Fatigue analysis services for parts that must last
Predict how many load cycles your part can survive, where the first crack will start and which design change adds the most life, using fatigue analysis built on the real stresses in your geometry.
A part can carry its maximum load comfortably and still fail after months in service. Repeated loading well below yield grows small cracks from the most stressed point, usually a fillet, a hole, a thread, a weld toe or a rough surface, until the remaining section can no longer carry the load. Static strength says nothing about how long that takes.
More about fatigue analysis
Fatigue analysis combines the stresses from FEA with the loading the part sees in service, and estimates life from the material’s S-N curve, the relationship between stress amplitude and cycles to failure. Mean stress matters as well as amplitude: a tensile mean shortens life, so a correction such as Goodman or Gerber is applied. Where loads vary, cycles are counted and damage is summed with Miner’s rule, D = Σ nᵢ/Nᵢ, with failure expected as D approaches 1. Surface finish, size and notch sensitivity adjust the curve to the actual part.
Life is very sensitive to stress. For a welded detail with an S-N slope of 3, cutting the stress range by 10 % extends life by about 37 %. That is why the changes F-CAD®’s design engineers propose, such as a larger radius, a relocated weld or a smoother transition, can be worth more than a thicker section. Life is assessed to the fatigue method or standard you specify.
Typical parts: shafts and axles, welded frames and chassis, suspension and linkage arms, lifting equipment, pressure-cycled housings, springs, and brackets on vibrating machinery.
What a fatigue study answers
How long will it last?
Predicted life in cycles, hours or years of service for the load history you specify, set against the design life you need.
Where will it crack first?
The locations with the highest fatigue damage, which are not always the locations of highest static stress.
Does it have an unlimited life?
For steels with an endurance limit, whether stress amplitudes stay below it. Aluminium alloys have no true endurance limit, so their life is always finite and must be stated.
How much does the mean stress matter?
The effect of preload, self-weight, pressure or an interference fit on life, which can be large for a part cycling in tension.
Which load events do the damage?
Damage broken down by event, so you know whether start-ups, occasional peaks or everyday cycles are using up the life.
What change adds the most life?
Geometry, weld, material and surface treatment options compared on the life they gain.
What we analyse
Stress-life (S-N) fatigue
High-cycle fatigue life from S-N data for the material, adjusted for surface finish, size and notch effects.
Mean stress correction
Goodman, Gerber or Soderberg corrections for parts carrying a static load, preload or pressure beneath the cycling load.
Variable amplitude loading
Life under a measured or specified load history in which cycle size and mean stress change from one cycle to the next.
Welded joints
Fatigue of welded details using the weld classifications and S-N curves of the code or standard you specify.
Combined load events
Operating, start-stop and occasional loads, each with its own number of cycles, brought together in one life estimate.
Crack initiation sites
The fillet, hole, thread or weld toe that limits the life, identified on the part.
Design for life
Radii, transitions, weld position, material and surface treatments such as shot peening compared on predicted life.
Before you start
What to send us
- CAD of the part or assembly
- Loads for each event, and how often each occurs over the design life
- Measured load or strain data, if you have it
- Material grade, and S-N data if available
- Surface finish, heat treatment and any surface treatment
- Required design life, and the code or standard you specify
What if we do not have a load history?
We can work from defined load cases with an estimated number of cycles for each, which is common at concept stage. Measured data later sharpens the estimate.
Can fatigue analysis replace a durability test?
No. It shows whether the design is likely to reach its life and where it is weakest, so physical testing starts from a stronger design. Testing remains the way to demonstrate life.
Why is our part cracking below its yield strength?
That is the signature of fatigue. Cycling stress, often concentrated at a notch or weld, can start a crack far below yield, and the analysis shows which feature and which load event is responsible.
Do welds need different treatment?
Yes. Residual stress and local weld geometry make the fatigue strength of a welded joint largely independent of the parent steel’s grade, so welds are assessed with weld-specific S-N curves from your chosen code.
Tell us how long it has to last
