Fatigue Life Prediction Using Abaqus

Fatigue is defined as the weakening and eventual failure of a material due to repeated or fluctuating stresses.

Fatigue life prediction is critical to ensuring safe service and structural integrity of mechanical structures.

Introduction to the concept of fatigue

To understand fatigue, we must first grasp the concept of a stress cycle. A stress cycle consists of the repeated loading and unloading of a material, much like bending a paperclip back and forth. Understanding the mechanisms of fatigue is crucial for engineers because it enables them to design structures and components that can withstand the rigors of repeated loading.

Fatigue fracture is a prevalent and critical failure mode in mechanical equipment and engineering structures as there are usually no visible signs of impending failure on the surface of an element or structure e. g. macroscopic plastic deformation or cracks. The sudden failure of fatigue-fractured components can have disastrous consequences for industrial projects causing significant economic losses and in some cases enormous casualties. In fact fatigue failure affects almost all metals and ~90% of mechanical service failures can be attributed to fatigue.
Therefore it is essential to conduct systematic investigations into fatigue behavior and understand its intrinsic failure mechanisms.

Types of Fatigue

There are two main types of fatigue:

Low cycle fatigue – When the material cracks and fails due to stress amplitudes that are slightly less than the material’s yield stress and the number of cycles is fairly low (less than 10,000 cycles) .low-cycle fatigue involves a smaller number of stress cycles, often in the hundreds or thousands, but with each cycle causing significant plastic deformation. This type of fatigue is common in structures that experience high loads or extreme temperature fluctuations, such as pressure vessels, nuclear reactors, and offshore platforms. LCF cracks often initiate on the surface and grow rapidly.

High cycle fatigue – When the material cracks and fails due to stress amplitudes much less than the material’s yield stress and the number of cycles is fairly high (greater than 10,000 cycles). This type of fatigue is often associated with components that experience vibrations or rotating motions, such as turbine blades, engine crankshafts, and gear teeth. The failure in HCF usually initiates below the surface and gradually propagates inward.

Fatigue life prediction methods

Generally, the methods for fatigue life prediction can be classified into physics-based, data-driven, and hybrid approaches

1-Phenomenologically based models

The proposed fatigue model is able to reasonably predict the fatigue lives of various asphalt mixtures without changing the model coefficients.

There are several methods involved in fatigue analysis, most prominently are:

The stress-life method
The strain-life method
Linear Elastic Fracture Mechanics (LEFM) method

The evaluation of a material’s fatigue characteristics is a crucial factor in determining its fatigue limit and S–N curve. There are various engineering methods for assessing fatigue parameters, including the single-point, lifting group, strain-controlled fatigue life, multi-amplitude measurement, and amplification measurement methods.

Read about Abaqus Material Models here

2-Physically based models

A physically based model of fatigue refers to a computational or mathematical framework that simulates the fatigue behavior of materials based on fundamental physical principles, rather than relying purely on empirical observations. These models aim to capture the underlying mechanisms of fatigue damage, such as crack initiation, microstructural evolution, and crack propagation, using principles from mechanics, thermodynamics, and materials science.

To improve the efficiency of predicting fatigue parameters, many researchers have expanded the application of evaluation methods by utilizing temperature data to predict fatigue S–N curves. However, traditional experimental approaches to evaluating material fatigue lifetime are costly, require numerous specimens, and take up a considerable amount of time. Therefore, alternative methods have been developed to obtain faster evaluations of fatigue life.

3-Thermodynamically based models

Thermal fatigue arises from repeated heating and cooling cycles, causing thermal expansion and contraction within a material. This can induce significant stresses, especially in components with constrained geometries or those made of materials with differing thermal expansion coefficients. Examples include turbine blades, heat exchangers, and brake discs.

4-Two-scale based models

In high cycle fatigue (HCF) tests, the applied cyclic stress is typically lower than the macroscopic yield strength, and the main deformation occurs through macroscopic elasticity shakedown.

How to do Fatigue Analysis in Abaqus?

Fatigue life prediction in Abaqus is a multi-step process that involves stress analysis, damage assessment, and fatigue life estimation. Below is a step-by-step guide to conducting fatigue analysis using Abaqus:

Step 1: Perform Static or Dynamic Stress Analysis

Fatigue analysis starts with a stress or strain analysis under cyclic loading.

1.1 Create the Model

Define the geometry (2D or 3D CAD model).
Assign material properties (Young’s modulus, Poisson’s ratio, density).
Apply boundary conditions (supports, constraints).

1.2 Apply Loads

Use cyclic loads (force, displacement, pressure) to represent real-world conditions.
Define load amplitude variations (constant amplitude or variable amplitude loading).

1.3 Mesh the Model

Use fine meshing in high-stress regions (critical areas prone to fatigue).
Choose tetrahedral, hexahedral, or quadratic elements for accuracy.

1.4 Run the Simulation

Solve for stress and strain results under cyclic loading.

Step 2: Extract Stress and Strain Data

Use Abaqus/CAE or Abaqus/Viewer to identify critical stress locations.
Export stress (σ) and strain (ε) results at high-stress regions.
Use time-history plots to check loading cycles.

Step 3: Fatigue Life Prediction Methods

Fatigue life can be estimated using two main approaches:

3.1 Stress-Life Approach (S-N Curve)

Suitable for high-cycle fatigue (HCF) (low stress, high cycle count).
Uses experimental S-N (Stress vs. Number of Cycles) curves.
Implement in Abaqus with fe-safe (Dassault Systèmes fatigue analysis tool).
Steps:
1. Obtain S-N curve data for the material.
2. Define stress amplitude and mean stress.
3. Use Miner’s Rule to estimate fatigue damage accumulation.

3.2 Strain-Life Approach (ε-N Curve)

Suitable for low-cycle fatigue (LCF) (high stress, low cycle count).
Uses Coffin-Manson equation to predict fatigue failure.
Requires plastic strain and cyclic properties.

Step 4: Using fe-safe for Fatigue Analysis

Abaqus does not have a built-in fatigue module, but you can use fe-safe, an advanced fatigue tool.

4.1 Export Stress Results from Abaqus

Use ODB file to extract stress data.
Import data into fe-safe.

4.2 Define Fatigue Parameters in fe-safe

Select material properties (S-N, ε-N curves).
Define loading cycles and fatigue model.
Choose mean stress correction (Goodman, Gerber, Smith-Watson-Topper).

4.3 Run Fatigue Simulation

Compute fatigue damage accumulation.
Predict fatigue life (Nf) at critical locations.

Step 5: Post-Processing and Results Interpretation

Use fe-safe to generate fatigue life contours (number of cycles to failure).
Identify critical failure locations.
Compare with experimental results.

Alternative Methods for Fatigue Analysis in Abaqus

Use Python Scripting to automate fatigue calculations.
Implement custom fatigue post-processing in Abaqus via user subroutines.
Read more about Fatigue Analysis in Abaqus

Conclusion

Perform stress analysis in Abaqus.
Extract stress and strain results.
Use S-N or ε-N fatigue models.
Use fe-safe for fatigue analysis.
Interpret fatigue life results.