Real-Time Cutting Temperature Measurement in Turning of AISI 1045 Steel through an Embedded Thermocouple—A Comparative Study with Infrared Thermography

: During machining processes, a high temperature is generated in the cutting zone due to deformation of the material and friction of the chip along the surface of the tool. This high temperature has a detrimental effect on the cutting tool, and for this reason, it is of the utmost importance to assess the cutting temperature in real time during these processes. Despite all the advances and investigation in this ﬁeld, accurately measuring the cutting temperature remains a great challenge. In this sense, this work intends to contribute to solving this problem by experimentally evaluating the potential of the developed approach for embedding thermocouples into the rake face of cutting tools for measuring cutting temperature in real time during dry turning of AISI 1045 steel for different cutting parameters and comparing the obtained results with infrared thermography measurements at the exact same point. A well-deﬁned, smooth micro-groove with good surface quality was produced by laser surface modiﬁcation. Then a laser-welded K-type thermocouple was ﬁxated in the micro-groove with a MgO ceramic adhesive, ensuring protection from wear and chips, which allowed the creation of WC-Co cutting inserts with the ability to measure cutting tool temperature with a maximum error of 0.96%. Results showed that, despite yielding the same trend, the tool temperature measured by the IR thermographic camera was always lower than the temperature measured by the K-type embedded thermocouple. The proposed embedded thermocouple method proved to be a reliable, precise, accurate, and cost-effective approach for real-time temperature measurement capable of providing useful information for cutting parameter optimization, thus allowing increased productivity and tool life.


Introduction
Machining processes are not yet completely understood due to the highly non-linear nature of the process and the complex interaction between deformation and temperature [1].During machining processes, it is estimated that approximately 90% of the mechanical work applied to the workpiece is transformed into thermal energy, generating a very high temperature in a small area of the cutting tool (cutting zone) [2][3][4].This high temperature strongly influences tribological phenomena and adhesion, tool wear, tool life, workpiece surface integrity and quality, and chip formation mechanisms, leading to high operating Heigel et al. [27] measured the temperature at the tool-chip interface with an infrared thermographic camera while machining Ti-6Al-4V with a transparent cutting tool.This allowed measurement of the radiant temperature near the cutting edge, as well as observation of the chip curl and breakage.Hao et al. [28] compared the cutting temperature of uncoated and TiAlN coated cutting tools in orthogonal cutting of H13 hardened steel by using an IR thermographic camera.Ramirez-Nunez et al. [29] used infrared thermography to develop a smart sensor for the timely detection of tool rupture in the milling process in dry and with coolant conditions, being the smart sensor capable of providing heat curves, the thermogram, and the thermogram's average acquired during the monitoring of the cutting process.Pratas et al. [30] monitored the temperature during face milling of an Inconel 718 workpiece with a diamond coated cemented carbide end mill by infrared thermography and compared it with a boron doped diamond-based thermistor fabricated by HFCVD.
Additionally, other methods have been employed for rake face cutting temperature measurement.Li et al. [20] reported a novel fabrication process for embedding K-type thinfilm thermocouples into the rake face of cemented carbide cutting tools.Kesriklioglu et al. [31] fabricated thin-film thermocouples in a commercially available tungsten carbide cutting insert rake face for accurately measuring the tool-chip interface temperature during turning of AISI 1040 steel.Han et al. [32] obtained the real-time temperature in the cutting edge during turning by using a near-infrared fiber-optic two-color pyrometer inserted into a micro-hole below the tool rake face.Saelzer et al. [33] utilized a small two-color fiber-optic pyrometer with a diameter of 330 µm to directly measure the tool rake face temperature in orthogonal cutting of AISI 1045 steel and Ti-6Al-4V by preparing a workpiece with three slots for locally and temporarily interrupting the chip and, consequently, allowing measurement of the rake face tool temperature.This strategy was then employed by Afrasiabi et al. [34] to validate the results obtained by numerical simulation.
In this study, an embedded K-type thermocouple approach was developed for rake face cutting tool temperature measurement during dry turning of AISI 1045 steel.This approach differs from other embedded thermocouple studies reported in the literature because it doesn't resort to the creation of holes through the cutting tool, that can weaken the overall tool strength, to measure the temperature in the rake face [16,22,24,35,36].Different cutting parameters were evaluated to assess tool temperature, with the obtained results being compared with infrared thermography measurements at the exact same point, thus providing useful information for cutting parameter optimization and, consequently, allowing increased tool life and productivity.

Concept Design
Cemented carbides, notably WC-Co, are the most widely used materials for cutting tools due to their inherent high hardness, toughness, and wear resistance [37,38].For this reason, uncoated WC-Co commercial cutting inserts (CNMG 120408-GS, Palbit S.A., Branca, Portugal) were chosen to be used in this study.A K-type thermocouple was selected to be embedded in the rake face as close as possible to the cutting edge, but outside the cutting zone so as to not affect the cutting process.As shown in Figure 1, the thermocouple tip was placed at 2.9 mm from the cutting edge in a micro-groove produced specially for this purpose.High-temperature ceramic adhesive was used for thermocouple protection, insulation, and fixation in the micro-groove.

Micro-Groove Production
A Nd:YVO4 fiber laser (XM-30D, XianMing Laser, Liaocheng, China) with a maximum working power of 30 W, a laser spot size of 10 µm, a pulse width of approximately 10 µs, and a wavelength of 1064 nm was used to produce a micro-groove on the rake face of the WC-Co cutting inserts.The micro-groove was defined by using a computer-aided design system, with the laser parameters being optimized to obtain suitable dimensions for further thermocouple embedding and a better surface quality and smoothness.This process led to the selection of the parameters presented in Table 1.During laser processing, a jet of air was used to remove the produced debris and then all samples were ultrasonically cleaned in isopropyl alcohol for 5 min and airdried to remove any remaining debris and contaminants resulting from the process.

Embedded Thermocouple Production and Calibration
K-type thermocouples were manufactured by laser welding (LM-DVO 100, Sisma, Vicenza, Italy) the positive and negative leg wires (0.13 mm diameter), thus creating the hot junction.To ensure the thermocouple fixation in the micro-groove, as well as insulation between the thermocouple and insert material and thermocouple protection from wear and chips, a MgO ceramic adhesive (Resbond 906, Final Advance Materials, Didenheim, France) with a maximum working temperature of 1650 °C was used.The characteristics of MgO, such as high dielectric strength, durability, malleability, quick response to temperature fluctuations, mechanical stability at high temperatures, and extremely high melting point (2852 °C), make it the preferred option for thermocouple insulation material in a multitude of temperature measurement applications [39,40].

Micro-Groove Production
A Nd:YVO4 fiber laser (XM-30D, XianMing Laser, Liaocheng, China) with a maximum working power of 30 W, a laser spot size of 10 µm, a pulse width of approximately 10 µs, and a wavelength of 1064 nm was used to produce a micro-groove on the rake face of the WC-Co cutting inserts.The micro-groove was defined by using a computer-aided design system, with the laser parameters being optimized to obtain suitable dimensions for further thermocouple embedding and a better surface quality and smoothness.This process led to the selection of the parameters presented in Table 1.During laser processing, a jet of air was used to remove the produced debris and then all samples were ultrasonically cleaned in isopropyl alcohol for 5 min and airdried to remove any remaining debris and contaminants resulting from the process.
Table 1.Laser parameters used for micro-groove production.

Embedded Thermocouple Production and Calibration
K-type thermocouples were manufactured by laser welding (LM-DVO 100, Sisma, Vicenza, Italy) the positive and negative leg wires (0.13 mm diameter), thus creating the hot junction.To ensure the thermocouple fixation in the micro-groove, as well as insulation between the thermocouple and insert material and thermocouple protection from wear and chips, a MgO ceramic adhesive (Resbond 906, Final Advance Materials, Didenheim, France) with a maximum working temperature of 1650 • C was used.The characteristics of MgO, such as high dielectric strength, durability, malleability, quick response to temperature fluctuations, mechanical stability at high temperatures, and extremely high melting point (2852 • C), make it the preferred option for thermocouple insulation material in a multitude of temperature measurement applications [39,40].
The WC-Co inserts with the laser-welded K-type embedded thermocouple in the rake face were then calibrated by using PREZYS T-25N equipment.For the calibration tests, heating from 25 to 125 • C and then cooling from 125 • C to 25 • C at intervals of 10 • C with a stabilization time of 240 s for each step were performed.

Tool Temperature Measurement
The WC-Co cutting inserts with a K-type embedded thermocouple were then used to assess the tool temperature during single pass dry turning of AISI 1045 steel for different cutting parameters (see Table 2).These tests were carried out in a lathe machine (Leopard I, Gornati, Legano, Italy) for an evaluation length of 150 mm; the thermocouples were connected to a National Instruments signal acquisition system module (NI 9213), and three tests were performed for each experiment.Simultaneously, the temperature was monitored by an IR thermographic camera (FLIR E60, Teledyne FLIR, Wilsonville, OR, USA) that was placed above the tool rake face to measure the same exact point as the embedded thermocouple.The emissivity of an object is a crucial parameter when determining its temperature by IR thermography, the emissivity being dependent on the composition of the material and its temperature.Based on the literature [41][42][43], and since, in this study, uncoated WC-Co cutting inserts were used, thus presenting a matt surface, a value of 0.90 was defined for the emissivity.The temperature registered by the IR thermographic camera was then correlated with the temperature measured by the embedded thermocouple over the machining time.

Results and Discussion
In the present study, a micro-groove was produced on the rake face of uncoated WC-Co cutting inserts by laser surface modification in order to embed a K-type thermocouple for tool temperature measurement during dry turning of AISI 1045 steel.As shown in Figure 2a,b, the produced micro-groove matched the defined design (see Figure 1), being well-defined, smooth, and with good surface quality.Additionally, no microcracks, spatter, and heat-affected zones were observed, allowing to conclude that the laser parameters for micro-groove production were properly selected so as to not affect the cutting insert properties.The micro-groove surface exhibited some scanning marks, due to the action of the laser, which is potentially beneficial for the embedding process.As reported in the literature, surface roughening plays an important role in the bonding strength, since a larger bonding area contributes to a strength increase of the adhesive joint [44,45].
Figure 2c depicts the micro-groove surface topography analyzed by 3D optical profilometry, and the micro-groove profile is shown in Figure 2d.The laser parameters, presented in Table 1, were chosen to obtain a groove capable of guaranteeing a successful embedding of the K-type thermocouple, with an achieved micro-groove width and depth of 468.0 ± 7.7 µm and 439.3 ± 19.9 µm, respectively.Figure 2c depicts the micro-groove surface topography analyzed by 3D optical profilometry, and the micro-groove profile is shown in Figure 2d.The laser parameters, presented in Table 1, were chosen to obtain a groove capable of guaranteeing a successful embedding of the K-type thermocouple, with an achieved micro-groove width and depth of 468.0 ± 7.7 µm and 439.3 ± 19.9 µm, respectively.
Laser welding proved to be a viable method for thermocouple production, since it was able to form a strong hot junction with a spherical shape, as shown in Figure 3a.This type of junction, exposed junction, allied with the high thermal conductivity of the MgO ceramic adhesive, allows greater sensitivity and quicker response time when compared to an insulated or grounded junction [46,47].A hot junction diameter of approximately 250 µm was obtained, which is in line with the desirable dimensions to be fitted in the produced micro-groove.In Figure 3b it is possible to observe the K-type embedded thermocouple in the WC-Co cutting insert with the high temperature MgO ceramic adhesive.Laser welding proved to be a viable method for thermocouple production, since it was able to form a strong hot junction with a spherical shape, as shown in Figure 3a.This type of junction, exposed junction, allied with the high thermal conductivity of the MgO ceramic adhesive, allows greater sensitivity and quicker response time when compared to an insulated or grounded junction [46,47].A hot junction diameter of approximately 250 µm was obtained, which is in line with the desirable dimensions to be fitted in the produced micro-groove.In Figure 3b it is possible to observe the K-type embedded thermocouple in the WC-Co cutting insert with the high temperature MgO ceramic adhesive.Figure 2c depicts the micro-groove surface topography analyzed by 3D optical profilometry, and the micro-groove profile is shown in Figure 2d.The laser parameters, presented in Table 1, were chosen to obtain a groove capable of guaranteeing a successful embedding of the K-type thermocouple, with an achieved micro-groove width and depth of 468.0 ± 7.7 µm and 439.3 ± 19.9 µm, respectively.
Laser welding proved to be a viable method for thermocouple production, since it was able to form a strong hot junction with a spherical shape, as shown in Figure 3a.This type of junction, exposed junction, allied with the high thermal conductivity of the MgO ceramic adhesive, allows greater sensitivity and quicker response time when compared to an insulated or grounded junction [46,47].A hot junction diameter of approximately 250 µm was obtained, which is in line with the desirable dimensions to be fitted in the produced micro-groove.In Figure 3b it is possible to observe the K-type embedded thermocouple in the WC-Co cutting insert with the high temperature MgO ceramic adhesive.Figure 4 shows the temperature measured by the K-type embedded thermocouple in comparison with the reference thermocouple during the calibration test.To evaluate the reproducibility, three tests were performed in different WC-Co cutting inserts, showing the measurements' repeatable results.For the evaluated temperature range, a maximum error of ±1.2 • C was found, representing an error of 0.96% of the maximum temperature measured by the system (125 • C), which attests to the reliability and accuracy of the laser welding and embedding processes for producing WC-Co cutting tools capable of measuring cutting temperature in real time.
Figure 4 shows the temperature measured by the K-type embedded thermocouple in comparison with the reference thermocouple during the calibration test.To evaluate the reproducibility, three tests were performed in different WC-Co cutting inserts, showing the measurements' repeatable results.For the evaluated temperature range, a maximum error of ±1.2 °C was found, representing an error of 0.96% of the maximum temperature measured by the system (125 °C), which attests to the reliability and accuracy of the laser welding and embedding processes for producing WC-Co cutting tools capable of measuring cutting temperature in real time.The influence of different cutting parameters, namely cutting speed and depth of cut, on the tool temperature during turning of AISI 1045 steel was measured by the developed K-type embedded thermocouple and the obtained results were compared with infrared thermography measurements at the exact same point.
Figure 5 depicts the effect of cutting speed variation on the tool temperature for both methods at a constant depth of cut of 1 mm.As expected, the higher the cutting speed, the higher the tool temperature generation, with both methods producing this outcome.However, despite yielding the same trend, as depicted by curve fitting, the tool temperature measured by the IR thermographic camera was always lower than the temperature measured by the K-type embedded thermocouple.This is due to some difficulties encountered during the measurement process, such as chip obstruction of the measuring point and/or a defined value of emissivity that was lower than the real one, which leads to a lower temperature reading by the IR thermographic camera [16].Additionally, there is a tool temperature increase with the machining time, but a tendency to reach an equilibrium temperature is observed if the cutting time is long enough.The influence of different cutting parameters, namely cutting speed and depth of cut, on the tool temperature during turning of AISI 1045 steel was measured by the developed K-type embedded thermocouple and the obtained results were compared with infrared thermography measurements at the exact same point.
Figure 5 depicts the effect of cutting speed variation on the tool temperature for both methods at a constant depth of cut of 1 mm.As expected, the higher the cutting speed, the higher the tool temperature generation, with both methods producing this outcome.However, despite yielding the same trend, as depicted by curve fitting, the tool temperature measured by the IR thermographic camera was always lower than the temperature measured by the K-type embedded thermocouple.This is due to some difficulties encountered during the measurement process, such as chip obstruction of the measuring point and/or a defined value of emissivity that was lower than the real one, which leads to a lower temperature reading by the IR thermographic camera [16].Additionally, there is a tool temperature increase with the machining time, but a tendency to reach an equilibrium temperature is observed if the cutting time is long enough.
At the end of the evaluation length, cutting is halted, leading to a sudden decrease in temperature, with the maximum tool temperatures measured by the K-type embedded thermocouple being 178.07 • C, 246.65 • C, 256.36 • C, and 327.25 • C, for cutting speeds of 40 m/min, 70 m/min, 130 m/min, and 160 m/min, respectively.The similar maximum tool temperature obtained for cutting speeds of 70 m/min and 130 m/min, was due to the evaluation length not being enough to reach the temperature equilibrium, since a higher cutting speed leads to a lower machining time, meaning that higher cutting speeds need more time to reach the maximum tool temperature.As shown in Figure 5, the maximum tool temperature was reached for experiment 2 at 86.2 s of machining time, while for experiment 3A, the same temperature occurred at 40.0 s of machining time.It is also worth mentioning that for experiment 4, only 15.0 s were necessary to reach this temperature, while for experiment 1, the evaluation length was not sufficient.
In Figure 6, the effect of depth of cut variation on the tool temperature, for both methods, is shown for a constant cutting speed of 130 m/min.It is possible to observe that the tool temperature increased with the depth of cut increase, since a higher cutting depth leads to more material being deformed, leading to a higher heat generation in the cutting zone.Maximum tool temperatures of 256.36 • C and 340.03 • C were measured by the K-type embedded thermocouple for a depth of cut of 1 mm and 2 mm, respectively.At the end of the evaluation length, cutting is halted, leading to a sudden decrease in temperature, with the maximum tool temperatures measured by the K-type embedded thermocouple being 178.07 °C, 246.65 °C, 256.36 °C, and 327.25 °C, for cutting speeds of 40 m/min, 70 m/min, 130 m/min, and 160 m/min, respectively.The similar maximum tool temperature obtained for cutting speeds of 70 m/min and 130 m/min, was due to the evaluation length not being enough to reach the temperature equilibrium, since a higher cutting speed leads to a lower machining time, meaning that higher cutting speeds need more time to reach the maximum tool temperature.As shown in Figure 5, the maximum tool temperature was reached for experiment 2 at 86.2 s of machining time, while for experiment 3A, the same temperature occurred at 40.0 s of machining time.It is also worth mentioning that for experiment 4, only 15.0 s were necessary to reach this temperature, while for experiment 1, the evaluation length was not sufficient.
In Figure 6, the effect of depth of cut variation on the tool temperature, for both methods, is shown for a constant cutting speed of 130 m/min.It is possible to observe that the tool temperature increased with the depth of cut increase, since a higher cutting depth leads to more material being deformed, leading to a higher heat generation in the cutting zone.Maximum tool temperatures of 256.36 °C and 340.03 °C were measured by the Ktype embedded thermocouple for a depth of cut of 1 mm and 2 mm, respectively.The IR thermographs obtained over some machining times and the embedded thermocouple measurements show that the cutting tool heated up very quickly in the initial seconds, but over the machining time, this rate tended to decrease (see Figure 7).Additionally, this rate of temperature rise increased as the cutting speed and cutting depth increased, as shown in Figures 5 and 6.In Figure 7, it is also possible to observe the temperature distribution over the tool's surface, the chip formation, the K-type embedded thermocouple, the workpiece, and the cutting zone.The IR thermographs obtained over some machining times and the embedded thermocouple measurements show that the cutting tool heated up very quickly in the initial seconds, but over the machining time, this rate tended to decrease (see Figure 7).Additionally, this rate of temperature rise increased as the cutting speed and cutting depth increased, as shown in Figures 5 and 6.In Figure 7, it is also possible to observe the temperature distribution over the tool's surface, the chip formation, the K-type embedded thermocouple, the workpiece, and the cutting zone.
By comparing both measurement methods, it is possible to conclude that the obtained results highlight some of the disadvantages of the infrared thermography method, since smooth, accurate, and reproducible measurements were difficult to obtain.On the other hand, the proposed embedded thermocouple method seems to be a reliable, precise, accurate, and cost-effective approach for real-time temperature measurement.These results also allow us to conclude that the MgO ceramic adhesive ensured efficient thermocouple fixation as well as insulation and protection from wear and chips.
The IR thermographs obtained over some machining times and the embedded thermocouple measurements show that the cutting tool heated up very quickly in the initial seconds, but over the machining time, this rate tended to decrease (see Figure 7).Additionally, this rate of temperature rise increased as the cutting speed and cutting depth increased, as shown in Figures 5 and 6.In Figure 7, it is also possible to observe the temperature distribution over the tool's surface, the chip formation, the K-type embedded thermocouple, the workpiece, and the cutting zone.By comparing both measurement methods, it is possible to conclude that the obtained results highlight some of the disadvantages of the infrared thermography method, since smooth, accurate, and reproducible measurements were difficult to obtain.On the other hand, the proposed embedded thermocouple method seems to be a reliable, precise, accurate, and cost-effective approach for real-time temperature measurement.These results also allow us to conclude that the MgO ceramic adhesive ensured efficient thermocouple fixation as well as insulation and protection from wear and chips.

•
Laser surface modification was able to produce a well-defined, smooth micro-groove with good surface quality matching the defined design for subsequent thermocouple embedding.

Conclusions
• Laser surface modification was able to produce a well-defined, smooth micro-groove with good surface quality matching the defined design for subsequent thermocouple embedding.

•
The laser parameters used in this work were properly selected to not affect the cutting insert properties, since no microcracks, spatter, or heat-affected zones were observed.

•
Laser welding proved to be a viable approach for producing reliable, accurate, and precise K-type thermocouples with a maximum error of 0.96% of the evaluated temperature and achieving a hot junction diameter of approximately 250 µm.

•
The IR thermographs obtained over some machining times and the embedded thermocouple measurements showed that the cutting tool heated up very quickly in the initial seconds of turning, but over the machining time, this rate tended to decrease.

•
The proposed embedded thermocouple method was shown to be a reliable, precise, accurate, and cost-effective approach for real-time temperature measurement, thus providing useful information for cutting parameter optimization, and allowing increased productivity and tool life.

•
Additional studies will be performed to develop a reliable approximation of the temperature in the cutting edge by correlating the obtained results in this study with numerical simulations and/or analytical modeling.Additionally, the influence of tool wear, tool coatings, and tool geometry on the tool temperature measured by the approach developed in this study are other research lines that can contribute to the knowledge in this field.

Figure 1 .
Figure 1.Schematic representation of the K-type embedded thermocouple WC-Co insert.

Figure 1 .
Figure 1.Schematic representation of the K-type embedded thermocouple WC-Co insert.

Figure 3 .
Figure 3. (a) SEM image of the laser-welded K-type thermocouple hot junction; (b) SEM image of the WC-Co cutting insert with the K-type embedded thermocouple.

Figure 3 .
Figure 3. (a) SEM image of the laser-welded K-type thermocouple hot junction; (b) SEM image of the WC-Co cutting insert with the K-type embedded thermocouple.Figure 3. (a) SEM image of the laser-welded K-type thermocouple hot junction; (b) SEM image of the WC-Co cutting insert with the K-type embedded thermocouple.

Figure 3 .
Figure 3. (a) SEM image of the laser-welded K-type thermocouple hot junction; (b) SEM image of the WC-Co cutting insert with the K-type embedded thermocouple.Figure 3. (a) SEM image of the laser-welded K-type thermocouple hot junction; (b) SEM image of the WC-Co cutting insert with the K-type embedded thermocouple.

Figure 4 .
Figure 4. Calibration test of the K-type embedded

Figure 4 .
Figure 4. Calibration test of the K-type embedded thermocouple.

Figure 5 .
Figure 5.Effect of cutting speed variation on the tool temperature over the machining time (ap = 1 mm).

Figure 5 .
Figure 5.Effect of cutting speed variation on the tool temperature over the machining time (a p = 1 mm).

J 12 Figure 6 .
Figure 6.Effect of depth of cut variation on the tool temperature over the machining time (vc = 130 m/min).

Figure 6 .
Figure 6.Effect of depth of cut variation on the tool temperature over the machining time (v c = 130 m/min).

Figure 7 .
Figure 7. Example of IR thermographs obtained for experiment 1 over some machining times.

Figure 7 .
Figure 7. Example of IR thermographs obtained for experiment 1 over some machining times.
• WC-Co cutting inserts with the ability to measure cutting tool temperature in real time, with great sensitivity, quick response time, as well as protection from wear and chips, were obtained.• Despite yielding the same trend, the tool temperature measured by the IR thermographic camera was always lower than the temperature measured by the K-type embedded thermocouple.• Maximum tool temperatures of 178.07 • C, 246.65 • C, 256.36 • C, and 327.25 • C were measured by the K-type embedded thermocouple for cutting speeds of 40 m/min, 70 m/min, 130 m/min, and 160 m/min, respectively.• Regarding the depth of cut, maximum tool temperatures of 256.36 • C and 340.03 • C were measured by the K-type embedded thermocouple for a depth of cut of 1 mm and 2 mm, respectively.

Table 1 .
Laser parameters used for micro-groove production.

Table 2 .
Dry turning tests cutting parameters.