PSI - Issue 47

Jan Patrick Sippel et al. / Procedia Structural Integrity 47 (2023) 608–616 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

615

8

, = · · √ · √

(1)

The generator data of the AISI 4140 specimen is displayed in Figure 8, with the excerpt regarding power, frequency and amplitude in Figure 8 a, before the failure analogously displayed to Figure 7 a.

Fig. 8. (a) Excerpt of the logger data of an AISI 4140 specimen before final failure; (b) Close up of the natural frequency and amplitude data during failure; (c) Defect size of the fisheye responsible for the beginning decrease in natural frequency and amplitude.

In Figure 8 b, focusing on the final failure of the AISI 4140 specimen, we observe a less pronounced and more prolonged decrease of the natural frequency as well as of the amplitude during several pulses without unstable crack propagation. The maximum SIF at the border of the fisheye while touching the surface can be calculated using Equation 1, with C = 0.65 for surface defects and the red area marked in Figure 8 c. In case of AISI 4140 this maximum SIF value of K max,FiE = 23.8 MPam 1/2 does not exceed the fracture toughness of K C,AISI 4140 = 32.7 MPam 1/2 and therefore, the unstable crack propagation is not expected to happen at this position of the crack tip. We can conclude that the SIF, as well as the crack propagation rate is altered due to the amplitude drop resulting in the observed beach marks as well as the heterogeneous fracture surface areas previously discussed. The alternation between force and fatigue fracture surface area indicates the interaction between the increasing defect area due to the crack propagation and the declining amplitude. The brighter surface area band in the middle of the specimen might indicate a maximum SIF close to the fracture toughness, however the growing crack seems to reduce the residual cross section of the specimen. Due to this cross-section reduction the frequency decreases resulting in a decreased amplitude to such an extent, that the maximum SIF drops resulting in the fatigue fracture surface area shown in Figure 6 h. Further planned investigations include an FEM analysis to derive the relationship between crack length and the natural frequency as well as tests with artificial defects and higher data acquisition rate regarding natural frequency as well as amplitude. Although these investigations are needed to derive a direct correlation between the maximum SIF and the specific fracture surface characteristics, the following conclusions can already be drawn at this stage. 4. Conclusion Fracture surfaces of specimens from AISI 52100 and AISI 4140, with martensitic microstructure, display significant differences after ultrasonic fatigue testing. After analyzing these fracture surfaces, as well as the generator data regarding power, frequency and amplitude of the ultrasonic test setup the following conclusions can be drawn: • Depending on the material, either an abrupt or a slower and prolonged decrease of natural frequency and amplitude is observed. • The (unstable) crack propagation and the resulting reduction of the natural frequency, which causes the reduction of the amplitude, seem to compete. • If the fracture toughness is reached before the natural frequency drop is significant, no influence of the resulting amplitude drop on late crack propagation, or the fracture surface is evident (AISI 52100). • However, if the fracture toughness of the material is higher, the natural frequency is significantly decreased due to the longer crack length before unstable crack propagation is reached and the further crack propagation is significantly altered due to the amplitude reduction (AISI 4140).