PSI - Issue 30

Anna Zykova et al. / Procedia Structural Integrity 30 (2020) 216–223

219

Anna Zykova et al. / Structural Integrity Procedia 00 (2020) 000–000

4

1 mm/min. The dimensions of the test samples’ working part were 2.7 mm  2.5 mm  12 mm. Structural studies were performed using an Altami MET 1C optical microscope and an Olympus LEXT 4100 confocal microscope. The chemical composition of the filament and additive products from AISI 304 austenitic steel was determined using a NitonXL3t X-ray fluorescence spectrometer.

Table 2. Combinations of technological parameters for a series of 8 experiments. Number of experiment series Current Linear speed

Layer fill factor

1 2 3 4 5 6 7 8

– + – + – + – +

– – + + – – + +

– – – – + + + +

3. Results and discussion Figure 2 presents images of the appearance of the walls obtained by electron-beam formation according to the combinations of technological parameters indicated in Table 2. The appearance of the obtained walls shows that modes 1, 2, 4-6 make it possible to obtain samples from AISI 304 austenitic steel wire with satisfactory shapes and sizes. At the modes 2, 4, and 6 the electron beam current does not change and amounts to 65 mA, while parameters such as surfacing speed and wire feed coefficient vary. The combination of a moderate linear energy value (368 kJ/m) with a low austenitic steel wire feed coefficient ( k = 0.9), implemented in mode 4, allows the stable formation of the product with a satisfactory appearance (Fig. 3d). The high value of linear energy (650 kJ/m), implemented in modes 2 and 6, allows the austenitic steel wire to be completely melted regardless of the wire feed coefficient ( k = 0.9 or k = 1.3). At the same time, excessive melting of previously formed layers occurs (Fig. 3b). In mode 2, this leads to the penetration of the wire into the substrate to a greater depth, and in subsequent layers – to the melting and spreading of the molten wire along the wall (Fig. 2b). As a result, geometrically low and wide samples are formed, and the boundaries of the layers are practically indistinguishable, which is clearly seen in the image of the wall macrostructure (Fig. 2b). In mode 6 the melting of the formed product’s material is partially compensated by the increased volume of the wire with a larger value of its feed coefficient, which avoids the spreading of the molten wire along the wall and obtain a wall with even geometric parameters (Fig. 3f, 2f). An even better result is achieved in mode 5 (Fig. 2e) with a combination of a high value of wire feed coefficient ( k = 1.3) and also a moderate but slightly larger value of linear energy (400 kJ/m). As follows from a comparison of the geometric dimensions (and appearance) of the products formed in modes 1 and 5, in the mode 5 the increased by 8.0% linear energy made it possible to completely melt the feed wire without melting the previously formed layers (Fig. 3e, 2e). Whereas at a similar linear energy value (400 kJ/m) in combination with a low value of the wire feed coefficient ( k = 0.9), there is excessive melting of previously formed layers takes place in mode 1 (Fig. 3a), which leads to a violation of the wall geometry (Fig. 2a). Combinations of parameters of the printing process for modes 7 and 8 do not allow to carry out the process of forming walls. In mode 7 the product is not formed (Fig. 2g), due to the low linear energy value (225 kJ/m) in combination with the high value of the wire feed coefficient ( k = 1.3). Austenitic steel wire does not melt and the additive manufacturing is not implemented. The combination of the same value of wire feed ( k = 1.3) with a moderate value of linear energy (368 kJ/m) does not provide a stable process of product formation in mode 8 (Fig. 2h). The wire melts partially and the additive process stops already on the third layer. The combination of low linear energy values (225 kJ/m) and wire feed coefficient k = 0.9, implemented in mode 3, essentially allows the formation of the product (Fig. 2c). However even in this case, the supplied linear energy is not enough to completely melt the wire (Fig. 3c).