A component may pass static tests without any problems and yet fail after a few months of operation. No overload, no operating error, no visible warning. This phenomenon is called material fatigue and is the main cause of mechanical failures during operation.
For a production manager or technical director, this means costs that are difficult to anticipate: unplanned downtime, warranty claims, emergency part replacements, and, in severe cases, safety risks. The good news: fatigue can be predicted and prevented as early as the design phase through fatigue analysis.
In this guide, we explain how this type of analysis works, what business decisions it supports, and how to integrate it into your product development process.
Why Do Components Fail Prematurely?
Most mechanical parts do not fail because the load has exceeded the material’s strength. They fail because they have been subjected to cyclic stress—millions of times—at stresses that appear to be safe.
Each loading cycle causes microscopic damage. This damage accumulates over time. At some point, a crack appears. The crack grows with each cycle until the remaining section can no longer bear the load. The final failure is sudden and often occurs without any visible warning signs.
A drive shaft, a robot arm, a welded frame, or a pressure vessel: all of them go through this process if the design did not take actual cyclic stresses into account. And the consequences aren’t just technical. A premature failure on a production line means hours or days of downtime, contractual penalties, and a loss of customer trust.
Conventional static analysis, based on safety factors, does not account for this risk. A component that is correctly sized from a static perspective may have a fatigue life ten times shorter than expected. That is why fatigue analysis is a separate step with its own methodology.
There is also a more subtle financial aspect. The cost of an in-service failure is not limited to the replaced part. It includes repair labor, emergency shipping, lost production during the downtime, and any collateral damage to adjacent equipment. In regulated industries, incident reports and additional inspections are also required. Taken together, these costs typically exceed the value of the part itself by a factor of ten or more. A timely fatigue analysis eliminates this risk for a fraction of the cost.
What Is Material Fatigue?
Fatigue is the progressive degradation of a material under repeated loads that are below the static breaking limit. The process consists of three stages: crack initiation, crack propagation, and final fracture.
Two fundamental tools are important for design decisions: S-N curves and stress concentration factors.
S-N Curves, in Brief
The S-N curve (also known as the Wöhler curve) relates the applied stress level (S) to the number of cycles the material can withstand before failure (N). It is a measure of each material’s durability.
The interpretation is intuitive. The higher the stress in the component, the lower the number of cycles to failure. Steels generally have a fatigue limit: below a certain stress level, the part can operate practically indefinitely. Aluminum alloys do not have this clear limit, so the service life must be calculated explicitly for each application.
What does this mean for you, as a decision-maker? It means that the choice of material is not based solely on static strength and price. Two materials with similar static strengths may exhibit completely different fatigue behaviors. The correct S-N curve, obtained from standardized tests, is the basis for any reliable prediction of service life.
Voltage Concentrators: Design Weaknesses
Fatigue cracks do not appear just anywhere. They appear where the part’s geometry locally increases stress: holes, keyways, abrupt changes in cross-section, threads, and weld beads.
This effect is quantified by the stress concentration factor. A connection radius that is too small can double or triple the local stress compared to the nominal value. The part appears to be correctly dimensioned on paper, but in reality it operates at much higher stresses precisely at the critical point.
The good news is that stress concentrators are a design issue, not a matter of fate. An increased radius, a smoother transition between cross-sections, or repositioning the hole can increase the service life several times over, without incurring additional material costs.
Types of cyclic loads: constant vs. variable amplitude
Not all cyclic requests look the same, and this difference completely changes the analysis methodology.
A constant-amplitude load consists of identical, repeating cycles. A shaft rotating at a constant speed under a constant load is the classic example. The analysis is straightforward: you read the service life from the S-N curve, with the necessary corrections.
In industrial reality, however, things are rarely so orderly. Production equipment starts up, stops, accelerates, and operates under varying loads. This is what we call variable-amplitude loading. Here, the analysis requires two additional steps: breaking down the load history into individual cycles using the “rainflow” counting method, and summing the fatigue damage caused by each cycle.
The “rainflow” method deserves a brief explanation, because it appears in every serious durability report. An actual load history looks chaotic: large and small peaks mixed together. The method sorts this chaos into complete cycles, each with its own amplitude and average stress. Only then can each cycle be evaluated on the S-N curve and accounted for as damage. Without this breakdown, the calculation cannot possibly be accurate.
Here’s a concrete example. A robotic arm in a robotic cell typically lifts 20-kilogram parts, but a few times per shift it also handles 60-kilogram parts. If the analysis considers only the usual load, the predicted service life appears reassuring. However, it is the rare but heavy cycles that take the heaviest toll on the structure’s lifespan. The load profile must account for them.
Why does this distinction matter at the decision-making level? Because an analysis based on a simplified load profile can seriously overestimate service life. A few infrequent high-load cycles can consume a disproportionate portion of a component’s service life. Actual operating data is more valuable here than any optimistic assumption. A key reference on this topic is the chapter on variable-amplitude loading published by eFatigue, a recognized technical resource in the field.
Fatigue at High vs. Low Cycle Counts: Two Methodologies
Fatigue analysis is divided into two categories, based on the number of cycles and the intensity of the load.
High-cycle fatigue occurs at relatively low stresses, below the yield strength, and over service lives exceeding 100,000 cycles. The material remains within the elastic range. This is typical for shafts, bearings, structures subject to vibration, and rotating components. The calculation method uses stresses and S-N curves.
Low-cycle fatigue occurs under intense loading that causes local plastic deformation. The service life is in the hundreds or thousands of cycles. This is the typical operating condition for components subjected to thermal cycling, repeated starts and stops, or controlled overloads. Here, the calculation is based on strains, not stresses, using the strain-life method.
Choosing the wrong methodology leads to inaccurate predictions. An experienced engineering partner identifies the correct operating mode based on your equipment’s actual operating profile, not on assumptions.
Factors Affecting Fatigue Life
The same part, made of the same material, can have very different lifespans, depending on several key factors.
Surface quality matters a great deal. Fatigue cracks almost always start at the surface. A ground surface lasts significantly longer than one that is roughly machined or corroded.
The average stress affects the result. A cycle that fluctuates around a tensile stress reduces the part’s service life more quickly than the same cycle centered at zero. Goodman or Gerber corrections adjust the calculation to account for this effect.
Surface treatments can extend service life. Shot peening, rolling, or nitriding introduce compressive residual stresses at the surface, which delay the initiation of cracks.
Welds are critical points. A weld bead combines stress concentration, residual tensile stresses, and potential internal defects. That is why standards include specific design rules for welded structures.
The operating environment accelerates degradation. Corrosion, high temperatures, and contact with chemicals drastically reduce fatigue strength compared to the catalog values.
Practical conclusion: The values in the tables are just a starting point. A proper analysis adjusts the material data to the actual conditions of your application.
How to Simulate Fatigue Before Production
Until a few decades ago, durability was tested through physical testing: you would build a prototype, subject it to cyclic stress for months on end, and hope it would hold up. The method works, but it is slow and expensive.
Numerical simulation has changed the game. The modern workflow looks like this: the 3D model of the part is analyzed using the finite element method, which calculates the stress distribution at every point in the geometry. If you want to better understand this step, we’ve explained the process in detail in our practical guide to finite element analysis (FEA).
Fatigue analysis is then applied to the FEA results: material curves, load profile, surface corrections, and average stress corrections. Specialized tools such as Ansys nCode DesignLife automate this process and generate durability maps: you can see exactly where on the part the first crack appears and after how many cycles.
The benefit for your business is immediate. You identify weaknesses in the virtual model, not on the equipment installed at the customer’s site. You compare design options in a matter of days, not months of testing. You reduce the number of physical prototypes to a minimum. And when physical testing is still necessary, you use it for final validation, not for exploration.
How to Interpret the Results: Lifespan Prediction
The result of a fatigue analysis is not simply a pass or fail verdict. It is a prediction of service life, expressed in cycles or years of operation, for each area of the part.
The theoretical basis for predicting fatigue under variable loads is the accumulation of damage. The Palmgren-Miner rule, the most widely used in practice, treats each load cycle as a draw on a total lifetime budget. When the sum of these deductions reaches the critical threshold, the part fails.
As a decision-maker, what should you look for in the analysis report?
Critical areas: where the most damage occurs and whether that location coincides with areas that are difficult to inspect or replace. Predicted service life, relative to the required service life: a comfortable ratio gives you a margin for variations in material and operating conditions. Sensitivity of the result: how much the prediction changes if the actual load exceeds the assumptions by 10 or 20 percent.
A word about the limits of prediction. Fatigue is a phenomenon with natural statistical variation: two identical parts, tested under identical conditions, do not fail after exactly the same number of cycles. That is why rigorous calculations rely on material curves defined for a given survival probability and safety factors applied to the service life. When you receive a report, ask what statistical assumptions underlie the final figure. The answer tells you a lot about the rigor of the analysis.
A well-documented prediction of service life becomes a selling point. You can use data to support the warranty period you offer, determine the appropriate level of spare parts inventory, and plan preventive maintenance based on realistic intervals—not on intuition.
Designing for Sustainability: Prevention Costs Less Than Failure
The economic rule of product development is simple: a problem discovered during the concept phase costs tens of times less than the same problem discovered during operation. When it comes to fatigue, this rule applies in full.
Designing for durability involves several specific practices. Generous fillet radii in high-stress areas. Gradual, rather than abrupt, changes in cross-section. Positioning welds outside high-stress areas. Selecting materials based on fatigue behavior, not just static strength. Specifying surface quality where it matters, not everywhere.
There is also a valuable side effect. Fatigue analysis shows you not only where the part is too weak, but also where it is unnecessarily oversized. Material in areas with almost no damage can be removed. I’ve covered this topic in detail in the article on structural optimization methods for weight reduction, which naturally complements the durability analysis.
The result: parts that are lighter, cheaper to manufacture, and have a controlled service life. Exactly the balance you’re looking for in a competitive industrial project.
When is the right time for analysis? Ideally, at two points in the project. The first review is conducted on the advanced concept, when the geometry is stable but changes are still relatively inexpensive. The second review is conducted on the final design, with the definitive material, surface treatments, and manufacturing technology. For equipment already in service, analysis serves a different purpose: it explains a recurring failure, determines whether a part can operate safely beyond its designed service life, or validates a change in operating conditions required by production.
What information do we need from you? A 3D model of the part or its working drawings, the material and any surface treatments applied, plus as detailed a description as possible of the operating conditions: loads, frequencies, daily operating cycles, and environmental conditions. The closer the load profile is to reality, the more valuable the prediction is for decision-making.
Applicable Standards and Regulations
Fatigue analysis cannot be improvised. High-risk fields have codified calculation rules, and compliance with them is often required by contract or by law.
For pressure vessels, the reference is the ASME Code, Section VIII, Division 2, which includes detailed fatigue evaluation procedures. Equivalent European standards have their own approaches, and the differences between them are not merely formal. A comparative analysis published by the British institute TWI shows that ASME, BS, and CEN rules can lead to different results for the same welded structure.
For civil and industrial steel structures, the fatigue section of Eurocode 3 applies. For the automotive industry and equipment, SAE guidelines and FKM recommendations from Germany are commonly used as references in design practice.
Key takeaway from this section: Choosing the right standard depends on the industry, the market, and the end customer’s requirements. An analysis report aligned with the appropriate standard will pass the customer’s technical audit and protect your position in the event of a dispute.
Next step: Turn sustainability into a competitive advantage
Fatigue strength analysis is not an additional design cost. It is a form of technical insurance: you pay a little during the design phase so you don’t have to pay a lot during operation.
The Centerline team performs FEA analyses for strength, durability, and fatigue of industrial components and equipment, ranging from the evaluation of an existing design to the complete optimization of a critical part. See what our engineering analysis and optimization services cover, or contact us directly via our contact page. We’ll let you know right away if your critical part needs a fatigue analysis and what data is required for an accurate evaluation.
Frequently Asked Questions About Fatigue Analysis
What is fatigue strength analysis?
This is the method used to estimate the service life of a mechanical component subjected to repeated stresses. The analysis combines the stress distribution in the part—typically obtained through FEA—with material data (S-N curves) and the actual load profile to predict where and after how many cycles a fatigue crack will occur.
Why does a part fail due to fatigue if it passed the static analysis?
Static analysis only verifies whether the part can withstand the maximum load applied a single time. Fatigue occurs at lower stresses, but these are repeated thousands or millions of times. Damage accumulates with each cycle, until a crack forms and propagates. That is why fatigue analysis is a separate calculation step.
What is the difference between fatigue caused by a high number of cycles and that caused by a low number of cycles?
High-cycle fatigue occurs at low stresses, within the elastic range, and over service lives exceeding 100,000 cycles; the calculation uses S-N curves. Low-cycle fatigue occurs under intense loading, with local plastic deformation, and over lifetimes in the hundreds or thousands of cycles; the calculation is based on deformation.
What factors reduce a component’s fatigue life?
The most significant factors are stress concentrators (holes, threads, sudden changes in cross-section, welds), poor surface quality, average tensile stress, corrosion, and high temperature. Some of these factors can be controlled through design and surface treatments, which can extend the service life several times over.
Can a fatigue assessment be performed without physical tests?
Yes, for the most part. FEA simulation, combined with specialized durability software, predicts critical areas and service life as early as the design phase. Physical testing remains useful for the final validation of critical components, but the number of prototypes and the duration of the testing program are significantly reduced.
