On an node-by-node bases, the Mechanical Fatigue Tool (FT) calculates the alternating stress and mean stress for the specified loading. If "Mean Stress Curves" is specified as the Mean Stress Theory, the FT uses the S-N curve for that mean stress along with the calculated alternating stress to determine the number of allowable cycles. If the calculated mean stress lies between specified mean stress S-N curves, the FT interpolates between the curves.

The mean stress correction curves shown in the FT graphics window are just generic curves shown for visualization. They are not the actual curves used in the analysis. The actual S-N mean stress curves used in the analysis can be viewed in Engineering Data.

Why do you have a S-N curve with a large negative mean stress? The effect of negative mean stress is usually ignored. Can you show an image from Engineering Data of your curves