4.1. Measured temperature during the micromilling operation
For each test, a graphic similar to the example shown in Fig. 8 was generated. It is possible to observe the temperature measured by the three thermocouples, in which it is noted that there is a peak, indicating the temperature reached during the passage of the tool. Despite the micromill passing through the thermocouple during the tests, the maximum temperature measured does not refer to the moment when the tool is cutting the region of the thermocouple. This is due to the response time, the measurement frequency and the fact that the thermocouple detaches from the workpiece after the tool passes. Therefore, the maximum temperature recorded refers to the average temperature of a region close to the tool's cutting zone, where the thermocouple is fixed.
For some tests, some thermocouples were welded outside the tool's path. That is because it is difficult to keep the alignment. Among all the results obtained, the lowest temperature measured was 20.346 ° C, for the second test (56 mm machined length), in which the thermocouple was welded very far from the tool path, with a temperature difference of 0.366 ° C. The highest temperature measured was 68.832 ° C, in the 37 ° machined slot (1036 mm machined length), with a welded thermocouple on the dow-nmilling side, obtaining a temperature difference of 45.597 ° C.
The initial observation was regarding to the behavior of the measured temperature with respect to the position of the thermocouple. It was investigated if there is any trend regarding the position of the thermocouple, comparing results obtained for thermocouples fixes on the down-milling side, up-milling side and in the center of the slot. This analysis was done by comparing results obtained in the same test, in order to reduce the influence of possible differences in the execution of various tests and tool wear.
In 18 tests out of 46, there were thermocouples fixed in different positions in the workpiece surface were the slot was machined (up-milling, center and down-milling sides), making it possible to compare the temperature for differentpositions. In the other tests, the thermocouples were fixed in similar positions, or outside the slot machined.
It was found that in 13 tests, a similar trend is observed for the temperature behavior. The temperature increases during the tool movement in the active cycle. The temperature of the down-milling side was higher than that of the up-milling side. As for the 5 other slots, in 2 of them, the temperature on the up-milling side was higher and, in three, the temperature in the center of the slot was higher. It was not possible to evaluate the errors because it was not possible to fix the thermocouple in the same positions to repeat the test.
Figure 9 shows an example of the general behavior of temperature obtained in the tests. The graph shows the results obtained test 1 (machined length of 28 mm). The welded thermocouple, approximately in the center of the channel (T1), showed the smallest temperature difference, 4.646°C. For the second thermocouple (T2), positioned on the down-milling side, the temperature increases to 8.903°C. Finally, the temperature obtained by the third thermocouple (T3), which was fixed close to the end of the down-milling side, the measured temperature difference 10.283°C. The maximum temperature measured were 25.294°C, 29.653°C and 30.953°C.
The graphic in Fig. 10 shows another test in which this trend was observed. It shows the temperature differences, during test 8 (224 mm machined). Note that the thermocouple welded at the up-milling side of the channel, measured the smallest temperature difference (7.67°C) compared to the other two thermocouples. The thermocouple T2, also on the up-milling side, but more farther from the side of the slot than T1, measured a temperature difference of 9.224°C, slightly higher. The third thermocouple, fixed close to the center of the slot showed a greater temperature difference among the three thermocouples (12.829°C). The maximum temperatures recorded by T1, T2 and T3 were, respectively, 28.274 ° C, 29.815 ° C and 33.468 ° C.
Figure 11 shows an example of two thermocouples welded in similar positions, however on opposite sides, that is, one at the end of the down-milling side (T3), and the other at the end of the up-milling side (T2), enabling a comparison between them. The temperature T3 (down-milling side) is higher (temperature difference of 18.564°C) than temperature T2, at the up-milling side (temperature difference of 15.995°C). This emphasizing the trend found in most of the 18 tests in which such comparison was possible. The temperature T1 is not shown in the graph because the thermocouple detached from the workpiece surface before the tool passed it. The maximum temperatures recorded by T2 and T3 were, respectively, 39.914°C and 42.443°C.
Some authors, when comparing the cutting temperature between the down-milling and up-milling sides, found higher temperatures generated in down-milling side. Toh , when comparing the chip surface temperature generated between up and down milling orientations, during milling of a hardened steel AISI H13, observed that the temperatures generated when up milling were lower in all conditions analyzed, this difference being 3–8%. Ueda et al.  also found, in conventional milling of carbon steel, that the temperature at the tool tip is higher in the down milling than in the up milling, this difference being approximately 70 ° C.
Different researchers, as mentioned in the first chapter of this paper, found that the maximum temperature occurs in the center of the slot. However, it is noteworthy that most studies in the literature involved only analysis by means of computer simulation or few experimental tests, without further analysis of the thermocouple position.
According to the trend presented in this study, there is a suggested explanation for the higher temperature that occurred on the down-milling side. When the tool starts cutting, on the up milling side, the cutting temperature increases due to the conversion of work into heat. Through the propagation of the generated heat, the temperature in the regions close to the cutting zone rises. The temperature increases with the uncut chip thickness. As this variable increase, the heat transfer from the cutting zone to the surroundings also increases. Thus, on the down-milling side, the temperature increases, as the cutting edge approaches.
Contributing to this increase in temperature in the down milling side, there is the formation of large burrs. With a greater volume of material being pulled to this side, the greater the amount of heat that is conducted to the edge of the slot. Thus, the sum of these factors can rise the temperature during the movement of the cutting edge from the up milling to the down milling side.
The second analysis carried out with the results obtained was related to the influence of wear on the measured workpiece temperature. It is known that in conventional machining, wear contributes to increase the cutting temperature, however for micro machining, this is a phenomenon yet to be investigated. In order to carry out this analysis more appropriately, focusing on the influence of the position of the thermocouple on the cutting temperature, the measured values corresponding to similar positions in the different channels were selected. Therefore, in order to verify whether there is an influence of wear on temperature during the process, the values corresponding to the thermocouple marks on the up milling side, close to the edge of the slots, were considered first, in positions that presented the maximum similarity.
In the graphic of Fig. 12, it is possible to observe the temperature difference in certain tests in which there were thermocouples on the up-milling side, with the image of their respective marks. Between the machined length of 224 mm and 756 mm, the trend is for the temperature to increase. The temperature difference is 9.224°C after 224 mm machined length and 42.551°C after 756 mm machined length. It can also be seen in this graph that, after a machine length of 392 mm there is a large increase in the temperature difference, from 12.056°C to 35.256°C (560 mm machined).
Similar analysis was also performed for the down-milling side. In the graphic in Fig. 13, it can be seen that there is also a tendency for the temperature to increase with the machined length, or tool wear. In addition, it can be seen that the increase becomes more evident after test 16 (448 mm machined). From the first test, to the sixteenth, the temperature difference increases from 8,903°C to 10.909°C and after, for the 21st test (588 mm machined), the temperature difference was 22.649°C.
The results for the thermocouples fixed outside the tool path also indicate that the temperature increases with microtool wear. In Fig. 14, this trend can be observed, in which it is noted that there was a large increase in the temperature difference when comparing the first result with the last. When machining channel five (140 mm machined), there was a difference temperature of 2,471°C and, when machining channel 43 (1204 mm machined), the difference is 20,538°C.
Thus, it is noted that the temperature tends to increase with the machined length. As tool wear increases, the edge radius becomes larger and the tool less sharp. The cut, then, requires greater efforts and, consequently, the cutting temperature is higher.
4.2. Results of the Numerical Simulation
The numerical simulation was carried out with the main objective of complementing information on the temperature distribution. In Fig. 15, the image of a given moment in the simulation can be seen. It shows the similarity between the geometry of the tool used in the experimental tests with the simulated one. It can also be observed that in the cutting region, the elements are smaller and the mesh isdenser.
To analyze the behavior presented by the temperatures, several points were observed, along the tool's path. It is worth mentioning that two turns revolutions simulated for each cutting edge, totaling four paths. First, the maximum temperatures presented in the system and their evolution during machining were analyzed. In Fig. 16, it is possible to observe the trajectory of a cutting edge during its first turn, with the formation of a chip. Initially, the first edge starts machining on the up-milling side (Fig. 16 (1)). The temperature starts to rise towards the center of the channel, where it becomes maximum (Fig. 16 (5)). Continuing the movement towards the down-milling side, where the cutting edge leaves and the chip detaches (Fig. 16 (8)), the temperature begins to decrease. However, it presents higher values. Thus, the second edge starts the cut and the same behavior was observed. Throughout the simulation, the distribution of maximum temperatures was similar.
The highest temperature during the action of the first cutting edge was 72.62°C, with a maximum temperature difference of 52.62°C. Considering the two turns of both edges (4 complete paths), the highest temperature obtained was 83.48°C, in the center of the slot, during the trajectory of the second cutting edge, on its first turn. In addition, it is observed that the temperature increases only in the cutting region of the tool, corroborating the results obtained in the experimental tests in which the temperature increased only when the cutting edge is close to the thermocouple. An important factor for this behavior is the low thermal conductivity of the machined material, which causes most of the heat to be concentrated in the cutting region.
The temperature difference between workpiece, tool and chip was also simulated. In Fig. 17 it is possible to observe the temperature distribution in each one of these components separately. Note that the highest temperature is in the cutting zone and, when looking at the components separately, the largest temperature gradients are presented by the chip and workpiece. In the tool, the highest temperature is concentrated at the tip of the tool. However, the temperature reached by this one is lower than that presented by the workpiece and chip. This may be due to the low volume of material removed in this process, which makes the contact length between tool and workpiece / chip short.
This result is different from conventional machining, for which the higher temperature is in the tool cutting edge. Ezugwu and Wang  explain that, because the low thermal conductivity of the Ti-6AI-4V alloy, about 80% of the heat generated during the cut is conducted to the tool, when using cemented carbide cutting tools.
The values for the temperatures in the workpiece were similar to those measured during the experimental work. In Fig. 18, an example is shown where there is only the temperature distribution in the workpiece. It can be observed, at a point close to the tool's path, that the temperature was 28°C, which is similar to that measured in the first test performed, as shown in Fig. 18. However, it is worth mentioning that the simulation uses information from the program database, the results obtained being only an approximation.