Spring Relaxation vs Fatigue: Two Failures, Two Different Design Fixes
Relaxation and fatigue are the two ways springs die, and they are opposites: relaxation is a spring that slowly loses force while sitting still under load, fatigue is a spring that snaps after many cycles of movement. If a spring no longer pushes hard enough, treat it for relaxation — lower the stress, preset it, cool it down. If a spring breaks into pieces, treat it for fatigue — lower the stress amplitude, remove surface damage, add shot peening.
Designers routinely apply the wrong fix because both failures show up as "the spring failed" and both get blamed on material quality. But the mechanisms are different enough that the cure for one can be useless — or harmful — for the other. Presetting, for example, is a powerful relaxation cure but does almost nothing for a fatigue crack that starts at a surface scratch. Learn to tell the two apart from the symptom, the part history and the stress state, and the design fix becomes obvious.
Relaxation: The Spring That Quietly Gives Up
Stress relaxation is the gradual loss of force in a spring held at a fixed deflection over time. Nothing breaks, nothing looks wrong — the free length may barely change, but the load at working height drifts down. The mechanism is micro-plastic creep: under sustained stress, atoms slowly rearrange and the elastic strain converts to permanent strain. Temperature accelerates it dramatically — a spring that holds load for years at 20°C can lose 20% of its force in weeks at 150°C.
| Driver | Effect on relaxation | Practical lever |
|---|---|---|
| Operating stress level | Loss rises steeply with stress | Design to lower % of tensile strength |
| Temperature | Loss roughly doubles per 30–40°C rise | Reduce temp or upgrade alloy |
| Time under load | Fast initial loss, then slowing | Preset to remove early settlement |
| Material | Nickel and stainless alloys relax less | Upgrade from carbon steel |
| Surface condition | Minor role | Shot peening adds modest benefit |
Takeaway: relaxation is controlled before the spring ever cycles — by stress level, temperature, presetting and alloy. The classic fix sequence is: check the operating stress against the material curve, add a preset operation so the spring is delivered already settled, and if it still drifts, move up one alloy family or cool the environment. A relaxation test — hold samples at working length and temperature for 100–1,000 hours and measure the load curve — is the only honest verification.
Fatigue: The Spring That Breaks in Pieces
Fatigue is crack growth under repeated loading. Each cycle puts the wire surface through tension and compression; a microscopic defect — a drawing scratch, a decarburized layer, a pit, a tool mark from coiling — becomes a crack that grows a little every cycle until the remaining cross-section snaps. Fatigue failures show a classic fracture face: smooth beach marks from crack growth and a rough final tear zone. The number that matters is stress amplitude (the swing between minimum and maximum stress), not the peak stress alone.
| Driver | Effect on fatigue | Practical lever |
|---|---|---|
| Stress amplitude | Life drops steeply as amplitude rises | Reduce deflection range or redesign rate |
| Mean stress | Higher mean stress shortens life | Lower preload where possible |
| Surface defects | Scratches act as crack starters | Shot peening, polishing, defect-free wire |
| Corrosion | Pits start cracks fast | Plating, stainless material, dry environment |
| Stress raisers | Hooks, sharp bends concentrate stress | Larger bend radii, redesigned ends |
Takeaway: fatigue is a surface and amplitude problem. The single most effective production fix is shot peening, which puts the surface in compression so a crack cannot open — it routinely multiplies fatigue life two to five times on dynamically loaded springs. The second fix is design: lower the amplitude by changing the rate or the working range, and enlarge bend radii at hooks and ends. If a spring has run millions of cycles and then breaks after a process change, look for what changed on the surface — a new plating bath, a rougher wire draw, a cheaper coil former.
Telling Them Apart on a Failed Part
Before changing anything, diagnose. The two failure signatures are different enough that a five-minute inspection decides the fix.
| Observation | Indicates relaxation | Indicates fatigue |
|---|---|---|
| Symptom | Loses force, still intact | Breaks, cracks, shortens suddenly |
| Part history | Static or slow cycling, often hot | Millions of cycles, dynamic service |
| Appearance | No fracture, maybe slight set | Beach marks + final tear zone on break |
| Load check | Force below spec at working height | Force often fine until the break |
| Location of damage | Whole body uniformly affected | Localized: surface scratch, hook bend, pit |
Takeaway: measure before you blame. A spring that lost 15% of its force but shows no crack is a relaxation case — redesign stress or material. A spring that snapped at a surface pit after 2 million cycles is a fatigue case — peen it, polish it, or lower the amplitude. Misdiagnosis is expensive: treating a fatigue failure with presetting alone wastes a production run, and treating relaxation with shot peening alone disappoints the customer twice.
Designing for Both at Once
Many real springs must survive both: a valve spring sits compressed for hours (relaxation) and also cycles millions of times (fatigue), often hot. The design sequence that covers both is: pick the material from the temperature and corrosion envelope first, set the stress from the fatigue amplitude requirement, add presetting for dimensional and load stability, then shot peen for fatigue. Verify with two tests — a load-retention test for relaxation and a cycle test to target life for fatigue. Worked together, these two disciplines are why a well-made spring outlives three cheap ones in the same machine.
At BQUQ the spring line treats the two failure modes separately from the first quotation: we ask about deflection range and cycles for fatigue, and about sustained load, temperature and required hours for relaxation. The compression spring and extension spring lines are set up to preset, heat-stabilize and shot peen to the application, and test reports cover both load retention and cycle results. Send the working envelope, the load, the temperature and the cycle count to sc@bquq.com or WhatsApp +86 13713157787, and the quotation comes back within 12 working hours with the failure mode already designed out.
Email sc@bquq.com or WhatsApp +86 137 1315 7787 with your PDF/DXF/STEP file. An engineer reviews it and replies with price, lead time and DFM notes on working days.
Q: How can I tell if my spring failed from relaxation or fatigue?
Check the part. Still in one piece but weak on the load tester means relaxation; broken with a two-zone fracture face (smooth beach marks plus rough tear) means fatigue. Correlate with history: hot and static points to relaxation, high cycle counts points to fatigue.
Q: Does shot peening fix stress relaxation too?
Partially. Peening adds compressive residual stress that slows relaxation slightly, but the dominant relaxation levers are operating stress, temperature and presetting. For a relaxation problem, preset and reduce stress first; treat peening as a bonus, not the cure.
Q: What stress level should I design to for long fatigue life?
As a rule of thumb for carbon and alloy spring steels, keeping the torsional stress amplitude below roughly 20–25% of tensile strength and the maximum stress below 45–50% gives lives in the millions of cycles, with peening extending that. Exact values belong on the material's fatigue diagram — ask the factory for the design point check.
Q: Why does presetting hurt fatigue performance?
Presetting plastically deforms the wire, which can create micro-damage at the surface and reduce the compressive residual stress benefit. For heavily cycled springs, shot peening after presetting restores surface compression. When both matter, the order is: coil, preset, peen, test.
Q: Can relaxation be recovered, or is the spring permanently damaged?
Permanent. Relaxation converts elastic strain into plastic strain; unloading and reloading does not restore the lost load. The only remedies are design changes — lower stress, better alloy, lower temperature — or replacing the spring. That is why relaxation testing before production approval is worth its cost.
Authored by the BQUQ Engineering Team. BQUQ is an ISO9001-certified source factory in Dongguan, China, running CNC machining, metal stamping, custom springs, heat sink and collet lines under one roof. Send drawings to sc@bquq.com or WhatsApp +86 13713157787 for a quote within 12 working hours. www.bquq.com
