Magnetic flux gradient observation during fatigue crack propagation: A case study of SAE 1045 carbon steel used for automotive transmission parts

Abstract

The objective of this study is to evaluate the application of the metal magnetic memory (MMM) technique for investigations on fatigue crack propagation in a ferromagnetic material. Fatigue failure caused by stress concentration is serious in practical engineering. However, early fatigue damages cannot be detected by using traditional nondestructive testing (NDT) methods. Therefore this paper study about NDT method called metal magnetic memory (MMM) that has potentials for evaluating the fatigue damage at the early damage and critical fracture stages. While its capacity to evaluate the distribution of self-magnetic leakage field signals on the component’s surface is well-established, there remains a need to scrutinize the physical mechanism and quantitative analysis aspects of this method. To begin with, a fatigue test involving a loading of 7kN was conducted on a SAE 1045 carbon steel specimen. This material is frequently used in the manufacturing of automotive transmission components that include the axle and spline shaft. MMM signals were measured along a scanning distance of 100 mm and analysed during the propagation stage. Other than revealing that the value of the magnetic flux gradient signals dH(y)/dx increased in tandem with the crack length, the results also led to the detection of the crack growth location. It was anticipated that the dH(y)/dx value will also exhibit an upward trend with a rise in the fatigue growth rate of da/dN . A modified Paris equation was utilized to correlate dH(y)/dx with da/dn through the replacement of the stress intensity factor range Δ K . This resulted in the log-log plot of da/dN versus dH(y)/dx portraying an inclination similar to the log-log plot of da/dN versus Δ K . A linear relationship was established between dH(y)/dx and Δ K with the R2 value as 0.96. Players in the automotive industry can benefit from the disclosure that dH(y)/dx can effectively replace Δ K for the monitoring of fatigue crack growth behaviour.