Jtam-A4.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 52, 3, pp. 781-791, Warsaw 2014 A METHOD OF MEASURING MOVING ELEMENT DISPLACEMENTS IN MICRO-HYDRAULIC VALVES WITH THE USE OF OPTICAL FIBERS Grzegorz Łomotowski Wrocław University of Technology, Department of Hydraulic Machines and Systems, Wrocław, Poland e-mail: grzegorz.lomotowski@gmail.com Elżbieta Bereś-Pawlik Wrocław University of Technology, Department of Telecommunications and Teleinformatics, Wrocław, Poland e-mail: elzbieta.pawlik@pwr.wroc.pl Thisworkpresents a fiberoptic technique used tomeasure displacements ofmoving elements in micro-hydraulic valves. Fiberoptic sensors coupled with a reflection method have proven to be perfect for touchlessmeasurement ofmovement for elements hard tomeasure by other methods due to their small size and presence of a hydraulic oil. The article presents results of experiments revealing correlationbetween themeasurement signal and elementmovement as well as sample displacement changes in time of the moving element in non-equilibrium states – for both, a stable action and self-exciting vibrations of the element. Keywords:microhydraulics,micro valves, fiber optic, displacement, vibration,measurement 1. Introduction In the presentworld ofminiaturization, it is a common trend to apply novel solutions to all sorts of machinery drive systems. One of them – hydraulic drive systems are characterized by high power density, i.e. ability to transfer high levels of power with a relatively small size and mass of the drive system elements. Traditional massive pneumatic or electromagnetic drive systems can be replaced by miniaturized hydraulic systems of equal power transfer capacity and broad automation potential. Considerations mentioned above contribute to dynamic growth of a new technical field known as high-pressure micro hydraulics–hydraulics, where the flow is less than 2dm3/min. Some examples ofmicro-hydraulic systems aswell as a global review ofminiaturized micro-hydraulic elements producedbyvarious companieswere described in detail in Łomotowski (2013). High-pressure micro hydraulics is a relatively new field of technology that still requires a rangeof scientific studies. It turnsout that someof the rules governing classic large-size hydraulic systems might not apply to micro hydraulics. Examples of effects possibly significant for high- pressuremicro hydraulics are non-isothermal flow (Kollek et al., 2011; Kollek and Łomotowski, 2012) and obliteration (Kudźma et al., 2011). The liquid pressure and rate of floware the two basic parametersmeasuredwhile researching micro-hydraulic systems and elements.When conducting research onmicro-hydraulic elements, otherparametersmayhave tobemeasuredaswell; for example,measuringdisplacementsofparts moving inside the studied micro valves in equilibrium and non-equilibrium states can broaden knowledge about processes taking place inside the elements. This should lead to eliminate the disadvantageous vibration. The vibration of moving parts inside valves, which may gradually destroy the cooperating elements, may derive from external factors; for example the vibration can be propagated from an element located in valve surroundings which have contact with the valve body. The effects of external mechanical vibrations on hydraulic valves and theoretical 782 G. Łomotowski, E. Bereś-Pawlik analysis of the contribution of selected vibration insulators to reduction in the hydraulic valve housing vibrations was carried out in Stosiak (2011). An effective method of reduction of the influence of external vibrations on the hydraulic valve is an application of special insulators for valve housing (Kudźma and Stosiak, 2013; Stosiak, 2011). The application of insulators leads to valve housing vibrations reduction in a wide frequency range (Stosiak, 2011). The other type of vibration is self-excited vibration which derives from unstability of dynamic systems (Misra et al., 2002). An example of a hydraulic element where stability loss can occur what leads to self-excited vibration, is a relief valve (Licsko et al., 2009; Łomotowski, 2013). Measuring the displacement of the working valve closing element allows for obtaining more information about the valve static and dynamic properties. The gained knowledge leads to designing correctly high-quality micro-hydraulic elements. Measurements of moving part displacements in micro-hydraulic components are difficult for several reasons. The basic problems are small hydraulic valve dimensions and the presence of the oil, which leads to the decision of including fiberoptic technology in the measurement methodology employed. The measurement method involving fiberoptic sensors founds its way to wide use in industrial, military and biomedical systems (Yasin et al., 2008; (Harun et al., 2010; Shen et al., 2010). Its advantages include but are not limited to: non-contactmeasurement ability for very little displacements, very good dynamic properties (widemeasurement frequency range) and very small dimensions of the sensors themselves (Cao et al., 2005; Yasin et al., 2008; Harun et al., 2010). These factors exactly, as well as the fact that light can be transmitted through hydraulic oil, contributed to making a decision on choice of the fiberoptic method for measuringmicro valve closing elements displacements. 2. Displacement and the mechanical vibration measurement method using fiberoptic technique According to Samian et al. (2009) there are three types of fiberoptic displacement sensors, and they take advantage of threedifferentphenomena: changingwavelength, changing light ray inten- sity (amplitude sensor) and change in phase shift (phase sensor).Thedisplacementmeasurement method associated with light intensity change is the best for conducting dynamicmeasurements as it is relatively cheap and the results are easy to interpret, even though themethod is a little less accurate.Most of the time, thismethod relies on amirror that reflects light towards a photo- sensor. Appropriate shifting of themirror changes the reflected light intensity, and consequently the measuring signal. This method has been used in many different variations as published in Cao et al. (2005), Yasin et al. (2008), Samian et al. (2009), Shribak et al. (1996), Harun et al. (2010), Shen et al. (2010) tomention a few.The authors of these publications propose applicable mathematical models (for the relationship between measuring signal and mirror shifting) and they confirm themodels bymeans of experiments. Both, single andmultimode optical fibers are used for the proposedmethods. The work by Cao et al. (2005) treats about two situations: one, where two optical fibers are connected so that the first fiber conducts the signal from transmitter and the second fiber directs the signal toward photosensor; two, where more transmitting and receiving fibers have been connected. The work by Harun et al. (2010) presents a setup with a single transmitting fiber and two receiving fibers, whereas the work by Yasin et al. (2008) depicts one transmitting fiber against sixteen receiving fibers. Figure 1 shows a simplifiedoptical systemdiagramof the construction described in theworks by Samian et al. (2009), Shribak et al. (1996) and Shen et al. (2010) that employs only one multimode optical fiber for both the sending and receiving signals. For this design, amultimode A method of measuring moving element displacements in micro-hydraulic valves... 783 coupler is used to connect the transmitter, receiver and fiber pointed at themovingmirror. The presented research was conductedwhile sensors were in a dry environment, however the authors of thework by Shribak et al. (1996) stated that the filling of the space between the fiber tip and the mirror with a light-absorbing liquid of an appropriately selected refraction index would be a goodmethod for increasing themeasurement range as well as the sensitivity of themeasuring system. That, in turn, evoked an idea that sensors immersed in the hydraulic oil should work very well as long as the oil meets specific requirements. Fig. 1. Simplified diagram of the optical measuring system According to Samian et al. (2009) and Shribak et al. (1996), the total correlation between the measuring signal, i.e. the reflected light power and the mirror displacement is not linear in the entire measurement range. However, the range where that correlation is very close to linear can be found. Fiberoptic displacement sensors work in that range. Figure 1 presents a simplified diagram of displacement sensors which can offermeasurement accuracy within millimetres, micrometres or even nanometres – depending on the optical fiber used and lights source intensity. An interesting configuration is described in the work by Shen et al. (2010), where a broad spectrum light source offering smooth intensity choice change was used along with a special collimator which enabled broadening the testing range. The sensor could work in the 5-30cm range. A layout described by Samian et al. (2009) and Shribak et al. (1996), Fig. 1, that is a part of distance measuring fiberoptic sensors, is used to measure moving element displacements inside one of the micro-hydraulic valves. This model allowed for developing a measuring technique, where displacements of an element having a few millimetres in size, entirely immersed in the hydraulic oil, could be measured in the 0-100µm range. Optical fiber GI 50/125 was used to evaluate the displacement. A micro pressure relief valve was the object of research with a fiberoptic-based technique attempted to be used tomeasure displacements of a few-millimetre-longmoving element entirely immersed in thehydraulicoil. Figure 2presentsa simplifieddraftof suchavalve and its operating principle. The adjusting screw x s controls the set pressure, i.e. the force the closing element exerts on the valve seat. The valve remains closed until the opening pressure set with the adjusting screw is reached in the inlet channel P. When the pressure is exceeded, the valve opens and the liquid is allowed to flow to the outlet channel T. As long as the defined rate of flow ismaintained, the pressure in the approach channel P remains close to the level set with the adjusting screw, and the valve remains open. Hence, this valve is a sort of a pressure regulator that can be unstable and succumb to self-exciting vibrations if construction parameters are maladjusted. The tested valve seat diameter is 2mm. The valve has been designed to work with liquids flowing at rates up to 1dm3/min and 16MPa of maximum setting pressure. Figure 2 presents some construction changes thatwere implemented in order tomakemeasu- rements of displacements of the closing element possible. An optical fiber embedded in a metal pipe is inserted into a hole drilled in adjusting screw 1. A mirror is glued to closing element 2. 784 G. Łomotowski, E. Bereś-Pawlik Fig. 2. Method of measuring displacements of the moving element: (a) simplified relief valve diagram; (b) fiberoptic testing probe 1 andmirror 2 installed in the relief valve for the displacement measurements; (c) relief valve, equipped with fiberoptic testing probe 1, mirror 2 and calibrating probe 3, during calibration Laser light sent from the transmitter is directed at the mirror, so it can travel across the valve and go back to the receiver – light sensor. A multimode 2× 1 coupler is used so that a single fiber can transport the transmitted and reflected beams simultaneously. The longer the distance between the closing element and the seat, the shorter the distance between the optical fiber and themirror, and the higher power of the laser beam returning to the detector that, in turn, produced a proportional electric signal. Undernormalworking conditions, the closing elementmoves away fromthe seat as a result of the liquid pressure acting on it. The fiberoptic technique gives ability to measure displacement changes in time. The required beforehand calibration can be done with separate calibrating shaft (3) presented in Fig. 2c. The shaftmoves the closing element away from the seat to create a gap of required size. Following the calibration, the calibration shaft can be removed from the inlet channel, and the channel can be reconnected to the system without affecting the optical fiber position. It isworthmentioning that the calibration should take place before eachmeasurement as not only the reflected signal power depends on the closing element position, but also on the optical fiber holding the testing probe setting. Figure 3 depicts two cooperating elements that play the key role in the optical fiber based measurements of the studiedmicro valve. Fig. 3. Cooperating elements: (a) mirror-equipped element whose movements aremeasured, (b) testing probe with optical fiber installed in its core A method of measuring moving element displacements in micro-hydraulic valves... 785 It is noteworthy that both themirror-equipped closing element and the optical fiber holding testing probe work in a low-pressure environment (that comes from the liquid flow resistance met on its direct way to the tank). This is very advantageous as it facilitates sealing off the testing probe, and itminimizes the risk of forcing the optical fiber out of the testing probewith the pressurized liquid. Measuring the closing element displacements involves: • specially adapted micro valve • optical fiber 50/125µm in diameter • 850nm-wavelength VCSEL diode of 0,5mW (transmitter) • Si-PIN diode with 500µmof active area • multimode 2×1 coupler with 50/50 split coefficient • micrometer gauge (for calibrating the measuring system) • electronic circuitry changing the light signal from the sensor to an electric signal readable by the oscilloscope • Tektronix TDS224 oscilloscope • Universal micro hydraulics testing station (Łomotowski, 2013) 3. Calibrating the measuring system Calibration is imperative in order to obtain information about displacement magnitude. While the valve is working at the micro valve testing station, calibration can be done before or after taking displacement measurements. The calibration should be done at the same conditions as duringmeasurements (e.g. the same temperature of thehydraulic oil). Figure 4 shows the studied micro valve during its calibration. Fig. 4. Studied micro valve during calibration: 1 – micro valve body, 2 – calibration shaft for displacing the closing element, 3 – micrometer gauge for measuring closing element displacements in relation to seat, 4 – testing probe (measuring shaft) containing optical fiber, 5 – fiberoptic coupler, 6 and 7 – coupler-transmitter and coupler-receiver connectors Table 1 contains a sample of measuring system calibration results for two different testing probepositions. Inboth cases, the increase of testing signal after opening thevalvewasmeasured in relation to its value at the time of valve closed position. Threemeasurement setswere performed in each case because ofmarked external interference with the testing signal. Among other things, some micro valve manufacturing flaws and impu- rities in the hydraulic oil adversely affected light travel through the valve causing interference. 786 G. Łomotowski, E. Bereś-Pawlik Table 1.Calibration results for two settings of the initial distance fromthemirror to the optical fiber Setting 1 Setting 2 x [µm] u1 [mV] u2 [mV] u3 [mV] x [µm] u1 [mV] u2 [mV] u3 [mV] 10 4 10 6 10 8 6 8 20 8 12 10 20 16 16 18 30 14 18 16 30 26 22 24 40 22 22 20 40 32 24 30 50 28 28 28 50 38 36 40 60 34 32 32 60 44 46 46 70 38 38 36 70 52 50 52 80 42 40 40 80 62 58 62 90 44 44 42 90 72 64 70 100 50 44 48 100 78 72 76 It has been noted that the calibration relationship between the closing element displacement and testing signal measured in volts is approximately linear. Figure 5 depicts calibration results for two different testing probe positions with associated linear approximations. Fig. 5. 5: Sample calibration results for two different testing probe settings (x axis – displacement in µm, y axis – testing signal in mV) For the calibration represented by results obtained during setting 1, the relationship can be expressed with formula (3.1)1; for the calibration represented by results obtained during setting 2, the relationship can be expressed with formula (3.1)2 x [µm] = 1.967u [mV] x [µm] = 1.322u [mV] (3.1) In order to estimate the repeatability and linearity of measurement results, based on three sets of results, the regression coefficient confidence level canbeestimatedbymeansof themethod of least squares. In the first case, for the confidence level of 95%, the regression coefficient falls between 1.9091 and 2.0243, therefore the coefficient deviation is 2.9%. In the second case, for the confidence level of 95% the regression coefficient falls between 1.2977 and 1.3464, therefore the coefficient deviation is 1.8%. These results confirmthe correct choice of themeasurementmethod.Accuracy of themeasu- rements is hard to evaluate since the setup for mirror-equipped closing element displacements was pretty rough itself. Although the testing probe position was adjusted with an accuracy of 10µm, setting the gap between the mirror and the optical fiber tip was not necessarily that A method of measuring moving element displacements in micro-hydraulic valves... 787 accurate. The valve has been manufactured by conventional standards. In spite of demand for high quality control, the valve was still produced in quality levels allowing for shape and size deviations ranging tens of micrometers. 4. Micro valve testing at the micro-hydraulic station Working in ahydraulic system, thedisplacementmeasurements of the relief valve closing element were taken at a universal micro-hydraulic element testing station (Łomotowski, 2013). Figure 6 depicts the hydraulic system put together for the purpose of conducting our study. Fig. 6. The hydraulic system for measurements of the micro valve closing element displacements Figure 7 shows the studied hydraulic system diagram. The system consists of the following elements (marked with numbers corresponding to Figs. 6 and 7): 1 – tested micro relief valve adapted to allow taking the closing element displacement me- asurements with the fiberoptic method 2 – testing probe holding the optical fiber used to transport the displacement measuring signal back to the measurement system 3 – hydraulic oil tank 4 – micro pumps motorized with a variable-speed revolution motor allowing for smooth pumped oil flow changes from 0 to 1.1dm3/min 5 – throttle valve 6 – electromagnetic shut off valve 7 – digital manometer of measuring range 40MPa 8 – pressure strain gauge of measuring range 25MPa 9 – flow quantity meter of measuring range 2dm3/min The very essence of the measurements is registering the displacement changes in time of the valve closing element lift following an appropriate step change in the flow directed at the micro valve. Different actions can be obtained depending on the opening pressure set and the 788 G. Łomotowski, E. Bereś-Pawlik Fig. 7. Diagram of the hydraulic system for measurements of the micro valve closing element displacement flow rate directed at it. In order to achieve the desired action, micro pump (4) has to be set to an appropriate revolution rate that yields a desired rate of flow, then electromagnetic shut off valve (6) has to be overrun to close. As a result, the entire pumped liquid flows toward the pressure relief valve. In addition, pressure strain gauge (8) enables recording of pressure changes in time. Figure 8 presents the recording system that consists of: (1) fiberoptic coupler, (2) transmitter electronic module power supply that also converts light signals from the sensor (receiver) to electric signals readable by the oscilloscope, (3) device that converts pressure sensor signals to electric signals readable by the oscilloscope, (4) oscilloscope, (5) computer recording signals from the oscilloscope in files. Fig. 8. Themeasuring system for the micro valve closing element displacement The studied valve was built in the way that allowed for changes in some construction para- meters in order to test for the best parameter choice for the best performance. Figure 9 presents valve closing element timing associated with correctly selected construction parameters. In this case, the valve action is stable. Figure 9 shows unaltered recording and the same recording filtered by the moving average method. Original recording show some deviations from the equ- ilibriumcaused bymeasurement systemnoises, impuritiesmovement in the hydraulic oil causing changes in the reflected light luminosity as well as slight closing element vibrations caused by the liquid flowing around it. A method of measuring moving element displacements in micro-hydraulic valves... 789 Fig. 9. Exemplary course of the micro valve closing element displacements during a step change in the flow directed at the valve – unaltered signal recording and refined recording (noises filtered out) Figure 10 presents a time interval record of the closing element displacement in the va- lve with construction parameters chosen incorrectly. In this case, the valve action is unstable. Considerable self-excited vibrations can be observed in the valve. Figure 11 presents results of vibrations Fast Fourier Transform – the original record and the transform. It is noteworthy that valve vibrations are so intensive that the closing element knocks the seat. Such operation is unacceptable because the cooperating parts can quickly get damaged, and because of powerful pressure pulsation. Unstable valve action can easily be recognized as it causes a characteristic noise. Fig. 10. Exemplary course of the micro valve closing element displacement during a step change in the flow directed at the valve – unstable valve action Fig. 11. Exemplary valve vibration frequency analysis during an unstable action It is clearly seen that the base frequency of self-excited vibrations is 1830Hz.The first higher harmonic can clearly be noticed. No vibrations occur above 6000Hz. 790 G. Łomotowski, E. Bereś-Pawlik 5. Conclusions Fiberoptic technology enables measuring displacements of small moving elements entirely sub- merged in ahydraulic oil.Thepresentedmethodcanbeapplied to studymicro-hydraulic systems elements –micro valves in particular. Thismethod of taking vibration frequencymeasurements is highly accurate. It is also important that the method provides ability to record the constant- displacement component; therefore, it is possible to measure the way the micro valve moving element displacements shape in time in valves working in both the equilibrium state and in non-equilibrium state. Researchers collecting information this way can broaden their knowledge about valves functioning and phenomena associated with it. That knowledge should contribute to establishing criteria for designing high-quality micro-hydraulic valves correctly. Thanks to very good dynamic properties of the measuring method, fiberoptic displacement sensors can be used, among others, in automatic control of elements inmicro-hydraulic systems. For example, a micro-hydraulic directional valve equippedwith a fiberoptic sensor can give out measurements that influencemore precise settings for the slider travels. A very good measurement method has been presented in this article. Nevertheless, the me- thod requires further development involving a system built entirely with high-quality compo- nents. The presented calibration results are based on a very rough displacement setup, yet the results repeatability is good. Future attempts should bemade on system components produced withhigher accuracy.Requireddisplacements shouldbe executedwithbetter precision, e.g. with a piezoelectric actuator (Harun et al., 2010). References 1. CaoH.,ChenY.,ZhouZ.,ZhangG., 2005,Generalmodels ofoptical-fiber-bundledisplacement sensors,Microwave and Optical Technology Letters, 47, 5, 494-497 2. Harun S., W., Yang H., Z., Yasin M., H. Ahmad H., 2010, Theoretical and experimental study on the fiber optic displacement sensor with two receiving fibers, Microwave and Optical Technology Letters, 52, 1, 373-375 3. KollekW.,KudźmaZ., ŁomotowskiG., StosiakM., 2011,Non-isothermal flow inmicrohy- draulic systems, [In:]Fundamentals of Design,Modeling andOperation ofMicrohydraulic Elements and Systems (in Polish),W. Kollek (Edit.), OficynaWydawnicza PolitechnikiWrocławskiej,Wro- cław, 65-72 4. Kollek W., Łomotowski G., 2012, Non-isothermal flow in microvalve clearances (in Po- lish),Proceedings of Hydraulic and Pneumatic Drives and Controls 2012. International Scientific- Technical Conference, Wrocław 2012, 205-217 5. Kudźma Z., Rutański J., Stosiak M., 2011, Fluid requirements in microhydraulic systems, [In:] Fundamentals of Design, Modeling and Operation of Microhydraulic Elements and Sys- tems (in Polish), W. Kollek (Edit.), Oficyna Wydawnicza Politechniki Wrocławskiej, Wrocław, 97-104 6. Kudźma Z., Stosiak M., 2013, Reduction of infrasounds in machines with hydrostatic drive, Acta of Bioengineering and Biomechanics, 15, 2, 51-64 7. Licsko G., Champneys A., Hos C., 2009, Nonlinear analysis of a single stage pressure relief valve, IAENG International Journal of Applied Mathematics, 39, 4, 286-299 8. Łomotowski G., 2013, The influence of construction parameters on the static and dynamic cha- racteristics of hydraulic microvalves, Ph.D. Thesis,WrocławUniversity of Technology,Wrocław 9. Misra A., Behdinan K., Cleghorn W. L., 2002, Self-excited vibration of a control valve due to fluid-structure interaction, Journal of Fluids and Structures, 16, 5, 649-665 A method of measuring moving element displacements in micro-hydraulic valves... 791 10. Samian, PramonoY.H., Rohedi A.Y., Rusydi F., ZaidanA.H., 2009,Theoretical and expe- rimental study of fiber-optic displacement sensor using multimode fiber coupler, Journal of Opto- electronics and Biomedical Materials, 1, 3, 303-308 11. ShenW.,WuX.,MengH.,ZhangG.,HuangX., 2010,Longdistancefiber-optic displacement sensor based on fiber collimator,Review of Scientific Instruments, 81, 12, 123104 12. Shribak M.J., Kolpashchikov V.L., Martynenko O.G., 1996, Fiber optic sensor of linear displacement,Proc. SPIE. 2895,Fiber Optic Sensors,V, 305 13. Stosiak M., 2011,Vibration insulation of hydraulic system control components,Archives of Civil and Mechanical Engineering, 11, 1, 237-248 14. Yasin M., Harun S. W., Kusminarto K., Ahmad H., 2008, Fiber-optic displacement sensor using a multimode bundle fiber,Microwave and Optical Technology Letters, 50, 3, 661-663 Manuscript received September 22, 2013; accepted for print February 26, 2014