PSI - Issue 42

Matěj Mžourek et al. / Procedia Structural Integrity 42 (2022) 457 – 464 Matěj Mžourek / Structural Integrity Procedia 00 (2019) 000 – 000

461

5

3.3. Roughness Manual grinding followed the turning process of the smooth specimens, as demanded surface roughness values were not initially met. Roughness measurements of two specimens per series were subsequentially conducted via a Mahr mechanic profilometer MarSurf LD 120. Detected mean Rz values were used to modify the fatigue limit based on the FKM-Guideline, through the k R coefficient: = 1 − 0.22 · ( ) · ⁡(2 · /400) (3) This coefficient was normalized according to the k R,A01 parameter obtained for the A01 series, and applied to the fatigue limit of smooth specimens, changing the mutual position of the curves, see Table 3. for summary. The slope of the curves is modified by this correction, as the original point on the S-N curve corresponding to the ultimate tensile strength R m is kept fixed (thus expecting zero influence of the roughness effect in the static response, see Fig. 4).

Table 3. Results of roughness measurements on smooth specimens Specimen A01 A02 A03

A04 Mean Ra [mm] 0.806 0.822 0.686 0.310 Mean Rz [mm] 4.659 5.843 4.563 2.427 k R /k R,A01 [-] 1.000 0.982 1.002 1.052

Fig. 4. The correction of fatigue curves by the roughness effect coefficient k R applied at the fatigue limit. S u corresponds to R m .

No modification based on roughness was performed on the S-N curves of the notched specimens at this time. It is expected that the impact of surface roughness on fatigue properties is more pronounced in unnotched specimens. 3.4. Hardness mapping Vickers HV10 hardness in the cross-section of the semi-product bar was measured using a Struers Duramin 40AC3 automatic hardness tester. From the obtained 2D hardness maps, an average sub-surface hardness of HV 10 = 315 was detected for the smooth (A01-4) specimen and HV 10 = 320 for the notched (A51-2), as a rising hardness gradient was found to be present when moving in the radial direction from the center of the cross-section. Garwood et al. (1951) tested an equivalent SAE 4140 steel, and an expression was derived for estimating fatigue limit from measured values of HV , see Fig. 5. From the relative difference of the estimated fatigue limit for the two values encountered in specimens used here, the experimentally obtained fatigue limit of the notched specimens was reduced by 1.1 %. Similarly, the paper by Pavlina et al. (2008) was used to derive the modification of the tensile strength based on the cross-section of the evaluated specimen.