Creep-Fatigue Interaction in High-Temperature Components

Components cycled at high temperature with sustained hold periods — power plant headers, turbine rotors, reactor internals — routinely fail sooner than either pure fatigue cycling or pure creep loading would predict alone, because the two damage mechanisms interact. This guide connects creep and fatigue testing into a single framework: the transgranular-versus-intergranular damage mechanisms, hold-time effects on cyclic life, the bilinear damage envelope used in design codes, and the time fraction, ductility exhaustion, and strain range partitioning methods used to predict life.

What Is Creep-Fatigue Interaction?

Creep-fatigue interaction occurs when a component experiences cyclic loading — mechanical, thermal, or both — while also spending significant time at sustained high temperature and stress within each cycle, most commonly a hold period at peak strain or load. Under these conditions, two distinct degradation mechanisms operate simultaneously and, critically, interact rather than simply adding independently: cycle-dependent fatigue damage from the repeated straining itself, and time-dependent creep damage accumulated during the sustained hold. The combined effect is generally more severe than either mechanism alone would predict, which is why creep-fatigue assessment is treated as a distinct discipline within high-temperature design codes rather than as a simple combination of separately evaluated creep and fatigue analyses.

Two Damage Mechanisms: Transgranular Fatigue vs Intergranular Creep Cavitation

Pure low-cycle fatigue damage in metals typically develops through transgranular mechanisms — slip band formation and crack growth through the interior of grains, largely independent of grain boundary condition. Creep damage, by contrast, develops predominantly at grain boundaries through the nucleation, growth, and eventual linkage of creep cavities, an inherently intergranular process. When both mechanisms operate together, fatigue crack growth accelerates as it increasingly follows creep-cavitated grain boundaries rather than propagating purely transgranularly, and the observed fracture surface typically shows a mix of transgranular and intergranular facets whose proportion shifts toward intergranular as the creep contribution to total damage grows.

Effect of Hold Time on Cyclic Life

Introducing a hold period at peak strain into an otherwise continuous-cycling low-cycle fatigue test consistently reduces cycles to failure compared with the same strain range cycled without a hold, for two related reasons. First, during a strain-controlled hold, stress relaxes over time as some of the elastic strain converts to inelastic (creep) strain at essentially constant total strain, which increases the effective plastic/inelastic strain range experienced by the material each cycle beyond what the nominal applied strain range alone would suggest. Second, the sustained stress and temperature during the hold itself drives creep cavitation at grain boundaries, directly seeding the intergranular damage described above. Life reduction from hold time is not unlimited, however: because stress relaxation is itself a time-dependent, decaying process, very long hold times can approach a saturation effect where further hold time adds comparatively little additional damage once stress has largely relaxed, an important nonlinearity that simple linear damage models do not always capture well.

The Damage Mechanism Map: Fatigue-Dominated, Mixed, and Creep-Dominated Regimes

The relative contribution of fatigue versus creep damage to total life depends strongly on strain amplitude and hold time, and can be organized into a damage mechanism map analogous in spirit to the fretting map used for contact fatigue. One useful way to express this is a damage ratio comparing the individually calculated creep and fatigue damage fractions.

Z = Dc / (Dc + Df)

  Z < ~0.33   :  fatigue-dominated regime (cyclic plasticity controls)
  ~0.33-0.67  :  mixed-mode regime (competitive interaction)
  Z > ~0.67   :  creep-dominated regime (time-dependent damage controls)

where Df = fatigue damage fraction, Dc = creep damage fraction

Short hold times and larger strain amplitudes tend to sit in the fatigue-dominated regime, where cyclic plasticity is the primary degradation driver and the interaction, while present, is comparatively modest. Long hold times, particularly at lower strain amplitude and higher temperature, push toward the creep-dominated regime, where time-dependent damage controls life and continuous-cycling fatigue data alone becomes a poor predictor of actual component life.

Life Prediction Methods

Three broad families of methods are used to predict creep-fatigue life, each with different data requirements and levels of mechanistic detail.

Time Fraction Rule (TFR)

Df + Dc  =  Σ(ni/Nfi)  +  Σ(tj/trj)   ≤  Envelope limit

where:
  ni/Nfi = applied cycles / cycles-to-failure, per cycle type
  tj/trj = hold time / creep rupture time at hold
                stress and temperature, per hold period

TFR is the approach used in ASME Subsection NH and RCC-MR, evaluating fatigue damage from continuous-cycling fatigue data and creep damage from independently determined creep rupture data, then combining both fractions and checking the result against the bilinear interaction diagram described below.

Ductility Exhaustion (DE)

The ductility exhaustion method, the primary approach in the UK R5 procedure, instead calculates creep damage as the accumulated creep strain divided by the material’s available creep ductility (strain to rupture) under the relevant conditions, rather than as a time fraction. A recognized limitation is that the calculated creep damage fraction can exceed unity for some materials and conditions, which produces conservative (sometimes overly conservative) predictions unless a modified ductility exhaustion approach is applied to correct for this behaviour.

Strain Range Partitioning (SRP)

Strain range partitioning, developed by Manson and Halford, takes a more mechanistic route by dividing the total inelastic strain range within a cycle into four possible components depending on whether straining is time-independent (plastic, P) or time-dependent (creep, C) in tension and compression: PP (fully plastic reversal), CP (creep in tension, plastic in compression), PC (plastic in tension, creep in compression), and CC (fully creep reversal). Each partition has its own characteristic strain-life relationship determined experimentally, and life for a general mixed cycle is predicted by combining the relevant partitions according to their proportion of the total inelastic strain range. SRP offers a more physically grounded description of creep-fatigue interaction than simple linear damage summation but requires substantially more experimental characterization to apply.

The Creep-Fatigue Interaction (Bilinear Damage) Diagram

Rather than treating a simple sum D_f + D_c ≤ 1 as universally acceptable, design codes such as ASME Subsection NH define a bilinear damage envelope in D_f-D_c space: two straight-line segments meeting at a material-specific intersection point, inside which the combination of calculated fatigue and creep damage is considered acceptable. The intersection point is placed well inside the simple D_f + D_c = 1 line for materials known to show strong creep-fatigue interaction, reflecting test data showing that such materials fail at combined damage fractions considerably below what independent, additive damage would suggest.

Governing Codes and Standards

Materials Most Affected

Testing Methods

Creep-fatigue interaction is characterized experimentally using strain-controlled low-cycle fatigue testing with a deliberately imposed hold period, typically in tension and sometimes also in compression, inserted at peak strain each cycle, following standardized procedures such as ASTM E2714 and E2760. Stress relaxation during the hold is recorded directly, and resulting cyclic life is compared against continuous-cycling (no hold) fatigue tests at the same total strain range to isolate and quantify the life reduction specifically attributable to the hold period, providing the experimental basis for both the time fraction and ductility exhaustion damage parameters used in design.

Design Implications

• Minimize unnecessary hold time at peak stress/strain and elevated temperature during operating cycles wherever process requirements allow, since hold time is often the single largest lever on creep-fatigue life.
• Recognize that continuous-cycling fatigue data alone will overpredict life for components with significant high-temperature hold periods; dedicated creep-fatigue data or code-based assessment is required.
• Pay particular attention to fine-grained HAZ regions in creep-resistant Cr-Mo and 9-12%Cr steel weldments, a location where creep-fatigue interaction and Type IV cracking both concentrate.
• Always use the material- and code-specific bilinear interaction diagram coordinates rather than a generic damage summation limit for final design acceptance.
• Account for the nonlinear, saturating nature of stress relaxation when extrapolating life predictions to very long hold times, since simple linear time-fraction extrapolation can be non-conservative or overly conservative depending on the material and regime.

Frequently Asked Questions

Recommended Reference Books

Disclosure: MetallurgyZone participates in the Amazon Associates programme. If you purchase through these links, we may earn a small commission at no extra cost to you. This helps support free technical content on this site.

Further Reading