What is Variable-Amplitude Fatigue Analysis?
🙋
What exactly is "variable-amplitude" fatigue? I thought engineers just tested parts with the same stress repeated over and over.
🎓
That's a great starting point. Basically, constant-amplitude testing is for labs. In practice, real-world loads are messy and change all the time. For instance, a car suspension experiences small vibrations from the road, medium bumps from potholes, and occasional huge jolts from curbs—all mixed together. That's variable amplitude. This simulator lets you create such a complex load history using the "Load Spectrum Type" dropdown and harmonic controls.
🙋
Wait, really? So how do you even start to predict failure from that random-looking load data?
🎓
You break it down into simple cycles we can understand, using a clever algorithm called Rainflow Counting. It scans the load history you create—try moving the "Fundamental Amplitude A1" and "2nd Harmonic Ratio r2" sliders to see the waveform change—and identifies individual stress cycles, each with its own amplitude ($\sigma_a$) and mean stress ($\sigma_m$). It's like sorting a bag of mixed candy by type and size before counting.
🙋
Okay, so you get a list of cycles. But how do you add up the damage from all these different cycles to see if the part fails?
🎓
That's where Miner's Rule comes in. It's a simple but powerful idea: damage from each cycle type is additive. For each cycle the Rainflow method found, we calculate how many of those cycles would cause failure ($N_i$), using the S-N curve parameters you set like the "Endurance Limit $S_e$". Then we sum the ratio of applied cycles to failure cycles. If the total $D$ reaches 1, failure is predicted. Try adjusting the "Basquin Exponent m" to see how material sensitivity changes the cumulative damage bar.
Physical Model & Key Equations
The core of the prediction is Miner's Rule for cumulative damage. It assumes the damage from each stress cycle is independent and can be summed linearly.
$$D = \sum_{i=1}^{k}\frac{n_i}{N_i}\leq 1$$
Here, $D$ is the total cumulative damage (failure occurs at $D \geq 1$). $n_i$ is the number of applied cycles at a specific stress level $i$, and $N_i$ is the number of cycles to failure at that same stress level, which comes from the material's S-N curve.
The S-N curve (Stress vs. Cycles to failure) is modeled using the Basquin equation, which is a power-law relationship in the high-cycle fatigue regime.
$$N_i = \left( \frac{\sigma_{a,i}}{S_e}\right)^{-m}\cdot N_e$$
In this equation, $\sigma_{a,i}$ is the stress amplitude of cycle $i$, $S_e$ is the material's endurance limit (the stress below which failure theoretically never occurs), $m$ is the Basquin exponent (defining the slope of the S-N curve on a log-log plot), and $N_e$ is a reference number of cycles, often $10^6$ or $10^7$, associated with the endurance limit.
Real-World Applications
Aerospace Component Design: Aircraft wings and landing gear endure a complex spectrum of loads from take-off, turbulence, and landing. Engineers use this exact analysis to schedule maintenance and ensure the airframe lasts for the required number of flight cycles without being overly heavy.
Automotive Suspension & Chassis: As mentioned, every pothole and curb applies a different load. Carmakers simulate decades of driving on virtual proving grounds, using Rainflow counting on the simulated stresses to predict the lifespan of suspension arms and chassis welds.
Wind Turbine Blade Analysis: Blades face constantly varying wind gusts, gravity cycles, and rotational forces. Fatigue analysis is critical for these massive, expensive structures to achieve a 20+ year life in a highly corrosive environment.
Electronic Circuit Board Reliability: Temperature changes cause materials to expand and contract at different rates, inducing cyclic stress in solder joints. This thermal cycling is a major cause of failure, and its variable nature (e.g., daily cycles vs. power on/off cycles) is analyzed using these methods.
Common Misconceptions and Points to Note
When you start using this tool, there are a few common pitfalls to be aware of. First and foremost, understand that rainflow counting is not the end of the process. A frequent mistake is applying Miner's rule directly to the counted stress cycles without performing mean stress correction. For instance, a tensile mean stress reduces the fatigue life even under the same stress amplitude. You can select options like the "Goodman rule" in the tool's "Mean Stress Correction" feature. Make sure to input the correct material tensile strength $\sigma_B$ to strive for a realistic assessment.
Next, be careful with selecting the S-N curve parameters $S_e$ and $m$. Using textbook standard values can lead to significant prediction errors due to the influence of the actual material, surface finish, and environment (e.g., corrosion). For example, the fatigue limit can differ by more than 20% between polished and machined specimens of the same S45C steel. For reliable life prediction, the golden rule is to determine parameters from your own material test data whenever possible.
Finally, let go of the assumption that failure is guaranteed at a total Miner's rule damage of $D=1$. While Miner's rule is a convenient linear approximation, actual failure can occur at $D=0.7$ or $D=1.3$. This is because it doesn't account for the influence of load sequence (whether large loads come first or later). Always treat it as an "estimate." For critical components, applying a safety factor and setting design allowable values at $D=0.3$ or $0.5$ is a practical rule of thumb.