How do you identify and manage the risks and uncertainties of fatigue and creep failure scenarios?

In composite materials, especially in the absence of stress raisers, fatigue damage can be distributed throughout the material, it is not necessarily localized. In case of low fatigue loads for a long period we can also encounter 'wide-spread fatigue damage', where the issue is the interaction between localized damages in different locations (e.g. cracking at multiple different holes along a riveted joint).

Fatigue refers to the weakening of a material due to repeated or cyclic loading below its ultimate strength. It is characterized by the accumulation of small cracks that grow incrementally under cyclic loading, eventually leading to failure, often at stress levels significantly lower than the material's static strength. Creep is the gradual deformation of a material under constant stress or load over an extended period, typically at elevated temperatures. Unlike fatigue, which involves cyclic loading, creep occurs under sustained or constant loading conditions. Creep is common in metals and alloys subjected to high temperatures but can also affect some polymers and ceramics.

Creep is permanent deformation that occurs due to long-term micromechanical processes which are active when a material is subjected to stress BELOW the elastic limit. Above the elastic limit is plasticity, not creep. Creep processes are typically thermally-activated, relating to applied stress skewing the distribution of random atomic motion in a particular direction, even when the applied stress does not produce enough strain energy to move atoms directly.

What is Fatigue and Creep? Fatigue: The progressive and localized structural damage that occurs when a material is subjected to cyclic loading. Creep: The slow, permanent deformation of a material under constant stress, usually at high temperature.

My knowledge of this leads me to the following perception: a. Visual inspection: Check regularly for visible signs of cracks and deformations. b. Non-destructive techniques (NDT): Use ultrasound and X-rays to identify internal flaws. c. Tension monitoring: Install tension sensors to detect changes in load. d. Practical example: In gas turbines, NDT is essential to identify early fatigue and creep.

This text is more applicable to metals. In composites we obviously don't have mechanisms such as dislocations or grain boundary cracking. Furthermore fatigue damage in the form of delaminations or transverse matrix cracking often occurs below the surface of the composite. NDT is essential to finding these kinds of damages. Selecting an appropriate NDT technique requires understanding what kind of damage you are looking for, and the physical characteristics of the material you are inspecting.

Fatigue Detection: Visual Inspection: Look for cracks and surface defects. Non-Destructive Testing (NDT): Methods like ultrasonic testing, radiography, and magnetic particle inspection. Monitoring Systems: Use of sensors to track changes in stress and strain. Creep Detection: Strain Gauges: Measure deformation over time. Periodic Inspection: Regular checks for elongation or deformation. Thermography: Detects temperature changes indicating material deformation.

Detecting fatigue and creep damage in materials typically involves a combination of techniques ranging from visual inspections to advanced analytical methods. For Fatigue detection: 1) Look for signs of cracking, such as visible cracks, corrosion pits, or changes in surface roughness. 2) Identify characteristic fatigue crack patterns, such as beach marks or ratchet marks, progression lines, smooth fracture surface etc, which indicate the direction and growth of cracks. UT/RT/ECT etc can also be helpful. For Creep damage detection: 1) Optical/SEM etc to look for microstructural changes, such as creep voids and grain boundary cracking. 2) Measure dimensional changes, such as elongation or deformation, over time

It is important to remember the mechanisms of fatigue failure. Generally, crack initiation is followed by crack growth and some final quasi-static fracture. S-N data gives you the total life for a given stress, stress concentration, and specimen geometry; which is not necessarily very useful. If fatigue damage is accumulating in your part, S-N data will not tell you whether you are pre-initiation or in a crack growth phase approaching fracture. Therefore, most simulation of fatigue and creep involves using design curves to find regions of "safe load" in which there is no damage accumulation. Formally, this is not simulation of fatigue or creep, it is simulation of the limit state beyond which they may occur.

Fatigue Modeling: S-N Curves (Stress-Life): Plots stress against the number of cycles to failure. Fracture Mechanics: Focuses on crack growth and propagation. Finite Element Analysis (FEA): Simulates cyclic loading and predicts fatigue life. Creep Modeling: Norton-Bailey Law: Describes the creep rate as a function of stress and temperature. Time-Temperature Parameters (e.g., Larson-Miller): Predicts creep life based on stress and temperature. Creep Constitutive Models: Describes material behavior under long-term loading.

Based on failure analysis experience i would suggest that preventing fatigue and creep damages in materials and structures may involves proactive design, material selection, good operational practices, and monitoring strategies. 1) by designing the components with smooth transitions and fillets to reduce stress concentration points. Select materials with high fatigue resistance. 2) Ensure proper dimensions and shapes to distribute stresses evenly 3) Apply surface treatments like shot peening to induce compressive residual stresses 4) Control operating temperatures to avoid exceeding material creep limits. Protect materials from corrosive environments that can accelerate creep degradation. Choose materials with good creep resistance.

While proper design against fatigue will consider more aspects than just this, a first rule of thumb can be to minimize any stress concentrations occurring in your structure, by avoiding sudden changes of stiffness where possible. E.g. use tapers instead of steps and use larger fillet radii

You can either design your product to be durable, or design it to be spareable. Classically, regulations (eg. aero regulations) require you to remove your product from service before the residual strength (the amount of strength remaining after in-service degradation) drops below the limit load (the maximum load expected). Obviously, you can design your product with conservative assumptions to maximise durability; but don't simply assume this is the smart move. Maybe saving weight and cost in the part but sparing it out at 5 years life is a better plan. In wind turbines, 10 years extra life can reduce cost of energy by 10-15%, but in aircraft it gains the customer nothing because after 5-10yr service all flight costs are recurring costs.

Simple is the best. First lesson what is the problem in a meanwhile observe the situation and collect all of them. When you catch the main problem rest of them will solve automatically.

Fatigue Prevention: Material Selection: Use high-strength, fatigue-resistant materials. Design Optimization: Avoid stress concentrations; use fillets and smooth transitions. Surface Treatments: Apply coatings or surface hardening techniques. Regular Maintenance: Schedule inspections and replace components before failure. Creep Prevention: Material Selection: Use creep-resistant materials such as superalloys. Temperature Control: Operate within safe temperature limits. Load Management: Reduce stress levels and avoid constant high loads. Protective Coatings: Use thermal barriers to reduce material temperature.

One point to be aware of is the tendency in concept design to ignore durability concerns. You may be told by a customer to "ignore fatigue for now". It is part of your job as a good engineer to resist that urge. It is relatively simple to come up with allowable stresses and strains which allow for fatigue and/or creep damage. They don't need to be perfect or accurate to the MPa. There is no good reason to design concepts based on quasi-static strengths and "leave fatigue for later".