Optical Temperature Sensor Evaluation in a Working Gear Motor: Application of Luminescence Thermometry in Industrial Technology

3D printing constitutes a technological advancement, revolutionizing contemporary industry by enabling manufacturers to fabricate intricate, customized components efficiently and precisely from digital blueprints. Moreover, the fusion of 3D printing with cutting‐edge materials has given rise to compelling elements boasting a diverse range of applications. For this reason, this work describes the incorporation of a luminescent material, NaYF4, doped with Yb3+ and Er3+, embedded in resin for 3D printing to create motorized luminescent gears. The fabricated luminescent gears take advantage of the intensity ratio between the Er3+ emissions at 525 nm (2H11/2 → 4I15/2) and 550 nm (4S3/2 → 4I15/2), which are thermally coupled, to detect the slight temperature variations that gears undergo through friction. This technique can be complementary to thermovision, proving especially valuable for monitoring temperature in elements where measurement with thermographic cameras or direct contact thermometers is hampered. The findings demonstrate that optical measurements provide enhanced (statistical) precision in temperature readings compared to thermovision, with δT = 0.07 K for luminescence thermometry as opposed to δT = 0.3 K for the thermal camera. This work can inspire new research directions using 3D printing and materials with exciting properties, fostering innovative solutions in contemporary industrial technologies.


Introduction
Currently, 3D structure printing is growing in multiple fields of life, from the most straightforward, such as the design of models and the creation of small elements for personal use.To uses with a much more significant impact on society, such as the design of prostheses for people with health problems, its application in architecture for scale models of large projects, or its implementation in manufacturing tools and industrial elements. [1,2]12] P. I. Martín-Hernández, F. Rivera  Among them, most are the remote optical (luminescent) sensors that use the unique spectroscopic properties of lanthanide ion-doped materials to estimate temperature remotely.This can be done by using diverse thermometric parameters such as bandwidth, emission band shift, and luminescence intensity ratio (LIR), which are often very sensitive to temperature changes, enabling high sensitivity and (statistical) precision optical thermometry. [13]][16][17] Moreover, the materials doped with Er 3+ ions can be excited with low-energy NIR light and emit a visible, bright green light associated with the energy upconversion phenomena occurring in various inorganic compounds. [5,18]herefore, in this work, we used the NaYF 4 nanoparticles doped with Yb 3+ and Er 3+ ions as a remote optical temperature sensor based on a non-linear, upconversion luminescence (anti-Stokes process) excited with 975 nm NIR laser, which is transparent to many materials and has high penetration depth.Once calibrated, the relationship between the intensity ratio of the 525 nm and 550 nm bands and the temperature (295-375 K range) can be used to monitor temperature values in diverse systems optically.We mixed a resin with the mentioned luminescent material to fabricate gears and determine the temperature that is reached due to the friction between the teeth in the gears (heat generation due to mechanical work), thus demonstrating the use of luminescent materials as remote temperature sensors in real-world applications, i.e., in the field of industrial technologies (Figure 1).Finally, we compared our results with temperature results from a thermographic camera, i.e., a complementary technique for contactless temperature detection.

Fabrication of 3D Gear
The resin employed was a Simple Siraya Tech, doped with a 5 wt% of NaYF 4 :Yb 3+ , Er 3+ and introduced in an ultrasonic cleaner to homogenize.The prototypes were designed using Fusion 360 software, and the digital file was introduced in Lychee Slicer to select the printing parameters.From this model, the gears were printed by Stereolithography using an Anycubic Photon Ultra 3D printer (see Figure 2).The final pieces were cleaned with a 99.9% isopropyl alcohol and dried before measurements.

Temperature Sensing
The excitation of the sample was made with a tunable continous wave (CW) Ti:Sapphire laser system, Spectra-Physics 3900S, pumped with a 532 nm Spectra-Physics Millenia, adjusted at 975 nm using a lens before the sample, that gave flux of ≈0.6 W cm −2 in the spot on the sample (on-target power density), that was measured using a power meter (Ophir StarLite).The changes in the bands associated with the 2 H 11/2 → 4 I 15/2 (≈ 525 nm) and 4 S 3/2 → 4 I 15/2 (≈ 550 nm) transitions are due to thermal coupling between the 2 H 11/2 and 4 S 3/2 levels, which are separated with energy of ca.800 cm −1 .The emission spectra were recorded in the visible range using an Ocean Optics HR4000 High-Resolution Fiber Optic Spectrometer, as seen in Figure 3a.As temperature increases (using a Linkam system, THMS600), the relative emission intensities associated with 525/550 nm bands change significantly.In order to use these changes as a thermometric parameter and develop a temperature sensor, the ratio of the integrated intensities of the emission bands centered at 525 and 550 nm has been calculated.The determined LIR values as a function of temperature are presented in Figure 3b.As the temperature increases, the population of the upper level 2 H 11/2 is growing; therefore, the corresponding emission intensity from that level rises with respect to level 4 S 3/2 .The dependence follows the Boltzmann distribution law: where LIR is the luminescence (band) intensity ratio (525/550 nm); ∆E is the energy difference between the analyzed excited states; k B is the Boltzmann constant; T is the absolute temperature; and A is a constant that depends on the states degeneracies, branching ratio of the transitions in relation to the ground state, spontaneous emission rates and energy of the transitions. [13]As can be seen in Figure 3b, the data fits correctly.In addition to compare the sensitivity of NaYF 4 with other possible materials that can be used as temperature sensors, the relative sensitivity (S r ) given by the formula is calculated: The obtained S r values change from 0.7 to 1.1% K −1 in the temperature range studied (Figure 3b).Each sensor is also characterized by the uncertainty of detection (resolution), for this reason an experimental estimation of this error was made by taking 100 measurements under the same conditions at different temperature values, 299, 323, and 373 K, which are representative for the T-range studied (Figure 3c).Based on these 100 measurements, the LIRs were calculated, and the expression that correlated the LIR value with the temperature (expressed in Figure 3b) was used.It is observed that the estimated values for all temperature values are distributed approximately in a Gaussian way around the expected/measured temperature value.The standard deviation obtained varies between 0.07 and 0.1 K, indicating that this sample can be used as an accurate, high-resolution optical sensor.As can be seen in Table 1, the relative sensitivity and temperature uncertainty values are reasonable in comparison with other materials.

Gear Measurements and Comparison of Sensing Results with Thermovision Data
The custom-built setup controlled by software in the LabView environment was used to measure the change in temperature of rotating gears.The measurement involved simultaneously powering the gear motors and recording data from the spectrometer and the thermal imaging camera.The tested gear had 12 teeth and was mounted between two gears (24 teeth each) connected to Maxon A-max 22 motors.The same laboratory power supply powered both motors, but the first motor was supplied with voltages of 10 and 15 V DC, and this corresponded to the rotation speeds of the test gear at ≈670 and 1000 rpm, respectively.The second, acting as a system load was provided with a reverse-biased voltage of 1 V DC.The tested gear was illuminated by a diode-pumped solid-state (DPSS) CW 980 nm 100 mW laser.Using the neutral density rotating filters, the laser beam power was reduced to 30 mW to avoid the gear's laser heating during the measurement.All measurements were carried out at this power, and the laser stability was checked using Thorlabs PM100D optical power meter for one hour of operation; the standard deviation was 52 μW.The gear temperature was simultaneously measured using a FLIR A35 60 Hz thermal imaging camera at the same point where the laser beam fell.The fabricated (3D printed) gear doped with 5 wt% of NaYF 4 :Yb 3+ , Er 3+ has been connected to a motor that allows its speed to be changed according to the applied voltage.Figure 4 shows a photograph taken in daylight showing the excitation and detection region for the optical (luminescence) method and with the thermographic camera, where a gradient of colors can be observed as a function of the detected temperature specified on the scale.In addition, the temperature reading spot (with an NIR laser and an optical setup) is indicated with a pointer in the central image.Due to the contact forces acting between the gears, a temperature increase is expected mainly on the face of the teeth.Therefore, the experiment has been carried out by analyzing the temperature variation on this surface for different revolutions per minute of the motor.It is worth mentioning that to obtain the gear temperature measurements using the optical sensor, the calibration obtained in Figure 3b is used.This way, the gear temperature can be obtained using the luminescence thermometry technique (i.e., analyzing the resulting band intensity ratio).In other words, the mechanically heated (due to friction) area was irradiated with a NIR 975 nm laser, and the resulting green up-conversion emission was analyzed with a spectrometer.
As shown in Figure 5a, there are three different situations in which the comparison has been made.At 0 rpm (stationary state), the gears do not generate any friction, and therefore, their temperature remains constant, as confirmed by both methods, i.e., the thermal camera and the luminescent sensor.On the other hand, when we increased the speed to 700 and 1000 rpm, the friction among the gears caused gradual heat generation and the resulting variation in the temperature readouts by both methods.The final temperature values obtained from the measurements increase as a function of the angular speed of the engine.The thermal camera has many oscillations when the gears turn, while these oscillations are much smaller for the optical sensor.The detected oscillations are associated with the error in the temperature readouts.Therefore, 100 measurements were made at 700 rpm with the thermal camera in a stationary regime, obtaining a deviation of 0.3 K, ≈4 times higher than the one obtained by the optical sensor in the same conditions, i.e., 0.07 K (see Figure S1, Supporting Information).
It is worth noting, that heating took a long time, and even with constant revolutions, the temperature rose gradually, which can be clearly seen in Video S1 (Supporting Information).However, we have also performed the additional measurements in order to estimate and compare the impact of revolutions on the temperature values of the working gear system at fixed time interval, i.e., after 100 s of rotations.The determined temperature values as a function of gears revolutions (rpm) are shown in Figure S2 (Supporting Information).As expected, with increasing revolutions of the working gear system from 0 to 2900 rpm, the temperature values increase from 294 to 308 K, respectively, which can be clearly seen in Figure S2 (Supporting Information).
The experimental setup for measuring the motor revolutions is based on using an incident laser light on the toothed surface of the wheel parallel to the gear axes.In this experiment, a laser at 532 nm was used to detect the scattered light by the teeth of the gear wheel when the light was focused on them.A photodiode detector captures some of the scattered light.In this way, if the motor is run for a certain voltage by connecting the photodiode to the oscilloscope, a periodic signal can be observed that gives information on the elapsed time in which the laser light is scattered on each tooth of the gear (see Figure 5b).

Conclusion
Here, we have successfully fabricated the luminescent gears by combining a 3D-printed polymer resin with an inorganic luminophore (upconverting) acting as the internal optical thermometer, allowing the detection of temperature elevation in the working gear motor system.The dependence of the visible upconversion emission of the NaYF 4 :Yb 3+ , Er 3+ nanoparticles, stimulated with a NIR laser light, as a function of temperature, has been studied.Using the luminescence thermometry technique, we successfully monitored (with high resolution) the temperature at the working gear surface.When the experiment was carried out, the temperature was simultaneously measured using a complementary technique, i.e., a thermographic camera.However, in that case, the oscillations in the temperature readouts were significantly more significant (0.3 K) than those obtained by luminescence thermometry (0.07 K).Through this, we demonstrated the applicability of the developed remote optical temperature sensor and the proposed method for temperature detection for industrial purposes.In summary, this type of measurement using intrinsic optical (luminescent) sensors has significant advantages over any other kind of temperature sensor, such as hightemperature resolution (of the order of 0.1 K), a high spatial resolution, since it is possible to analyze ultra-small areas of mechanical parts in the micro-sized and submicron-sized regions.It should be noted that this method is insensitive to electromagnetic interference due to the electrical connections of the detectors and allows measurements in places where contact measurements are inaccessible.

Figure 1 .
Figure 1.Scheme of the gear system used for heat generation and optical setup applied for detection (top); energy level diagram showing the transitions involved in the luminescent material used during the experiments, emphasizing the thermally-coupled levels of Er 3+ and the energy transfer up-conversion processes occurring in the lanthanide-doped luminescent material.

Figure 2 .
Figure 2. Photograph showing a 3D printed gear wheel used in the experiments.

Figure 3 .
Figure 3. a) Upconversion emission spectra of the NaYF 4 :Yb 3+ , Er 3+ as a function of temperature.b) LIR and S r versus T. c) Histograms showing temperature distribution over the series of measurements, together with the calculated uncertainties, expressed as standard deviations.

Figure 4 .
Figure 4. Photography of the working gear system in daylight (up) and using the thermal camera (down), taken at 297 and 303 K.

Figure 5 .
Figure 5. a) Comparison of the temperature readouts by the thermal camera (black curve) and the optical/luminescent sensor (orange curve) in three situations: at 0 rpm (stationary state), 700 rpm, and 1000 rpm.b) Scheme for the laser-assisted gear rpm detection.

Table 1 .
Comparison of performance of different luminescent thermometers (doped with Er 3+ ).