Can we describe kinetics of fatigue crack growth without influence of R-ratio?