3. Results

3.1 Fatigue life

Fatigue test results are presented in Fig. 7-a and Fig. 7-b as semi-log S-N curves (Wöhler curves), with S corresponding to the maximum stress reached in the working section of the specimen during a fatigue cycle and N corresponding to the number of cycles leading to failure. Power trend curves are plotted on these graphs. For reasons of confidentiality, the stress level S on the S-N graphs, as well as the values of the surface integrity parameters presented in the following sections (roughness, microhardness and nanohardness), are normalized by their maximum value.
S-N curves obtained for D = 6.35 mm show no difference in fatigue strength between the two drilling processes. In contrast, S-N curves obtained for D = 9.53 mm show a significant increase in fatigue life for the specimens drilled by the axial technique. This gain in fatigue life can be observed for all stress levels but is greater for the lower ones.
In the aircraft industry, the fatigue performance of a structure is commonly evaluated by the Airbus Fatigue Index (AFI), which corresponds to the stress S associated with a forecast fatigue life of 100 000 cycles. Thus, AFI values were determined for all configurations from S-N curves. They are shown in Fig. 7-c and Fig. 7-d (as values normalized by the maximum AFI value obtained).
For D = 6.35 mm, similar AFI values can be observed for both drilling processes whereas, for D = 9.53 mm, a gain in AFI of 15% is visible with the axial drilling process.
The fracture surfaces were observed for all drilling configurations. No difference in failure mode was noted between the different configurations. For all specimens, the crack initiated in the hole edge area and the fracture surface was divided into two distinct zones: the crack propagation zone and the final fracture zone (Fig. 8).
Thus, the drilling configuration can have a significant influence on the fatigue strength of the drilled part but it has no impact on the failure mode of the specimen. The differences in fatigue life observed between the drilling configurations were probably related to differences in the hole surface integrity. Characterization test results of the hole surface integrity are presented in the following sections.

3.2 Roughness measurements

Among all the roughness parameters measured, only the results of arithmetic average height of profile (Ra), which gives an average indication of the overall surface topological state, are presented here. They are shown in Fig. 9.
For all drilling configurations, the measured Ra value complies with the aeronautical specification. No correlation can be established between the Ra results and the fatigue test results. A significant difference in Ra is observed between the two drilling processes for D = 6.35 mm whereas no difference in fatigue life is noticed between the two drilling techniques for this hole diameter. This is probably related to the relatively low Ra values measured compared to the aeronautical specification. The Ra values are not high enough to have an impact on the fatigue strength of the part.
The possible relationship between the hole roughness and the fatigue life results was also studied for all the other 2D and 3D roughness parameters measured. No obvious correlation between the roughness parameters and the fatigue strength of the drilled part can be highlighted from this analysis. As for Ra, this is explained by the low roughness values measured.
Thus, in this study, the hole roughness is not a major factor influencing the fatigue life of the drilled part.

3.3 Hardness measurements

After analysis of the topographic aspects of the hole surface integrity, the metallurgical and mechanical aspects were investigated. The Vickers microhardness measurements carried out in this framework are presented in Fig. 10.
For D = 6.35 mm, holes obtained by axial and orbital drilling processes have similar microhardness levels whereas, for D = 9.53 mm, a significant difference in microhardness level is observed between the two drilling techniques. For this diameter, a gain of 29% in microhardness level can be seen for axial drilling.
These results show a correlation of fatigue life results with microhardness measurements: the increase in hole microhardness for axial drilling at D = 9.53 mm seems to explain the gain in fatigue life observed for the same configuration (Fig. 7).
To study this result further, nanoindentation tests were performed. The test results are presented in Fig. 11 as graphs showing the variation of the nanohardness according to the distance from the hole edge.
For D = 6.35 mm, nanohardness results show a similar material depth affected by hardness variation, around 15 µm for both drilling processes. In contrast, for D = 9.53 mm, a significant difference in material depth affected is observed between the two drilling techniques. The material depth affected for orbital drilling seems to be very small (less than 5µm) whereas it is around 40 µm for axial drilling. Also, during indentation tests for the specimen corresponding to axial drilling at D = 9.53 mm, the first indents column induced quite small, irregular prints from which no hardness value could be calculated. This indicates that the nanohardness value of the hole subsurface obtained for this configuration seems to be high.
These results highlight the thinness of the subsurface material layer affected by the drilling operation of a 2024-T351 aluminium part, in agreement with the numerical model. The nanohardness results are in accordance with the microhardness results. The hole hardness seems to be a major factor influencing the fatigue behaviour of a drilled part with a positive influence.

3.4 SEM-EBSD analysis

In order to identify the phenomena responsible for the difference in hole surface hardness, a metallographic analysis was carried out. In this framework, GROD maps of the hole edge area obtained through SEM-EBSD analysis (Fig. 12) were analysed.
The GROD map obtained for axial drilling at D = 9.53 mm shows a significant material depth affected by grain misorientations, around 40 µm depth. This affected material layer varies slightly according to the grains. In contrast, almost no grain misorientation, or only some a few microns deep, is observed for the other drilling configurations. A material layer affected by grain misorientation corresponds to a strain hardened material layer. This means that axial drilling at D = 9.53 mm induces a significant strain hardening of the hole subsurface compared to other drilling configurations.
These results explain the microhardness and nanohardness results. The increase in hardness in the hole surface and subsurface, for axial drilling at D = 9.53 mm, seems to be due to the significant strain hardening of the material in the hole edge area. This assumption is strengthened by the fact that the affected material depths identified with the nanoindentation tests and with the SEM-EBSD tests are the same.

3.5 Residual stress analysis

The residual stress state of the drilled part, which, according to the literature, can also have a significant impact on the fatigue life, was investigated through the novel HOCT technique. Specimen deformations obtained with HOCT are presented in Fig. 13.
No deformation of the specimen is noticed after the material removal for the case of axial and orbital drilling at D = 6.35 mm and the case of orbital drilling at D = 9.53 mm. However, significant specimen deformation is observed for axial drilling at D = 9.53 mm. This means that this drilling configuration induces significant residual stress fields in the hole edge area compared to the others.
These results show a correlation of the fatigue test results with the residual stress state of the material. A high fatigue quality index is associated with the presence of significant residual stresses in the hole edge area. Furthermore, a high residual stress state is associated with high hole hardness.