A Validity of Estimation Methods of Total Strain-Fatigue Life Curve about Copper and Copper Alloys

Abstract

Before now, several estimation methods of total strain-fatigue life curve (Δεt-Nf curve) formulated by mechanical properties have been proposed. In these estimation methods, Universal slopes method, Modified universal slopes method and Medians method can be utilized effectively for preliminary fatigue design. However, it appears that the validity of three estimation methods about copper and copper alloys has been not investigated. Accordingly, in this report, actual fatigue life (Nf) of copper and copper alloys were compared to predicted fatigue life (Np) obtained by each three estimation methods in order to investigate their validity.The conclusions are obtained as follows :(1) Equation approximated on Medians method by strain controlled fatigue test results of copper and copper alloys was suggested as follows.Δεt=(2.70σB/E)Nf-0.0805+0.543Nf-0.529(2) Precision of Np obtained by Universal slopes method was the worst in three estimation methods. In addition, this Np was non-conservative prediction under 104cycles and conservative prediction over 105cycles. Consequently, It is decided that Universal slopes method was not practical for preliminary fatigue design.(3) Precision of Np obtained by Modified universal slopes method was better than one of Np obtained by Medians method. But, the advantages and disadvantages of two estimation methods could not be judged, because precisions of two may be change according to increase of experimental results. At the present time, It is thought that there is no difference of a practical use for preliminary fatigue design between Modified universal slopes method and Medians method.