Mesocale stresses in Additively Manufactured Ti-6Al-4V

Strantza, M., Vrancken, B., Prime, M. B., Truman, C., Rombouts, M., Brown, D. W., Guillaume, P., and Van Hemelrijck, D., 2019, "Directional and oscillating residual stress on the mesoscale in additively manufactured Ti-6Al-4V," Acta Materialia, 168, 299-308. https://doi.org/10.1016/j.actamat.2019.01.050. preprint (pdf).

RESULTS: First ever measurement of layer to layer stress oscillations in AM part Unprecedented resolution of oscillating stresses on the mesocale! The scale of the 0.5 mm thick build layers In fact, we were able to reslve the oscillations over nearly 30 layers:
Titanium AM specimens: Ti-6Al-4V Produced with Laser Metal Deposition process The contour of each layer was deposited first and then filled via a continuous bidirectional scan strategy with a spacing of 600 µm between neighbouring scan tracks The layer thickness was 500 μm and the scanning direction of successive layers was rotated by an angle of 90°
Most slit depth increments ever for a sliting measurement: 161 total increments Cut slot using 100 µm diameter brass wire Slot cut in 51 µm increments (0.002") initially, then coarser Strain at each depth was measured with strain gage on back surface opposite the cut
Strain data shows mesoscale: Strains show long scale variations: strains tend towards compression Looking closely, there is superimposed a sinusoidal variaiton at 1 mm wavelength Can see this by fitting macroscale variation with smooth curve and then plotting misfit to data Strains alternate every 0.5 mm: the size of the build layers
Only "pulse method" able to fit data: These are the strains that result from the stress results at the top of the page Quality of inverse solution for stress can be judged by how well it fits (reproduces) strain data. This stress variation cannot be well fit using series expansion Because of mesoscale variations Was able to fit well using "pulse" method. Misfit between inverse solution and data is mostly less than 1 microstrain - just noise!