Creep in Metals: Mechanisms, Stages, and Design Implications for High-Temperature Service

By WeldFabWorld June 11, 2026 37 min read
Creep in Metals — Mechanisms & Design Implications | WeldFabWorld
By WeldFabWorld Category: Welding Metallurgy Updated: March 2026

Creep in Metals: Mechanisms, Stages, and Design Implications for High-Temperature Service

Creep in metals is the time-dependent plastic deformation that occurs under sustained mechanical stress at elevated temperatures, and it is among the most consequential material behaviours that welding engineers, pressure vessel designers, and power plant operators must understand and design against. Unlike yield or fracture — which occur rapidly at a critical stress threshold — creep accumulates slowly and silently over thousands of operating hours, eventually leading to dimensional changes, loss of pre-stress, or sudden rupture without prior warning. Every high-temperature pressure component, from boiler headers and turbine rotors to refinery reactors and steam piping, is subject to creep throughout its service life.

The significance of creep goes beyond simple deformation. In welded joints, creep behaviour is complicated by the heterogeneous microstructure created by the welding thermal cycle. The base metal, the weld metal, and the various sub-zones of the heat-affected zone (HAZ) all have different microstructures and therefore different creep rates. Under sustained loading, stress redistribution occurs across these mismatched zones, concentrating creep damage in the weakest region. In Creep-Strength-Enhanced Ferritic (CSEF) steels such as P91 and P92 — now widely used in ultra-supercritical power plants — this leads to a well-documented and potentially catastrophic failure mode known as Type IV cracking, which occurs in the fine-grained intercritical HAZ of circumferential welds.

This article provides a complete engineering reference on creep in metals: the temperature threshold at which creep becomes relevant, the three stages of the creep curve and the physical mechanisms driving them, the mathematical models used for design and life assessment, the ASME code framework for creep-limited pressure design, and the critical implications for welded structures in high-temperature service. Whether you are preparing a welding procedure for P91 headers, reviewing fitness-for-service of aged boiler components, or studying for professional certification, this guide provides the technical depth you need.

Temperature Threshold: When Does Creep Begin?

Creep is not unique to extreme environments — it occurs at all temperatures in all crystalline materials, but at low temperatures the creep rate is negligibly small for engineering purposes. The classical criterion for the onset of engineering-significant creep is when the homologous temperature T/Tm exceeds approximately 0.4 (where both T and Tm are expressed in Kelvin). Below this threshold, dislocation mobility and atomic diffusion are too slow to produce measurable time-dependent deformation under normal engineering stresses and timescales.

Material Melting Point (Tm) Creep Onset (~0.4 Tm) Design Code Creep Range Start
Carbon Steel (P-1) ~1500°C (1773 K) ~437°C (710 K) 371°C (700°F) — ASME Sec. II-D
Low-Alloy Cr-Mo Steel (P-4/P-5A) ~1480°C ~420°C 371°C (700°F)
Austenitic SS (P-8, 304/316) ~1400°C (1673 K) ~397°C (670 K) 538°C (1000°F) — ASME Sec. II-D
P91 / P92 CSEF Steel (P-15E/F) ~1490°C ~423°C 371°C — creep governs above ~520°C
Nickel Superalloy (IN718) ~1260°C (1533 K) ~340°C Typically above 650°C for design
Lead (Pb) 327°C (600 K) ~-33°C (240 K) Creeps measurably at room temperature
Copper (Cu) 1085°C (1358 K) ~270°C Above 200°C for sustained loads
Key Point: The homologous temperature rule (0.4 Tm) is a guideline, not an absolute threshold. At very high stresses, creep can occur at lower homologous temperatures. ASME BPVC Section II-D determines actual design transition points by comparing time-independent and time-dependent allowable stresses and using the more conservative value at each temperature.

The Creep Curve: Three Stages of Creep Deformation

The standard characterisation of creep behaviour is the engineering creep curve, obtained by applying a constant tensile load to a specimen held at constant temperature and recording the resulting strain as a function of time until fracture. The characteristic shape of this curve reveals three distinct stages, each governed by different physical processes operating at the microstructural scale.

I
Primary Creep
Decreasing creep rate. Work hardening raises resistance faster than thermally-activated recovery. Transient, short-duration phase.
II
Secondary Creep
Minimum constant creep rate. Work hardening and recovery reach dynamic equilibrium. The dominant design-critical stage.
III
Tertiary Creep
Accelerating creep rate. Void nucleation, grain boundary cavitation, and necking lead rapidly to rupture.

Stage I — Primary Creep

Primary creep begins at the moment a load is applied at elevated temperature. The creep rate is initially high but decelerates continuously as deformation progresses. The controlling mechanism is work hardening: as the material deforms, dislocation density increases, dislocations tangle and pile up at grain boundaries and obstacles, and the material’s resistance to further deformation rises. At the same time, thermally-activated recovery processes (dislocation climb, cross-slip) begin to partially anneal the work hardened structure, but during primary creep the hardening rate exceeds the recovery rate. Primary creep is characterised by a rapidly falling creep rate and constitutes a relatively small fraction of the total creep life in most engineering applications.

Stage II — Secondary (Steady-State) Creep

Secondary creep represents the most important stage from a design perspective. The creep rate reaches a minimum constant value — called the minimum creep rate or steady-state creep rate (ἐ̇min) — because work hardening and thermally-activated recovery have reached a dynamic equilibrium. For every dislocation that becomes pinned, another is freed by thermal activation. The material deforms at a steady, constant rate that can persist for the majority of the component’s design life. The minimum creep rate is a direct function of stress and temperature and is the quantity measured in creep testing for design data. It is expressed mathematically by the Norton power law (described in detail in the mathematical models section).

Stage III — Tertiary Creep and Rupture

Tertiary creep is marked by a rapidly accelerating creep rate that leads quickly to fracture. Several concurrent damage mechanisms are responsible: the nucleation and growth of grain boundary voids and microcavities (grain boundary cavitation), the coalescence of these cavities into intergranular cracks, and macroscopic necking which increases the local true stress on the reduced cross-section. At the microstructural level, precipitate coarsening during tertiary creep reduces the pinning effect on dislocations and grain boundaries, further accelerating the deformation rate. In well-designed components, tertiary creep should not be reached during the intended design life, as significant dimensional changes and cracking may occur well before final rupture.

Design Criterion: The minimum creep rate (secondary stage) is the parameter of primary importance in ASME design. ASME BPVC uses the stress to produce a secondary creep rate of 0.01% per 1,000 hours (equivalent to 1% in 100,000 hours) as one of the criteria governing allowable stress in the creep temperature range. Engineers should not rely on tertiary creep as a safety margin in design — it represents the beginning of component failure, not a reserve.

Atomic-Scale Mechanisms of Creep Deformation

Creep deformation occurs by several distinct physical mechanisms, each dominant in a different region of the stress-temperature space for a given material. Understanding which mechanism governs your design conditions allows you to make rational material selection decisions and predict how a material will behave when conditions change. In practice, more than one mechanism can operate simultaneously, and transitions between mechanisms occur smoothly across boundaries in stress-temperature space (visualised in deformation mechanism maps).

Mechanism 1
Dislocation Glide
Dislocations move along slip planes with thermal activation helping them overcome barriers (precipitates, solute atoms). Dominant at relatively high stresses. Stress exponent n typically 3-5.
Mechanism 2
Dislocation Climb (Power-Law Creep)
Dislocations climb out of their slip plane by absorbing or emitting vacancies, allowing them to bypass obstacles. The dominant mechanism in the secondary stage of most engineering alloys. Requires high temperature for vacancy diffusion.
Mechanism 3
Nabarro-Herring Creep
Diffusion of vacancies through the bulk grain lattice from high-stress to low-stress regions. Dominant at very high temperatures and low stresses. Creep rate proportional to 1/d² (grain size). Stress exponent n = 1.
Mechanism 4
Coble Creep
Vacancy diffusion along grain boundaries rather than through the lattice. Dominant at lower temperatures than Nabarro-Herring creep. Creep rate proportional to 1/d³. Important for fine-grained materials including HAZ sub-zones in CSEF steels.
Mechanism 5
Grain Boundary Sliding
Grains slide relative to each other at grain boundaries. Contributes significantly to creep in the high-temperature regime. Leads to void nucleation at triple junctions and grain boundary facets, initiating Stage III damage. Enhanced in fine-grained materials.
Mechanism 6
Power-Law Breakdown
At very high stresses (relative to yield strength), the power-law relationship breaks down and creep rate increases more steeply with stress. Associated with obstacles to dislocation glide being overcome by stress alone rather than thermal activation.

The relative importance of these mechanisms depends critically on the material’s grain size, because diffusion-based mechanisms (Nabarro-Herring and Coble creep) and grain boundary sliding all become faster as grain size decreases (grain boundary area per unit volume increases). This is why engineering alloys intended for high-temperature service are often specified in a coarse-grained condition, and why the fine-grained HAZ sub-zones in CSEF steel welds are so much weaker in creep than the parent base metal.

Mathematical Models: Norton Power Law and Larson-Miller Parameter

Norton Power Law (Steady-State Creep Rate)

The most widely used model for the steady-state (secondary) creep rate is the Norton-Bailey power law. It expresses the minimum creep rate as a function of applied stress and temperature:

Norton Power Law — Steady-State Creep Rate
ἐ̇ = A · σn · exp(−Q / RT)

where:
ἐ̇ = steady-state (minimum) creep strain rate [s−¹ or h−¹]
A = material constant (determined experimentally)
σ = applied stress [MPa or psi]
n = stress exponent (typically 1 for diffusion creep; 3–8 for dislocation creep)
Q = activation energy for the dominant creep mechanism [J/mol]
R = universal gas constant = 8.314 J/(mol·K)
T = absolute temperature [K]

Interpretation of stress exponent n:
n = 1 → diffusion-controlled (Nabarro-Herring / Coble creep)
n = 3–5 → dislocation creep (climb-controlled) — most steels
n > 8 → power-law breakdown (high-stress regime)

The exponential temperature dependence means a 10–20°C increase in operating temperature can double the creep rate of a steel component.

Larson-Miller Parameter (Creep Rupture Life Prediction)

The Larson-Miller Parameter (LMP) is a practical engineering tool that allows short-duration, high-temperature creep rupture test data to be extrapolated to predict rupture lives at lower temperatures and longer service times — the conditions that actually govern pressure equipment design. It is based on the observation that time to rupture and temperature are related through an Arrhenius-type dependence, so that the same rupture result can be achieved either by increasing temperature or by increasing time proportionally.

Larson-Miller Parameter
P = T · (C + log tr) × 10−3

where:
P = Larson-Miller Parameter [K · (log h + C) × 10−³]
T = absolute temperature [K]
tr = time to rupture [hours]
C = material constant (≈ 20 for carbon and low-alloy steels; 15–25 depending on material)

Worked Example — Predicting Rupture Life:
Problem: A P22 header is tested to rupture at 650°C (923 K) in 1,000 hours at 100 MPa.
Step 1: Calculate LMP from test data (C = 20)
P = 923 × (20 + log 1000) × 10−³
P = 923 × (20 + 3) × 10−³ = 923 × 23 × 10−³ = 21.23
Step 2: At same stress (100 MPa), find rupture time at 565°C (838 K):
21.23 = 838 × (20 + log tr) × 10−³
20 + log tr = 21.23 / 0.838 = 25.33
log tr = 5.33 → tr = 105.33 = ~213,800 hours
Reducing temperature from 650°C to 565°C at the same stress increases predicted rupture life by over 200×.
Engineering Tip: The Larson-Miller method is an extrapolation tool and carries significant uncertainty when extrapolating beyond the temperature range of the underlying test database. ASME Section II-D allowable stresses are based on large, statistically-analysed databases of creep rupture data. For fitness-for-service assessments of components with mixed operating histories, consult API 579-1/ASME FFS-1, which provides validated procedures including the MPC Omega method for remaining life estimation.

Monkman-Grant Relationship

The Monkman-Grant relationship provides a simple empirical link between the minimum creep rate from the secondary stage and the total time to rupture. For many engineering alloys, the product of minimum creep rate and rupture time is approximately constant for a given material:

Monkman-Grant Relationship
ἐ̇min × trm = CMG

where CMG is the Monkman-Grant constant and m ≈ 1 for many steels
This means: if you measure ἐ̇min in a short test, you can estimate tr without running a full rupture test.

Application: Useful for in-service creep rate measurements to estimate remaining life.

Material Factors Governing Creep Resistance

Creep resistance is not an intrinsic property of a pure metal alone — it is strongly influenced by alloying strategy, microstructural condition, grain size, and heat treatment. The following factors are the primary levers available to alloy designers and fabricators for controlling creep behaviour:

Melting Point and Homologous Temperature

Materials with higher melting points have higher activation energies for diffusion and dislocation climb, which directly retards all thermally-activated creep mechanisms. Refractory metals (W, Mo, Ta, Nb) and nickel-based superalloys exploit this principle for the most demanding high-temperature applications such as gas turbine blades operating above 1000°C.

Grain Size

For diffusion-dominated creep mechanisms (Nabarro-Herring and Coble creep) and grain boundary sliding, the creep rate varies inversely with the square or cube of the grain size. Larger grains mean less grain boundary area per unit volume and a lower creep rate. This is why large-grain or directionally-solidified castings are used in applications like turbine blades where grain boundary sliding would otherwise govern. Conversely, the fine-grained HAZ sub-zones in CSEF steel welds are inherently more susceptible to creep by exactly this mechanism.

Precipitation and Dispersion Strengthening

Stable precipitates act as barriers to dislocation movement. In Cr-Mo steels (P-5A to P-15F), the creep strength comes primarily from M23C6 carbides and MX carbonitrides (where M is Cr or V and X is C or N) dispersed on grain boundaries and within grains. These precipitates pin dislocations and inhibit grain boundary sliding. The critical requirement is that these precipitates remain stable (do not coarsen or dissolve) throughout the intended service life — this is exactly what happens in the ICHAZ during welding (precipitate dissolution) and during long-term thermal exposure at the service temperature (Ostwald ripening / coarsening).

Solid-Solution Strengthening

Alloying elements in solid solution (Mo, W in steels; Cr in nickel superalloys) slow atomic diffusion and raise the activation energy for dislocation climb. This is why P91 (9Cr-1Mo-V-Nb) has dramatically better high-temperature creep strength than P22 (2.25Cr-1Mo) despite both being ferritic Cr-Mo steels — the combination of higher chromium content, vanadium, niobium, and nitrogen produces a much more potent and stable creep-resistant microstructure.

ASME Design Allowables in the Creep Temperature Range

ASME BPVC Section II Part D provides temperature-dependent allowable stress values for all listed pressure vessel and piping materials. As temperature increases into the creep range, the governing criterion shifts from static strength-based values (fraction of yield or ultimate tensile strength) to creep- and rupture-based values. The design stress value at any temperature is the minimum of all applicable criteria.

Design Criterion Basis Temperature Range Typical Governing Material
2/3 × Yield Strength (Sy) Short-term tensile / yield Below creep range Carbon steel at ambient to ~370°C
1/3 × UTS (Su) Static fracture prevention Below creep range All materials at lower temperatures
1% / 100,000 hr creep Creep rate criterion (0.01%/1000h) Creep range Carbon and low-alloy steels >370°C
80% of min. rupture strength at 100,000 hr Creep rupture prevention Creep range All materials in creep range
67% of avg. rupture strength at 100,000 hr Creep rupture (mean life) Creep range All materials in creep range
ASME Section II-D Reference: Allowable stress values (S values) for creep-range service are tabulated in ASME Section II-D, Table 1A (ferrous materials) and Table 1B (non-ferrous materials). The transition from strength-governed to creep-governed allowables typically occurs between 370°C and 550°C for steels, depending on the material group. These values are used directly in ASME Section VIII Division 1, ASME Section I (boilers), and ASME B31.1/B31.3 piping code calculations. For fitness-for-service of existing equipment with known creep damage, see API 579-1/ASME FFS-1.

In addition to design life considerations, the ASME approach recognises that operating temperature and stress history are rarely constant over a component’s life. Power plants cycle between full load and shutdown; refinery reactors experience temperature swings during catalyst regeneration. Cumulative creep damage is assessed using the linear damage rule (Robinson’s rule), analogous to Miner’s rule in fatigue:

Robinson’s Linear Creep Damage Rule
D = ∑ (ti / tr,i) ≤ 1.0

where:
D = cumulative creep damage fraction
ti = time spent at stress σi and temperature Ti
tr,i = time to rupture at stress σi and temperature Ti (from Larson-Miller or material data)
When D ≥ 1.0, the component has exhausted its creep life fraction

In practice, API 579-1 recommends D ≤ 0.1 for conservative life assessment in the absence of detailed inspection data.

Creep in Welded Joints: Stress Redistribution and Type IV Cracking

Welded joints in high-temperature service present a fundamentally different and more complex creep scenario than the homogeneous base metal. The welding thermal cycle creates a structure with at least four distinct zones — weld metal, coarse-grain HAZ (CGHAZ), fine-grain HAZ (FGHAZ), intercritical HAZ (ICHAZ), and subcritical HAZ (SCHAZ) — each with different creep properties. When this heterogeneous structure carries a sustained load at elevated temperature, stress redistribution occurs continuously as the softer zones creep faster and shed load to the stronger zones. This redistribution drives damage accumulation preferentially in the zone with the lowest creep ductility or the most rapid creep rate.

Classification of Creep Damage in Weldments

A widely used classification scheme for creep damage location in circumferential weldments was developed by Schuller, Hagn, and Wotischeck (1974) and remains the industry standard for damage assessment:

Type Damage Location Description Governing Materials
Type I Weld Metal Longitudinal or transverse cracks within the weld metal, remaining within weld metal Any material; weld metal undermatching
Type II Weld Metal / HAZ interface Cracks originating in weld metal but growing into adjacent HAZ Moderate-strength welds
Type III Coarse Grain HAZ (CGHAZ) Creep cavitation and cracking in the high-temperature HAZ adjacent to the fusion line Lower Cr-Mo steels (P-3, P-4, P-5A)
Type IV Fine Grain / Intercritical HAZ Premature creep failure in the soft, fine-grained zone at the outer edge of the HAZ. Life-limiting failure mode in CSEF steels P91, P92, P122, E911 (CSEF steels)
Type IV Cracking — The Primary Threat to P91 Welds: Type IV cracking in CSEF steel welds such as SA-335 P91 piping occurs in the intercritical HAZ, a narrow band of material that experiences peak temperatures during welding between the Ac1 and Ac3 temperatures (approximately 820–920°C for P91). In this zone, the prior austenite grain structure is refined, the strengthening MX precipitates partially dissolve, and coarser M23C6 carbides form on boundaries. After PWHT, this zone is left with a finer, weaker microstructure than the adjacent base metal or CG-HAZ. Under sustained high-temperature loading, creep voids nucleate on grain boundaries and facets in this zone, coalesce into cracks, and eventually cause premature rupture — often at lives well below those predicted by base metal data alone.

Implications for Welding Procedure and Design

Understanding weld creep behaviour has direct implications for how P91 welding procedures are qualified and applied in service:

The width of the ICHAZ is strongly influenced by welding heat input — high heat input creates a wider HAZ and a wider Type IV zone, while lower heat input narrows the HAZ but may increase hydrogen cracking risk during the pre-heat/interpass cycle. Optimal control of heat input within the qualified WPS range is therefore critical. Post Weld Heat Treatment (PWHT) at the correct temperature (730–800°C for P91) helps restore some of the precipitate structure in the ICHAZ but cannot fully replicate the base metal microstructure in this zone. PWHT temperature is critical: undertemperature PWHT leaves excess martensite hardness; overtemperature PWHT over-tempers the martensitic matrix and can dissolve beneficial precipitates.

Design strategies to mitigate Type IV failures include: avoiding placing welds in high-stress locations; specifying weld end preparations that minimise the HAZ contribution to the critical cross-section; using controlled heat input welding procedures; and implementing in-service inspection programmes that include replica metallography and hardness survey of weld heat-affected zones.

In-Service Monitoring and NDT for Creep Damage

Detecting and quantifying creep damage before it reaches the tertiary stage is essential for safe continued operation of high-temperature pressure equipment. Several complementary inspection techniques are used, each with different sensitivity to different stages of creep damage progression:

Inspection Method Creep Damage Stage Detectable Sensitivity Notes
Replica Metallography Stage I/II (early cavitation) High Can detect isolated creep voids. Requires polished surface; semi-invasive. Most reliable early-stage technique.
Hardness Survey (Vickers/Brinell) Stage I/II (softening) Medium Detects PWHT softening and over-tempering. Quick and cost-effective. Does not directly detect voids.
Ultrasonic Testing (UT / TOFD) Stage II/III (linked cracks) Medium Detects subsurface cracking and wall thinning. Cannot detect isolated voids reliably. TOFD preferred for weld creep cracking.
Phased Array UT (PAUT) Stage II/III High Improved resolution and coverage over conventional UT. Can characterise crack size and orientation. Required for most power plant weld inspections.
Dimensional Monitoring Accumulated strain (any stage) Low Measures total creep elongation or diameter change. Requires baseline readings from installation. Slow to detect local damage.
Magnetic Particle / Dye Penetrant Stage III (surface cracks only) Low Only detects surface-breaking cracks. Type IV cracks typically initiate mid-wall. NOT suitable as primary creep monitoring tool.
Inspection Strategy for P91 Welds: The industry best practice for in-service monitoring of P91 and P92 circumferential welds in power plant headers combines replica metallography at the outer weld surface with TOFD or PAUT volumetric inspection. Replica metallography identifies the onset of surface creep cavitation (classified A/B/C/D/E on the standard scale). PAUT/TOFD surveys detect sub-surface crack growth. Inspection intervals are defined based on remaining life calculations using the Larson-Miller parameter and accumulated operating temperature-time history from plant records.

High-Temperature Materials Comparison

Selecting the right material for a high-temperature pressure application requires comparing creep strength at the design temperature, long-term microstructural stability, weldability, and relevant code coverage. The table below summarises the key engineering alloys used in creep service and their practical design limits:

Material ASME P-No. Max Design Temp. Creep Mechanism Key Risk Typical Application
Carbon Steel (P-1) P-1 ~425°C Dislocation climb Creep onset above 370°C; spheroidisation Low-pressure boilers, general vessels
2.25Cr-1Mo (P22) P-5A ~565°C Dislocation creep + grain boundary sliding Type III HAZ cracking; hydrogen attack risk Hydroprocessing reactors, boiler headers
9Cr-1Mo-V (P91) P-15E ~620°C Dislocation creep + precipitation pinning Type IV HAZ cracking; soft zone from improper PWHT USC boiler headers, steam piping
9Cr-2W-Mo (P92) P-15F ~650°C Dislocation creep + W solid-solution Type IV cracking; Laves phase embrittlement Advanced USC power plant
TP304/316 Austenitic SS P-8 ~700°C Dislocation climb + grain boundary sliding Sensitisation; sigma phase embrittlement Chemical plant, heat exchangers, cryogenic
Alloy 800H (20Cr-32Ni) P-45 ~900°C Dislocation creep in FCC matrix Carburisation; metal dusting Steam reformers, ethylene crackers
IN625 (UNS N06625) P-43 ~980°C Solid-solution + gamma prime strengthening Stress relaxation cracking in aged condition Severe corrosion + high-temp environments

For more on material selection and P-Number groupings for high-temperature service, see our complete ASME P-Number reference guide. For the specific welding and PWHT requirements for P91 in power plant applications, see our detailed article on P91 welding requirements.

Recommended Reference Books on Creep and High-Temperature Metallurgy

Creep of Metals and Alloys — R.W. Evans & B. Wilshire
Comprehensive reference covering all creep mechanisms, the three-stage creep curve, and creep design approaches including Norton power law and Larson-Miller methods for engineering alloys.
View on Amazon
Metallurgy of Welding — J.F. Lancaster
Authoritative text on the metallurgical aspects of welded joints in all material groups, with detailed coverage of HAZ microstructure, Type IV cracking, and creep in Cr-Mo and CSEF steel welds.
View on Amazon
Welding Metallurgy — Sindo Kou
Definitive academic text on solidification, HAZ transformations, and elevated-temperature behaviour of weld metals. Essential reading for understanding Type IV cracking mechanisms in CSEF steels.
View on Amazon
Mechanical Behaviour of Materials — Norman E. Dowling
Covers creep, fatigue, fracture, and combined loading from an engineering design perspective, with worked examples for high-temperature equipment and pressure vessel applications.
View on Amazon
Disclosure: WeldFabWorld participates in the Amazon Associates programme (StoreID: neha0fe8-21). 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.

Creep Deformation Mechanism Domains

Different creep mechanisms dominate in different regions of the stress-temperature space. Engineers designing for creep service need to know which mechanism will govern their application so they can select appropriate materials and predict behaviour correctly. The following schematic deformation mechanism map illustrates the approximate boundaries for the most important mechanisms in a typical engineering steel:

Frequently Asked Questions

At what temperature does creep become significant in carbon steel?
Creep becomes significant in metals when the operating temperature exceeds approximately 0.4 times the absolute melting temperature (0.4 Tm in Kelvin). For carbon steel with a melting point of around 1500°C (1773 K), this threshold is approximately 437°C (710 K). In practice, ASME BPVC design codes begin applying time-dependent (creep) allowable stresses for carbon steel above approximately 371°C (700°F), and for austenitic stainless steels at higher temperatures consistent with their respective melting points. Lead, by contrast, creeps measurably at room temperature because its homologous temperature at 20°C exceeds 0.5 Tm.
What are the three stages of creep?
The three stages of creep are: (1) Primary creep, where the creep rate decreases over time as work hardening and strain hardening raise the material’s resistance to further deformation; (2) Secondary (steady-state) creep, where the creep rate reaches a minimum constant value as work hardening and thermally-activated recovery processes reach equilibrium — this minimum creep rate is the primary parameter used in engineering design; and (3) Tertiary creep, where the creep rate accelerates rapidly due to microstructural damage such as void nucleation, grain boundary cavitation, and macroscopic necking, leading rapidly to rupture. The minimum creep rate in the secondary stage is expressed by the Norton power law and is used to define ASME design allowable stresses in the creep temperature range.
What is the Norton power law for creep?
The Norton power law (Norton-Bailey equation) describes the steady-state creep rate as a function of applied stress and temperature: strain rate = A × sigma^n × exp(−Q/RT), where A is a material constant, sigma is the applied stress, n is the stress exponent (typically 3–8 for metals), Q is the activation energy for the dominant creep mechanism, R is the universal gas constant, and T is absolute temperature. The stress exponent n identifies which mechanism governs: n = 1 indicates diffusion-dominated creep; n = 3–5 indicates dislocation-controlled creep; n above 8 suggests power-law breakdown. For engineering steels in normal pressure equipment service, n is typically 4–7, confirming dislocation creep as the dominant mechanism.
What is the Larson-Miller parameter and how is it used?
The Larson-Miller parameter (LMP) is a time-temperature combination parameter that allows creep rupture data obtained at high temperatures over short durations to be extrapolated to predict behaviour at lower temperatures over longer service lives. The LMP is defined as: P = T × (C + log tr) × 10^−3, where T is the absolute temperature in Kelvin, tr is the time to rupture in hours, and C is a material constant (approximately 20 for most steels). By plotting applied stress against the LMP for a given material using test data, engineers can read off the predicted time to rupture at any combination of stress and temperature. It is widely used for establishing creep design allowables in codes such as ASME BPVC and for remaining life calculations in fitness-for-service assessments per API 579-1/ASME FFS-1.
What is Type IV cracking in welds and why is it dangerous?
Type IV cracking is a premature creep failure mode that occurs in the fine-grained or intercritical heat-affected zone (FGHAZ/ICHAZ) of welds in creep-strength-enhanced ferritic (CSEF) steels such as P91 and P92. During welding, the thermal cycle in this narrow HAZ sub-zone produces a fine-grained microstructure with coarsened or dissolved precipitates, reducing creep strength relative to the base metal. Under sustained high-temperature service, creep damage accumulates preferentially in this weaker zone, eventually forming a band of interconnected cavities that leads to premature fracture. Type IV cracking is particularly dangerous because it typically initiates subsurface (mid-wall) and is not detectable by surface NDT methods such as dye penetrant or magnetic particle testing. Periodic replica metallography and phased array ultrasonic testing are required for in-service monitoring. See our guide on P91 welding requirements for mitigation strategies.
How does ASME BPVC account for creep in pressure vessel design?
ASME BPVC Section II Part D provides temperature-dependent allowable stress values for all listed materials. For temperatures in the creep range, the allowable stress is governed by the lower of: 67% of the average stress to cause rupture in 100,000 hours, 80% of the minimum stress to cause rupture in 100,000 hours, or 100% of the stress to produce a creep rate of 0.01% per 1,000 hours (the secondary creep criterion). These values are established from large databases of creep rupture and creep rate data and are tabulated in ASME Section II-D for direct use in pressure vessel and piping design calculations. For fitness-for-service assessment of components with suspected creep damage, API 579-1/ASME FFS-1 provides specific creep assessment Level 1, 2, and 3 procedures.
What material design strategies improve creep resistance?
Several strategies improve creep resistance: (1) Using materials with high melting points to raise the temperature threshold for creep; (2) Promoting coarse grain structures to reduce grain boundary area and minimise grain boundary sliding and diffusion creep; (3) Adding precipitation-forming alloying elements (V, Nb, Ti, Mo) to create stable carbide or nitride precipitates that pin dislocations and grain boundaries against dislocation climb and recovery; (4) Using solid-solution strengthening elements (Cr, Mo, W) that slow diffusion and raise activation energy; (5) Directional solidification or single-crystal casting for turbine blades to eliminate transverse grain boundaries; and (6) Minimising exposure to the creep range by design (lower stress levels, reduced operating temperatures where feasible). The progression from P22 to P91 to P92 in power plant headers illustrates how progressive alloy design improvements advance each of these strategies.
What non-destructive testing methods detect creep damage in service?
Several NDT methods detect creep damage: (1) Replica metallography, where a surface impression is taken from a polished and etched weld area and examined under a microscope to identify creep voids and cavities — this is the most direct method for detecting early-stage creep damage and is standard practice for P91 welds in power plant maintenance; (2) Hardness testing and mapping, which detects softening in over-tempered HAZ regions; (3) Phased array ultrasonic testing (PAUT) and TOFD, which detect subsurface cracking and wall thinning; and (4) Dimensional monitoring to track accumulated creep strain. Dye penetrant and magnetic particle testing are only effective once cracks reach the surface and are insufficient as primary creep monitoring tools for CSEF steels where Type IV cracking initiates mid-wall. API 579-1/ASME FFS-1 provides fitness-for-service procedures for quantifying remaining life.

Related Articles on WeldFabWorld

P91 Steel Welding Requirements Complete guide to P91 (9Cr-1Mo-V) welding procedures, PWHT, and Type IV cracking prevention in power plant headers. Heat Treatment for Fabricators PWHT temperature ranges, holding times, and code requirements for all P-Number groups including CSEF steels in creep service. ASME P-Number Reference Guide Complete P-Number table from P-1 to P-15F with material classifications, typical grades, and PWHT requirements. Mechanical Testing in Welding Tensile, impact, hardness, and bend test procedures used in WPS/PQR qualification and high-temperature material evaluation. UG-84 Charpy Impact Testing When and how Charpy impact testing is required for pressure vessel welding procedures under ASME Section VIII Division 1. Welding Duplex Stainless Steels Microstructure, ferrite control, and welding procedure considerations for duplex and super-duplex stainless steels in high-temperature and corrosive service. Corrosion Guide: Types and Causes Comprehensive reference on oxidation, high-temperature corrosion, and other degradation mechanisms that interact with creep in elevated-temperature service. ASME Section IX Practice Quiz Test your knowledge of WPS/PQR qualification, essential variables, and P-Number classification for welding engineers and inspectors.