Probabilistic Fatigue Assessment for Machine Component Inspection Scheduling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current deterministic approaches to estimating fatigue crack growth in machine components under cyclic stresses result in overly conservative designs and premature component replacements, leading to increased costs and reduced service life due to the assumption of worst-case material properties and flaw locations, whereas probabilistic methods can provide more realistic service life values but require extensive simulations.

Innovation Solution

The implementation of a direct simulation probabilistic fracture mechanics (DSPFM) method that uses Monte Carlo simulations to distribute material properties and flaw sizes across the component, allowing for a risk-based approach to determine the probability of failure and optimize inspection schedules and component service life decisions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If deterministic fracture mechanics calculations are used to estimate component service life, then safety is improved by compensating for unknowns, but component service life is reduced and costs increase due to overly conservative assumptions

Engineering Contradiction:
ImprovesafetyVSAvoidcomponent service life
Core Design Contradiction:
ReliabilityVSDuration of action of stationary object

Solution Approach 1:

The patent transitions from deterministic parameter estimation (using minimum material properties and maximum flaw sizes) to probabilistic parameter distribution (using statistical distributions of material properties and flaw sizes). This allows the model to account for variability and uncertainty while avoiding overly conservative assumptions, thereby extending component service life while maintaining safety

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent creates multiple virtual representations (copies) of the component with different material properties and flaw characteristics through Monte Carlo simulations. These virtual copies allow for comprehensive assessment of potential failure scenarios without requiring physical testing of each extreme case, resolving the contradiction between safety and service life

Inventive Principle:
Principle #26Copying

2Ease of manufacture

If deterministic fracture mechanics calculations are used, then design and inspection decisions can be made with simple methods, but component service life is reduced due to conservative estimates

Engineering Contradiction:
Improvedesign simplicityVSAvoidcomponent service life
Core Design Contradiction:
Ease of manufactureVSDuration of action of stationary object

Solution Approach 1:

The patent changes the computational approach from simple deterministic calculations to probabilistic Monte Carlo simulations. While this increases computational complexity, it provides more accurate service life predictions that avoid premature replacement, ultimately improving the overall efficiency and economic viability of component management

Inventive Principle:
Principle #35Parameter changes

3Reliability

If conservative material properties and flaw sizes are assumed, then safety factors are increased, but component service life is reduced and replacement costs increase

Engineering Contradiction:
Improvesafety factorVSAvoidcomponent replacement
Core Design Contradiction:
ReliabilityVSLoss of substance

Solution Approach 1:

The patent transforms fixed conservative parameter values into probabilistic distributions that reflect actual material variability and flaw characteristics. This allows for more accurate prediction of when components will actually fail, reducing premature replacements while maintaining appropriate safety factors

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent incorporates inspection results and actual component performance data into the probabilistic model, allowing for continuous refinement of predictions. This feedback mechanism enables more accurate determination of replacement timing, reducing unnecessary replacements while maintaining safety

Inventive Principle:
Principle #23Feedback

4Reliability

If deterministic calculations with worst-case scenarios are used, then risk coverage is improved, but inspection frequency increases and productivity decreases

Engineering Contradiction:
Improverisk coverageVSAvoidinspection frequency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent changes from uniform deterministic assumptions to location-specific probabilistic assessments. By identifying critical regions and assigning appropriate probability distributions to different locations, the model can prioritize inspections in high-risk areas while reducing inspection frequency in low-risk areas, thereby improving productivity while maintaining comprehensive risk coverage

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3304034B1Scheduling inspections and predicting end-of-life for machine components
Publication Date: 2020.08.12 SIEMENS ENERGY INC
  • EP3304034B1 patent drawingFigure 1~3
  • EP3304034B1 patent drawingFigure 2
  • EP3304034B1 patent drawingFigure 4~5

AI summary

A method for operating a machine component under stress. The method comprises determining a probability of failure PoF(N) of the component as a function of N cycles, selecting a time-based acceptable risk limit for the component and selecting an operational profile for the component, converting the time-based acceptable risk limit to a cycle-based acceptable risk limit using the operational profile, comparing the cycle-based acceptable risk limit with the PoF(N) values to determine an operational status of the component, comparing the cycle-based acceptable risk limit with the PoF(N) values, and operating the machine component responsive to results of the comparing step.