Endurance Limit
Endurance limit is the maximum completely reversed cyclic stress a material can withstand for an effectively infinite number of cycles without fatigue failure, often approximated at about 10^6 cycles in testing. For steels, it is commonly estimated as roughly half the ultimate tensile strength, ranging from 35–60% of UTS, under laboratory conditions. This baseline value applies to idealized test specimens, not real machined parts.
In a CNC cell, endurance limit governs how a part behaves under repeated bending, torsion, clamp cycling, spindle loading, rotating shaft duty, and vibration-driven stress reversals. The practical goal is to keep the real component's alternating stress below the modified endurance limit, which is obtained by applying correction factors for surface finish, size, loading mode, temperature, and reliability. Machined surfaces with visible tool marks or rough Ra can suffer a significant fatigue penalty versus polished test coupons. This is why as-cut finishes, edge radii, and residual stresses are handled with adjustment factors rather than assuming the raw laboratory value. On the floor, this matters for spindle components, shafts, brackets, fixtures, saw arms, cyclically loaded hinges, lift mechanisms, drawer slides, and moving machine attachments. The critical fatigue detail is often the local notch geometry, fastener hole quality, keyways, thread roots, and undercut transitions rather than the bulk material.
- Ignoring surface finish penalties: A part machined with visible tool marks or poor Ra can have materially lower fatigue capacity than a polished or shot-peened part, so a design safe on paper may crack at the finish-affected surface.
- Using the base specimen value for a real part: Laboratory endurance limits come from standard specimens, so skipping Marin factors, notch effects, and reliability correction will overestimate life and underdesign the part for actual service conditions.
- Missing stress concentrators: Sharp shoulders, broached keyways, thread runouts, and unsupported holes create local peak stress far above nominal stress, causing failure even when average stress is below the quoted endurance limit.
Is endurance limit the same as fatigue strength?
Not always. For materials with a true fatigue plateau, endurance limit is the stress below which failure is not expected for very large cycle counts; for materials without a distinct plateau, engineers often use fatigue strength at a specified number of cycles instead.
Why is it usually stated for "completely reversed" loading?
Because standard endurance-limit data are commonly generated under fully reversed bending where the mean stress is zero, giving a clean baseline for comparison and correction factors.
Why do real parts need a "modified endurance limit"?
Because real components are affected by surface finish, size, load type, temperature, and reliability; these reduce the ideal specimen value and are combined through correction factors such as those used in Marin-style equations.