Tamuang, W., Olarnrithinun, S., and Aue-u-lan, Y., "Evaluation of Low Cycle Fatigue Models of 4-Point Bending Test for Die Life Prediction in Forging Process ," SAE Technical Paper 2017-01-1739, 2017, doi:10.4271/2017-01-1739.
A forging process is a process used for producing automotive components such as power train components due to advantages of a high production volume and superior part’s strength. During repeated forging sequence, the forming dies are undergone high forming load which is normally closed to or higher than the yield stress of the die materials especially in the local area. That would be a major cause of a local fatigue crack formation and as a result limit life service of the forming dies. This type of the failure is known as a low cycle fatigue. Normally, the life service is less than 104. To improve the capability of the forging processes, the tool life needs to be known. Thus, to achieve reliable estimation of the tool life during the design, the material testing is required. Normally, the standard testing method with a strain controlled fatigue testing would cover the determination of the fatigue properties of the materials by using a tested specimen subjected to the Uniaxial stress. This kind of test is not suitable for estimating the die life in the forging dies due to the fully reversed loading conditions which in fact the forging die would only experience the half loading condition. So, the four point bending fatigue test was proposed and used in this study to evaluate the life service of the material. Furthermore, the reliable fatigue models to evaluate the testing results are still under the development. In this study, 2 models, namely Wohler’s based on the stress life approach and Manson-Coffin-Baskin’s based on the strain life approach would be discussed and applied to predict the life service based on the four point bending test. Both methods would be compared and evaluated with the existing experimental data presented by [Brondsted, et. al, 1997] and [Kim, et. al., 2010] to verify the accuracy of the models.