PSI - Issue 14

Digendranath Swain et al. / Procedia Structural Integrity 14 (2019) 337–344 Swain et al. / Structural Integrity Procedia 00 (2018) 000 – 000

341

5

3. Results and Discussion

3.1. Residual stress behavior without heat-treatment

To understand the RS relaxation behavior, the strains measured during drilling of the sample-V3 before heat treatment is plotted in Fig. 2. Figure 2a compares the strain readings of the rosettes bonded on the two xz sides of the sample namely xz1 and xz2 (strain rosette names are 1 and 2, respectively). During drilling on the xz1 side, the strain rosette had some issues; hence the strain readings, up to a depth of 0.145 mm, were not considered. However, the drilling was continued at the same location after solving the issue. Since, there was less confidence on this measurement; drilling on the xz2 side was carried. Figure 2a shows that the strain readings were repeatable with similar directional behavior on both the xz1 and xz2 sides. Figure 2b compares the strain data on xz-2 plane and at the top xy built layer of the block on sample-V3 again without any heat treatment. An almost isotropic compressive strain distribution is seen on the xy top layer whereas the compressive strain distribution was anisotropic on the xz sides with the largest component along the z-directions. This comparison, in principle, indicates the tensile nature of RS induced in the DLMS process in various directions prior to heat treatment.

Fig. 2. Strains measured due to RS relaxation during the drilling of holes on specimen-V3 before heat treatment (a) comparison of strains on the two xz surfaces namely xz-1 and xz-2 and (b) comparison of the xz-2 and xy (top layer). The orientation of strain gage on the xz and xy plane is shown in the inset of (a) also. The direction of a-gage is along the z-axis whereas c-gage is aligning with x-axis. The RS evaluated using the strains in Fig. 2a are plotted in Fig. 3a, which shows a strikingly similar character of RS on the xz1 and xz2 faces. The peak stresses were induced at 0.2 mm depth. The RS magnitudes on xz1 side appear much larger than xz2 which could be due to the loss of the initial readings. The RS on xz1 surface would be discarded for any further discussions due to the rosette anomaly. It was also noticed that RS measured even on xz2 surfaces are of magnitude more than the ultimate strength (UTS) of the material and does not adhere to ASTM E387-13a standard. As far as the direction of the major principal RS was concerned, it was along the z-direction on both xz1 and xz2 sides, hence good repeatability could be seen (Fig. 3b). This implies that on the xz sides of a DMLS sample peak RS was along the built direction (z-direction). The repeatable nature of the major principal RS directions provides a good confidence on the measurements data obtained. As per literature the measured large magnitudes of RS are expected in Ti-6Al-4V AM components (Yadroitsava and Yadroitsev (2015)). Hence, the RS magnitudes on the xz sides can be taken as an indicator and a preliminary assessment of RS fields. It seems that hole drilling method of measuring such large magnitude RS may not be a viable procedure, since the maximum range of measurement is limited to 70% of yield strength (ASTM E387-13a non-uniform stresses). Therefore, one might have to resort upon alternative techniques of measuring RS in such 3-D printed components, cf. the contouring method. In Fig. 4a, a comparison of RS on the xz2 side and xy top layer is shown. RS on the xy layer was almost half of xz2 side. On the xy top layer the peak stresses are again seen at around 0.2 to 0.25 mm depth like the xz sides. However, both the major and minor RS magnitudes were similar in nature. As previously mentioned, the isotropic nature of strains measured during hole drilling is an a-priori indication of the equi-biaxial RS field. The surface stresses on xy top layer were compressive, whereas the xz sides had mostly tensile stresses throughout the 1mm