milling depth from the bottom surface of the plate. The Z posi- tions of the thermocouple sensors are shown in Figure 2 and given in Table 1. This build plate design was used for two different builds. Build 1 includes cylindrical samples built vertically above each thermocouple sensor. Build 2 includes a cylindrical sample built horizontally, such that the two ends are located above thermocouple sensors of equal depth. The schematics of Build 1 and Build 2 are shown in Figure 3. Build 1 is used to establish a baseline and monitor temperature accumula- tion during the process. Cylindrical columns of Inconel 718 2.54 cm (1 in.) in diameter are built above the first four ther- mocouple sensors (TC1–TC4), of varying Z positions, and 1.27 cm (0.5 in.) diameter columns above the second set of four sensors (TC5–TC8), of equal Z position as the first four. The purpose for this design is to identify the thermal environ- ment at the different four Z positions, representing the distance from the surface of the build plate. This will determine if the temperatures are hostile for other sensors, such as ultrasonic transducers, if they were to be attached at these depths for future experiments. More precisely, appropriate transduc- ers and Z position placements will be selected based on the findings in this experiment. There are two factors contribut- ing to the heat accumulation environment that is ultimately measured by the thermocouple sensors. The first is related to the cross-sectional area of the large-diameter cylinders, which is 4× larger than that of the small cylinders. Consequently, the duration of the heating and melting process is approximately 4× longer. Therefore, the heat intake at the large-diameter cyl- inders is also higher than the small cylinders by a factor of four. The second factor, however, is related to the heat dissipation rate, which is expected to be higher in the larger mass and heat sink effect associated with the large cylinders. This experiment will resolve which of the two competing factors will prevail. The temperature gradient across the part being built can result in residual stresses high enough to cause distortions and sepa- ration of the part from the build plate. The design of Build 2 is used to evaluate the effect of the part separation from the build plate on the temperature profile measured by a thermocouple ME | THERMOCOUPLEMONITORING TC1 TC1 TC5 TC6 TC7 TC8 TC5 TC6 TC7 TC8 Figure 3. Schematic demonstrating layup: (a) eight vertical cylinders in Build 1 (b) four horizontal cylinders in Build 2, with supports shown in red. The positions of some thermocouple sensors are shown for reference. Wiring channel TC1 TC2 TC3 TC4 TC5 TC6 TC7 TC8 Z position Build surface Figure 2. The design of the build plate and eight positions for the thermocouples. T A B L E 1 Z positions for all eight thermocouple (TC) sensors TC pair Z position TC1 and TC5 3.81 mm TC2 and TC6 7.62 mm TC3 and TC7 11.43 mm TC4 and TC8 15.24 mm 32 M A T E R I A L S E V A L U A T I O N • A P R I L 2 0 2 2 Z = 0.38 cm Z = 0.76 cm Z = 1.14 cm Z = 1.52 cm

sensor placed directly under it. The cylinder in Build 2 is built horizontally and has a diameter of 2.16 cm (0.85 in.). They are offset from the build plate by 5 mm (0.20 in.) tall fully dense volume supports. A grid support (50 hollow) was not used as it would provide an inferior thermal conductivity and weaker support against distortions from built up residual stresses. These cylinders are 13.46 cm (5.3 in.) in length, spanning the gap between a pair of thermocouple sensors of equal Z position (TC1 and TC5, TC2 and TC6, TC3 and TC7, and TC4 and TC8). They are used to monitor the temperature develop- ment at both ends of the cylinder and their anticipated separa- tion from the build plate during production. Results The temperature profile collected from thermocouple TC1 is shown in Figure 4. It is representative of the pattern observed through all the thermocouple sensors during the 30 h pro- duction of Build 1. More detail is shown in a 0.5 h (30 min) close-up segment from Figure 4, shown in Figure 5. When examining Figures 4 and 5 closely, the following observations can be made regarding the heat pattern detected by the TC1 sensor, which is representative of all eight sensors. The temperature reading begins at room temperature and rises rapidly as the build progresses. The initial rapid rise is attributed to the proximity of the melt layer to the sensors when the part is being built, the short height of the cylinders, and the cold build plate temperature at the start of production. As the build continues and the part’s height increases, the heat source is continuously moved farther away from the thermocouple sensors, causing the temperature to drop at the build plate. The temperature peaks between the 1 and 3 h marks, plateaus at the 5 h mark, and then begins to drop gradually until the end of production at the 30 h mark. This is demonstrated in the unaveraged data shown in Figure 4. The cool-down factor is relatively slow. At the conclusion of the 30 h production process, it takes the build plate approx- imately 5 h to return to room temperature, as shown in Figure 4. The rise from and drop to room temperature are both exponential. The peaks seen in Figure 5 are caused by the heat from the melt process, and the troughs are from the pause during the powder recoating cycle. Each data point is represented by a small square on the plot, collected at 12 s intervals. Five data points are encompassed between two consecutive peaks in Figure 5. This means that each layer is deposited in 60 s. The unique condition provided by the geometry of Build 1 is that the surface areas of the layers at all heights throughout the build are identical and will have the same melt duration and elapsed time. This will not be the case with the Build 2 design. Once every hour, an additional pause associated with the filter cleaning process increases the pause time and causes the temperature trough to drop a little lower. These drops are seen in Figure 4, as well as in Figure 5. The number of data points between the two peaks that encompass the filter cleaning trough in Figure 5 is eight, which accounts for three additional data points or approximately a 36 s cool-down time. The only difference the remaining seven thermocouple sensor measurements may exhibit from these observations derived from TC1 is in the peak temperature experienced at each location. Figure 6 provides an overview of the peak tem- peratures with respect to the Z position for each sensor and cylinder diameter. A few observations are made from the peak temperature measurements in Figure 6: The peak temperature is highest at the 3.81 mm (0.15 in.) Z position, as expected, because of its proximity to the heat source. Figure 4. Temperature readings from TC1 for Build 1: (a) unaveraged readings (b) with a moving average of five readings. 70 60 50 5 10 15 Time (h) 20 TC1 (unaveraged))g F 25 3530 40 30 20 ( lter cleaning 70 60 50 5 10 15 Time (h) 20 TC1 (average of 5) Filter cleaning 25 3530 40 30 20 50 51 52 53 54 55 56 57 58 5.0 5.1 5.2 5.3 5.4 5.5 Time (h) Filter cleaning pause Figure 5. Closeup of the unaveraged temperature readings from Figure 4. Arrow indicates filter cleaning pause, which results in a slightly lower drop in temperature. A P R I L 2 0 2 2 • M A T E R I A L S E V A L U A T I O N 33 T (°) T (°) T (°)