Fatigue Load Spectrum Measurement

A program that analyzes data from inertia measurement units installed in aircraft and generates g-exceedance curves.

The main file to run the tool is main.py.

Load monitoring is a method of determining load spectrum which can be considered an aircraft experiences during its various missions. This load spectrum is used for estimating the fatigue life of the aircraft and is of great importance. A large category of airplanes is fatigue-critical as opposed to static-critical, therefore, it is crucially important to have reasonable and meaningful load spectra when designing for fatigue life.

This tool, summarizes the recorded acceleration data into various family of cycles with a particular range of mean and amplitude values, known as the mean-amplitude matrix. Using this matrix, a simplified N_Z spectrum is formed over which the level-cross counting is performed to produce a g-exceedance plot.

Despite having several flavors, rainflow algorithm is an industry standard for counting the hysteresis cycles. The version of the algorithm that is used in the current program to count the stress cycles is an implementation of ASTM E1049-85 in Python.

The image above shows the example presented in "ASTM E1049-85". The resulting cycle count is shown below.

Using the current rainflow counting tool, the same analysis results in the following graph. Comparing the last two images above and below shows that the current tool reproduces the same results as the ASTM. Although this is not a comprehensive validation, but can be relied on for now.

Since the data is proprietary, they cannot be presented here but below, a sample of the g-exceedance plot that is generated by the tool based on 3 flights with the average length of 4 hours is shown.

Additional plots can be produced using this tool as well including "Cumulative Range Count", "from-to matrix", "range-mean matrix" and the raw data as shown below.