PSI - Issue 76

Vladimír Mára et al. / Procedia Structural Integrity 76 (2026) 123–130

127

the hardness is decreasing regardless of Si concentration due to the lower amount of Fe intermetalics in the microstructure.

Fig. 3. (a) IPF map of as-built state illustrating <001> epitaxial growth; (b) IPF map of T300 showing negligible impact of heat treatment; (c) IPF map of T6 showing influence of SRX on microstructure; (d) The effect of heat treatment on Si content in the α -Al matrix and Vickers hardness.

3.2. High-cycle fatigue

The results of HCF testing are summarized by S-N (Stress - Life cycle) curves shown in Fig.4. Kohout & V ě chet (K&V) model (Kohout and V ě chet, 2001) was used for fatigue regression analysis, since it effectively captures the transition between the low-cycle fatigue (LCF), HCF and fatigue limit domain. K&V parameters are provided in Table 3. Majority of the S-N curves suffer from the high scatter of experimental data points. This is further intensified by the high failure rate in the threaded heads. Also, many points reside in the finite life region (area of 10 5 - 10 6 cycle). The worst fatigue performance exhibits specimens after T240 (see Fig.4a) and T300 heat treatment (see Fig.4b), where K&V regression only covers the steep slope, while the proper transitions between the quasistatic, LCF, HCF and fatigue limit domains are completely missing. Also, none of the annealed specimens at lower stress amplitudes were able to reach a runout condition.

Fig. 4. (a) S-N curves of batch 1 displaying noHT and T240; (b) S-N curves of batch 2 displaying T200 and T300; (c) S-N curves of batch 3 displaying T6 and T6mod.

Although the experimental data scatter is high, the K&V regression of as-built state (noHT) and T300 treatment reliably captures the data trend and displays the proper S-N curve. In terms of fatigue performance, noHT exhibits superior σ FL fatigue limit values (~95.9 MPa) in comparison to the T200, T240 and T300 specimens (79.8 MPa, 83.4 and 83.8 MPa respectively). Results are significantly different in comparison to previous study with R=0.1, where noHT exhibits the lowest values (Matuš ů et al., 2024a). Compared to other configurations, data scatter and thread break occurrence is significantly lower for T6 heat treatment (see Fig.4c). S-N curves are uniform for standard T6, while at HCF domain of T6mod, higher data scatter can be observed. In both T6 configurations, the cases of thread breaks are observed only above the threshold of 10 6 cycles. The σ FL values are nearly identical for both treatments: 85.8 MPa and 87.9 MPa for T6 and T6mod respectively. Although fatigue strength of T6 specimens in the HCF