Skip to main content

Fatigue vs. Static Strength: Why Heavy Machines Crack at Low Stress

A heavy machine frame can pass a static strength check and still crack in service, because fatigue damage accumulates under repeated loads that may be well below the yield point. UTEC Industrial designs, engineers, machines, fabricates, and installs custom material handling systems for aerospace and heavy industry from its Spokane Valley, WA facility, integrating Allen-Bradley PLC and motion control with in-house CNC machining, heat treating, and stress relief. This article explains how fatigue differs from static failure, why welded details crack first, why stress range and cycle count govern fatigue life, what documented failures show, and how cracks are found and duty is tracked in service. Fatigue sits at the weld-fatigue link of the build chain, design → engineering → parts machining → fabrication → assembly → weld fatigue → stress relief → drives → controls → tuning → monitoring: the detail chosen at design sets the fatigue resistance, and the drives, controls, and monitoring downstream set how many stress cycles the detail sees and how early a crack is found.

What is the difference between static strength and fatigue strength?​

Static strength answers whether one application of the load will yield or break the part. Fatigue answers a different question. Roylance's MIT module defines fatigue as a process in which damage accumulates due to the repetitive application of loads that may be well below the yield point. It calls the process dangerous because a single application of the load would not produce any ill effects, and a conventional stress analysis might lead to an assumption of safety that does not exist.

The Steel Bridge Design Handbook makes the same point for fabricated steel structures. It says the fatigue process can take place at stress levels, calculated on the initial cross-section, that are substantially less than those associated with failure under static loading conditions, and that the usual condition that produces fatigue cracking is the application of a large number of load cycles. Among the civil structures it lists as susceptible to fatigue cracking are crane support structures. Shigley's machine-design text treats the two in separate chapters: its 2024 release has one chapter on failures resulting from static loading (Ch. 5) and another on fatigue failure resulting from variable loading (Ch. 6) (Roylance 2001, p. 1; Mertz rev. Grubb 2022, §1.1 p. 1; Nisbett and Budynas 2024, Ch. 5–6).

Why does steel crack below its yield strength?​

In what Roylance calls one popular view of fatigue in metals, the process begins at an internal or surface flaw where the stresses are concentrated. It starts as shear flow along slip planes; over a number of cycles that slip generates intrusions and extrusions that begin to resemble a crack, and the crack eventually turns to run transverse to the principal normal stress. The fracture surface records the history:

  • Slow growth. A region of slow crack growth is, in Roylance's words, usually evident as a "clamshell" around the initial flaw, and it often contains concentric "beach marks" where the crack was arrested for some number of cycles.
  • Fast fracture. The crack may become large enough to satisfy the fracture-mechanics criteria for rapid propagation, and that final phase produces the rough surface of fast fracture.
  • Post-mortem. Roylance writes that it is often possible to correlate the beach marks with specific instances of overstress.

Roylance writes that the modern study of fatigue is generally dated from the work of Wöhler, a technologist in the German railroad system in the mid-nineteenth century, who was concerned by the failure of axles after various times in service at loads considerably less than expected. A railcar axle is a beam in four-point bending; after each half turn its bottom becomes its top, so the stress at a point on its surface varies from tension to compression and back again, now known as fully reversed loading. The Steel Bridge Design Handbook adds that Wöhler's studies showed that stress concentrations and sharp angles in the axle configuration resulted in failures even though the stress in the material was well below its yield strength (Roylance 2001, pp. 1–2; Mertz rev. Grubb 2022, §1.2 p. 2).

What does an S-N curve show, and what is an endurance limit?​

An S-N diagram plots a constant cyclic stress amplitude against the number of loading cycles to failure. Roylance notes that millions of cycles might be required to cause failure at lower loading levels, and the cycle axis is usually plotted logarithmically. Three properties of the curve matter for a machine designer:

  • Endurance limit. In some materials, notably ferrous alloys, the curve flattens out, and below a certain endurance limit failure does not occur no matter how long the loads are cycled. For some other materials such as aluminum, no endurance limit exists, and the designer must arrange for the planned lifetime to be less than the failure point on the curve.
  • Scatter. Statistical variability is troublesome in fatigue testing: Roylance puts the need at perhaps twenty specimens at each of ten or so load levels to define the curve with statistical confidence, and fatigue lifetimes vary over orders of magnitude at a given stress.
  • Test time. It is generally impossible to cycle a specimen at more than about 10 Hz, and at that speed it takes 11.6 days to reach 10⁷ cycles.

The IIW recommendations for welded joints use the term fatigue limit for the fatigue strength under constant amplitude loading at a number of cycles large enough to be considered infinite. The Steel Bridge Design Handbook describes the bridge version of the idea: many bridge details exhibit a fatigue threshold such that, if all applied stress ranges are kept below it, the detail will not crack during its design life (Roylance 2001, pp. 3–4; Hobbacher and Baumgartner 2024, §1.3; Mertz rev. Grubb 2022, §3.1.3 p. 28).

Why do welded joints crack first?​

The Steel Bridge Design Handbook says the fatigue behavior of a fabricated steel structure is affected by pre-existing cracks or crack-like discontinuities, which most often occur at welded connections or other areas of stress concentration. It identifies where the smallest of them sit: small crack-like discontinuities, such as those at the intersection of the fusion line of the weld and the plate surface at the weld toe, are typically too small for detection, and their effects are incorporated into the design specifications. Broadly speaking, the handbook says, any mechanical detail has a better fatigue life than its equivalent welded detail. It points to the flaws and large stress concentrations at weld toes transverse to the direction of stress, to the fact that more flaws will be present when welding is used, and to inspection and repair of welded details being typically more difficult.

The IIW definitions use the weld toe as their example of a local notch: a localized geometric feature, such as the toe of a weld, that causes stress concentration; it does not alter the structural stress but generates a nonlinear stress peak. A hot spot is a point where a fatigue crack may initiate due to the combined effect of structural stress fluctuation and the weld geometry or a similar notch. For crane runway girders, Hectors, Chaudhuri and De Waele summarize the literature this way: most fatigue cracks in crane runway girders occur at the toes of the fillet welds that join the flanges and stiffeners to the web. Ricker's 1982 runway paper warns that welding the rail to the crane beam may cause cracking due to the fatigue stresses (Mertz rev. Grubb 2022, §1.1 p. 1, §1.2 p. 3, §2.1.3 p. 10; Hobbacher and Baumgartner 2024, §1.3; Hectors et al. 2022, §1; Ricker 1982, p. 188).

Why does stress range matter more than peak stress in a welded frame?​

The answer is residual stress. The Steel Bridge Design Handbook explains that, as a weld cools and contracts, the plate restrains it, leaving the weld and a relatively small volume of adjacent plate in tension, and that the magnitude of the tensile residual stress can reach the yield strength of the material. Built-up welded members carry high residual tensile stress at the flange-to-web junction, which is also where the flaws that start fatigue cracks are likely to be. For the usual condition, in which this initial stress is at or near the yield stress level, the handbook concludes that the stress range is the governing condition affecting fatigue crack growth, rather than the maximum applied stress or the stress ratio. In larger welded structures, it says, the residual stresses significantly reduce the effects of the mean stress and of the steel grade upon crack propagation for standard weldable structural steels, and it is generally agreed that stress range is the dominant stress parameter for fatigue design.

For mean stress in general, Roylance describes the Goodman diagram, in which a straight lifeline from the endurance limit on the alternating-stress axis to the ultimate tensile stress on the mean-stress axis gives the effective endurance limit for a combination of mean and alternating stress. As engineering reasoning, the practical consequence for a welded machine frame is that a dead load that raises the mean stress at a weld toe matters less than the live load that cycles it, and specifying a higher-strength steel does not by itself raise the fatigue resistance of the welded detail; the handbook states its conclusion for welded steel structures, and its own context is bridges (Mertz rev. Grubb 2022, §2.2.1 pp. 10–12; Roylance 2001, p. 4).

What are weld detail categories, and why don't bridge values transfer to machines?​

In the AASHTO approach, standard structural details are arranged into categories relative to their expected fatigue life, based on nominal stress ranges. The AASHTO bridge specification, as the Steel Bridge Design Handbook reproduces it, defines eight Detail Categories for fatigue: A, B, B′, C, C′, D, E and E′, with Category A the highest fatigue resistance and E′ the lowest. Over part of its range each category's curve is a sloping straight line on log-log axes, and the slope of all the design curves was found to be very close to −3.0, which was imposed on the equations; beyond a point that depends on the category, the line turns horizontal at the constant-amplitude fatigue threshold. The design curves are based on a 98 percent confidence limit, a lower bound of fatigue resistance.

The handbook gives one comparison that shows how far detailing moves life: at a stress range of 100 MPa (about 15 ksi), the average fatigue life of a cover-plated rolled beam would be 1 million cycles, while a built-up girder's would be over 8 million. It also says a small change in stress range produces a significant change in the nominal fatigue resistance. As UTEC Industrial's own arithmetic on a slope of 3, life varies with the inverse cube of stress range: raising the stress range by 10 percent cuts life to 1 ÷ 1.1³ ≈ 0.75 of its value, and halving life takes only a 26 percent rise in stress range (2^(1/3) ≈ 1.26).

These are bridge values and are attributed to the bridge context here; they are not machine-design allowables. For welded machinery, the IIW recommendations give fatigue resistance data for welded components made of wrought or extruded products of ferritic/pearlitic or bainitic structural steels up to a yield strength of 960 MPa, of austenitic stainless steels, and of aluminium alloys commonly used for welded structures, but state that they are not applicable to low cycle fatigue (defined in the text by limits on the nominal stress range of 1.5 times the yield strength and on the maximum nominal stress of the yield strength, with corresponding limits for shear stress) and are generally not applicable for corrosive conditions or for elevated temperature operation in the creep range. AWS D14.4/D14.4M:2019, Specification for the Design of Welded Joints in Machinery and Equipment, is described on the ANSI webstore as establishing acceptance criteria for classifying and applying carbon and low-alloy steel welded joints used in the manufacture of machines and equipment; the webstore labels that edition HISTORICAL. No fatigue category value from either document is quoted in this article (Mertz rev. Grubb 2022, §2.2.2 p. 12, §3.1.2 pp. 25–26, §3.1.3 p. 28, §3.1.4 p. 30; Hobbacher and Baumgartner 2024, §1.2; AWS D14.4/D14.4M:2019).

How many stress cycles does a continuously running machine accumulate?​

As engineering reasoning, the cycle count is set by the process rate and the operating hours, not by the rated load. The arithmetic below is UTEC Industrial's own, for an illustrative machine that applies one significant stress cycle to a weld detail every 6 seconds (10 cycles per minute):

Operating patternHours per yearCycles per year
One shift: 8 h/day, 5 days/week, 50 weeks2,0001,200,000
Two shifts: 16 h/day, 5 days/week, 50 weeks4,0002,400,000
Continuous: 24 h/day, 365 days8,7605,256,000

At 10 cycles per minute in continuous service the detail passes 1 million cycles, the handbook's average life for a cover-plated rolled beam at 100 MPa, in about 69 days (1,000,000 ÷ 14,400 cycles per day). The same machine on one shift takes about 10 months. The comparison is arithmetic on a bridge figure, not a prediction for any machine detail.

Heavy cranes show why continuous operation matters. Hectors, Chaudhuri and De Waele write that, compared to other types of civil structures, most load cycles on crane runway girders occur close to the maximum design loads (citing earlier work), that in the steel industry many plants rely on heavy-duty cranes that have continuous operational hours, and that this continuous heavy cyclic loading makes crane runway girders especially prone to fatigue damage. Ricker wrote in 1982 that few other structures suffer such an extreme range of stresses and as high an incidence of maximum loadings and fatigue as crane runways (Mertz rev. Grubb 2022, §2.2.2 p. 12; Hectors et al. 2022, §1; Ricker 1982, p. 181).

How is a variable duty load turned into a fatigue check?​

A machine whose load varies from cycle to cycle does not match a constant-amplitude S-N test. Three tools, each with stated limits, convert a duty history into a fatigue assessment:

  • Cycle counting. The IIW recommendations define cycle counting as the procedure of converting the history of variable amplitude loading into an equivalent spectrum or transition matrix, for example by the "Rainflow" or "Reservoir" methods.
  • Damage sum. The IIW define the damage sum as the ratio of fatigue damage sustained to fatigue damage required to cause failure, defined as the ratio of the number of applied stress cycles and the corresponding fatigue life at constant amplitude loading. Roylance presents Miner's law, Σ nᵢ/Nᵢ = 1, using absolute cycles. He adds that Miner's law should be viewed like many other material "laws", a useful approximation that might be accurate enough to use in design, but that its assumption of linear damage accumulation should be viewed skeptically and that it often fails to capture the essential physics of the fatigue process.
  • Crack growth. Once a crack exists, Roylance writes that a great deal of experimental evidence supports correlating the crack growth rate per cycle with the stress intensity factor range, da/dN = A(ΔK)^m. Roylance writes that the exponent m is often near 4 for metallic systems, and his Table 1 lists m = 3 and A = 10⁻¹¹ for steel; A and m depend on the material, environment, frequency, temperature and stress ratio.

As engineering reasoning, these tools are only as good as the load history fed into them. A duty spectrum assumed at design and never measured carries the same uncertainty into the damage sum that the scatter in S-N testing already carries into the fatigue curve (Hobbacher and Baumgartner 2024, §1.3; Roylance 2001, pp. 4–7).

What do documented fatigue failures in heavy handling equipment look like?​

Published failure records show fatigue alongside other causes, and the records state how certain each finding is:

  • Ladle-crane wire rope. Panda, Mohapatra and Dabbiru report a 32 mm, 1960 MPa grade galvanized steel wire rope on a hot-metal ladle crane that fractured after only six months of service. In the abstract, SEM fractography confirmed fatigue crack initiation, propagation and final overload fracture, and the authors attribute the premature failure to the combined effects of pre-existing surface defects, MnS inclusions, degradation of the galvanized coating, hardness variation and cyclic bending fatigue in service.
  • Crane boom collapse (federal record). An MSHA investigation of a crane boom that collapsed while lifting a crusher during demolition work found that the accident occurred because the crane was used beyond the manufacturer's design capacity: the total pick was 21.7 tons against a chart capacity of 16.2 tons for the configuration. The report records that the crane had reportedly been used as a dragline, which would subject a boom to repetitive loading and possible fatigue cracking of the lacings, and it found repaired lacing welds, poor-quality repair welds, corrosion and no maintenance logs. Investigators determined that either pre-existing fatigue-related cracks were present in lacing connections near the failure hinge or those lacings fractured during three pre-lift attempts. Fatigue was possible; the determined cause was overload.
  • Crane runway girder (assessment, not a failure). Hectors, Chaudhuri and De Waele assessed a welded overhead crane runway girder with a full-shell finite element model validated using strain measurements, and used load spectra based on real operational data to assess the fatigue properties of two critical joints.

As engineering reasoning, the MSHA case shows why a fatigue finding in a failure record needs to be read with its qualifier: repair welds, corrosion and overload all appear in the same report, and its conclusion says the accident occurred because the crane was used beyond the manufacturer's design capacity (Panda et al. 2026, Abstract; MSHA MAI-2009-05, Investigation and Root Cause Analysis; Hectors et al. 2022, Abstract).

How are fatigue cracks found during inspection?​

The Steel Bridge Design Handbook says steel structures are inspected for cracks both during fabrication, to limit the size of initial flaws, and during service, to ascertain any crack growth. Its laboratory observation sets the problem: as much as 80% of the fatigue life has been consumed by the time a fatigue crack emanating from an internal flaw reaches the surface and can be observed. It adds that criteria for in-service inspections are not yet well developed in design specifications, and that surface defects are subject to visual and magnetic-particle inspection (in its bridge context, where it adds that they are not a concern for bridges).

For overhead and gantry cranes in regular service, OSHA 29 CFR 1910.179 sets inspection intervals that, in its words, are dependent upon the nature of the critical components of the crane and the degree of their exposure to wear, deterioration, or malfunction. Periodic inspection, at intervals depending upon the crane's activity, severity of service, and environment, or as the rule specifically indicates, includes deformed, cracked, or corroded members, and worn, cracked or distorted parts such as pins, bearings, shafts, gears, rollers, locking and clamping devices. The rule is crane-specific, not a general machinery inspection rule. For one federal owner's weld inspection requirements on long-life cranes, see Powerhouse Crane and Gate-Hoist Duty Cycle Requirements. UTEC Industrial performs NDT and CMM inspection on the welded structures it builds (Mertz rev. Grubb 2022, §1.1 p. 1, §2.1.2 p. 6, §2.1.3 pp. 7 and 10; OSHA 29 CFR 1910.179-2016, (j)(1)(ii) and (j)(3)).

How can sensors and controls track fatigue duty on a running machine?​

A fatigue design rests on an assumed duty, and the controls are where the real duty can be measured. The Hectors runway study is one published example of measurement feeding an assessment: its global finite element model was validated using strain measurements, and its load spectra were based on real operational data. ISO 17359:2018 gives guidelines for the general procedures to be considered when setting up a condition monitoring programme for machines, and states that it is applicable to all machines; it gives guidelines, not requirements.

The intelligence layer below is engineering practice built on the definitions above, not a requirement of the cited sources:

  • Cycle and load logging. The PLC counts working cycles and records the peak load from a load cell or drive torque on each one, binning them into a measured load spectrum that can be compared with the design spectrum and cycle-counted (for example by rainflow) offline.
  • Strain at critical details. Strain gauges or optical fibre sensors near a weld detail the design identified as critical give stress ranges directly, for comparison with the design stress range.
  • Overload prevention. Load-cell or drive-torque limits interlocked with the motion stop a lift or move above rating, the condition the MSHA record gives as the reason its accident occurred.
  • Dynamic load control. Tuned acceleration and deceleration ramps on VFD or servo drives reduce the dynamic part of each stress cycle; as the bridge slope-3 arithmetic shows, a modest cut in stress range has a large effect on fatigue life.
  • Inspection prompts. Accumulated cycle counts trigger inspection of the critical welds, rather than the calendar alone.

UTEC Industrial, a Rockwell Automation Recognized System Integrator, integrates Allen-Bradley ControlLogix and CompactLogix controllers with VFD and servo drives, which is where cycle logging and load interlocks of this kind are built (Hectors et al. 2022, Abstract; ISO 17359:2018; Hobbacher and Baumgartner 2024, §1.3).

Where does fatigue sit in the design-to-monitoring chain?​

Each link of the build chain can add or remove fatigue life:

  • Design and engineering. As engineering reasoning, the detail sets the resistance. Ricker advised in 1982 that, because of the fatigue factor associated with intermittent welding, continuous welds (AISC stress category B vs. E) be considered for runway reinforcing members even where strength alone does not warrant them, and that reduced allowable stresses be used where cyclical loading would result in structural fatigue; that is 1982 AISC practice, not a current allowable.
  • Fabrication and weld quality. As engineering reasoning, weld-toe geometry and the micro-discontinuities whose effects, the bridge handbook says, are incorporated into the design specifications are set on the shop floor, and the MSHA record shows what poor repair welds add to a structure already under cyclic load.
  • Stress relief. Welding leaves tensile residual stress that can reach yield, as the bridge handbook says. What thermal and vibratory stress relief do to a welded frame before final machining is covered in Stress Relief for Machine Bases and Frames Before Final Machining and VSR Applications: Weldments, Machine Frames, and Oversize Assemblies. Whether relief changes the fatigue life of a given welded detail is, in this article, engineering reasoning only; the IIW recommendations treat stress ratio (§3.5.1) and improvement techniques (§3.5.3) as their own sections, and none of their values are quoted here.
  • Drives, controls, tuning, and monitoring. As engineering reasoning, these links set the stress range the detail actually sees and how early a crack is found, as the previous answer describes.

UTEC Industrial stress-relieves welded frames in a 6 × 10 × 17 ft, 1,800 °F car-bottom furnace or by automated vibratory stress relief before machining them (Ricker 1982, pp. 193 and 205; Mertz rev. Grubb 2022, §2.2.1 p. 11; Hobbacher and Baumgartner 2024, §3.5.1 and §3.5.3).

What should a specification for a fatigue-loaded machine state?​

A specification that states only the rated load leaves the fatigue design to assumption. As engineering practice drawn from the sources above, a buyer's specification for a cyclically loaded handling machine should state:

  • Duty. Operating hours per day and days per year, cycles per hour, and the expected spread of loads per cycle, not only the maximum.
  • Design life. The service life in years, from which the designer derives a cycle count.
  • Environment. Corrosive or elevated-temperature service, since the IIW recommendations say they are generally not applicable for corrosive conditions or for elevated temperature operation in the creep range.
  • Fatigue method. Which recommendations or code the fatigue check follows, with its edition, and that bridge-specification values are not used as machine allowables.
  • Inspection. Fabrication NDT, the critical welds identified for in-service inspection, and their access.
  • Monitoring. Cycle and load logging, overload interlocks, and how the measured spectrum is reported against the design spectrum.

Without the duty and life inputs, the fatigue check cannot be run, and a frame sized only for static strength might, in Roylance's words about conventional stress analysis, rest on an assumption of safety that does not exist (Roylance 2001, p. 1; Hobbacher and Baumgartner 2024, §1.2).

Related Articles

References​

  • Roylance, D. Fatigue (3.11 Mechanics of Materials module). Massachusetts Institute of Technology, 2001.
  • Mertz, D.; rev. Grubb, M.A. Steel Bridge Design Handbook, Chapter 12: Design for Fatigue, AISC Pub. B912-22. National Steel Bridge Alliance / American Institute of Steel Construction, 2022.
  • Hobbacher, A.F., Baumgartner, J. Recommendations for Fatigue Design of Welded Joints and Components, 3rd ed. Springer (IIW Collection), 2024. DOI 10.1007/978-3-031-57667-6.
  • Hectors K, Chaudhuri S, De Waele W (2022). "Fracture mechanics and hot spot stress-based fatigue life calculation: Case study for a crane runway girder." Fatigue & Fracture of Engineering Materials & Structures, 45(9), 2662-2675. DOI 10.1111/ffe.13729
  • Nisbett, K.J., Budynas, R.G. Shigley's Mechanical Engineering Design, 2024 release. McGraw Hill, 2024. ISBN 9781265472696.
  • Ricker DT (1982). "Tips for Avoiding Crane Runway Problems." Engineering Journal (AISC), 19(4), 181-205. DOI 10.62913/engj.v19i4.388
  • Panda, S., Mohapatra, J.N., Dabbiru, S.K. (2026). "Failure Analysis of a Prematurely Failed EOT Crane Wire Rope Used for Hot Metal Ladle Handling." Journal of Failure Analysis and Prevention, 2026. DOI 10.1007/s11668-026-02585-5.
  • MSHA MAI-2009-05: Report of Investigation, Fatal Machinery Accident, February 19, 2009. Mine Safety and Health Administration, 2009.
  • AWS D14.4/D14.4M:2019: Specification for the Design of Welded Joints in Machinery and Equipment. AWS, 2019.
  • OSHA 29 CFR 1910.179-2016: Overhead and Gantry Cranes. U.S. Department of Labor, 2016.
  • ISO 17359:2018: Condition Monitoring and Diagnostics of Machines — General Guidelines. International Organization for Standardization, 2018.

Ready to Discuss a Material Handling System?​

UTEC Industrial designs, engineers, machines, fabricates, and installs custom material handling systems for heavy industry, from the stress-relieved structure and drives to the Allen-Bradley PLC controls, tuning, and monitoring that run them, at its Spokane Valley, WA facility. Send UTEC the application, loads, and duty cycle to start a system review.

Request a Quote →

Questions? Call (509) 922-1832 or email sales@utec.co

Last updated on