PSI - Issue 17

Ludvík Kunz et al. / Procedia Structural Integrity 17 (2019) 222–229 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

228

7

Comparison of da/dN vs  K curves from Figs. 3-5 for heat treatments at 380, 740 and 900 °C is brought in Fig. 10. In the Paris region (for the  K > 6 MPam 1/2 ) all curves for loading with the stress ratio R = 0.1 are practically identical. The experimentally determined crack growth rates are consistent with the results published by Cain et al. (2015) for heat treatment at 650 °C. They are also in agreement with data published by Leuders et al. (2013) for material heat treated at 800 C and Brüggemann et al. (2018) for orientation c and temperature 800 °C. The influence of the stress asymmetry in Paris region is quite weak. The crack growth rate for R = 0.8 is only twice of that at R = 0.1. Significant influence of the heat treatment and the loading asymmetry appears in the threshold region. For material heat treated at 740 °C the value 2.7 MPam 1/2 was found which is lower than 3.7 MPam 1/2 for the heat treatment at 380 °C. The explanation of lower value of the threshold may be explained either by differences in microstructure or on the basis of residual stresses. The microstructure after heat treatment at 740 °C consists of fine needles of  /  ʹ phase in the  matrix, Fig. 11. The morphology, as revealed on metallographic images, seems to be like that after the heat treatment at 380 °C, Fig. 8. The explanation of the lower threshold may rely on two mechanisms. The first one is the difference in microstructure, the second one are residual stresses. The microstructure as revealed by metallographic observations does not point to major differences. More detailed microstructural study e.g. by EBSD would be necessary to elucidate the possible differences which can explain the observed behaviour. Also fractographic observations, which are however not presented in this work, do not indicate differences in the near threshold propagation in material heat treated at 380 and 740 °C. The second possible influence on the crack growth is related to the residual stresses. The large scatter of experimental points for the as built material is attributed to the residual stresses, e.g. Leuders et al. (2013). The long range residual stresses, however, were completely removed by the heat treatment at 380 °C according to the X-ray measurement. On the other hand, the scatter of data in Fig. 4 is lower than that in Fig. 3. This can be interpreted in such a way that not only the long range residual stresses but also the short range residual stresses can influence the crack growth. To support this explanation determination of the short range residual stresses would be necessary. The crack growth curves for material heat treated at 380 and 900 °C are practically identical. For the  K > 10 MPam 1/2 , the da/dN curves coincide with the curve published by Cain et al. (2015) for the orientation c and heat treatment at 890 °C, see the line with squares in Fig. 10. The microstructure of the material heat treated at 900 °C is shown in Fig. 12. It consists of lamellas of (  + ) phases with locally coarse  grains. This structure is softer than structures after heat treatment at lower temperatures. The tensile strength of the alloy after tress relieving at 380 °C is 1227 ± 29 MPa in all directions. After heat treatment at 740 °C the tensile strength remains in this scatter. Only after heat treatment at 900 °C the tensile strength decreased by 200 MPa. The increase of the threshold is in line with the general understanding that with lower strength and higher plasticity thresholds are usually higher. The  K th value 3.5 MPam 1/2 is in excellent agreement with  K th = 3.48 MPam 1/2 found for the alloy heat treated at 650 °C for 3 h in vacuum with subsequent cooling in argon by Wycisk et al. (2014) on CT specimens produced by 200 W laser and 30  m layer thickness. From the paper by Leuders et al. (2013) it follows that there is a scatter in the threshold values. For the orientation c and heat treatment at 800 °C for 2 h the threshold value 3.9 MPam 1/2 was found and two measurements indicate values 3.7 and 4.8 MPam 1/2 for the orientation a . From the recent work by Brüggemann et al. (2018) it follows that for material processed at 800 °C for 2 h and orientation c the threshold  K th = 3.6 MPam 1/2 which is in excellent agreement with measurement performed in this work. The stress relieving heat treatment, which is a necessary post-processing of Ti6Al4V alloy manufactured by DMLS influences the microstructure and residual stresses which affect the characteristics of the propagation of long fatigue cracks. It has been found that under suitably chosen processing DMLS parameters the growth of long fatigue cracks in the Paris region is nearly identical after heat treating at 380, 740 and 900 °C. The influence of details of microstructure appears in the near the threshold region, below  K ~ 10 MPam 1/2 at R = 0.1. The lowest threshold value 2.7 MPam 1/2 was found for the heat treatment at 740 °C. The heat treatment at 380 and 900 °C results in nearly identical threshold  K th 3.7 and 3.5 MPam 1/2 . 5. Conclusions