The comparison between laboratory results and industrial performances is particularly inaccurate in situations
where the industrial application involves considerable impact (or repeated impact), because laboratory test can
hardly simulate these impacts