Creep Rupture vs Stress Relaxation — PatSnap Eureka
Creep Rupture vs. Stress Relaxation in High-Temperature Bolted Joint Design
Two distinct time-dependent failure modes govern high-temperature bolted joints in power generation, aerospace, and process industries. Creep rupture ends in structural fracture; stress relaxation ends in preload loss and leakage. Understanding the boundary condition that separates them is the foundation of safe joint design.
One Mechanism, Two Boundary Conditions — and Very Different Consequences
Both creep rupture and stress relaxation are thermally activated and share the same underlying creep constitutive behavior of the bolt material. The critical difference lies in the boundary condition imposed on the joint. Creep rupture operates under load-controlled (constant stress) conditions: the bolt is the load-bearing element, and as it accumulates inelastic strain through primary, secondary, and tertiary creep stages, it terminates in structural fracture. Stress relaxation operates under displacement-controlled (strain-controlled) conditions: the joint geometry is externally constrained, elastic strain converts progressively to creep strain, and the stored elastic energy — the preload — dissipates.
The engineering consequence of this distinction is profound. Stress relaxation governs leak-before-break scenarios in flanged connections, where the failure criterion is loss of sealing integrity rather than immediate fracture. Creep rupture governs structural integrity of load-bearing studs, where fracture is material-consuming and irreversible. Conflating the two leads to either over-conservative or non-conservative designs. According to the 2016 bolt assembly optimization study, the creep curve (constant load, strain output) and the stress relaxation curve (fixed displacement, stress decay output) are formally distinct, and design calculations must treat them separately. Learn more about IP analytics for engineering materials or explore the PatSnap platform for cross-domain life prediction data.
The three classical creep stages — primary (decelerating strain rate), secondary (quasi-steady minimum strain rate), and tertiary (accelerating strain rate ending in fracture) — are relevant to both phenomena, but the damage metric differs. For creep rupture, the key measurable is elongation of the bolt shank relative to rupture elongation. For stress relaxation, the key measurable is residual clamping force relative to minimum design preload. Standards bodies including ASME and VDI have developed separate calculation frameworks for each mode, reflecting this mechanistic split.
Creep Rupture vs. Stress Relaxation: Design Parameter Comparison
A structured comparison of the two failure modes across the parameters that matter most for bolted joint design, inspection, and remediation.
| Parameter | Creep Rupture | Stress Relaxation |
|---|---|---|
| Boundary Condition & Driving Physics | ||
| Boundary condition | Load-controlled (constant stress) | Displacement-controlled (fixed total strain) |
| Driving variable | Sustained tensile load on bolt shank | Fixed elongation from initial tightening preload |
| Failure criterion | Structural fracture of bolt (irreversible) | Preload drops below minimum design threshold |
| Damage Evolution & Measurable | ||
| Damage progression | Primary → secondary → tertiary creep stages; void nucleation at grain boundaries | Elastic strain converts progressively to creep strain; preload decays |
| Key measurable | Bolt shank elongation rate (Larson-Miller / life-elongation parameter) | Residual clamping force; retightening torque required |
| Fracture morphology | Intergranular; main crack perpendicular to bolt axis; secondary 45° cracks from torsional assembly stress | No fracture — gradual load decay only |
| Engineering Consequence & Remediation | ||
| Engineering consequence | Structural failure; bolt must be replaced | Joint leakage, loosening, or loss of sealing integrity |
| Primary application domain | Turbine casing studs, pressure vessel bolting above 500–600°C | Bolted flanged connections, gasket sealing joints in process industry |
| Key remediation strategy | Remaining-life assessment via elongation measurement (Mitsubishi Power methodology) | Retightening at primary-to-secondary creep transition (2016 optimization study) |
| Geometric design lever | Multiaxial stress correction at thread roots (Hitachi, 2014) | L/N ratio optimization — larger free shank length tolerates more creep before preload exhaustion (British Thomson-Houston, 1957) |
Four Proven Approaches to Managing Both Failure Modes
The patent and literature dataset identifies distinct engineering strategies for each failure mode, with some approaches addressing both simultaneously.
Retightening at the Primary-to-Secondary Creep Transition
The 2016 bolt assembly optimization study demonstrates that applying a second tightening at the inflection point from rapid to steady-state creep produces a new, improved stress relaxation curve with a longer residual preload life. This defines an optimal maintenance interval tied to material creep behavior rather than calendar time — a critical distinction for high-temperature flange management. The interval is material- and temperature-dependent, creating space for proprietary material characterization services.
Operational — maintenance intervalL/N Ratio Optimization for Creep Strain Tolerance
British Thomson-Houston Company (1957, GB) analytically derived that the ratio of free shank length (L) to loaded nut engagement length (N) governs how much creep strain the bolt can accumulate before relaxation causes load to drop below design threshold. Designs with larger L/N ratios tolerate more creep strain before preload is exhausted. This foundational geometric insight remains embedded in modern design standards and underpins current bolt proportioning rules for high-temperature service, as recognized by bodies such as ISO.
Geometric design — L/N ratioShape Memory Alloy Bolts for Active Preload Recovery
Efremov, Anatoly (2008, US) proposes manufacturing bolt elements from shape memory alloys (SMAs) that, when heated to operational temperature, attempt to recover their memorized shape. Because they are constrained by the joint, they generate reactive stresses in the direction opposite to relaxation-induced elongation, effectively maintaining or recovering clamping load through “negative creep” effects. The patent has since lapsed, suggesting commercial adoption did not follow, but the approach remains mechanistically sound and is referenced by NIST materials research on active fastener systems.
Active compensation — SMA boltsLife-Elongation Rate Method for Remaining-Life Management
The Mitsubishi Power patent family (US 2010, US 2012, AU 2011, CA 2013) establishes a unified methodology: (1) conduct creep tests on bolt material to determine life-elongation rate as a function of a time-temperature parameter (Larson-Miller type); (2) create a lifespan-assessment diagram; (3) measure in-service bolt elongation; (4) map to remaining life. The method is specifically noted to be applicable to high-alloy bolts with limited microstructural changes, where conventional metallographic techniques have low discrimination. This represents a defensible and well-protected IP position across US, AU, and CA jurisdictions. See also PatSnap IP analytics for competitive landscape mapping.
Quantitative remaining-life managementFiling Activity and Assignee Concentration: 1957–2025
The dataset of 14 patent records reveals a field with mature conceptual foundations (1957), a mid-phase assessment tooling push (2008–2014), and a current wave of coupled creep–fatigue modelling from Chinese institutions (2020–2025).
Assignee Patent Activity (Records in Dataset)
Mitsubishi Power holds the largest block with 5 records focused on turbine bolt remaining-life assessment. Chinese institutions are the most active recent filers (2020–2025).
Innovation Wave Timeline: Three Eras
The field progressed from geometric design foundations (1957) through active assessment tooling (2008–2014) to coupled creep–fatigue digital modelling (2020–2025).
Mitsubishi Power’s Unified Creep Rupture Remaining-Life Framework
The Mitsubishi Power patent family (US, AU, CA, 2010–2013) establishes a four-step methodology for quantitative remaining-life management of high-temperature turbine bolts subject to creep rupture.
What the Patent Landscape Tells Practitioners and IP Teams
Four evidence-based strategic observations derived from the 14-record patent and literature dataset, relevant to design engineers, IP strategists, and procurement teams.
Design Codes Must Distinguish Boundary Conditions Explicitly
The mechanistic difference between creep rupture (load-controlled, ends in fracture) and stress relaxation (displacement-controlled, ends in preload loss) requires separate design criteria. Stress relaxation governs leak-before-break scenarios in flanged joints; creep rupture governs structural integrity of load-bearing studs. Conflating the two leads to either over-conservative or non-conservative designs.
Retightening Interval Should Be Derived from Material Creep Behavior, Not Calendar Time
The 2016 bolt assembly optimization study demonstrates that retightening at the inflection point of the creep curve produces the maximum preload life extension. This interval is material- and temperature-dependent, creating space for proprietary material characterization services that can differentiate maintenance contracts.
Grain Boundary Cavitation, Ductility Exhaustion, and the Microstructural Basis of Creep Rupture
Stress relaxation is a macroscopic phenomenon — preload decay — with no inherent microstructural damage accumulation in the bolt itself. Creep rupture, by contrast, involves nucleation, growth, and coalescence of grain boundary voids, ultimately producing intergranular fracture. This microstructural distinction is what makes the two failure modes fundamentally different in their damage irreversibility.
In ferritic heat-resistant steels (P91, P92, Grade 91, Grade 92) — the dominant bolt materials for ultra-supercritical (USC) turbine applications — creep rupture proceeds by void nucleation at grain boundaries, particularly at carbide precipitate clusters and triple-point junctions. Research published in 2019 demonstrates that prior residual stress — directly analogous to assembly preload — accelerates cavitation and reduces rupture life at 650°C for P92 steel. This finding has direct implications for bolt-stretching versus torque-tightening protocols: assembly methods that minimize residual torsional stress extend creep rupture life. Organizations including EPRI have published guidance on this topic for power plant operators, and PatSnap’s materials intelligence tools can support alloy selection for high-temperature service.
Notch geometry — analogous to bolt thread root stress concentration — reduces creep ductility and can either strengthen (notch strengthening) or weaken (notch weakening) the joint depending on stress state, as shown by the 2019 ductility exhaustion study for Grade 92 steel. Hitachi’s 2014 US and EP patents correct uniaxial creep damage parameters by a time-dependent multiaxiality factor, recognizing that bolt thread roots and shank-to-head transitions create stress concentrations that elevate triaxiality and reduce ductility, accelerating rupture relative to uniaxial predictions. Stress relaxation, by contrast, does not involve ductility exhaustion — the material is not approaching fracture, merely redistributing its strain budget.
The Nanjing University of Aeronautics and Astronautics 2024 US patent introduces a damage tolerance factor (λ) that classifies creep failure modes: grain boundary cavitation (1 < λ < 2.5), necking (2.5 < λ < 5), and unstable microstructure (λ > 5). This framework ensures that accelerated test conditions preserve the same failure mode as service conditions — a critical requirement when using short-term data to predict long-term creep rupture life.
without fatigue coupling
with fatigue coupling
Creep Rupture vs. Stress Relaxation — key questions answered
Creep rupture operates under load-controlled (constant stress) conditions, where the bolt accumulates progressive plastic strain until fracture. Stress relaxation operates under displacement-controlled (strain-controlled) conditions, where the bolt’s clamped load diminishes over time as elastic strain converts to creep strain under fixed displacement constraints. The failure consequence also differs: creep rupture ends in structural fracture, while stress relaxation results in joint leakage, loosening, or loss of sealing integrity.
Patents from Mitsubishi Power establish that turbine bolts operating above 500°C, and especially above 600°C, are principal candidates for creep rupture failure. The life-elongation rate of the bolt shank serves as the key measurable damage proxy for this failure mode.
The 2016 bolt assembly optimization study shows that retightening at the inflection point from rapid to steady-state creep (the primary-to-secondary creep transition) produces a new, improved stress relaxation curve with a longer residual preload life. This defines an optimal maintenance interval tied to material creep behavior rather than calendar time.
The turbine bolt fracture analysis (2019) and P92 residual stress study (2019) both confirm that torsional assembly stresses reduce effective creep rupture life. The fracture morphology shows secondary 45° cracks indicating superimposed torsional assembly stresses. Joint design protocols that minimize torque-induced residual stress, such as bolt-stretching rather than torque-tightening, are mechanistically justified.
The British Thomson-Houston Company (1957) established that the ratio of free shank length (L) to loaded nut engagement length (N) governs how much creep strain the bolt can accumulate before relaxation causes load to drop below design threshold. Designs with larger L/N ratios tolerate more creep strain before preload is exhausted.
China National Nuclear Corporation’s 2025 patents establish that ignoring prior fatigue damage underestimates creep deformation by more than a factor of two. In the studied pressure vessel component, maximum creep deformation increased from 0.72% to 1.62% when fatigue damage coupling was included.
PatSnap Eureka searches patents and research literature to answer instantly.