科研 Skill
金属疲劳寿命评估
fatigue-analysis-workflow
Builds, calibrates, and validates fatigue assessments for metals and components: stress-life, strain-life, multiaxial critical-plane. Use when predicting fatigue life or crack growth.
SKILL.md
Description
Build, calibrate, and validate fatigue assessments for metallic materials and components. Select among stress-life, strain-life, local-strain, multiaxial critical-plane, defect-sensitive, crystal-plasticity, and fracture-mechanics methods according to the physical stage being predicted and the available evidence.
This skill treats fatigue initiation, microstructurally short-crack behavior, long-crack propagation, and total life as distinct modeling targets. It supports scientific and engineering analysis; it does not replace applicable design codes, certified material allowables, inspection planning, or safety review.
Domain Context
Fatigue is progressive damage under repeated or fluctuating loading. A useful workflow begins by defining the prediction target:
- Initiation life: cycles to a specified detectable crack, stiffness change, or other operational criterion. The definition must accompany every reported value.
- Crack nucleation: formation of a crack embryo at a persistent slip band, inclusion, pore, surface feature, interface, or other microstructural site. This is not automatically equivalent to an engineering initiation-life measurement.
- Short-crack life: growth while crack dimensions are comparable to microstructural or plastic-zone length scales. Long-crack threshold and closure assumptions may not apply.
- Long-crack life: propagation from a characterized crack under conditions where continuum fracture-mechanics parameters are valid.
- Total life: an explicitly stated combination of initiation and propagation models, with a consistent handoff crack size and no double counting.
The nominal HCF, LCF, VHCF, and oligocyclic labels are useful descriptors, not universal mechanism boundaries. Cycle-count thresholds vary with material, environment, geometry, temperature, loading mode, and the life definition. [EXPERT REVIEW NEEDED]
Thermomechanical fatigue (TMF) additionally couples cyclic strain and temperature histories. In-phase and out-of-phase labels do not by themselves define the relevant oxidation, creep, phase-transformation, or constitutive response.
When to Use This Skill
Use this skill when:
- selecting a defensible fatigue model for coupon, subcomponent, or component data;
- reducing constant-amplitude or variable-amplitude histories into auditable cycle measures;
- fitting S-N, strain-life, crack-growth, defect-sensitive, or statistical fatigue models;
- transferring calibrated material behavior to local finite-element fields;
- ranking fatigue-critical regions using multiaxial criteria or fatigue indicator parameters (FIPs);
- assessing how pores, inclusions, surface finish, residual stress, texture, or grain structure change risk;
- designing verification experiments, sensitivity studies, and uncertainty propagation;
- comparing conventional continuum and crystal-plasticity predictions without conflating numerical convergence with physical validation.
When Not to Use This Skill
Do not use this skill as the sole basis when:
- a regulated design requires a specific code, certified database, safety factor, or damage-tolerance procedure;
- creep rupture, monotonic overload, brittle cleavage, corrosion cracking, fretting, irradiation, or wear dominates and is not represented;
- the load history, temperature, environment, surface state, residual stress, or material condition is unknown;
- a crack is outside the domain where the chosen stress-intensity, J-integral, or small-scale-yielding assumptions hold;
- a fitted relation is being extrapolated beyond its stress, strain, life, crack-size, frequency, temperature, or environment range;
- a visually plausible finite-element or CPFE field is being treated as validated life prediction without independent fatigue evidence.
Core Concepts
Method selection
| Prediction target and evidence | Preferred starting point | Required boundary statement |
|---|---|---|
| Nominally elastic, constant-amplitude coupon or component loading | Stress-life (S-N/Basquin) | Stress definition, load ratio, runout treatment, surface and environment |
| Local cyclic plasticity or low-cycle response | Strain-life (Coffin-Manson-Basquin) | Stabilized cyclic response, strain measure, mean stress, failure criterion |
| Elastic global model with a localized notch | Local-strain method with cyclic Ramberg-Osgood and notch correction | Applicability of Neuber or alternative correction and multiaxiality |
| Nonproportional or multiaxial local histories | Critical-plane or invariant method | Plane search, stress/strain tensors, criterion calibration domain |
| Measured variable-amplitude history | Rainflow counting plus an explicitly justified damage rule | Filtering, residual cycles, sequence effects, mean-stress treatment |
| Characterized long crack | Fracture-mechanics crack-growth model | Geometry factor, crack closure/ratio, threshold, validity limits |
| Microstructure-sensitive initiation ranking | CPFE plus calibrated FIP | RVE representativeness, averaging rule, life mapping, uncertainty |
| Pore- or inclusion-controlled fatigue | Defect-sensitive model or explicit defect simulation | Defect metric, location, morphology, residual stress, population statistics |
| TMF | Temperature-dependent constitutive and fatigue model | Phase relation, dwell, oxidation/creep interactions, thermal gradients |
Stress-life and mean stress
For a Basquin convention written in reversals,
sigma_a = sigma_f_prime * (2 N_f)^b
record whether N_f means cycles or reversals, whether stress is nominal or local, and how runouts enter the fit. Goodman, Gerber, Soderberg, and Walker relations encode different assumptions; none is a universal mean-stress correction. SWT is instead a fatigue damage parameter based on maximum normal stress and normal strain amplitude. Select and validate the correction for the material, life regime, load ratio, and stress/strain formulation. [EXPERT REVIEW NEEDED]
Strain-life and local notch response
A common uniaxial strain-life form is
epsilon_a = (sigma_f_prime / E) * (2 N_f)^b
+ epsilon_f_prime * (2 N_f)^c
Use stabilized cyclic properties where the material response stabilizes. If cyclic softening, hardening, ratcheting, mean-stress relaxation, or non-Masing behavior matters, retain the cycle evolution or use a suitable constitutive model. Cyclic Ramberg-Osgood parameters are not interchangeable with monotonic parameters.
Neuber's rule is an approximate local elastic-plastic correction, commonly applied to notch problems under constrained conditions. Its accuracy depends on geometry, loading, material behavior, and implementation; compare it with an elastic-plastic finite-element solution for consequential decisions. [EXPERT REVIEW NEEDED]
Multiaxial fatigue
The resolved shear stress on a slip system is
tau = s . sigma . n
where s is the slip direction and
` the plane normal expressed in the same frame as the Cauchy stress tensor. The Schmid factor is a uniaxial special case, not a complete cyclic multiaxial fatigue model.
Fatemi-Socie combines cyclic shear with the maximum normal stress on the candidate plane. Smith-Watson-Topper combines maximum normal stress and normal strain amplitude. Brown-Miller combines shear strain amplitude with normal strain. Criterion choice, plane-search resolution, nonproportional hardening, and the definition of amplitudes materially affect results. [EXPERT REVIEW NEEDED]
Variable amplitude and cumulative damage
Rainflow counting converts a scalar history into cycle ranges and means. Document turning-point extraction, filtering, binning, treatment of half cycles, and the scalar channel used for multiaxial loading. Palmgren-Miner summation,
D = sum(n_i / N_i)
is a linear accounting model. It does not inherently represent sequence, overload retardation, load interaction, evolving residual stress, or a fatigue limit. Treat D = 1 as a calibrated convention rather than a physical law. [EXPERT REVIEW NEEDED]
Crack growth and total life
For long-crack growth, a Paris-region model is
da/dN = C * (Delta K)^m
Use it only over the fitted range. Near-threshold behavior, crack closure, load ratio, overloads, environment, corrosion, temperature, and rapid fracture require appropriate extensions and validity checks. Define Delta K, effective driving force if used, geometry factors, crack-front representation, and termination criterion.
Do not start a Paris-law integration at an arbitrary microscopic flaw and call the result total life. The transition from nucleation or short-crack behavior to long-crack behavior must use an explicit handoff size and compatible life definitions.
Microstructure, defects, and statistics
CPFE resolves orientation-dependent slip and can generate FIPs based on cyclic plastic strain, plastic work, resolved shear, or related measures. FIP averaging volume, hot-spot rule, boundary treatment, RVE ensemble, and empirical mapping to life are model choices requiring calibration and sensitivity analysis. [EXPERT REVIEW NEEDED]
Murakami-style sqrt(area) relations are empirical defect-sensitive frameworks whose constants and applicability depend on material class, hardness, defect location, load ratio, and environment. Do not transfer a surface-defect fit to internal defects or a new alloy without validation. [EXPERT REVIEW NEEDED]
Fatigue scatter often requires censored-data treatment for runouts, specimen-level covariates, batch effects, and probabilistic reporting. A deterministic best-fit curve is not a lower-bound design curve.
Thermomechanical fatigue
TMF analysis must retain synchronized temperature and mechanical histories, thermal expansion, temperature-dependent cyclic constitutive response, and the selected life metric. Oxidation, creep-fatigue interaction, phase change, coatings, and dwell effects may dominate in different regimes. Their omission or coupling requires system-specific justification. [EXPERT REVIEW NEEDED]
Key Workflows
Workflow 1: Define the fatigue data contract
Create a machine-readable record before fitting any model:
material:
alloy: "Ti-6Al-4V"
processing: "laser powder bed fusion, stress relieved"
orientation: "build direction parallel to loading axis"
surface: "as-built"
loading:
control: "strain"
waveform: "triangular"
ratio_definition: "R_epsilon = epsilon_min / epsilon_max"
temperature_K: 293
environment: "laboratory air"
life_definition:
event: "25 percent peak-load drop"
unit: "cycles"
data:
runouts_are_censored: true
specimen_geometry: "documented separately"
raw_history_retained: true
Add calibration/test partition, specimen identifiers, uncertainty, machine compliance correction, and links to raw data. Reject rows whose units, life criterion, or loading ratio cannot be recovered.
Workflow 2: Fit and audit a stress-life model
Use a log-likelihood or regression appropriate to the variance model and include runouts as right-censored observations when possible. A minimal calculation from already calibrated Basquin parameters is:
import numpy as np
def basquin_cycles(sigma_a, sigma_f_prime, b, calibration_bounds):
"""Return cycles to the calibrated event; Basquin form uses reversals."""
sigma_a = np.asarray(sigma_a, dtype=float)
if np.any(sigma_a <= 0) or sigma_f_prime <= 0 or b >= 0:
raise ValueError("Require positive stresses and b < 0")
cycles = 0.5 * (sigma_a / sigma_f_prime) ** (1.0 / b)
lo, hi = calibration_bounds
if np.any((sigma_a < lo) | (sigma_a > hi)):
raise ValueError("Stress amplitude is outside the calibration range")
return cycles
Plot observations, runouts, fitted median, uncertainty or tolerance bounds, and residuals against stress, batch, surface state, and load ratio. Do not infer an endurance limit merely because the test stopped.
Workflow 3: Solve a strain-life relation with an explicit bracket
import numpy as np
from scipy.optimize import brentq
def strain_life_cycles(
epsilon_a, E, sigma_f_prime, b, epsilon_f_prime, c, reversal_bounds
):
"""Solve Coffin-Manson-Basquin for cycles to the calibrated event."""
if min(epsilon_a, E, sigma_f_prime, epsilon_f_prime) <= 0:
raise ValueError("Positive strain and material parameters are required")
if b >= 0 or c >= 0:
raise ValueError("This convention requires b < 0 and c < 0")
def residual(log_reversals):
reversals = np.exp(log_reversals)
predicted = (
sigma_f_prime / E * reversals**b
+ epsilon_f_prime * reversals**c
)
return predicted - epsilon_a
lo, hi = reversal_bounds
if not (0 < lo < hi):
raise ValueError("Provide positive ordered reversal bounds")
reversals = np.exp(brentq(residual, np.log(lo), np.log(hi)))
return 0.5 * reversals
Validate the root bracket and parameter covariance. Compare predictions against held-out hysteresis loops and life data, not only the fitted strain-life points. For notches, compare local-strain approximations with elastic-plastic finite-element fields at representative amplitudes.
Workflow 4: Reduce a variable-amplitude history without hiding assumptions
- Preserve the original timestamped signal and units.
- Remove offsets or noise only with a documented physical rationale.
- Extract reversals and apply a specified rainflow implementation.
- Retain range, mean, count, and any dwell or temperature labels.
- Apply mean-stress and life models inside their calibration domains.
- Compare linear damage with block or spectrum tests that expose sequence effects.
import numpy as np
import pandas as pd
def miner_damage(counted_cycles, life_function):
"""counted_cycles columns: range, mean, count."""
required = {"range", "mean", "count"}
if not required.issubset(counted_cycles.columns):
raise ValueError(f"Missing columns: {required - set(counted_cycles.columns)}")
if counted_cycles.empty:
raise ValueError("At least one counted cycle is required")
table = counted_cycles.copy()
numeric = table[["range", "mean", "count"]].apply(pd.to_numeric, errors="raise")
if not np.isfinite(numeric.to_numpy()).all():
raise ValueError("Cycle ranges, means, and counts must be finite")
if (numeric["range"] <= 0).any() or (numeric["count"] <= 0).any():
raise ValueError("Cycle ranges and counts must be positive")
table[["range", "mean", "count"]] = numeric
table["amplitude"] = 0.5 * table["range"]
life = np.asarray(life_function(table["amplitude"], table["mean"]), dtype=float)
if life.shape != (len(table),) or not np.isfinite(life).all() or (life <= 0).any():
raise ValueError("The life model must return one positive finite life per row")
table["predicted_life"] = life
table["damage"] = table["count"] / table["predicted_life"]
return float(table["damage"].sum()), table
Record the rainflow package and version. Validate its output against a published or hand-checked history before production use.
Workflow 5: Transfer loads through an elastic-plastic finite-element model
- Define geometry, contact, preload, residual stress, temperature, and load sequence.
- Calibrate a cyclic constitutive model against hysteresis loops covering relevant amplitudes and temperatures.
- Refine the mesh at notches while tracking local stress, strain, plastic strain, and the fatigue metric.
- Converge both spatial mesh and load/time increments; check equilibrium and constitutive integration.
- Extract tensor histories in a declared coordinate frame, not only extrema.
- Compute the fatigue metric outside or inside the solver with a version-controlled implementation.
- Validate local fields against strain gauges, digital image correlation, diffraction, or another independent measurement where available.
Use FEM multiphysics for weak-form, solver, boundary-condition, and mesh-convergence practice.
Workflow 6: Evaluate multiaxial critical planes
Transform the full stress and strain histories to candidate material planes. Refine the angular grid until the maximum damage parameter and critical-plane orientation are stable. Preserve shear sign conventions and distinguish tensor shear from engineering shear strain. For nonproportional loading, use a constitutive model capable of representing the observed additional hardening before applying Fatemi-Socie, SWT, Brown-Miller, or another calibrated criterion.
Validate against held-out proportional and nonproportional tests. A criterion that ranks one loading path correctly may still fail under a different phase angle or mean stress.
Workflow 7: Build a CPFE-to-FIP workflow
- Construct an ensemble of RVEs with measured or justified grain size, morphology, texture, phases, and boundary conditions.
- Verify orientation conventions, slip systems, constitutive parameters, and crystal/laboratory frame transformations.
- Converge mesh, cycle discretization, number of simulated cycles, and RVE ensemble size.
- Compare aggregate stress-strain loops and texture-sensitive response against experiment.
- Compute multiple physically motivated FIPs over declared grains, bands, or averaging volumes.
- Calibrate the FIP-to-initiation relation on separate specimens and quantify uncertainty.
- Test whether hot-spot location and life ranking persist across statistically equivalent RVEs.
Use crystal plasticity workflow for kinematics, slip, hardening, RVE construction, and texture validation.
Workflow 8: Assess defect-sensitive and statistical fatigue
Characterize defects in three dimensions when morphology or connectivity matters. Separate surface, near-surface, and internal populations; retain inspection resolution and missed-detection limits. Compare empirical sqrt(area) predictions, explicit local-field simulations, and available defect-containing specimen tests. Fit population and life models with specimen or volume effects, censored runouts, and batch structure. Report prediction intervals and sensitivity to the largest-defect assumption rather than a single deterministic life.
Common Failure Modes
| Failure mode | Why it fails | Detection and correction |
|---|---|---|
| Mixing cycles and reversals | Introduces factor-of-two errors in fitted and predicted life | State the convention in equations, data, code, and plots; test a known point |
| Undefined initiation criterion | Makes life values incomparable | Record the event, detection method, and threshold for every dataset |
| Calling nucleation, initiation, and total life equivalent | Combines physically different stages | Define each stage and the handoff crack size |
| Applying an HCF/LCF/VHCF cycle boundary universally | Mechanisms do not change at universal cycle counts | Classify by observed response and document system-dependent labels |
| Fitting runouts as failures | Biases long-life behavior | Use censored-data methods or report the limitation |
| Inferring an endurance limit from test termination | A runout is not proof of infinite life | Retain censoring and test beyond the proposed knee when justified |
| Using nominal stress at a notch | Misses local concentration and plasticity | Use a validated notch correction or local finite-element solution |
| Using monotonic Ramberg-Osgood parameters cyclically | Misrepresents stabilized hysteresis | Fit cyclic stress-strain data and inspect loop evolution |
| Applying Neuber without checking its domain | Local approximation may fail under multiaxiality or extensive plasticity | Compare with elastic-plastic FE at representative cases |
| Ignoring mean stress or mean-strain relaxation | Distorts fatigue driving force | Measure or model relaxation and validate the selected correction |
| Treating Goodman, Gerber, or SWT as universal | Mean-stress models have different assumptions | Select by material and regime; compare held-out load ratios |
| Using a scalar rainflow history for multiaxial loading without justification | Loses path and phase information | Use a validated multiaxial counting/critical-plane procedure |
| Filtering out damaging peaks | Alters counted ranges and sequence | Archive raw histories and quantify filter sensitivity |
| Dropping rainflow half cycles silently | Changes spectrum damage | State residual-cycle policy and test the implementation |
| Assuming Miner damage is sequence-independent physics | Linear summation omits interaction and retardation | Validate with representative spectrum or block tests |
| Extrapolating an S-N curve below its evidence | Produces unsupported long-life estimates | Enforce calibration bounds and show extrapolated regions |
| Applying Paris law outside the Paris regime | Misses threshold and rapid-growth behavior | Use regime-appropriate growth laws and validity checks |
| Ignoring load ratio, closure, or overloads in crack growth | Misstates effective driving force | Calibrate with relevant histories and closure treatment |
| Starting long-crack integration at a microscopic flaw | Violates continuum and long-crack assumptions | Model or bound short-crack behavior and define a handoff size |
| Double counting initiation and propagation | Inflates total life | Align stage definitions and count each interval once |
| Using a 2D defect metric for a connected 3D defect without qualification | Misrepresents morphology and local fields | Compare sections with tomography or explicit simulations |
Transferring a sqrt(area) fit across defect locations or alloys | Empirical constants and mechanisms differ | Recalibrate by material, surface/internal class, and environment |
| Using one RVE as a statistical prediction | Confounds one realization with population behavior | Run statistically equivalent ensembles and report spread |
| Selecting the maximum element-level FIP | Creates mesh-sensitive hot spots | Define a physical averaging rule and converge it |
| Calibrating and validating a FIP on the same specimens | Hides mapping error and overfit | Reserve independent specimens or microstructures |
| Incorrect orientation or tensor-shear convention | Rotates slip systems or changes amplitudes | Unit-test transformations and tensor/engineering shear factors |
| Unconverged FE mesh or load increments | Produces numerical hot spots or missed cyclic response | Converge fields and final fatigue metric, not only displacement |
| Treating solver convergence as physical validation | A stable numerical answer can represent the wrong model | Compare independent fields, loops, sites, and lives with data |
| Omitting surface state, residual stress, or environment | Removes major fatigue drivers | Match specimen/component state and propagate uncertainty |
| Combining incompatible datasets | Mixes life definitions, units, ratios, and processing states | Audit metadata and fit hierarchical or separate models |
| Ignoring TMF phase, dwell, and thermal gradients | Misses creep, oxidation, and constraint interactions | Retain synchronized histories and validate representative cycles |
| Reporting only median life | Conceals fatigue scatter and reliability | Report data, censoring, uncertainty, and lower-tail metrics as appropriate |
Best Practices
- Define the modeled life stage and experimental event before selecting an equation.
- Preserve raw load, strain, temperature, and environment histories with units and timestamps.
- Keep nominal, structural, notch, and local stresses distinct.
- Write every amplitude, range, ratio, tensor-shear, and cycle/reversal convention explicitly.
- Partition data by specimen or batch before fitting transforms, descriptors, or models.
- Retain runouts as censored observations and show them on plots.
- Fit only within experimentally supported loading, temperature, environment, and microstructure domains.
- Compare against simple baselines before adding high-dimensional FE, CPFE, or surrogate models.
- Converge the final fatigue quantity with mesh, increment, plane grid, RVE size, and cycle resolution.
- Validate cyclic constitutive behavior before validating a derived fatigue indicator.
- Separate parameter calibration, numerical verification, physical validation, and uncertainty assessment.
- Use held-out load ratios, spectra, microstructures, or geometries to expose distribution shift.
- Report sensitivity to notch treatment, mean-stress model, damage rule, defect metric, and FIP averaging.
- Preserve software versions, solver inputs, random seeds, scripts, and intermediate counted cycles.
- Treat commercial-solver fatigue modules as version-sensitive implementations; verify conventions, defaults, supported criteria, and field extraction against documentation and benchmark cases. [EXPERT REVIEW NEEDED]
Integration with Other Skills
- Use FEM multiphysics to construct and verify local continuum fields.
- Use crystal plasticity workflow for microstructure-resolved cyclic constitutive response.
- Use phase-field workflow when fatigue damage is coupled to explicit microstructure evolution, without assuming that a qualitative field is a calibrated crack model.
- Use reproducibility workflow for provenance, environments, seeds, and executable records.
- Use scientific visualization for hysteresis loops, S-N data, residuals, crack-growth curves, FIP maps, and uncertainty displays.
- Use small-data ML for regularized models when fatigue datasets are sparse and structured.
- Use Gaussian process materials for calibrated interpolation and uncertainty-aware surrogate studies.
- Use Bayesian optimization only after defining physically admissible design spaces and validation gates.
- Use surrogate validation to test grouped splits, extrapolation, calibration, and hidden duplicates.
Key References
- Suresh, S. (1998). Fatigue of Materials (2nd ed.). Cambridge University Press. https://doi.org/10.1017/CBO9780511806575
- Schijve, J. (2009). Fatigue of Structures and Materials (2nd ed.). Springer. https://doi.org/10.1007/978-1-4020-6808-9
- Basquin, O. H. (1910). The exponential law of endurance tests. Proceedings of the American Society for Testing and Materials, 10, 625-630. No DOI listed in the bibliographic source used for verification.
- Coffin, L. F., Jr. (1954). A study of the effects of cyclic thermal stresses on a ductile metal. Transactions of the ASME, 76, 931-950. No DOI listed in the bibliographic source used for verification.
- Manson, S. S. (1953). Behavior of Materials Under Conditions of Thermal Stress (NACA-TN-2933). National Advisory Committee for Aeronautics. https://ntrs.nasa.gov/citations/19930083626
- Miner, M. A. (1945). Cumulative damage in fatigue. Journal of Applied Mechanics, 12(3), A159-A164. https://doi.org/10.1115/1.4009458
- Fatemi, A., & Socie, D. F. (1988). A critical plane approach to multiaxial fatigue damage including out-of-phase loading. Fatigue & Fracture of Engineering Materials & Structures, 11(3), 149-165. https://doi.org/10.1111/j.1460-2695.1988.tb01169.x
- Paris, P., & Erdogan, F. (1963). A critical analysis of crack propagation laws. Journal of Basic Engineering, 85(4), 528-533. https://doi.org/10.1115/1.3656900
- McDowell, D. L., & Dunne, F. P. E. (2010). Microstructure-sensitive computational modeling of fatigue crack formation. International Journal of Fatigue, 32(9), 1521-1542. https://doi.org/10.1016/j.ijfatigue.2010.01.003
- Murakami, Y. (2019). Metal Fatigue: Effects of Small Defects and Nonmetallic Inclusions (2nd ed.). Elsevier. https://shop.elsevier.com/books/metal-fatigue-effects-of-small-defects-and-nonmetallic-inclusions/murakami/978-0-12-813876-2
- ASTM International. (2021). ASTM E466-21: Standard Practice for Conducting Force Controlled Constant Amplitude Axial Fatigue Tests of Metallic Materials. https://doi.org/10.1520/E0466-21
- ASTM International. (2021). ASTM E606/E606M-21: Standard Test Method for Strain-Controlled Fatigue Testing. https://doi.org/10.1520/E0606_E0606M-21
- ASTM International. (2023). ASTM E1049-85(2023): Standard Practices for Cycle Counting in Fatigue Analysis. https://doi.org/10.1520/E1049-85R23
- ASTM International. (2024). ASTM E647-24: Standard Test Method for Measurement of Fatigue Crack Growth Rates. https://doi.org/10.1520/E0647-24
- ASTM International. (2025). ASTM E2368-25: Standard Practice for Strain Controlled Thermomechanical Fatigue Testing. https://doi.org/10.1520/E2368-25
- International Organization for Standardization. (2012). ISO 12107:2012 Metallic materials - Fatigue testing - Statistical planning and analysis of data (2nd ed.). https://www.iso.org/standard/50242.html
示例提示词
帮我用应变-寿命方法评估这个构件在多轴载荷下的疲劳寿命。
所属插件
安装该插件会一并获得同领域的 42 个 Skill / MCP。
同插件的其他科研 Skill
- 贝叶斯优化材料发现Covers Bayesian optimization for materials and chemistry design: surrogate models, acquisition functions, batch/constrained BO. Use when planning sample-efficient experiments or simulations.
- CALPHAD热力学计算工作流Covers CALPHAD computational thermodynamics: database selection, equilibrium/metastable calculations, binary/ternary phase diagrams. Use when predicting phase stability or solidification paths.
- 材料第一性原理与分子动力学模拟Routes first-principles and classical materials simulation (DFT, MD, ML potentials) to the right specialized workflow. Use as an entry point for any simulation-based materials task.
- 蠕变力学建模与验证Selects, calibrates, and validates creep deformation and rupture models under sustained thermomechanical load, covering rate prediction. Use when assessing long-term high-temperature components.
- 晶体塑性有限元建模Covers crystal plasticity and CPFE modeling of metals: slip systems for FCC/BCC/HCP, Schmid factors. Use when simulating texture- and orientation-dependent plastic deformation.
- 衍射结构表征Covers diffraction-based structure characterization (XRD, neutron, SAED): space group determination, Rietveld refinement, phase identification. Use for any diffraction-pattern analysis technique.
- EBSD织构分析Analyzes EBSD orientation maps for grain size, misorientation, phase, and texture, with pole figures and twin-boundary identification. Use for electron backscatter diffraction workflows.
- 电化学性能测试分析Covers electrochemical characterization: cyclic voltammetry, battery cycling analysis, EIS. Use for electrochemistry-specific data analysis distinct from general materials characterization.
- 多物理场有限元建模Covers finite element modeling for coupled materials, mechanics, and transport problems: weak-form construction, boundary conditions, mesh choice. Use for multiphysics FEM simulations.
- 连续尺度有限元力学建模Covers continuum-scale mechanical, thermal, and multiphysics simulation with Abaqus/COMSOL conventions, mesh generation. Use for macroscale FEM consuming measured or computed material properties.
- 断裂力学评估与建模Selects, constructs, and validates fracture-mechanics assessments: LEFM checks, elastic-plastic fracture mechanics. Use when assessing cracked materials or components.
- 高斯过程材料回归建模Covers Gaussian process regression for materials and chemistry datasets: descriptor selection, kernel design. Use when a probabilistic surrogate model with uncertainty is needed.
- 自驱动实验室闭环材料发现Covers combinatorial synthesis, robotic experimentation. Use for autonomous or robotics-driven experimental campaigns, a less standardized domain.
- 材料数据库可复现访问Covers reproducible API access to materials databases (Materials Project, AFLOW, OQMD, NOMAD, COD, ICSD-style sources): authentication. Use when programmatically querying these databases.
- 材料数据库跨库查询Covers querying public materials databases beyond Materials Project: AFLOW, OQMD, NOMAD, JARVIS. Use when pulling or comparing data from these live, API-integrated resources.
- 材料数据标准与互操作Covers data interoperability standards for materials science: CIF crystal files, the OPTIMADE query API. Use when structuring or exchanging materials data between tools.
- 材料结构特征化Covers converting structures, compositions, and trajectories into ML feature vectors: Magpie-style elemental statistics. Use when building inputs for materials ML models.
- 材料科学图表绘制规范Covers publication-quality figure conventions for materials science plots: phase diagrams, pole figures, stress-strain curves. Use when the deliverable is a figure, not just an analysis result.
- 材料信息学性质预测Covers ML for materials property prediction and generative design: descriptor-based models, crystal GNNs (CGCNN/MEGNet-class). Use for materials informatics distinct from interatomic potentials.
- 材料生命周期评估Covers life-cycle assessment for materials and manufacturing routes: goal/scope definition, functional units, inventory construction. Use for environmental-impact studies of materials.
- 材料科学文献检索Covers formulating literature search queries for materials science questions and synthesizing/citing results responsibly. Requires a web search tool; it cannot substitute for one.
- 工程材料选型设计Selects engineering materials via material indices and Ashby charts, ranking candidates by performance per cost or mass. Use when picking a material class for a design constraint.
- 材料统计方法与实验设计Covers statistics shared across materials characterization and simulation: design of experiments. Use for cross-cutting statistical treatment of materials data.
- 材料结构与成分核心分析Covers core structure and composition analysis: reading/writing crystal structures, composition handling, symmetry analysis, phase diagrams. Use as a general-purpose structure-analysis toolkit.
常见问题
金属疲劳寿命评估能做什么?
Builds, calibrates, and validates fatigue assessments for metals and components: stress-life, strain-life, multiaxial critical-plane. Use when predicting fatigue life or crack growth。
如何安装金属疲劳寿命评估?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「金属疲劳寿命评估」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。
金属疲劳寿命评估需要哪些权限?
该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。
金属疲劳寿命评估经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。