PSI - Issue 5

Marek Smaga et al. / Procedia Structural Integrity 5 (2017) 989–996 Marek Smaga et al. / Structural Integrity Procedia 00 (2017) 000 – 000

994

6

from the surface was followed by a continuous decrease up to zero detected at the depth of around 200 µm. For both feeds only γ - austenite and α´ -martensite were detected at the surface while ε -martensite was detected only below the surface. The increase of feed led to an increase of the degree of deformation and an increase of the thermal load during the turning process. On one hand, the higher degree of deformation results in higher martensite fraction. On the other hand, with increasing thermal load the tendency to martensite formation is lowered Angel (1954), Hahnenberger et al. (2014). The shift of maximum of α´ -martensite fraction for feed of 0.35 mm/rev to a greater depth is thus a consequence of the higher thermal loads at higher feed. Figure 4d shows the residual stress distribution in the MSL t, f=0.15 , MSL t, f=0.35, and ASL t specimens. Generally, relatively high tensile residual stresses in the range of 800MPa were detected in the surface of all specimens, independent of the turning parameters and processes. The drop to zero residual stresses occurred within a surface distance of about 50 µm. For higher distances from the surface, residual stresses close to zero were measured. This distribution of macroscopic residual stresses is, generally, typical for turning processes and does not correlate with the austenite-to-martensite phase transformation. Figure 4e shows the development of micro hardness near the surface. The maximum micro hardness was measured for all three surface layers close to the surface. With increasing surface distance the micro hardness approaches to the core hardness of 231 HV0.01. The surface and near surface hardening can be attributed both to the formation of  - and/or  ’ -martensite but also to an increase in dislocation density. Nevertheless, comparing Figs 4b and 4e a clear correlation between the micro hardness and  ´-martensite content can be found within a group of various surface morphologies up to the depth of about 150 µm,. The MSL t, f=0.35 specimen shows in this depth range the highest values of micro hardness, the ASL t specimen the smallest values of micro hardness and micro hardness distribution in the MSL t, f=0.15 specimen is lying between above surface morphologies. Fig. 5: 3D micrographs of different surface morphologies as documented by CM for fatigue testing specimens (martensitic surface layer after mechanical and electrolytical polishing turned with CO 2 snow-cooling (MSL p ), austenitic surface layer after mechanical and electrolytical polishing (ASL p ), martensitic surface layer after turning with the CO 2 snow-cooling (MSL t, f=0.15 ), austenitic surface layer after turning without CO 2 snow-cooling (ASL t, f=0.15 ) and martensitic surface layer after turning with the CO 2 snow-cooling (MSL t, f=0.35 )). Besides phase distribution, residual stresses and micro hardness, among important factors influencing fatigue life belongs the surface topography. The surface topographies of all investigated specimen variants measured by confocal microscope are displyed in three dimensions in Fig. 5. The polished reference specimens with (MSL p ) and without martensite (ASL p ) possess a similar, very flat surface while a considerable roughness is charateristic for the as-turned samples. Topography profiles measured in a surface area of 4.8 mm × 0.8 mm were used to determine the roughness parameter R z (maximum height of profile) according to ISO 4287:1997. Comparing micrographs obtained for ASL t, f=0.15 and MSL t, f=0.15 samples the effect of CO 2 snow cooling on the surface topography can be clearly seen. In case of CO 2 snow cooling, the same turning parameters result in smaller surface roughness. During turning without CO 2 snow cooling, chip adhesion takes place on the cutting edge and the process is generally not stable. An increase of feed leads to an increase of the roughness parameter R z . Fatigue life of specimens cyclically loaded at ambient temperature with the constant stress amplitude  a = 270 MPa is presented for all five investigated surface morphologies in Fig. 6a. Results of fatigue tests performed for all surface morphologies at 300°C with the constant stress amplitude of 180 MPa are shown in Fig. 6b. The stress amplitudes were chosen from S-N curves of conventionally turned (without the CO 2 snow cooling) and mechanically / electrolytically polished specimens published in Skorupski et al. (2014) where at AT and T = 300°C, stress amplitudes of  a = 270 MPa and 180 MPa, respectively, led to fatigue failure at numbers of cycles in the range of N f > 10 4 , i.e. at 3.3. Fatigue life