Fatigue performance of superfinish hard turned surfaces in rolling contact

Salah Ramadan Agha, Purdue University


The purpose of this research is to study and model the performance of superfinish hard turned surfaces in rolling contact fatigue. The rolling contact fatigue is first modeled by taking into account the residual stresses generated by the hard turning process. Then, micro-hardness along with residual stresses produced were used for modeling. Results show that changing the cutting conditions will change the residual stresses and the micro-hardness produced and consequently change the fatigue life. Within the range of this study, it is shown that changing the cutting parameters will change the fatigue life by a factor of 40 times. It was also shown that high speeds and/or large flank wear will result in structural changes known as white layer, the presence of which was shown to drastically decrease the fatigue life. Moreover, it is shown that the rolling contact fatigue life of superfinish hard turned surfaces has a highly consistent repeatability as compared with ground surfaces. Finally, it is shown that hard turning results in softening the surface layers to varying degrees depending on the cutting conditions. In modeling the rolling contact fatigue life of the superfinish hard turned surfaces, first, the residual stresses were superimposed on the service stresses due to the loading, and then the fatigue life was modeled using the maximum equivalent stress (Von-Mises). It is shown that including the residual stress in the fatigue life model gives better prediction than that obtained by the maximum Hertzian stress which does not account for the residual stresses. Moreover, the inclusion of the maximum equivalent stress, its location and the volume-at-risk gives better results than using the maximum equivalent stress alone. The fact that the above models could not accurately predict the rolling contact fatigue life, prompted the author to include the micro-hardness distributions along with the residual stress distributions. The author used Linear Elastic Fracture Mechanics to develop a model that includes loading, residual stresses, and micro-hardness. The predicted lives using this model were found to correlate very well with the experimental data.




Liu, Purdue University.

Subject Area

Industrial engineering|Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server