Fatigue Analysis

Fatigue Tools and Stress-Life

• Architects and designers measure structural failure with fatigue tools. Usually designers create in-house fatigue tools that test the fatigue on a building or other structure. However, ANSYS is a new fatigue analysis tool that tests loading, mean-stress effects, and other issues using a stress-life approach. With ANSYS systems, designers test fatigue data saved in XML format that shows the mean-stress dependent or with multiple r-ratio curves, as long as data is available. Fatigue material data is then stored through ANSYS as tabular, alternating stress vs. life points. Fatigue results generate after a fatigue tool inserts into the structure, then creates an analysis of loading type, handling of mean stress and more.

Total-Life Approach

• As opposed to the stress-life approach, the total-life approach uses S-N curves to show the number of cycles to failure for each load case. Miner's rule combines the damage of each case to give a total amount of damage analysis to the entire structure after the loading test. S-N curves use log values, because the variations are larger. Fatigue spectra determines the loading sequence after the loads are performed on the structure, each of which generates a load factor and number of cycles. Another fatigue analysis program, LUSAS, uses the total-life approach to calculate fatigue results. You may view results in loglife, which shows the life expectancy of a structure, depending on the level of loads measured in cycles to failure. The other result shows in damage, which is the amount of damage that a structure takes under fatigue loading.

Loading

• Stress creates fatigue on the object. Loading stress is the most common type of fatigue. Fatigue changes over time, depending on the load as it changes over time, meaning that stress becomes greater over long periods of time after significant loading on the structure. Different load types to analyze are the following, according to ANSYS. Constant amplitude, proportional loading, called "back of the envelope," with a loading ratio that shows the ratio of the second load to the first load as "LR = L2/L1," based on a proportional loading that uses only one set of finite element stress results, thus eliminating cumulative damage. The second type of loading is nonconstant amplitude, proportional loading, which means that, like constant amplitude, only one set of results is needed, but the load ratios vary over time, which creates a cumulative damage analysis. See Resources for figures and further explanation.

Mean-Stress Correction

• As fatigue analysis shows stress from loading, the mean stress must be accounted for. Interpolation is one way to account for mean stress by looking at the material curves. Static material also generates S-N data to show mean stress in total life approaches. Other mean-stress handling theories are Gerber, Goodman and Soderberg. According to ANSYS, the Gerber theory treats negative and positive mean stresses as the same, but Goodman and Soderberg do not account for any correction in negative mean stresses. Goodman and Soderberg do not use correction in these cases because they are more conservative theories and do not acknowledge compressive mean stress, which shows slower fatigue-crack growth in structures.