PSI - Issue 28

Mohamed Ali Bouaziz et al. / Procedia Structural Integrity 28 (2020) 1039–1046 M.A. BOUAZIZ et al/ Structural Integrity Procedia 00 (2019) 000–000

1045

7

to ≈ 40 µm, and r y ≈ 20 µm. Conversely, for the material made of 125 µm layers, the average pattern radius in the transverse direction is equal to ≈ 65 µm, and r y ≈ 40 µm. This result shows that the strain field fluctuations are correlated to the layer thickness (i.e., the larger the thickness, the larger the pattern radii).

(a)

(b)

Fig. 5. Pattern radii histories for the (a) 50 µm and (b) 125 µm samples.

5. Conclusion In this paper, an experimental investigation was performed to study the behaviour of 17-4PH stainless steel obtained by ADAM. Two micro single edge notch tension (µ-SENT) specimens were fabricated with the same dimensions but with two different layer thicknesses (i.e., 50 µm and 125 µm). The micro speckle pattern deposited onto the surface of the specimen enabled for the quantification of kinematic fields through digital image correlation. Thanks to the speckle sizes and high resolution images, refined discretisations were possible, which resulted in detailed measurements of the kinematic fields. For both tested samples, ellipsoidal concentration zones appeared in the strain fields. A significant difference was noticed in the size of the strain patterns between both materials. Effects of layer thickness were investigated by estimating the pattern radii from the autocorrelation of the strain fields measured at each loading step of both tests. It was found that the pattern radii were virtually constant, which means that these patterns are a characteristic feature of 17-4PH stainless steel obtained by ADAM. When using 125 µm thick layers at the fabrication process, the pattern radii were about two times larger than those for 50 µm layers. This result correlates the layer thickness and the strain field patterns observed at the surface of tested materials. Additional tests are desirable to further investigate the correlation reported herein. Further, in the present study, a unique regularization length was selected to penalise the correlation cost function. Other lengths may be considered, and their effect on the reported results should also be investigated. Acknowledgements The authors wish to thank the FabAdd Platform of EPF Troyes for manufacturing the samples. The ‘Troyes Champagne Métropole’ and the Aube Department Council (France) are also acknowledged for the financial support of Markforged-Metal X System. This work was supported by Region Grand Est (France) via the funding NANO– FAB (18–GES–027) Soutien aux jeunes chercheurs, and its financial support for the Metal X System acquisition. References

Agarwala, M.K., Weeren, R.V., Bandyopadhyay, A., Safari, A., Danforth, S.C., Priedeman, W.R. 1996. Filament feed materials for fused deposition processing of ceramics and metals. 1996 International Solid Freeform Fabrication Symposium , 451-458.