Mayorkinos 1

Structural integrity

Applied loads can be:

  • Static
  • Cyclic
  • Stochastic (random)
  • Impact

Deterioration

Stochastic loads can be difficult to predict

Structural integrity management

Operate until failure then repair/replace

Run for period of time then carry out maintenance

Operate for period of time and then inspect

Monitor continuously and diagnose problems

Predict damage initiation and evolution and plan maintenance based on this

Maintenance Strategies

Corrective

Preventive

Condition-based

Risk-based or predictive

Data, model or both driven

Higher uncertainty and variability, particularly for stochastic loading

Carried out at pre-determined intervals

Fault is identified and then repaired/replaced

Maintanence is conducted after condition has deteriorated beyond an acceptable level

Reliability-centred maintenance

Based on a combination of different maintanence strategies

May employ preventative measures if catastrophic failure is unnaceptable

Good understanding of failure mechanisms is required

Uses risk-based inspection

Part of overall plant management

Various NDT methodologies may be employed

Requires wide range of information

Reasons for failure

Lack of understanding

Loading conditions are not understood or are underestimated

Additional factors having bigger effect than considered

Innaccurate modelling

Presence of defects or material used outside of specifications

Incorrect installation or maintanence

Normal accident theory

Accidents are unavoidable in complex systems therefore called normal accidents

System is susceptible to normal accidents if it is complex, tightly coupled and has catastophic potential

E.g. Chernobyl

Swiss cheese model

Gaps in defence can align and failure can occur

Safety order in complex systems

  1. Design for minimum risk - eliminate risks in design
  1. Incorporate safety into design - use safety features or devices
  1. Provide warning devices - detect faults
  1. Develop procedures, standards and training

Failure patterns

image

A. Bathtub, B. Wear out, C. Fatigue, D. Initial break-in, E. Random, F. Infant mortality

Fault tolerant design

Design considerations:
-Estimated flaws
-Tolerance limit set
-Initial flaw size is assumed
-Time or cycles for crack growth defined
-Inspection requirements defined

Fail safe and safe life

Safe life: no crack initiation, or cracks do not grow to critical dimension to cause failure during operational lifetime

Fail safe: Cracks are not allowed to grow to point of causing failure. Partial failure should not compromise overall safety

Equations

Factor of safety

Structural capacity of a system to sustain loads beyond those predicted

FS = UTS/DS

Ratio of yield stress (YS) or ultimate tensile stress (UTS) to max. design stress (DS)

Margin of safety

Excess capacity of a structure to sustain the max stresses it could ever sustain

MS= (Failure load/Design load)-1

Paris-Erdogan Law

Empirical law used to estimate crack growth rates within load range

image

a is crack length (mm), n is number of load cycles, C and m are material constants, deltaK is stress intensity factor range (MPa m )

Stress, strain etc

image

Fatigue limit

Lab testing shows steels have a safe stress at which failure will not occur, regardless of cycles

Only found in certain materials

Affected by:
-Periodic overloads
-Corrosion
-High temperatures

image