Why the n parameter should be lower than 1 in Johnson-Cook model?

Hi all,

In the comments of the /mat/law2 Johnson-Cook, it shows The plastic hardening exponent, n must be less than or equal to 1. Why this limit? From my test data, I do the fitting, and found that the n was large than 1. 

<?xml version="1.0" encoding="UTF-8"?>