sensors Article

Design of a Sensor System for On-Line Monitoring of Contact Pressure in Chalcographic Printing José Antonio Jiménez *, Francisco Javier Meca, Enrique Santiso and Pedro Martín Department of Electronics, University of Alcalá, Alcalá de Henares, Madrid 28805, Spain; [email protected] (F.J.M.); [email protected] (E.S.); [email protected] (P.M.) * Correspondence: [email protected]; Tel.: +34-91-885-6547 Received: 1 August 2017; Accepted: 31 August 2017; Published: 5 September 2017

Abstract: Chalcographic printer is the name given to a specific type of press which is used to transfer the printing of a metal-based engraved plate onto paper. The printing system consists of two rollers for pressing and carrying a metal plate onto which an engraved inked plate is placed. When the driving mechanism is operated, the pressure exerted by the rollers, also called contact pressure, allows the engraved image to be transferred into paper, thereby obtaining the final image. With the aim of ensuring the quality of the result, in terms of good and even transfer of ink, the contact pressure must be uniform. Nowadays, the strategies utilized to measure the pressure are implemented off-line, i.e., when the press machines are shut down for maintenance, which poses limitations. This paper proposes a novel sensor system aimed at monitoring the pressure exerted by the rollers on the engraved plate while chalcographic printer is operating, i.e., on-line. The purpose is two-fold: firstly, real-time monitoring reduces the number of breakdown repairs required, reduces machine downtime and reduces the number of low-quality engravings, which increases productivity and revenues; and secondly, the on-line monitoring and register of the process parameters allows the printing process to be reproducible even with changes in the environmental conditions or other factors such as the wear of the parts that constitute the mechanical system and a change in the dimensions of the printing materials. The proposed system consists of a strain gauge-based load cell and conditioning electronics to sense and treat the signals. Keywords: chalcography printer; roller alignment; contact pressure; gauge measurement; nip pressure; load cell; strain measurement

1. Introduction Chalcography is the art of relief printing using rollers and metal plates among other components. The traditional chalcographic printers pose the problem that the contact pressure or mechanical stress in the roller-plate contact area, known as nip pressure, can be non-uniform along the roller generatrix. Contact pressure is applied by lowering the upper pressure drum onto the press bed using pressure adjusting screws [1]. Currently, there are no printing systems that include on-line monitoring of the contact pressure. It is always necessary to shut down the machines in order to make the appropriate adjustments to guarantee a high quality of printing. In this paper, the design of a sensor system aimed at monitoring the contact pressure in chalcographic printers for printing proofs is described. The system implemented allows the performance of various ink and paper types to be modeled and evaluated so that they can be used in production processes requiring gravure on an industrial scale, such as the printing of paper money and others. The proper roller alignment ensures a uniform contact pressure, which is vital not only to obtain a high quality product but also to guarantee an efficient process, increasing yield and reducing waste in terms of ink and paper. An incorrect alignment can result in high costs due to the damage caused to

Sensors 2017, 17, 2029; doi:10.3390/s17092029

www.mdpi.com/journal/sensors

Sensors 2017, 17, 2029

2 of 16

both the chalcographic printers and the final product. The chalcographic printers include mechanical systems for micrometer alignment. However, they lack systems that enable the real pressure to be determined, and as a result, such pressure cannot be effectively controlled. For this reason, short-term reproducible pressure conditions are only possible and, therefore, when parts become worn or there is variation in the dimensions or properties of the printing materials, mainly ink and paper, the impression changes. By including the proposed monitoring sensor system, the adjustment does not depend on the above-mentioned factors, which allows the alignment condition of the chalcographic printer to be determined. As a result, the reproducibility of the impressions improves, i.e., there are hardly any deviations from one another, due to the fact that the long-term conditions (temperature and contact pressure) can be easily replicated. Therefore, in order to ensure a high-quality and efficient printing process, the measurement and control of the contact pressure generated by the rollers against the engraved plate is critical. Unfortunately, obtaining accurate measurements of the contact pressure is a challenging task, especially when the chalcographic printer is operating. As stated above, the chalcographic printers currently available on the market include systems for precision adjustment [1]. However, there are no printers which have built-in electronics in order to online monitor the pressure exerted by the rollers on the engraved plate. Currently to this aim, external sensors are used instead. Nowadays, there are two ways of measuring the contact pressure, namely: static and dynamic. The former is straightforward and it is carried out when the rollers are stationary. A static sensor is placed between the rollers or between the roller and the engraved plate and then some pressure is applied. If necessary, the pressure can be adjusted until the desired value is obtained and then it is measured in both sides and in the center of the roller. There are two systems based on static sensors: the traditional sensor films or nipimpression paper and the feeler gauges.

•

•

The sensor films or nip impression paper [2–4] are based on a film that is placed between the roller and the engraved plate to characterize the contact pressure statically. Then, some pressure is applied with the rollers at rest. This strategy is not expensive although it can turn out to be costly depending on the size of the contact area and the number of measurements required until a proper alignment is achieved. The feeler gauges, as a mechanical tool, are usually used for thickness measurement purposes although they can also be used to measure the uniformity of the roller-plate contact area. Depending on the size of the contact area, this approach can be time-consuming. Besides, the information is incomplete: the feeler gauges can only identify potential defects between the roller and the plate and they cannot provide information about roller-plate misalignments.

Both the nip impression paper and the feeler gauges are not entirely accurate as they introduce a systematic error on account of the additional thickness of the sensor itself. There are also commercial digital meters which allow static measurements of the contact pressure. The strategy employed is similar to that of the Impression Paper, but they can show the corresponding measurement in display-based devices [5]. The dynamic measurement, on the other hand, is made with the rollers slowly rotating. Unlike the static measurement, the dynamic approach is more realistic since the shape and characteristics of the rollers change when they are subjected to pressure. Hence, this is the preferred method to employ when it comes to precisely controlling the contact pressure. In [6,7], two similar approaches to dynamic contact pressure assessment are presented, named sigma-nip and nip pressure alignment tool (NPAT), which are based on an array of pressure sensors. Both the static and dynamic techniques, explained previously, reveal an important limitation: they are procedures that are carried out when the printer is at rest, under programmed maintenance and hence they halt production. Besides, these systems do not take into account the thickness of the

Sensors 2017, 17, 2029

3 of 16

engraved plate and the paper, and some of them cannot work within the operating temperatures of the chalcographic printers. In this paper, a novel sensor system aimed at obtaining the pressure profile between two rollers or between one roller and the engraved plate is presented. The main contribution of this paper is that the pressure profile is generated while the printer is working, which guarantees a correct roller-plate alignment. This, in turn, results in a significant improvement in terms of printing quality, cost savings and long-term reproducibility of the engravings. The designed sensor is a custom load cell based on 16 strain gauges arranged in a 4 × 4 matrix configuration. It must be noted that locating strain gauges on the two outermost edges of the engraved plate would be enough to be able to detect the roller misalignment. However, in this work, the four edges have been used with the aim of obtaining additional information. This information can be used to detect potential deformations and the wear of the rollers and the engraved plate, which can produce wrong profiles in the contact pressure. The transducers based on strain gauges are inexpensive and provide high precision, long-term stability and sufficient response time to allow high-speed measurements [8]. Therefore, the strain gauge-based transducers are the best option for both high-scale industrial processes and more simple tasks, especially when it comes to high precision measurements. These features truly justify the widespread use of this kind of sensor in fields ranging from the healthcare sector [9–11] to the industrial sector [12–15] and civil engineering [16], among others. In the last few years, several works have reported the use of pressure sensors based upon fiber optic sensors with promising results. The main drawback of fiber optic sensors, however, is the high cost associated with them. Hence, their use is restricted to special applications usually related to the construction and infrastructure sector [17–20]. Taking into account that the ink characteristics vary with the temperature, the proposed sensor includes a control system aimed at maintaining a constant temperature in the plate surface, usually 80 ◦ C, which is the normal process temperature in the industrial chalcographic printers. However, the current test regime for chalcographic printers does not allow tests of chalcographic printing at a temperature of 80 ◦ C to be carried out. The proposed system presented in this paper includes a heating module for the engraved plate with a corresponding temperature control unit. One of the most relevant features of the heating module is the low energy cost required to maintain the normal operating temperature of the engraved plate. This is possible on account of the module size, which allows the module to be located underneath the engraved plate, thereby minimizing heat loss. When implemented, the traditional technique is based on cartridge heaters of considerable dimensions, which cannot be placed close to the engraved plate. This fact causes substantial heat loss in comparison with the heating strategy developed in this paper, which is described in the following section. To conclude, the proposed system automatically generates a detailed report with the data captured in each impression, namely: the pressure profile of the roller against the engraved plate, the temperature of the engraved plate and hence that of the ink used in the printing process, environmental temperature and humidity and finally the rotation speed under operating conditions. After the introduction this paper is organized as follows. Section 2 gives a detailed description of the structure of the proposed sensor system used to monitor the pressure profile in chalcographic printers. This section analyzes the sensor module, the heating module, the signal conditioning module, the acquisition system and the data representation. The experimental results obtained are presented in Section 3. Finally, some conclusions are drawn in the final section. 2. Sensor System Description As a preliminary step before the proposed sensor system architecture is described, Figure 1 depicts the printer structure with a pressure profile to be measured by our system. The aim of the sensor system is to determine whether the pressure or mechanical stress distribution on the surface of the engraved plate is symmetrical and uniform, thereby ensuring the quality of the engraved printing.

Sensors 2017, 17, 2029

4 of 16

In a chalcographic printer, the upper roller is a carbon steel cylinder having a diameter of 200 mm Sensors 2017, 17, 2029 4 of 16 with an outer coating based on several carbon layers and a rubber blanket. The mechanical system Sensors 2017, 17, 2029 4 ofplate. 16 makes the rollers rotate, which produces movement the structure the engraved system makes the rollers rotate, whichthe produces theofmovement of containing the structure containing the This plate is a flat surface of 140 × 100 mm and has a thickness ranging from 3 to 4 mm, coated with engraved plate. This plate is a flat surface of 140 × 100 mm and has a thickness ranging from 3 to system makes the rollers rotate, which produces the movement of the structure containing the a4 thin of with chrome andlayer made copperand andmade sometimes nickel. mm,layer coated a thin ofof chrome of copper and sometimes nickel. engraved plate. This plate is aunderneath flat surface the of 140 × 100 mm andhas has been a thickness ranging 3 to of The plate located engraved plate, withfrom the aim The coated plate holder, holder, located themade engraved plate, been modified modified 4 mm, with a thin layerunderneath of chrome and of copper andhas sometimes nickel. with the aim of placing the strain gauges. From the of system, it be that the placingThe the plate strainholder, gauges. From underneath the perspective perspective of the the sensor sensor system, it can can be claimed claimed the plate plate located the engraved plate, has been modified with that the aim of holder has been redesigned to sense the pressure, thereby working as a load cell. Therefore, the plate holder hasthe been redesigned to sense pressure, working load Therefore, plate placing strain gauges. From the the perspective ofthereby the sensor system,asitacan becell. claimed that thethe plate holder is the essential part the sensor module included in system. holder is has the been essential part of ofto the sensor in the the monitoring monitoring system. holder redesigned sense the module pressure,included thereby working as a load cell. Therefore, the plate holder is the essential part of the sensor module included in the monitoring system.

Figure 1. System structure of the chalcographic printer. printer. Figure of the the chalcographic chalcographic Figure1.1.System System structure structure of printer.

Figure 2 shows the architecture of the proposed sensor system which consists of the following Figure 2 showsthe thearchitecture architectureof ofthe the proposed proposed sensor the following Figure 2 shows sensor system systemwhich whichconsists consistsofof the following parts: sensor module, Acquisition Data System (ADS), ADS ADSadaptation adaptationmodule, module,temperature temperature control parts: sensor module, Acquisition Data System (ADS), control parts: sensor module, Acquisition Data System (ADS), ADS adaptation module, temperature control module and the sensors to capture the environmental parameters, namely temperature and module sensors to capture the environmental parameters, temperature and module andand the the sensors to capture the environmental parameters, namelynamely temperature and humidity. humidity. humidity.

Figure proposedsensor sensorsystem. system. Figure2.2.Architecture Architecture of of the the proposed Figure 2. Architecture of the proposed sensor system.

Briefly, the sensor system works as follows: the measurements obtained by the sensor module areBriefly, sent to the conversion and works conditioning module. module filters and transforms the current sensor system as follows: the This measurements obtained by the sensor module loops into voltages before the signals are sent to the acquisition module, which is based on a are sent to the conversion and conditioning module. This module filters and transforms the current commercial board before installed a personal computer The temperature controlismodule is a loops into voltages theinsignals are sent to the(PC). acquisition module, which based on responsibleboard for raising and maintaining the plate holder(PC). to theThe ink temperature optimal temperature, commercial installed in a personal computer control normally module is

responsible for raising and maintaining the plate holder to the ink optimal temperature, normally

Sensors 2017, 17, 2029

5 of 16

Briefly, the sensor system works as follows: the measurements obtained by the sensor module are sent to the conversion and conditioning module. This module filters and transforms the current loops into voltages before the signals are sent to the acquisition module, which is based on a commercial board installed in a personal computer (PC). The temperature control module is responsible for Sensorsand 2017,maintaining 17, 2029 16 raising the plate holder to the ink optimal temperature, normally 80 ◦ C.5 of Finally, the system includes a commercial temperature probe used to measure the environmental temperature 80 °C. Finally, the system includes a commercial temperature probe used to measure the and humidity. These parameters are of interest to the reports generated by the system. Following on, environmental temperature and humidity. These parameters are of interest to the reports generated for each of the modules listed before, the specifications, design and implementation and the results by the system. Following on, for each of the modules listed before, the specifications, design and obtained are detailed implementation andand the analyzed. results obtained are detailed and analyzed. 2.1. Sensor Module

2.1. Sensor Module

Figure 3 depicts the plate holderholder structure designed to capture the mechanical stress distribution. Figure 3 depicts the plate structure designed to capture the mechanical stress As distribution. can be seen, the rectangular prism-shaped plate holder with dimensions 160 × 160dimensions × 30 mm has As can be seen, the rectangular prism-shaped plate holder with two160 grooves on the edgesand of the grooves, strain have16been placed (4 have on each × 160 ×and 30 mm has square two grooves on the square16 edges of gauges the grooves, strain gauges edge, seeplaced Figure(43b). Likewise, the Figure signal conditioning the strain gauges have been located been on each edge, see 3b). Likewise, modules the signalfor conditioning modules for the strain gauges been located(see on the base3c). of the grooves (see Figure 3c). on the basehave of the grooves Figure Strain Gauges

(a)

(b)

(c) Figure 3. (a) Structure of the designed plate holder; (b) The plate holder (160 × 160 × 30 mm) with the

Figure 3. (a) Structure of the designed plate holder; (b) The plate holder (160 × 160 × 30 mm) with the strain gauges adhered to the edges of the grooves and (c) Signal conditioning electronics integrated strain gauges adhered to the edges of the grooves and (c) Signal conditioning electronics integrated on on the plate holder. the plate holder.

According to the measurements taken by two load cells placed on the shaft ends of the rollers, the maximumtoforce to the engraved around kg. On thethe linear According the applied measurements taken byplate twoisload cells2000 placed on the other shaft hand, ends of rollers, of displacement holder plate can vary in a range 0.10the to 0.28 Thethe plate thevelocity maximum force appliedoftothe theplate engraved is around 2000from kg. On otherm/s. hand, linear holder has been made of a 7075-T7351 aluminum-zinc alloy. As a result, it exhibits a low Young’s velocity of displacement of the plate holder can vary in a range from 0.10 to 0.28 m/s. The plate holder of about 72 GPa, providing strain higher of other metals asuch steel ormodulus copper, of hasmodulus been made of a 7075-T7351 aluminum-zinc alloy.than As that a result, it exhibits lowasYoung’s and72 a high limit ofstrain 400 MPa, which is that higher that of thesuch engraved plate. fact ensures about GPa,elastic providing higher than ofthan other metals as steel or This copper, and a high that the holder plate can work in elastic mode. elastic limit of 400 MPa, which is higher than that of the engraved plate. This fact ensures that the In order to calculate the maximum mechanical stress between the roller and the plate holder holder plate can work in elastic mode. ,analytical cylindrical contact force models [21,22] and the plane strain problem of a curved elastic body pressed against an elastic half-space [23,24] have been considered . Finally, applying the Hertz equations to the static contact between the roller and the plate holder yields a maximum mechanical stress of 164 MPa, for an expected force of 2000 kg (applied on the system). Clearly, this force is below the elastic limit of the plate holder. When the maximum force is exerted, the contact surface is

Sensors 2017, 17, 2029

6 of 16

In order to calculate the maximum mechanical stress between the roller and the plate holder ,analytical cylindrical contact force models [21,22] and the plane strain problem of a curved elastic body pressed against an elastic half-space [23,24] have been considered. Finally, applying the Hertz equations to the static contact between the roller and the plate holder yields a maximum mechanical stress of 164 MPa, for an expected force of 2000 kg (applied on the system). Clearly, this force is below the elastic limit of the plate holder. When the maximum force is exerted, the contact surface Sensors 2017, 17, 2029 6 of 16 is 1.1 mm width. In practice, this width is greater on account of the engraved plate between both elements, roller the plate thegreater machined groovesofonthe theengraved plate holder, roller coating 1.1 mmi.e., width. Inand practice, this holder, width is on account platethe between both and its displacement. Therefore, the maximum mechanical stress is lower than that obtained from elements, i.e., roller and the plate holder, the machined grooves on the plate holder, the rollerthe Hertz equations. This is proved by the measurements in thestress final issystem concluding that the coating and its displacement. Therefore, the maximum taken mechanical lower than that obtained mechanical stress equations. endured by theisplate holder is below its elastictaken limit,inwhich ensures that the system from the Hertz This proved by the measurements the final system concluding is working without permanent strains. that the mechanical stress endured by the plate holder is below its elastic limit, which ensures that theWith system working withoutand permanent strains. theis aim of raising regulating the system temperature to 80 ◦ C (see Figure 4a), theheaters aim of and raising and regulating the system to 80 °C in (seethe Figure 4a), silicon silicon With rubber a temperature sensor Pt100 temperature have been installed grooves together rubber heaters circuit and a temperature sensor Pt100 have been installed inelectronics the grooves with the with the printed board that contains the signal conditioning fortogether the strain gauges. printed circuit board that contains the signal conditioning electronics for the strain gauges. To To prevent thermal drifts due to the close proximity between the rubber heaters and the conditioning prevent thermal drifts due to the close proximity between the rubber heaters and the conditioning electronics, and therefore, protect the electronic components from damage caused by overheating, electronics, and therefore, protect the electronic components damage caused by overheating, thermal fiber (superwool) has been used for insulation of bothfrom parts. Despite its minimum thickness, thermal fiber (Superwool) has been used for insulation of both parts. Despite its minimum thickness, this type of thermal fiber exhibits excellent insulating properties (see Figure 4b). this type of thermal fiber exhibits excellent insulating properties (see Figure 4b).

(a)

(b)

Figure 4. (a) Heating module for the plate holder: PT100 and silicon rubber heater y; (b) Thermal Figure 4. (a) Heating module for the plate holder: PT100 and silicon rubber heater; (b) Thermal insulating arrangement. insulating arrangement.

Considering both the Young’s modulus of the 7075-T7351 aluminum alloy (E = 72 GPa) and the Considering both Young’smechanical modulus stress of theof 7075-T7351 = 72 GPa) previously worked outthe maximum 164 MPa, thealuminum maximum alloy strain (E produced is and theμε previously mechanical of 164 the maximum 2278 [21]. The worked model ofout themaximum strain gauge used in stress the design of MPa, the sensor system is strain the produced is 2278 µε [21]. The model the strain gauge in the factor designof of 2.1 the sensor system EA-13-031CE-350/E by vishay [25]. ofThese gauges haveused a gauge and feature self-temperature-compensation for the 7075-T7351 aluminum alloy. Since the strain gauges are not is the EA-13-031CE-350/E by vishay [25]. These gauges have a gauge factor of 2.1 and feature embedded in a Wheatstone bridge configuration, compensation required minimize theare self-temperature-compensation for the 7075-T7351this aluminum alloy. isSince the to strain gauges of theinthermal expansion coefficient on measurement, equivalentistorequired +17.5 με/°C. This, noteffects embedded a Wheatstone bridge configuration, this compensation to minimize ◦ C. This, wide range of operating temperatures, produces a equivalent maximum total strainµε/ of around thetogether effects with of thethe thermal expansion coefficient on measurement, to +17.5 +1000 με. This value is comparable with the values of the strain measured as a consequence the together with the wide range of operating temperatures, produces a maximum total strain ofof around mechanical stress. The thermal expansion coefficient is the same for all the strain gauges in the plate +1000 µε. This value is comparable with the values of the strain measured as a consequence of the holder and hence,The no thermal error in the mechanical stress measurement is introduced. However, there a mechanical stress. expansion coefficient is the same for all the strain gauges in theisplate drastic reduction in the dynamic range, which negatively influences the measurements in the signal holder and hence, no error in the mechanical stress measurement is introduced. However, there is conditioning electronics module. a drastic reduction in the dynamic range, which negatively influences the measurements in the signal A circuit based on the 4–20 mA current loop transmitter XTR105U by Texas Instruments [26] conditioning electronics module. has been designed to condition the signals from each of the 16 strain gauges. The transmitter exhibits an output sensitivity of −8 μA/με. Since it is a custom design, there is no need to meet the 4–20 mA current loop standard. Bearing in mind that the transmitter XTR105U operates in the linear region for currents ranging from 2.6 mA to 24 mA, the maximum dynamic range for the measurements is 2675 με. The system performs an autocalibration as a prerequisite before monitoring the printing process. The method of autocalibration in this work involves the measurement of the signal

Sensors 2017, 17, 2029

7 of 16

A circuit based on the 4–20 mA current loop transmitter XTR105U by Texas Instruments [26] has been designed to condition the signals from each of the 16 strain gauges. The transmitter exhibits an output sensitivity of −8 µA/µε. Since it is a custom design, there is no need to meet the 4–20 mA current loop standard. Bearing in mind that the transmitter XTR105U operates in the linear region for currents ranging from 2.6 mA to 24 mA, the maximum dynamic range for the measurements is 2675 µε. The system performs an autocalibration as a prerequisite before monitoring the printing process. The method of autocalibration in this work involves the measurement of the signal amplitude provided by each strain gauge when they are at rest. These amplitudes are then used as a reference to determine the variations found when the roller is rotating. Therefore, the system offset thermal drift does not introduce measurement uncertainty although it does reduce the acceptable dynamic range of the deformation. The designed circuit has a thermal drift offset at the output of ±1.2 mA under the following condition: maximum expected temperature variations of 60 ◦ C from the set temperature of around 20 ◦ C to the printing maximum temperature of 80 ◦ C. This reduces the measurement dynamic range by ±150 µε from both ends of the measurement scale. Taking this into account, the system has been calibrated at a room temperature of 22 ◦ C so that it can provide 4 mA, without exerting mechanical stress and leaving a safety margin of 1.4 mA for the offset thermal drift. This margin of 1.4 mA comes from the difference result of the difference between the 4mA setting and the linear operating limit of 2.6 mA. The sensitivity thermal drift is extremely important since it modifies the sensitivity of each of the channels of the conditioning electronics, having an influence on the mechanical stress profile obtained. For this parameter, due to the gauge factor thermal drift (+100 ppm/◦ C), each channel is affected by the same amount. As a result, it will not have any effect on the measurement. The sensitivity thermal drift is also influenced by the electronic components that constitute the 4–20 mA current loop transmitters. With the device used in the design, a drift in the sensitivity as a function of the temperature of ±18 ppm/◦ C (±1σ) can be achieved. When evaluating within the operating temperature range of 60 ◦ C and for ±3σ, the maximum sensitivity deviation is ±0.32%. Similarly, the design guaranteed maximum sensitivity tolerance due to the real characteristics of the components is ±0.3%. The combined effect of such tolerance and the thermal drift yields a maximum uncertainty of ±0.44% (±3σ) in the sensitivity, and therefore, in the mechanical stress measurements. The analysis of the intrinsic noise generated at the output yields a value of 16 µApp, equivalent to a maximum uncertainty of ±1 µε, which can be negligible. This noise was calculated over −3 dB bandwidth of 100 Hz, being the bandwidth determined by the XTR105U (Texas Instruments, TX, USA) and the electronics of the ADS adaptation module. 2.2. ADS Adaptation Module This module low-pass filters the input signal and converts the current from the 4–20 mA current loop transmitters into voltage, before being adapted to the input range of the data acquisition board. The current-voltage conversion factor is 352 mV/mA, hence the output sensitivity will be −2.816 mV/µε. Since the maximum input current is 24 mA, the maximum output voltage is set to a value of 8.45 V. The bandwidth of the first-order low-pass filter is 160 Hz. Therefore, the final bandwidth of the system considering the response frequency of the XTR105U transmitter is on average 100 Hz, which is sufficient to capture the temporal variations of the information to measure. This fact will be proved later in the results sections. As in the case of the 4–20 mA current loop transmitters, the important contributions to uncertainty are produced by the sensitivity thermal drift which, for the devices employed in the design, takes a typical value of ±14 ppm/◦ C (±1σ). This value is slighty below that obtained with the 4–20 mA current loop transmitters of ±18 ppm/◦ C. As for the sensitivity tolerance due to the real characteristics

Sensors 2017, 17, 2029

8 of 16

of the components, a design guaranteed value of ±0.14% has been calculated. The combined effect of the ADS adaptation module and the 4–20 mA current loop transmitters introduces a maximum deviation in the sensitivity and, therefore, in the mechanical stress measurements, of ±0.53% when evaluating within the operating temperature range of 60 ◦ C and for ±3σ. Finally, due to the high-level input signal, the electronic noise introduced by the ADS adaptation module is negligible. 2.3. ADS Module The ADS is based on the PCI-6284 data acquisition board by National Instruments [27]. This board provides 32 analog inputs with sequential sampling. The sampling rate for each of the 16 channels used to capture the information from the strain gauges is 1 KS/s. Considering a maximum plate holderSensors displacement 2017, 17, 2029 speed of 0.28 m/s, the spatial resolution for the measurement system 8is ofabove 16 3 samples/mm. Both the sampling rate and the system bandwidth are sufficient to capture the plateof holder displacement speed of 0.28 as m/s, thebe spatial resolution variations the mechanical stress exerted, will proved later. for the measurement system is above 3 samples/mm. Both the sampling rate and the system bandwidth are sufficient to capture the The acquisition board includes 18-bit ADCs. Since the input range per channel has been set to variations of the mechanical stress exerted, as will be proved later. ±10 V, the quantification step is q = 76 µV. Taking into account the output sensitivity of the ADS The acquisition board includes 18-bit ADCs. Since the input range per channel has been set to adaptation −2.816 mV/µε, inTaking an equivalent quantification step of −27.1 × ADS 10−3 µε. ±10 V,module, the quantification step is qresults = 76 μV. into account the output sensitivity of the The Absolute Accuracy Full Scale that was in found in the datasheet is ±980 which× corresponds adaptation module,at−2.816 mV/με, results an equivalent quantification stepµV, of −27.1 10−3 με. The to ±0.35 Absolute µε. This is even below theScale value introduced mA current loopwhich transmitters. Therefore, Accuracy at Full that was foundby in the the 4–20 datasheet is ±980 μV, corresponds to the ADS influence system is introduced negligible. by the 4–20 mA current loop transmitters. ±0.35 με. Thisonisthe even belowaccuracy the value Therefore, the ADS influence on the accuracy negligible. As an example, Figure 5 shows thesystem signals from 4isstrain gauges located on the same edge of the As an example, Figure 5 shows the signals from 4 strain located on the edge of the same groove in the plate holder, captured from the ADS. The gauges time axis is given in same milliseconds while same groove in the plate holder, captured from the ADS. The time axis is given in milliseconds while is the amplitude axis is in volts. As can be seen in Figure 3a, the distance between the strain gauges the amplitude axis is in volts. As can be seen in Figure 3a, the distance between the strain gauges is constant and takes a value of 25 mm. Therefore, the distance between maxima in Figure 5 shows that constant and takes a value of 25 mm. Therefore, the distance between maxima in Figure 5 shows that the displacement speed of the plate holder is not uniform, although this issue is not critical for the the displacement speed of the plate holder is not uniform, although this issue is not critical for the purposes of this TheThe time distance andthe thelast lastpeaks peaks (maximum peaks) purposes ofsystem. this system. time distancebetween betweenthe the first first and (maximum peaks) is is 400 ms. This indicates that the average displacement speed in this experiment is 0.19 m/s, which 400 ms. This indicates that the average displacement speed in this experiment is 0.19 m/s, which is is withinwithin the expected range of 0.1 to 0.28 m/s. the expected range of 0.1 to 0.28 m/s. Considering the value forADS the input ADS input sensitivity of −2.816 mV/με, maximum strain Considering the value for the sensitivity of −2.816 mV/µε, the the maximum strain values values extracted Figure 5 that occur thesurface contact surface of the and rollerthe andstrain the strain gauges extracted from Figurefrom 5 that occur when thewhen contact of the roller gauges exactly exactly coincide range from −700µε. to −900 με. coincide range from − 700 to −900 0

Row 1 Row 2 Row 3 Row 4

Voltage (V)

-0.5 -1 -1.5 -2 -2.5 600

800

1000

1200

1400

1600

Time (ms) Figure 5. Signal captured from ADS of four consecutive strain gauges (Row 1, Row 2, Row 3 and

Figure 5. Signal captured from ADS of four consecutive strain gauges (Row 1, Row 2, Row 3 and Row 4) in the same groove. The voltage sign has been changed with respect to the voltages sampled Row 4)byinthe theADS same groove. The voltage sign has strain. been changed with respect to the voltages sampled by with the aim of reflecting the real the ADS with the aim of reflecting the real strain.

As can be observed in Figure 5, the signal voltage approximately changes following a triangular curve. In this case, the in a first-order low-pass filter takes the value of the gradient As can be observed inmaximum Figure 5, error the signal voltage approximately changes following a triangular of the triangle wave. This gradient is a direct function of the the strain and the plate holder speed. curve. In this case, the maximum error in a first-order low-pass filter takes the value of the gradient The maximum error is this gradient multiplied by the filter time constant with a value of of the triangle wave. This gradient is a direct function of the the strain and the plate holder 1.59 ms = 1/(100 Hz·2π), considering a bandwidth of 100 Hz. For the signals depicted in Figure 5, a maximum error of about ±20 mV must be taken into account. This corresponds to ±7 με when measuring a strain of about 1000 με, i.e., an error of ±0.7%. This error is virtually the same for each strain gauge, assuming that they are subjected to similar strain and for a constant displacement speed of the plate holder. Therefore, the effect of this error on the measurement of the mechanical stress uniformity will be below ±0.7%, which is acceptable. Finally, the curves in Figure 5 apparently do not show a noise contribution. When the signals

Sensors 2017, 17, 2029

9 of 16

speed. The maximum error is this gradient multiplied by the filter time constant with a value of 1.59 ms = 1/(100 Hz·2π), considering a bandwidth of 100 Hz. For the signals depicted in Figure 5, a maximum error of about ±20 mV must be taken into account. This corresponds to ±7 µε when measuring a strain of about 1000 µε, i.e., an error of ±0.7%. This error is virtually the same for each strain gauge, assuming that they are subjected to similar strain and for a constant displacement speed of the plate holder. Therefore, the effect of this error on the measurement of the mechanical stress uniformity will be below ±0.7%, which is acceptable. Finally, the curves in Figure 5 apparently do not show a noise contribution. When the signals from the ADS are analyzed at rest, i.e., with the printer shut down, yielded values for the noise range from 3.6 mVpp to 5 mVpp, depending on the acquisition channel, which corresponds to strains from ±0.64 to ±0.89 µε, respectively. These values are negligible, in line with the expected values after Sensors 2017, 17, 2029 9 of for 16 the the theoretical analysis of the designed system. Therefore, the chosen bandwidth of 100 Hz measured system proves to be adequate for both accurately obtaining the information and effectively theoretical analysis of the designed system. Therefore, the chosen bandwidth of 100 Hz for the limiting the noise to negligible levels. measured system proves to be adequate for both accurately obtaining the information and effectively limiting the noise to negligible levels.

2.4. Temperature Control Module

2.4. Temperature Control Module This module consists of the temperature controller aiming to maintain the plate holder temperature a constant value.ofThe can selectcontroller this temperature an the application installed Thisat module consists theuser temperature aiming tothrough maintain plate holder temperature constant value. can selectsensor this temperature through an the application on a PC. By using at thea information fromThe theuser temperature Pt100 and considering temperature installed on a PC. using the information from the temperature Pt100 considering reference selected, theBycontroller provides the necessary powersensor so that theand silicon rubberthe heater temperature reference selected, the controller provides the necessary power so that the silicon produces the heat energy to regulate the temperature to the desired value. rubber heater produces the heat energy to regulate the temperature to the desired value. A switch mode power supply unit of 27 V and 200 W is used as a power source for the silicon A switch mode power supply unit of 27 V and 200 W is used as a power source for the silicon rubber heaters. Finally, an RS485-to-RS232 converter allows the temperature controller and the PC rubber heaters. Finally, an RS485-to-RS232 converter allows the temperature controller and the PC application to communicate. Figure 6 shows the front and the interior of the device, which incorporates application to communicate. Figure 6 shows the front and the interior of the device, which the temperature system and thesystem ADS adaptation incorporatescontrol the temperature control and the ADSmodules. adaptation modules. The resistance heaters embedded holderexhibit exhibita nominal a nominal resistance of 10.9 The resistance heaters embeddedon onthe the plate plate holder resistance of 10.9 Ω. Ω. WhenWhen they they are powered with a voltage powersupply supply unit, dissipated power are powered with a voltageofof27 27VVfrom from the the power unit, thethe dissipated power for for each resistance is This 67 W. This ispower is sufficient to raise the temperature toapproximately 80 °C in each resistance heaterheater is 67 W. power sufficient to raise the temperature to 80 ◦ C in approximately 15 min. 15 min.

Figure 6. Interior andand frontal part ofofthe thetemperature temperature controller the ADS Figure 6. Interior frontal part thedevice deviceintegrating integrating the controller andand the ADS adaptation module. adaptation module.

3. Results 3. Results With the aim of verifying the operation of the system, several tests have been conducted. Before

With the aim of verifying the operation of the system, several tests have been conducted. Before performing the tests on the chalcographic printer, a hydraulic press, designed to provide a pressure performing the tests on the chalcographic printer, a hydraulic press, designed to provide a pressure of of 6000 psi, was used as a test bench. Likewise, several tools were developed to exert pressure over 6000 psi, was used a test bench. several tools were to exert over the the sensor on as different points. Likewise, In this way, the strain gaugedeveloped deformations werepressure qualitatively sensor on different points. In this way, the strain gauge deformations were qualitatively measured measured by using a graph paper template to discover the coordinates of the points on the plate by using a graph template to discover the7 coordinates of the points on the holder where holder wherepaper the pressure is exerted. Figure shows the hydraulic press with theplate sensor system integrated and the tools developed.

Sensors 2017, 17, 2029

10 of 16

the pressure is exerted. Figure 7 shows the hydraulic press with the sensor system integrated and the Sensors 2017, 17, 2029 10 of 16 tools developed.

Figure 7. 7. Hydraulic Hydraulic press press used used in in the the tests. tests. Figure

The tests conducted by using the hydraulic press demonstrated that it is not possible to The tests conducted by using the hydraulic press demonstrated that it is not possible to calibrate calibrate the response of the different strain gauges. For that aim, it is necessary to ensure high the response of the different strain gauges. For that aim, it is necessary to ensure high accuracy of accuracy of the position of the tools over the plate holder, which is not feasible with the available the position of the tools over the plate holder, which is not feasible with the available equipment. equipment. Failing that, the uniform response of the strain gauges has been guaranteed during the Failing that, the uniform response of the strain gauges has been guaranteed during the design phase by design phase by both calculating the uncertainty introduced by the electronics and taking into both calculating the uncertainty introduced by the electronics and taking into account the characteristics account the characteristics of the strain gauges. The total uncertainty introduced by both of the strain gauges. The total uncertainty introduced by both contributions, calculated in Section 2, contributions, calculated in Section 2, is ±0.53% (±3σ). In addition, since the process of placing the is ±0.53% (±3σ). In addition, since the process of placing the plate holder is not perfect, the contribution plate holder is not perfect, the contribution to the uncertainty due to the misalignment of the strain to the uncertainty due to the misalignment of the strain gauges was also considered. The strain gauges was also considered. The strain gauges placing was carried out by the technical support of gauges placing was carried out by the technical support of the firm called Vishay, who ensured the firm called Vishay, who ensured a maximum angular deviation of ±5° with respect to the a maximum angular deviation of ±5◦ with respect to the perpendicular to the plane of the plate holder. perpendicular to the plane of the plate holder. Considering this angular deviation, the maximum Considering this angular deviation, the maximum relative error for the deformation suffered by the relative error for the deformation suffered by the strain gauges is +1% [28], with this value being the strain gauges is +1% [28], with this value being the maximum difference in sensitivity among the maximum difference in sensitivity among the different strain gauges. Therefore, under the design different strain gauges. Therefore, under the design conditions stated above, the sensitivity error of the conditions stated above, the sensitivity error of the system ranges from −0.53% to +1.53%, which is system ranges from −0.53% to +1.53%, which is significantly below the adjustment threshold of ±10% significantly below the adjustment threshold of ±10% (see Section 3) so that the roller can be (see Section 3) so that the roller can be considered to be aligned. considered to be aligned. 3.1. Tests on the Chalcographic Printer 3.1. Tests on the Chalcographic Printer As has already been mentioned, the press consists of two rollers which move the plate holder As has already mentioned, press of which two rollers move the holder synchronistically. Thebeen upper roller has athe screw at consists both ends, allowswhich both control andplate adjustment synchronistically. uppertheroller has a screw at both ends, which allows both control and of the pressure usedThe to make impression. adjustment of the pressure used to make the impression. In this type of press, the adjustment is made by using a trial-and-error approach to verify the In this type of press, the adjustment is made by using a trial-and-error approach to verify the printing quality. Therefore, several impressions are made resulting in a waste of material. The proposed printing quality. Therefore, several impressions are made resulting in a waste of material. The system allows the adjustment to be made without wasting either paper or ink. This is simply done by proposed system allows the adjustment to be made without wasting either paper or ink. This is moving the roller over the plate holder. simply done by moving the roller over the plate holder. Several tests were carried out with an aligned roller and with the roller intentionally misaligned to Several tests were carried out with an aligned roller and with the roller intentionally misaligned different degrees: the roller with misalignment from the left and from the right, and small or significant to different degrees: the roller with misalignment from the left and from the right, and small or misalignment. For each test and for each strain gauge, the deformations were measured and registered, significant misalignment. For each test and for each strain gauge, the deformations were measured and, finally, the printing quality was evaluated by staff with printing skills and experience. and registered, and, finally, the printing quality was evaluated by staff with printing skills and In order to identify the acquired data, Figure 8 depicts the spatial arrangement of the gauges with experience. respect to the movement direction of the roller. Row 1 is, therefore, the first one to be subjected to the In order to identify the acquired data, Figure 8 depicts the spatial arrangement of the gauges with respect to the movement direction of the roller. Row 1 is, therefore, the first one to be subjected to the pressure exerted by the roller. The measured deformations for each strain gauge are shown in Tables 1 and 2 and Figures 9–13, in which the same colour code has been used.

Sensors 2017, 17, 2029

11 of 16

pressure exerted by the roller. The measured deformations for each strain gauge are shown in Tables 1 and 2 and Figures 9–13, in which the same colour code has been used. Sensors 2017, 17, 2029 11 of Sensors 2017, 17, 2029 1116 of 16

Figure 8. Strain Strain gauges position with respect totothe roller movement direction. Figure 8. gauges position with respect roller movement direction. Figure 8. Strain gauges position with respect tothe the roller movement direction.

3.1.1. Test with Misaligned Roller 3.1.1. Test with Misaligned Roller 3.1.1. Test with Misaligned Roller Figure shows theroller roller position forthis thisthis test, andand Figure 9b shows shows thethe deformations caused Figure 9a shows the roller position for test, Figure 9b shows deformations caused Figure 9a9ashows the position for test, and Figure 9b the deformations caused by by the roller and the time when they happen. In Figure 9b, the x-axis represents the time in in by the roller and the time when they happen. In Figure 9b, the x-axis represents the time the roller and the time when they happen. In Figure 9b, the x-axis represents the time in milliseconds milliseconds (ms) andand thethe y-axis thethe deformations (με). (ms) y-axis deformations (με). (ms)milliseconds and the y-axis the deformations (µε). 500500

Col.1 Col.1 Col.2 Col.2 Col.3 Col.3 Col.4 Col.4

Microstrains Microstrains

0 0

-500-500

-1000 -1000

-1500 -1500 0 0

(a) (a)

500500 1000 1000 1500 1500 2000 2000 2500 2500

Time (ms) Time (ms)

(b)(b)

Figure 9. (a) Diagram when thethe roller is inismisalignment andand (b) (b) corresponding signal captured. Figure 9. (a) Diagram when roller in misalignment corresponding signal captured. Figure 9. (a) Diagram when the roller is in misalignment and (b) corresponding signal captured.

As As cancan be be seen in Figure 9, the deformations suffered by by each strain gauge on on thethe same column seen in Figure 9, the deformations suffered each strain gauge same column have the same order of magnitude and the maxima happen sequentially in time, synchronized with As can be seen in Figure 9, the deformations suffered by each strain gauge on the same column have the same order of magnitude and the maxima happen sequentially in time, synchronized with the displacement of the roller. havethe thedisplacement same order of magnitude the roller. and the maxima happen sequentially in time, synchronized with To To better illustrate thethe roller misalignment, Figure 10 10 depicts thethe deformation measurements the displacement ofillustrate the roller. better roller misalignment, Figure depicts deformation measurements corresponding to the four strain gauges in row 3. In this figure, it is noticeable that the peak values corresponding to thethe fourroller strainmisalignment, gauges in row Figure 3. In this it is noticeable that the peak values To better illustrate 10figure, depicts the deformation measurements happen simultaneously but the higher values are obtained from the area subjected to more pressure. happen simultaneously but thegauges higher in values obtained fromitthe area subjected more pressure. corresponding to the four strain row are 3. In this figure, is noticeable thattothe peak values In order to determine misalignment, has been found thatthat best indicators thethe peaks In order to determine misalignment, it obtained has been found the best indicators are peaks happen simultaneously butthe thethe higher valuesitare from thethe area subjected toare more pressure. reached by by each strain gauge. Hence, thisthis hashas been thethe data used in the analysis. ForFor thethe sake of of reached each strain gauge. Hence, been data used in the analysis. sake In order to determine the misalignment, it has been found that the best indicators are the peaks clarity in thethe graphical representations, the maximum values depicted in the figures areare given in in clarity graphical representations, maximum in the figures given reached by in each strain gauge. Hence, this hasthe been the datavalues used indepicted the analysis. For the sake of clarity absolute values although the deformations have a negative value. absolute values although the deformations have a negative value. in the graphical representations, the maximum values depicted in the figures are given in absolute Figure 11 shows thethe absolute magnitude of the maximum deformations suffered by by each strain Figure 11 shows absolute magnitude of the maximum deformations suffered each strain values although the deformations have a negative value. | |. gauge on the sensor It is observed that the deformation is at a maximum in column 1 and gauge on the sensor | á á |. It is observed that the deformation is at a maximum in column 1 and minimum in row 4, which clearly indicates a roller misalignment. minimum in row 4, which clearly indicates a roller misalignment.

Sensors 2017, 17, 2029

12 of 16

Figure 11 shows the absolute magnitude of the maximum deformations suffered by each strain gauge on the sensor |SRCmáx |. It is observed that the deformation is at a maximum in column 1 and Sensors 12 Sensors2017, 2017,17, 17,2029 2029 12ofof16 16 minimum in row 4, which clearly indicates a roller misalignment. 200 200

Col.1 Col.1 Col.2 Col.2 Col.3 Col.3 Col.4 Col.4

Microstrains Microstrains

00 -200 -200 -400 -400 -600 -600 -800 -800

-1000 -1000 -1200 -1200 00

500 500

1000 1000

1500 1500

2000 2000

Time Time(ms) (ms)

2500 2500

Figure 10. Deformations of the strain gauge on row 3.

Figure Deformationsofofthe thestrain straingauge gauge on on row Figure 10. 10. Deformations row3.3.

|Microstrains| |Microstrains|

1500 1500 1000 1000 500 500 00 11 22 33

Sensor Sensorrow row

44

22

11

33

44

Sensor Sensorcolumn column

Figure 11. Maximum deformations of the strain gauges when there is a significant misalignment Figura 11. Figura 11.Maximum Maximumdeformations deformationsofofthe thestrain straingauges gaugeswhen whenthere thereisisaasignificant significantmisalignment misalignment (absolute values). (absolute (absolutevalues). values).

Table 1 shows the maximum absolute values for the deformations of each strain gauge in µε and shows maximum absolute values for ofofeach ininμε the meanTable value them.the The mean value of the deformations is also included, asstrain it willgauge be used in Table11of shows the maximum absolute values forthe thedeformations deformations each strain gauge μεthe and the mean value of them. The mean value of the deformations is also included, as it will be and the mean value of them. The mean value of the deformations is also included, as it will beused used final section. ininthe thefinal finalsection. section. Table 1. Maximum deformations suffered by each strain gauge, given in µε, under significant Table each strain Table 1.1. Maximum Maximum deformations suffered by each strain gauge, gauge, given given inin με, με, under under significant significant misalignment on the leftdeformations and the meansuffered value ofby the maxima. misalignment misalignmenton onthe theleft leftand andthe themean meanvalue valueofofthe themaxima. maxima.

Column 1

Row 1

Row Row11 Row 2 Row Row22 Row 3 Row Row33 Row 4 Row 4 Row 4 MEANT MEAN MEANT T

Column Column11 1138.55 1138.55 1138.55 1223.59 1223.59 1223.59 1163.09 1163.09 1163.09 1174.24 1174.24 1174.24

Column 2

Column Column22 468.78 468.78 468.78 541.81 541.81 541.81 491.48 491.48 491.48 456.43 456.43 456.43

Column 3

Column Column33 294.93 294.93 294.93 325.73 325.73 325.73 321.36 321.36 321.36 268.16 268.16 268.16 514.89 514.89 514.89

Column 4

Column Column44

113.73 113.73 113.73 95.34 95.34 95.34 88.76 88.76 88.76 72.23 72.23 72.23

Sensors 2017, 17, 2029

13 of 16

Sensors 2017, 17, 2029

13 of 16

The total mean value, MEANT , is worked out by averaging the maximum values of the deformations by the strain gauges as follows: Thecaptured total mean value, MEAN T, is worked out by averaging the maximum values of the deformations captured by the strain gauges as follows:

∑ R ∑C SRCmax (1) ∑ ∑16 (1) 16 The condition of alignment is determined from the relative differences with respect to the mean The condition alignment is determined from theMEAN relativeCidifferences with by respect to Equation the mean (2) value. For such purpose,ofthe mean value of each column, , is calculated using value. For such purpose, the mean value of each column, MEAN Ci, is calculated by using Equation (2) and then the deviation from the total mean value, DESCi , is determined by using the expression (3). MEANT =

and then the deviation from the total mean value, DESCi, is determined by using the expression (3).

∑ R SRimax (2) ∑ 4 (2) 4 MEANci − MEANT ·100 (3) DESCi = MEANT (3) 100 Table 2, shows the deviations corresponding to the previous test considering a significant misalignment. Table 2, shows the deviations corresponding to the previous test considering a significant MEANCi =

misalignment. Table 2. Maximum deformations (µε) suffered by each strain gauge in test 1: significant misalignment Table Maximum deformations (με) suffered by each strain gauge in test 1: significant on the left and2.total mean value of the maxima. misalignment on the left and total mean value of the maxima.

MEAN MEANC C DES DESCC

Column Column 11 1174.87 1174.87 128.2% 128.2%

Column Column 2 2 489.63 489.63 −4.9% −4.9%

Column Column33

Column 4 4 Column

302.54 302.54 −−41.2% 41.2%

92.5292.52 −82.0% −82.0%

3.1.2. Test with Aligned Roller

3.1.2. Test with Aligned Roller

The following test (Figure 12), illustrates an example of impression when the roller is in

The following test (Figure 12), illustrates an example of impression when the roller is in alignment. alignment. 200 Col.1 Col.2 Col.3 Col.4

0

Microstrains

-200 -400 -600 -800 -1000

0

500

1000

1500

2000

2500

Time (ms)

(a)

(b)

|Microstrains|

1000

500

0 1 2

4

3

Sensor row

(c)

3

4

2 1

Sensor column

(d)

Figura 12. Test with aligned roller. (a) Diagram of the roller position, (b) Strain gauge deformation,

Figure (c) 12.Maximum Test withabsolute alignedvalues roller.of(a) Diagram of the roller position, (b) Strain gauge deformation, the gauge deformations, means and deviations. (d) Representation (c) Maximum absolute values of the gauge deformations, means and deviations. (d) Representation of of the maximum absolute values. the maximum absolute values.

Sensors 2017, 17, 2029

14 of 16

Sensors 2017, 17, 2029

14 of 16

It can be concluded that, after performing multiple tests, the roller can be considered to be aligned It can be concluded that, after performing multiple tests, the roller can be considered to be when all the deviation values DESCi are below ±10%. This value has been determined by consulting aligned when all the deviation values DESCi are below ±10%. This value has been determined by with staff with expertise in printing about the quality of the images produced for the different degrees consulting with staff with expertise in printing about the quality of the images produced for the of deviation. different degrees of deviation. Figure 13a,b shows anengraving engravinghas has been made with a roller Figure 13a,b showtwo twoexamples examples where where an been made with a roller withwith slightslight misalignment and an roller,roller, respectively. The differences are hardly noticeable for a non-expert misalignment andaligned an aligned respectively. The differences are hardly noticeable for a in thenon-expert domain ofinprinting techniques, sincetechniques, the roller misalignment is close to the adjustment the domain of printing since the roller misalignment is close tothreshold the adjustment threshold indicated above. However, the printing expert considers valid the impression indicated above. However, the printing expert considers valid the impression on the right (Figure 13b), onallows the rightchoosing (Figure 13b), which for allows a value for the adjustment ±10% for the which a value thechoosing adjustment threshold of ±10%threshold for the ofDES Ci parameter, DES Ci parameter, as stated previously. as stated previously.

1000

|Microstrains|

|Microstrains|

1000

500

500

0

0

1

1 2

4

3

Sensor row

3

4

2 1

(a)

Sensor column

2

4

3

Sensor row

3

4

2 1

Sensor column

(b)

Figure 13. Engraving comparison. (a) Roller with minor misalignment and (b) aligned roller.

Figure 13. Engraving comparison. (a) Roller with minor misalignment and (b) aligned roller.

4. Conclusions

4. Conclusions

This paper presents a novel sensor system aimed at on-line monitoring of contact pressure with

application the field of testing.at The designed system looks the contact This papertopresents a chalcographic novel sensorengraving system aimed on-line monitoring of at contact pressure with application to the field of chalcographic engraving testing. The designed system looks at the contact pressure distribution between a cylinder and a flat surface. In order to monitor the contact

Sensors 2017, 17, 2029

15 of 16

pressure profile over the engraved plate, the proposed system uses an electronically integrated strain gauge-based load cell, which consists of a grid of 16 strain gauges arranged in a 4 × 4 matrix. The monitoring system developed can be applied to any printing equipment, e.g., flat-flat and cylinder-flat, provided that the contact pressure profile over the engraved surface is required. The results obtained within the experimental setting prove that the developed prototype operates effectively and efficiently. The representation of the contact pressure profile can be obtained while the printer is on production, i.e., on-line. This profile allows information about the cylinder-plate alignment to be obtained for each impression. To conclude, the proposed system constitutes a remarkable improvement in the printing process in terms of cost savings, in materials and avoiding breakdowns, as well as in the quality and reproducibility of the engravings under similar printing conditions over time. Acknowledgments: This work has been supported by the Real Casa de la Moneda-Fabrica de Moneda y Timbre (FNMT-RCM) through the MODICA Project (Ref. 166/09). Author Contributions: José Antonio Jiménez, Francisco Javier Meca and Enrique Santiso conceived and designed the sensor module of monitorization system. Francisco Javier Meca and Enrique Santiso conceived and designed the conditioning electronics. José Antonio Jiménez and Pedro Martín developed the software applications. All of the authors analyzed the results and checked the manuscript. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5. 6. 7. 8.

9. 10.

11. 12. 13. 14.

15.

Takach Press Corporation. Available online: https://www.takachpress.com/ (accessed on 27 June 2017). Fujifilm Prescale. Available online: http://www.fujifilm.com/products/prescale/ (accessed on 27 June 2017). Ez-Nip. Available online: http://www.sensorprod.com/eznip.php (accessed on 27 June 2017). Nip Impression Paper, an Easy-To-Use, Cost-Effective Tool for Trouble-Shooting Nip Systems. Available online: http://mengesroller.com/products/nip-impression-paper/ (accessed on 27 June 2017). Nip Control. Available online: http://www.nipcontrol.com/ (accessed on 27 June 2017). Sigma-Nip. Available online: http://www.sensorprod.com/sigmanip.php (accessed on 27 June 2017). Tekscan-Nip Pressure Alignment Tool (NPAT). Available online: https://www.tekscan.com/ (accessed on 27 June 2017). Ágoston, K. Vibration Detection of the Electrical Motors Using Strain Gauges. In Proceedings of the 9th International Conference Interdisciplinarity in Engineering, Procedia Technology, Tirgu-Mure, Romania, 8–9 October 2015; pp. 767–772. James Raso, V.; Durdle, N.G.; Hill, D.L.; Lou, E. An Electronically Integrated Load Cell [for Monitoring Pressures of Orthotic Braces]. IEEE Trans. Instrum. Meas. 1998, 47, 551–553. Stefanescu, D.M.; Farcasiu, A.T.; Toader, A. Strain Gauge Force Transducer and Virtual Instrumentation Used in a Measurement System for Retention Forces of Palatal Plates or Removable Dentures. IEEE Sens. J. 2012, 12, 2968–2973. [CrossRef] Demetropoulosa, C.K.; Morganb, C.R.; Sengupta, D.K.; Herkowitzd, H.N. Development of a 4-axis load cell used for lumbar interbody load measurements. Med. Eng. Phys. 2009, 31, 846–851. [CrossRef] [PubMed] Stephen, R.J.; Rajanna, K.; Dhar, V.; Kumar, K.G.K.; Nagabushanam, S. Thin-Film Strain Gauge Sensors for Ion Thrust Measurement. IEEE Sens. J. 2004, 4, 373–376. [CrossRef] Tavakolpour-Saleh, A.R.; Setoodeh, A.R.; Gholamzadeh, M. A novel multi-component strain-gauge external balance for windtunnel tests: Simulation and experiment. Sens. Actuators A Phys. 2016, 247, 172–186. [CrossRef] Niedworok, A. Dynamic Torque Measurement of Drive Transmission in Vehicles Being in Motion Using Strain Gauges Measurement Technique and Wireless Data Transmission. In Proceedings of the 16th European Conference on Power Electronics and Applications (EPE'14-ECCE Europe), Lappeenranta, Finland, 26–28 August 2014; pp. 1–8. Shi, H.; Ikezawa, S.; Ueda, T. A novel method for evaluating triaxial strain gages used in printed circuit board assemblies (PCBA) strain monitoring. In Proceedings of the IEEE Sensors Conference, Limerick, Ireland, 28–31 October 2011; pp. 1697–1700.

Sensors 2017, 17, 2029

16. 17.

18.

19. 20.

21. 22. 23. 24. 25. 26. 27. 28.

16 of 16

Kovacic, B.; Kamnik, R.; Štrukelja, A.; Vatin, N. Processing of Signals Produced by Strain Gauges in Testing Measurements of the Bridges. Proc. Eng. 2015, 117, 795–801. [CrossRef] Yan, C.; Ferraris, E.; Reynaerts, D. A pressure mapping device with Bragg Grating sensors inscribed in bow-tie fibres. In Proceedings of the IEEE Sensors Conference, Limerick, Ireland, 28–31 October 2011; pp. 394–397. Scott, R.H.; Pradipta Banerji, P.; Chikermane, S.; Srinivasan, S.; Basheer, P.A.M.; Surre, F.; Sun, T.; Grattan, K.T.V. Commissioning and Evaluation of a Fiber-Optic Sensor System for Bridge Monitoring. IEEE Sens. J. 2013, 13, 2555–2562. [CrossRef] Tang, Y.; Ren, Z. Dynamic Method of Neutral Axis Position Determination and Damage Identification with Distributed Long-Gauge FBG Sensors. Sensors 2017, 17, 411. [CrossRef] [PubMed] Strathman, J.; Watkins, S.E.; Kaur, A. Strain Monitoring of a Composite Wing. In Proceedings of the SPIE 9803, Sensors and Smart Structures Technologies for Civil, Mechanical, and Aerospace Systems, Las Vegas, NV, USA, 20 March 2016. Budynas, R.; Nisbett, K. Shigley’s Mechanical Engineering Design, 10th ed.; McGraw-Hill: New York, NY, USA, 2014; ISBN 978-0-07-339820-4. Pereira, C.; Ramalho, A.; Ambrosio, J. An enhanced cylindrical contact force model. Multibody Syst. Dyn. 2015, 35, 277–298. [CrossRef] Adams, G.G. Stick, partial slip and sliding in the plane strain micro contact of two elastic bodies. R. Soc. Open Sci. 2014, 1, 140363. [CrossRef] [PubMed] Zhupanska, O.I.; Ulitko, A.F. Contact with friction of a rigid cylinder with an elastic half-space. J. Mech. Phys. Solids 2005, 53, 975–999. [CrossRef] Vishay-Vpgsensors. Available online: http://www.vishaypg.com/docs/11070/031ce.pdf (accessed on 27 June 2017). Texas Instruments. Available online: http://www.ti.com/lit/ds/symlink/xtr105.pdf (accessed on 27 June 2017). National Instruments. Available online: http://sine.ni.com/nips/cds/view/p/lang/es/nid/14686 (accessed on 27 June 2017). Vishay-Vpgsensors, Errors Due to Misalignment of Strain Gages, Tech Note TN-511. Available online: http://www.vishaypg.com/docs/11061/tn511tn5.pdf (accessed on 27 June 2017). © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).