ISO 21657:2017 Crane Limit State Design Method Explained

Fundamentals of Limit States Design

Kelude Heavy Industry strictly follows national standards in its production processes. The key difference between the Limit States Design (LSD) method introduced by ISO 21657:2017 and the Allowable Stress Design (ASD) method used in the GB/T 3811 Crane Design Standard is as follows: The ASD method keeps the maximum calculated stress from each load combination within the material's allowable stress. Allowable stress = material yield strength ÷ safety factor. The safety factor is fixed (e.g., 1.5), and no distinction is made between different load types (permanent vs. variable loads) regarding their differing impact on safety. The LSD method, by contrast, keeps the design load effects (internal structural forces) under each limit state within the structure's load-bearing capacity. Different load types (dead weight, suspended load, wind load, inertia force) are each multiplied by different partial safety factors (load partial factors), determined by their influence and safety significance. The load-bearing capacity (resistance) is also multiplied by a resistance partial factor (<1), providing a more refined reflection of the actual reliability of materials and structures. The LSD method classifies limit states into two categories: the Ultimate Limit State (ULS) — the structure must not fail, buckle, or overturn under maximum design loads, with the highest safety level requirement. The Serviceability Limit State (SLS) — the structure must not exhibit excessive deformation, vibration, or cracking under normal service loads, which would affect normal operation without causing safety incidents. The standard requires that crane structures satisfy both ULS and SLS requirements, with each verified separately, and the more stringent condition governs the design.

Load Partial Factors and Combinations

The load partial factors and combinations specified in the standard are: Permanent loads (G) — crane dead weight, counterweights, and fixed attachments (partial factor γG=1.2). Variable loads (Q) — lifting loads (suspended loads, including the lifting spreader and material dead weight) with partial factor γQ=1.5 (for lifting loads) or γQ=1.35 (for wind loads). Accidental loads (A) — such as collision loads and seismic loads (partial factor γA=1.0). Load combination principles — the most unfavorable combination is taken (the combination that produces the maximum internal forces in the structure). Basic combination (working condition) — ULS combination = 1.2 × permanent load + 1.5 × lifting load + 1.35 × wind load × combination factor, where the combination factor Ψ accounts for the probability of wind and lifting loads occurring simultaneously (0.6–0.7). Extreme combination (non-working condition) — ULS combination = 1.0 × permanent load + 1.25 × maximum wind load (non-working condition wind pressure); lifting loads are not considered in the non-working condition (hook at no-load). Fatigue load combination — used for fatigue life verification — load partial factor = 1.0 — employs a fatigue load spectrum (the load level with the highest frequency of use, not the maximum load). The standard also introduces the "reliability index" β — a higher β value corresponds to a lower probability of failure — with a target reliability of β=3.0–4.5 (corresponding to a probability of failure of approximately 0.001–0.00003). The β value depends on the consequences of structural failure (safety level): Safety Level I (severe consequences — such as ladle cranes handling molten metal and large tower cranes) — β≥4.0; Safety Level II (generally severe consequences — general-purpose cranes) — β≥3.5; Safety Level III (relatively minor consequences — small maintenance cranes) — β≥3.0.

Ultimate Limit State Verification

The ULS verification items specified in the standard include: Strength verification — cross-section normal stress σ≤f_y/γM (f_y is the steel yield strength, γM is the resistance partial factor (1.0–1.15, depending on material type and failure mode); shear stress τ≤f_y/(√3·γM)). Stability check — overall stability (in-plane and out-of-plane stability of beam-columns under bending moment, calculated using Euler's formula and equivalent moment factors) — local stability (plate width-to-thickness ratios must meet limits so that the compression flange and web plate do not undergo local buckling). Fatigue verification — detail fatigue class — classified by welded joint detail type (e.g., butt welds, fillet welds, bolt holes, base metal — each detail type corresponds to a different fatigue strength S-N curve) — fatigue load spectrum (stress amplitude Δσ at each fatigue load level and the corresponding number of cycles n, evaluated using the Palmgren-Miner linear cumulative damage criterion, with total damage D≤1.0). Stiffness verification (SLS) — maximum mid-span deflection of the main girder under rated load ≤ L/700–L/1000 (determined by crane service rating and positioning accuracy requirements). Fracture mechanics verification — for cranes in service, when cracks are detected at critical locations, fracture mechanics methods are used to calculate crack growth rate and critical crack length under specific loads, determining the number of safe operating cycles or time before the next inspection. The standard emphasizes that the accuracy of LSD calculations depends on accurate estimation of load and material parameters; designers must adequately account for parameter uncertainty and should adopt more conservative partial factors when necessary.

LSD vs. ASD: A Comparative Analysis

The standard's comparative analysis of LSD and ASD methods: Safety margin allocation differs — the ASD method concentrates the entire safety factor (1.5) on the resistance side (allowable stress = yield strength / 1.5), without distinguishing load types — all loads are treated equally. The LSD method distributes the safety margin between the load side (load partial factors >1) and the resistance side (resistance partial factor <1) — the more unpredictable the load (e.g., wind load), the higher the partial factor; the more controllable the load (e.g., dead weight), the lower the partial factor — a more scientific approach. The LSD method can yield more economical designs at the same target reliability because the probability of all loads in a combination reaching their maximum simultaneously is very low, which is reflected through the combination factor (Ψ) — design loads are closer to actual load levels than in the ASD method, allowing structural weight reductions of 5%–15%. Fatigue verification methods differ — the ASD method uses a single overall safety factor for fatigue checks, while the LSD method uses a damage accumulation approach based on S-N curves, providing more precise quantification of fatigue damage. The standard concludes that for large and critical cranes (high capacity, high work duty), the LSD method is more accurate and safer than ASD, and recommends its use in the design of important cranes. For small and medium general-purpose cranes, the ASD method offers simpler calculations and meets service requirements. The standard notes that international crane design is currently transitioning from ASD to LSD, and ISO 21657 provides a technical framework for updating national standards.

FAQ