Abstrad-Fatigue crack growth behaviour under block loading sequences has been studied on the aluminium alloy 2024 T351, using four different equivalent constant amplitude concepts. The root mean squared method gives acceptable results only for relatively long blocks. The equivalent method based on the Pans law (modified to take into account crack closure) gives good predictions for the observed growth rates as does Elbers' method. Finally, a new method based on energy considerations gives excellent results for the studied test conditions. NOMENCLATURE a = crack length BL = block loading da/dN =crack growth rate i = cycle number ECA =equivalent constant amplitude K = stress intensity factor KO = stress intensity factor at crack opening Kmus = maximum stress intensity factor in a block K,, K-= maximum and minimum K levels in a cycle m, rn' = slopes of Pans' law and Elbers' relation respectively N = number of cycles P = load Nq = number of equivalent cycles P , , = maximum load in a block P , , Pd =maximum and minimum loads at cycle i P = crack opening load B = effective mean level Q = energy dissipated per cycle Qi =energy dissipated at cycle i in a block Q,-, = enclosed energy RMS = root mean squared R = load ratio %p = equivalent R ratio R, = load ratio corresponding to cycle i in a block LIE = Elbers' efficiency factor VA = variable amplitude a = specimen compliance AK = stress intensity factor range Qm, = total energy dissipated per block AKem = effective AK 6 =crack mouth opening displacement 998 N. RANGANATHAN AND J. R. DESFORGES