers should expect when they back-translate one rating to the other.
Limitations and Known Failure Modes
ISO 6336 is a cylindrical-involute-only standard; bevel, hypoid, worm, and planetary carriers are excluded and require ISO 10300, ISO 14521, and AGMA 6123 respectively [S2]. Inside its scope, the standard's Method C for equivalent misalignment can produce non-conservative face-load factors on highly loaded helical gears with low-shaft stiffness, and the 2026 finite-element comparison paper recommends augmenting the standard's elastic model rather than relying on Method C alone [S3].
AGMA 2001-D04 is also cylindrical-involute-only, and it assumes lubricant cleanliness and viscosity are sufficient; the standard does not natively model scuffing or micropitting, which ISO 6336-4 and ISO/TR 13989 cover separately [S7]. Both standards assume adequate gear-case rigidity, and both fail on polymer gears, which are handled in VDI 2736 and ISO 6336-3 annexes outside the core calculation.
Engineering Practice and Sourcing
The practical workflow at most European-OEM gearboxes now reads: ISO 6336 for the certified rating plate, AGMA 2001 as a secondary cross-check for U.S. customers, and FEM analysis on the housing-shaft-gear stack to resolve large face-load-factor disagreements [S3][S4].
For procurement engineers comparing a gear coupling or gear pump rating from a U.S. supplier against an EU data sheet, the rule of thumb is to request both ISO and AGMA numbers and check that the lower one governs, rather than accepting whichever the supplier publishes first.
Trackable next signals include the 2026 publication of the Method C improvement paper (Mechanism and Machine Theory, vol. 227, article 106502, October 2026) feeding into the next ISO 6336 maintenance cycle [S3], and the pending harmonisation discussion between AGMA 2001 and ISO 6336 on helix-load-distribution factors that has appeared in the gear-engineering press since 2019 [S5]. Engineers specifying new helical gear sets in 2026 should re-check the face-load factor against finite-element results before finalising the AGMA Q-number or ISO 1328 grade.
The underlying component specifications are covered under electronic load.
See also our earlier report, Forklift Stability Triangle: Load Center Math for Safe Capacity Derating.