Multi-Parameter Sensing Device to Detect Liquid Layers Using ... - MDPI

sensors Article

Multi-Parameter Sensing Device to Detect Liquid Layers Using Long-Period Fiber Gratings Zhihui Pan 1,2 , Ying Huang 2, * and Hai Xiao 3 1 2 3

*

School of Civil Engineering, Guangzhou University, Guangzhou 510006, China; [email protected] Department of Civil and Environmental Engineering, North Dakota State University, 1340 Administration Ave, Fargo, ND 58108, USA Department of Electrical and Computer Engineering, Clemson University, Clemson, SC 29634, USA; [email protected] Correspondence: [email protected]; Tel.: +1-701-231-7651

Received: 8 August 2018; Accepted: 12 September 2018; Published: 14 September 2018







Abstract: Insoluble liquids show layers such as water and oil. The detection of the exact interface locations and the level changes for layered liquids are of paramount importance for chemistry purifications, liquid storage in reservoirs, oil transportation, and chemical engineering. However, accurately measuring liquid layers is challenging. This paper introduces a multi-parameter sensing device based on a long-period fiber grating (LPFG) sensor simultaneously detecting boundary and level changes of layered liquids. Laboratory experiments demonstrated that the sensor device would respond to the liquid interface change as a sharp and sudden resonant wavelength change, while it would show a gradual and steady resonant wavelength change to the level changes of layered liquids. The lab experiments also showed that the sensor device has a higher sensitivity when a higher LPFG cladding mode is used. Keywords: layered liquid boundary; liquid level; long-period fiber grating sensor

1. Introduction For chemical processing, purification, storage, and transportation, it is critical to know whether unwanted liquids are present. If unwanted insoluble liquids exist, it becomes essential to locate the interface between the layered liquids and monitor the level change of each layer in real time. The measurement of liquid level change is widely applied. For instance, a fuel level sensor is commonly used in a car, truck or motorcycle to measure the gasoline level in the fuel tank. Sensor devices to detect level changes of liquid have been in place for decades. Various technologies are available including mechanical devices [1], electrical approaches [2–4], ultrasonic and acoustic technology [5]. A mechanical device estimates the liquid level through gauging the length of a wire on a floating ball or the gas pressure change inside an enclosed space [1]. Due to the implementation limitation of the mechanical device, it has a relatively low measurement accuracy. To measure more accurately, an electrical device can be used, such as magnetic, electrical conductivity, and capacitive sensing devices, through counting the changes of electrical-magnetic properties of the liquid [2–4]. These electrical devices can only be applied to liquids with significant changes of electrical or magnetic property towards liquid level changes. In addition, for some hazardous liquids, the use of these electrical devices may induce fire or explosion hazard. To have a safer field practice, an ultrasonic or acoustic sensor is an alternative to measure interface distance through finding the travel time of ultrasonic or acoustic waves [5]. However, due to the possible delays in sound wave measurements, an accurate level measurement requires high costs. Although the level sensors have been widely investigated and, to date, limited technologies can detect interfaces between layered insoluble liquids. Sensors 2018, 18, 3094; doi:10.3390/s18093094

www.mdpi.com/journal/sensors

Sensors 2018, 18, 3094

2 of 14

A fiber optic sensor, due to its unique advantages of compactness, immunity to electro-magnetic noise and moisture, no fire or explosion hazard, and long life cycle, can be a potential candidate for detecting layers of liquids. Various optic fiber sensors have been investigated for liquid level sensing such as intensity-based sensors [6], fiber Bragg grating [7,8], and long-period fiber gratings (LPFG). Among these fiber optic sensors, intensity-based sensors and fiber Bragg gratings are usually applied as on/off sensor for level warning. LPFG couples the core mode to a co-propagating cladding mode which decays rapidly as they propagate along the fiber axis from the scattering losses at the cladding-air interface in the fiber [9]. One cladding mode experiences total internal reflection at the interface between the cladding and the surrounding environment, resulting in a dependence of the effective index of the cladding mode on the surroundings environments. The LPFG sensors have been widely applied in chemical, temperature and strain sensing [10–12]. Due to high sensitivity to environmental changes, LPFGs can potentially be a continuous liquid level sensor. Khaliq et al. [13] first applied LPFGs to roughly demonstrate oil level detection. The sensor showed a large linear range with sensitivity of 4.8% change in transmission per millimeter. Grice et al. [14] repeated the experiment and found that the level change of the liquid with a specific refractive index would induce both a shift in wavelength and a change in the attenuation level of the selected loss band. Hu et al. [15] and Xue et al. [16] formed LPFGs into interferometer sensors to detect liquid level changes. Recently, LPFG sensors also have been investigated for multi-parameter sensing in liquids. In 2012, Wang et al. [17] used LPFG sensor to simultaneously detect liquid level and flow velocity. Later, Huang et al. [18] developed LPFG sensors for simultaneous measurements of liquid level and refractive index. Yu et al. [19] combined LPFG and tilted FBG for two-parameter sensing in liquids. More recently, Jiang et al. [20] has advanced the LPFG sensor to simultaneously measure liquid refractive index and temperature. Although LPFG sensors have been investigated for detecting liquid levels and multiple parameter sensing in liquids, limited research is done for liquid layer detection. The common approaches used for detecting the interfaces among various liquids are still highly dependent on visible examination or electrical capacitive sensors [21], which is inaccurate and in most cases not safe to operate. Although recently the authors introduced LPFGs for layered liquid detection [22], limited tests were performed to show its reliability. To meet this challenge, this study (1) systematically investigated the operational principle and feasibility of the LPFG-based sensing device for boundary detection between layered liquids, (2) advanced the measurements to multiple parameter detection of layered liquids with both the interfaces and the level changes of layered liquids, (3) validated the multi-parameter sensing through well-designed laboratory tests; and (4) performed sensitivity study on the sensor device for the rod diameter, material, and LPFG cladding modes selection. The paper is organized as follows: Section 2 discusses the sensor operational principle and device design; Section 3 explains the experimental setup; Section 4 conducts experiments to validate feasibility and effectiveness of multi-parameter sensing; Section 5 performs a sensitivity study of the sensor device through laboratory experiments; and Section 6 concludes the study and lay out its potential future work. 2. Operational Principles and Device Design 2.1. Principles of Operation The LPFGs used in this paper were fabricated by CO2 laser irradiation technique following reference [23]. The fiber used to fabricate the LPFG is single mode optical fiber (supplied by Corning). The LPFG has a sensor length of approximately 30–32 mm with 80–90 points of CO2 laser radiation and a period of 0.363–0.375 mm. Figure 1 shows a typical spectrum of one CO2 laser-induced LPFG with five different cladding modes (LP01 –LP05 ). Each valley in Figure 1 corresponds to one cladding mode. The resonant wavelength (λres ) of a LPFG sensor can be expressed as a linear function of its

Sensors 2018, 18, 3094

3 of 14

Sensors 2018, 18, x FOR PEER REVIEW

3 of 14

grating period (Λ) and effective refractive indices of the core (ne f f , co ) and the cladding mode LP0m (ne f f , cl, m ) as follows [9]: λres== -n , mm)Λ (1)(1) λres (n(en )Λ co − neefff f, ,clcl, f eff f , ,co

Figure response. Figure1. 1. Representative Representative LPFG transmission response.

In Equation (1), the effective refractive indices of the core, ne f f , co , is a constant which will not In Equation (1), the effective refractive indices of the core, neff , co , is a constant which will not be be affected by changing surrounding medium. The grating period, Λ, although responds to physical affectedsuch by changing surrounding medium. Thefiber, grating period,does Λ, although responds to physical changes as temperature and stress on the it usually not change with surrounding changes changes. such as temperature and effective stress onrefractive the fiber, it usually does not change with medium However, the index of the cladding mode LPsurrounding 0m , ne f f , cl, m , is highly dependent on the refractive index of the surrounding mediums, ns . A change in nns will bring medium changes. However, the effective refractive index of the cladding mode LP0m, eff , cl , m , is in a corresponding change of ne f f , cl, m in Equation (1), leading to a shift of the resonant wavelength of A change in ns will bring highly dependent the refractive of the surrounding mediums, the cladding modes,onλres . The LPFG index resonant wavelength sensitivity on nnss.can be obtained as [24]: in a corresponding change of

neff ,cl , m

in Equation (1), leading to a shift of the resonant wavelength dλres = λres ·γ·Γs (2) of the cladding modes, λres. The LPFG resonant dns wavelength sensitivity on ns can be obtained as [24]:

dλwhich where, γ describe the waveguide dispersion is positive for lower cladding modes and negative res = λ ⋅ γ⋅ Γs dependent on the resonant wavelength (2) for higher cladding modes and the turning d point ns of γresis highly range [24]. To avoid turning points and have a high sensitivity on the refractive index changes of surrounding mediums, an appropriate cladding needs be selected. 5.3 performs where, γ describe the waveguide dispersion whichmode is positive fortolower claddingSection modes and negative higher cladding modes and the turning point of is highly dependent on the resonant afor detail experimental investigation on influence ofγLPFG cladding modes on dλ and selects res /dnswavelength range To avoid turning points and have a high sensitivity on the index of the best[24]. cladding modes for the targeted application. Γs expresses therefractive dependence of changes the grating surrounding mediums, an appropriate cladding mode needs to be selected. Section 5.3 performs a to the surrounding refractive index and is defined by reference [24], which is negative for all the detail experimental investigation on influence of LPFG cladding modes on dλres/dns and selects the cladding modes. bestThe cladding for themedium targetedchanges application. Γs expresses dependence of location the grating to the LPFG modes surrounding from one liquid to the another along any during the surrounding refractive index and is defined by reference [24], which is negative for all the cladding modes. length of the sensing unit will introduce a sudden change of Γs in Equation (2), resulting in a change in The LPFG surrounding medium one in liquid to another of along any location duringis the sensor sensitivity, dλres /dn a suddenfrom change the sensitivity the sensor unit, which s . Thus, changes the length of the sensing unit will introduce a sudden change of Γ s in Equation (2), resulting in a the slope of the level measurements, indicates the occurrence of boundaries in layered insoluble liquids. change in the sensor sensitivity, dλ res /dn s . Thus, a sudden change in the sensitivity of the sensor unit, The interfaces between layered liquids locate at wherever the change of sensitivity change occurs along which is unit. the slope the level indicates the of occurrence of boundaries layered the sensor Fromof Equation (2),measurements, it shows that the resolution interface between liquidsin can be one insoluble liquids. The interfaces between layered liquids locate at wherever the change of sensitivity grating period, which is in between 0.363 mm to 0.375 mm. change occurs(2) along sensor From Equation it shows thatremain the resolution of interface Equation also the shows thatunit. if the properties of (2), layered liquids unchanged after the between liquids can be one grating mmsensitivity to 0.375 mm. occurrence of insolvable liquids, theperiod, Γs will which remainisasinabetween constant0.363 and the of the LPFG will (2) also shows that (1) if the properties offrom layered liquids(2)remain after the remainEquation unchanged. Then, Equation with sensitivity Equation can be unchanged used to continuously occurrence of insolvable liquids, the Γ s will remain as a constant and the sensitivity of the LPFG will monitor level changes until the liquids emerging the entire sensor unit. For more details of the remain unchanged. Then, Equation (1)level with changes, sensitivityitfrom Equation (2)to can be used to continuously LPFG sensing principle about liquid can be referred [24]. Therefore, based on monitor level changes thedevice liquidsbased emerging the entire sensor more details of the Equations (1) and (2), auntil sensor on LPFG sensing unitunit. can For potentially be used to LPFG detect sensing principle about liquid level changes, it can be referred to [24]. Therefore, based on Equations (1) and (2), a sensor device based on LPFG sensing unit can potentially be used to detect multiple

Sensors 2018, 18, 3094

4 of 14

multiple parameters of layered liquids, including the interface locations and the level changes of each 4 of 14 it can be seen that if a higher resolution for liquid level measurement is desired, a smaller LPFG period, Λ, may yield a better result [24]. However, a smaller period will parameters of layered liquids, including the interface locations and the level changes of each layer. From result in a short LPFG length leading to a smaller measurement range. In this study, all the LPFG Equations (1) and (2), it can be seen that if a higher resolution for liquid level measurement is desired, a sensors used share the same LPFG period and length to exclude the impacts from LPFG fabrication, smaller LPFG period, Λ, may yield a better result [24]. However, a smaller period will result in a short which will be investigated in future. LPFG length leading to a smaller measurement range. In this study, all the LPFG sensors used share the same LPFG period and length to exclude the impacts from LPFG fabrication, which will be 2.2. Sensor Device Design investigated in future. To measure multiple parameters for layered liquids as described in Section 2.1, the sensor device 2.2. Sensor Device is expected to beDesign robust, mobile, and ease to use. However, a single LPFG sensing unit made up of glassTo fiber with a diameter of 125 µmfor is layered too weak to be as a probe for any possible fielddevice detection. measure multiple parameters liquids described in Section 2.1,lab theand sensor This paper to designs a sensor device LPFGs to perform robust measurements. is expected be robust, mobile, and based ease toon use. However, a single LPFGlayered sensingliquid unit made up of Figure 2 shows a schematic view of the sensor device. The sensor device consists of two components: glass fiber with a diameter of 125 µm is too weak to be a probe for any possible lab and field detection. a LPFG sensing and adevice holding rod. The LPFG sensingrobust unit is attached onmeasurements. the surface of the This paper designsunit a sensor based on LPFGs to perform layered liquid holding rod using epoxy on the of both of device. the sensing unit. To avoidconsists the phenomenon of capillarity, Figure 2 shows a schematic view theends sensor The sensor device of two components: the epoxy on both ends of the sensing unit has a thickness of more than 2 mm. The material of the a LPFG sensing unit and a holding rod. The LPFG sensing unit is attached on the surface of the holdingrod rodusing can epoxy be glass or metal. Theofdiameter ofunit. the To holding cannot be too small to avoid holding on the both ends the sensing avoid rod the phenomenon of capillarity, capillarity effect measurements. A unit detail experimental investigation the The sensitivity of the the epoxy on bothon ends of the sensing has a thickness of more than 2on mm. materialstudy of the sensor device material andThe diameter of the holding rodrod willcannot be discussed in Section 5.1 and holding rod canonbethe glass or metal. diameter of the holding be too small to avoid capillarity effect on measurements. A detail experimental investigation on the sensitivity study of the Section 5.2. sensorAs device the material and2,diameter of thethe holding rod will be discussed in Sections andthe 5.2.level also on shown in Figure to measure interface in between layered liquids5.1and As also shown in Figurelight 2, to measure interface inlight between layered and the the leveltransmission changes changes simultaneously, from thethe broadband source willliquids go through simultaneously, lightsensing from theunit broadband light sourceon will goholding throughrod. the transmission fiber and fiber and the LPFG which is attached the The light will then bethe picked LPFG sensing unit which is attached on the holding rod. The light will then be picked up and recorded up and recorded continuously using an optical spectrum analyzer. The optical spectrum analyzer continuously using an spectrum analyzer. The optical spectrum analyzer willthe be monitored connected todata a in will be connected to optical a personal computer to perform data analysis and plot personal computer to perform data analysis and plot the monitored data in user-friendly view. user-friendly view. To from external damages during transportation of the Tofurther furtherprotect protectthe thesensing sensingunit unit from external damages during transportation of sensor the sensor device, a protection plastic or glass tube with an open end and larger diameter than the holding rod device, a protection plastic or glass tube with an open end and larger diameter than the holding rod can be applied to host the holding rod. The open end of the tube allows layered liquids to come into can be applied to host the holding rod. The open end of the tube allows layered liquids to come into the tube for measurement. An example of the protection tube can be seen in Figure 3a. For practical the tube for measurement. An example of the protection tube can be seen in Figure 3a. For practical application, the protected sensor device will be mounted or attached on the storage tanks or application, the protected sensor device will be mounted or attached on the storage tanks or containers containers at the locations where potential layered liquids want to be monitored. Thus, if any at the locations where potential layered liquids want to be monitored. Thus, if any interfaces of layered interfaces of layered liquids occur or any changes in liquid level during the length of the sensing unit, liquids occur or any changes in liquid level during the length of the sensing unit, spectrum changed of spectrum changed of the transmitted light, which will be recorded by the optical spectrum analyzer, the transmitted light, will becomputer recordedto bythe theinterested optical spectrum plotted and showed in which the personal users. analyzer, plotted and showed in the personal computer to the interested users. Sensors 18,Equations x FOR PEER (1) REVIEW layer.2018, From and (2),

Figure 2. Schematic of the sensor device. Figure 2. Schematic of the sensor device.

Sensors 2018, 18, 3094

5 of 14

Sensors 2018, 18, x FOR PEER REVIEW

(a)

5 of 14

(b)

layered liquids. liquids. Figure 3. (a) Experimental setup; (b) Photo for the layered

3. 3. Experiment Experiment Setup Setup Laboratory Laboratory experiments experiments were were conducted conducted to to validate validate the the developed developed sensor sensor device device for for multi-parameter measurements of layered liquids on liquid interface location and level changes. multi-parameter measurements of layered liquids on liquid interface location and level changes. More More importantly, experiments performed to investigate the influences of several factors importantly, experiments werewere also also performed to investigate the influences of several keykey factors on on the sensitivity of the sensor device, including the materials and diameter of the holding rod, and the the sensitivity of the sensor device, including the materials and diameter of the holding rod, and the cladding cladding mode mode of of the the LPFG LPFG sensors sensors for for different different needs needsin inpractical practicalapplications. applications. Three this paper paper including including water, water, decane, decane, Three layered layered liquids liquids were were used used for for laboratory laboratory validation validation in in this and propylene glycol. All these three liquids are colorless and odorless. Decane is an and propylene glycol. All these three liquids are colorless and odorless. Decane is an alkane alkane hydrocarbon CH33,, which which is is one one of of the the components hydrocarbon with with the the chemical chemicalformula formulaCH CH33(CH (CH22))88CH components in in gasoline gasoline (petroleum). refractive (petroleum).Decane Decaneisisnonpolar nonpolarand andwill willnot not dissolve dissolvein in polar polar liquids liquids such such as as water. water. It It has has aa refractive index of 1.41, which is larger than water of 1.33 in refractive index [25]. Propylene glycol is a index of 1.41, which is larger than water of 1.33 in refractive index [25]. Propylene glycol is a synthetic synthetic organic with the the chemical chemical formula formula CC33HH8O , which is primarily used in the production 8O organic compound compound with 2,2which is primarily used in the production of of polymers also seesuse useininfood foodprocessing processing[26]. [26]. Propylene Propyleneglycol glycol isis miscible miscible with with water polymers butbut also sees water but but insoluble with decane. It has a refractive index of 1.43, which is larger than that of decane (1.41). insoluble with decane. It has a refractive index of 1.43, which is larger than that of decane (1.41). It is It is known based Equations and thattemperature temperaturemay mayintroduce introduce significant significant measurement known based on on Equations (1)(1) and (2)(2)that measurement variances which is not related to any liquid layer or level measurements. Thus, in this variances which is not related to any liquid layer or level measurements. Thus, in this study, study, all all the the ◦ C). laboratory tests were performed in a controlled temperature environment at room temperature (22 laboratory tests were performed in a controlled temperature environment at room temperature (22 °C). For For practical practical applications, applications, aa temperature temperature compensation compensation LPFG LPFG sensor sensor is is recommended recommended to to be be installed installed parallel to the LPFG unit on the sensor device to eliminate the temperature effects on the sensor device. parallel to the LPFG unit on the sensor device to eliminate the temperature effects on the sensor device. Figure 3a shows the experimental setup to detect the layered liquids using the developed Figure 3a shows the experimental setup to detect the layered liquids using the developed sensor sensor device. The developed sensor device was protected inside a glass tube with a larger diameter. device. The developed sensor device was protected inside a glass tube with a larger diameter. The The protected sensor device was then placed a graduated cylinder. To perform a systematic protected sensor device was then placed insideinside a graduated cylinder. To perform a systematic study study the detection the liquid interface for various insolvable liquids understand on theon detection of theofliquid layerlayer interface for various insolvable liquids and and understand the the sensitivity of the device towards various parameters which may influence the detection, ten sets sensitivity of the device towards various parameters which may influence the detection, ten sets of of laboratory experiments had been performed controlled room temperature environments laboratory experiments had been performed in in thethe controlled room temperature environments as as shown in Table 1. Among experiments, three of laboratory performed on shown in Table 1. Among the the ten ten experiments, three sets sets of laboratory teststests werewere performed on only only one liquid as surrounding environment including Experiment for water for decane one liquid as surrounding environment including Experiment #1 for#1 water only, only, #2 for#2 decane only, only, and #3 for propylene glycol only. In addition, Experiment #1 was compared with and #3 for propylene glycol only. In addition, Experiment #1 was compared with numerical numerical simulations Two more simulations to to validate validate the the effectiveness effectiveness of of the the experimental experimental setup. setup. Two more sets sets of of experiments experiments were conducted on different combinations of liquids with Experiment #4 for water and were conducted on different combinations of liquids with Experiment #4 for water and decane decane and and Experiment #1 to to #5 Experiment #5 #5 for for propylene propylene glycol glycol and and decane. decane. Experiments Experiments #1 #5 used used the the same same rod rod material material (glass) (glass) and and size size (outer (outer diameter diameter of of 33 mm), mm), and and the the same sameLPFG LPFGcladding claddingmode modeof ofLP LP0707.. In addition to the laboratory validation experiments, sensitivity tests were In addition to the laboratory validation experiments, sensitivity tests were also also performed performed on on device device including including rod rod materials materials (Experiment (Experiment #6), #6), rod rod sizes sizes (Experiment (Experiment #7), #7), and and LPFG LPFG cladding cladding modes modes (Experiment (Experiment #8 #8 and and #9). #9). Two Two materials materials were were compared compared including including glass glass (Experiment (Experiment #4) #4) and and metal metal

(Experiment #6) using the same outer diameter of 3 mm and LPFG cladding mode of LP07. Two sizes of rod including outer diameter of 3 mm (Experiment #4) and 2 mm (Experiment #7) were compared

Sensors 2018, 18, 3094

6 of 14

(Experiment #6) using the same outer diameter of 3 mm and LPFG cladding mode of LP07 . Two sizes of rod including outer diameter of 3 mm (Experiment #4) and 2 mm (Experiment #7) were compared using the glass rod and same LPFG cladding mode of LP07 . Three LPFG cladding modes were analyzed including of LP06 (Experiment #8), LP07 (Experiment #4), and of LP08 (Experiment #9) using glass rod with outer diameter of 3 mm. All the liquid was gradually added into the graduated cylinder as shown in Figure 3b. Since the resolution of the LPFG for liquid interface changes is one LPFG grating period, which follows in between 0.363 mm to 0.375 mm in this study, a gradual change of 1 mm was used in laboratory experiments to ensure the device can detect the liquid interface accurately. The LPFG sensor was connected in series to a broadband light source ranging from 1520 to 1620 nm and to an optical spectrum analyzer (OSA), followed by a personal computer. The spectrum changes of the LPFG sensor were recorded using the OSA and personal computer throughout all the experiments. Table 1. Laboratory Test Matrix. Experiment No.

Liquids Used

Rod Material, Rod Size, and LPFG Cladding Modes

#1 #2 #3 #4 #5 #6 #7 #8 #9

Water Only Decane Only Propylene Glycol Only Water + Decane Propylene glycol + Decane Water + Decane Water + Decane Water + Decane Water + Decane

Glass rod with 3 mm outer diameter, LP07 Glass rod with 3 mm outer diameter, LP07 Glass rod with 3 mm outer diameter, LP07 Glass rod with 3 mm outer diameter, LP07 Glass rod with 3 mm outer diameter, LP07 Metal rod with 3 mm outer diameter, LP07 Glass rod with 2 mm outer diameter, LP07 Glass rod with 3 mm outer diameter, LP06 Glass rod with 3 mm outer diameter, LP08

4. Multi-Parameter Sensing Experimental Results and Discussions 4.1. Liquid Level Measurements for One Liquid Only (Experiment #1 to #3) In Experiment #1 to #3, water (ns = 1.33), decane (ns = 1.41), and propylene glycol (ns = 1.43) was gradually added into the graduated cylinder, respectively as shown in Figure 3a. Figure 4a shows the example spectrum changes of the LPFG sensing unit as the water level increases. In addition to laboratory experiments, numerical simulation was also performed for the cladding mode of LP07 following the procedures in reference [18]. Figure 4b compares the experimental resonant wavelength changes of the cladding mode of LP07 to simulations for a water level from 0 to 30 mm (full length of the grating) according to Equation (1). Figure 4b shows that the experimental and simulated results on the resonant wavelength changes agree with each other very well. Overall, the water level sensitivity of the grating from the test data between 5 mm and 27 mm is −0.220 nm/mm for the cladding mode of LP07 . The simulation yielded a water level sensitivity of −0.222 nm/mm for LP07 . The difference between the experimental data and simulations is likely due to the human error of variance for uneven water level addition during the tests.

the grating) according to Equation (1). Figure 4b shows that the experimental and simulated results on the resonant wavelength changes agree with each other very well. Overall, the water level sensitivity of the grating from the test data between 5 mm and 27 mm is −0.220 nm/mm for the cladding mode of LP07. The simulation yielded a water level sensitivity of −0.222 nm/mm for LP07. The difference between the experimental data and simulations is likely due to the human error of Sensors 2018, 18, 3094 7 of 14 variance for uneven water level addition during the tests.

Resonant wavelength change (nm)

-10

Intensity (dB)

-15 -20 -25 -30 -35 1555

Water level increases

1560

1565

Wavelength (nm)

1570

0.0

-2.5

-5.0

Simulated results Experimental results

-7.5

-10.0

1575

0

5

10

15

20

Water level change (mm)

Sensors 2018, 18, x FOR PEER REVIEW

(a)

25

30

(b)

7 of 14

Figure 4. 4. (a)(a)Transmission wavelengthchanges changesofofLP LP water level changes. Figure Transmissionspectrum; spectrum;(b) (b) Resonant Resonant wavelength 07 07 to to water level changes.

Figure 5 shows thethe comparison resonant Figure 5 shows comparison resonantwavelength wavelengthchanges changesofofthe thecladding claddingmode modeLP LP0707 to to the the change of levels in water, decane, and propylene glycol for Experiments #1–3. The slope changes of the change of levels in water, decane, and propylene glycol for Experiments #1–3. The slope changes of LPFG wavelength changeschanges can distinguish differentdifferent types of types liquidsofand the gradual the resonant LPFG resonant wavelength can distinguish liquids and theresonant gradual wavelength changes canchanges providecan information on liquid level changes. resonant wavelength provide information on liquid level changes.

Resonant Wavelength Change (nm)

2 0 -2 -4 -6 -8 -10

Water:n=1.33 Decane:n=1.41 PG:n=1.43

-12 -14 -16 -18 0

5

10

15

20

25

30

Liquid Level (mm)

Figure 5. 5. Resonant wavelength changes ofof LPLP Figure Resonant wavelength changes tothe thechange changeof oflevels levelsin inwater, water,decane, decane,and and propylene propylene 0707to glycol (PG). glycol (PG).

4.2.4.2. Layered Liquid Interface and Level Layered Liquid Interface and LevelDetection Detection(Experiment (Experiment#4#4toto#5) #5) Before Experiment #4#4 toto #5,#5,simulations Before Experiment simulationsand andexperiments experimentswere wereperformed performedon onthe theLPFG LPFGcladding cladding mode of LP with a surrounding liquid of refractive index changing from 1.33 to 1.44 by an an interval of 07 07 with a surrounding liquid of refractive index changing from 1.33 to 1.44 by mode of LP interval 0.005. The simulation was performed following the procedure in reference [27]. Figure 6a shows the of 0.005. The simulation was performed following the procedure in reference [27]. Figure 6a shows corresponding simulated resonant wavelength changes to the change ofofthe the corresponding simulated resonant wavelength changes to the change thesurrounding surroundingliquids liquids referred to water (n = 1.33) as the base liquid, according to Equations (1) and The between relation referred to water (n = 1.33) as the base liquid, according to Equations (1) and (2). The(2). relation between the resonant wavelength changes and theindex refractive index changes of thelayered surrounding the resonant wavelength changes and the refractive changes of the surrounding liquids layered liquidsparticularly is nonlinear,towards particularly a large value. Inexperiments, the laboratory is nonlinear, a largetowards index value. In index the laboratory theexperiments, three liquids theused three liquidsthe used include used in and #5, which are water s= include liquids usedthe in liquids Experiments #4 Experiments and #5, which#4are water (ns = 1.33), decane(n(n s =1.33), 1.41) decane (ns = 1.41) glycol and propylene The unfilled circles in Figure 6a and propylene (PG) (ns glycol = 1.43).(PG) The(nunfilled in Figure 6a represent therepresent resonant s = 1.43).circles thewavelength resonant wavelength changes measured the LPFG devicemodes for theLP07 cladding modes changes measured from the LPFGfrom sensor device forsensor the cladding and Figure 6b LP07 and Figure 6b shows the recorded spectrums for the interface changes between three different shows the recorded spectrums for the interface changes between three different layered liquids. It layered It Figure can be 6a seen Figure 6a results that the results arewith in good agreement can beliquids. seen from thatfrom the simulated aresimulated in good agreement the experimental with theThe experimental data. Thethat simulation predicted that change −7.500 nmliquid will occur the data. simulation predicted a change of −7.500 nm willa occur theofsurrounding changing from waterliquid to Decane, and the experiment showed and to bethe −6.700 nm. Theshowed slight difference between surrounding changing from water to Decane, experiment to be −6.700 nm. them is likely attributable to the potential variation of concentration in the tested liquids.

Sensors 2018, 18, 3094

8 of 14

0 0

Water Water

Decane Decane

-5

-5

-10 -10

Propylene glycol Propylene glycol

-15 -15 -20 -20

Simulated results for LP07 Simulated resultsresults for LP07 Experimental for LP07 Experimental results for LP07

-25 0.00 0.02 0.04 0.06 0.08 0.10 0.12 -25 0.00 0.02 0.04 0.06 0.08 0.10 0.12

Refractive Index (RI) Changes of Layered Liquids Refractive Index (RI) Changes (a) of Layered Liquids

-5 -5 -10 -10 -15 -15 -20 -20 -25 water -25 water ns=1.33 Decane -30 ns=1.33 Decane ns=1.41 propylene glycol -30 ns=1.41 propylene glycol ns=1.43 -35 ns=1.43 -35 1540 1545 1550 1555 1560 1565 1570 1575 1540 1545 1550 1555 1560 1565 1570 1575 Wavelength (nm)

Intensity (dB) Intensity (dB)

Resonant Wavelength Change (nm) Resonant Wavelength Change (nm)

The slight difference between them is likely attributable to the potential variation of concentration in Sensors 2018, 18, x FOR PEER REVIEW 8 of 14 theSensors tested liquids. 2018, 18, x FOR PEER REVIEW 8 of 14

Wavelength (nm) (b)

(a)

(b)

Figure 6. (a) Experimental and simulated resonant wavelength changes of LP07 in responding to Figure 6. 6. (a)(a) Experimental and simulated resonant changes ofofLP Figure Experimental and simulated resonant wavelength changeschanges. LP0707 in in responding responding to to changes in liquid interfaces; (b) Spectrum changeswavelength with liquid interface changes in in liquid interfaces; (b)(b) Spectrum changes changes liquid interfaces; Spectrum changeswith withliquid liquidinterface interfacechanges. changes.

Resonant Wavelength Change (nm) Resonant Wavelength Change (nm)

With the expectation of sudden resonant wavelength changes when the liquid boundary of With thethe expectation of sudden resonant wavelength changes whenwhen the liquid boundary of layers With expectation of sudden wavelength the boundary of layers presented during the length ofresonant LPFG, Experiment #4changes was conducted to liquid validate the detection presented during the length of LPFG, Experiment #4 was conducted to validate the detection of layers presentedbetween during the lengthand of LPFG, #4 was the conducted to validate the detection of boundary decane water.Experiment Figure 7a shows recorded transmission spectrum boundary between decanedecane and water. Figure 7a shows7a theshows recorded transmission spectrum changes of of changes boundary between and water. Figure the recorded transmission spectrum of Experiment #4 with boundary changes between water and decane for the cladding mode Experiment #4 with boundary changes between water and decane for the cladding mode LP and changes of Experiment #4 with boundary changes between water and decane for the cladding mode 07 LP07 and Figure 7b shows its resonant wavelength changes with adding water and decane. The Figure its resonant wavelength changes with adding water and decane. The interface in LPinterface 07 7b andshows Figure 7b shows its resonant wavelength changes with adding water and decane. The in between the two different layered liquids is clearly identified at the location of 12.5 mm between layered is clearly at identified thecenter location of 12.5 mm away interface intwo between two liquids is clearly at the location of 12.5 mmnm awaythe from thedifferent startthe point ofdifferent theliquids LPFGlayered sensing unitidentified with a sudden wavelength change offrom 6.6 from the start point of the LPFG sensing unit with a sudden center wavelength change of 6.6 nmthe theaway start point of the LPFG sensing unit with a sudden center wavelength change of 6.6 nm on on the boundary between the water and Decane. When compared to Figure 6a and Figure the 5, on the boundary between the water and Decane. When compared to Figure 6a and Figure 5, thethe boundary between the water and Decane. When to Figures 6aBefore and 5, the showed experiments showed great consistence with compared multiple experiments. the experiments decane was added, experiments showed great consistence with multiple experiments. Before thethe decane was added, the great consistence with of multiple experiments. Before the was added, wavelength center wavelength LPFG shifts to left with a sloop of decane −0.220 nm/mm. After thecenter liquid layer interface wavelength of LPFG shifts to left with a sloop of −0.220 nm/mm. After the liquid layer interface of center LPFG shifts to left with a sloop of − 0.220 nm/mm. After the liquid layer interface presents, presents, the resonant wavelength of the LPFG continued shifting to left as the level of decane the wavelength of the LPFG shifting to of left as theincreases level thepresents, resonant wavelength the LPFG continued shifting to leftshow as the level decane atdecane a ratecan increases at resonant a rate of of −0.504 nm/mm. Figure 7a,bcontinued clearly that the developed sensorofdevice at a rate of −0.504 nm/mm. Figure 7a,b clearly show that the developed sensor device can of increases −detect 0.504 nm/mm. Figure 7a,b clearly show that the developed sensor device can detect the interfaces the interfaces between layered liquids and at the same time the level changes of the associated detect the interfaces between layered liquids and at the same time the level changes of the associated between layered liquids and at the same time the level changes of the associated layered liquids. layered liquids. layered liquids. 2 0

-2 -4 -6 -8

-10 -12

2 0 -2 -4 -6

water: n=1.3 water: n=1.3

-8 -10

Decane: n=1.41 Decane: n=1.41

-12 -14

-14

-16 0 3 6 9 12 15 18 21 24 27 30 -16 0 3 6 9 12 15 18 21 24 27 30

(a) (a)

Liquid Level Change (mm) Liquid Level Change (mm)

(b) (b)

Figure (a) Transmission spectrum; (b) Resonant wavelength changes vs simultaneous liquid Figure 7. (a)7.Transmission spectrum; (b) Resonant wavelength changes of LPof07LP vs.07 simultaneous liquid Figure 7. (a) Transmission spectrum; (b) Resonant wavelength changes of LP 07 vs simultaneous liquid interface and level changes (water decane). interface and level changes (water and and decane). interface and level changes (water and decane).

Figure 8a shows the recorded transmission spectrum changes of Experiment #5 #5 with boundary Figure 8a shows the recorded transmission spectrum changes of Experiment with boundary Figure 8a shows the recorded transmission spectrum changes of Experiment #5 with boundary changes between PG and decane for the cladding mode LP07LPand Figure 8b 8b shows thethe resonant changes between PG and decane for the cladding mode 07 and Figure shows resonant changes between PG and decane for the mode LP07 water and Figure 8b by shows the and resonant wavelength changes for Experiments #4 cladding and #5 with adding followed decane adding wavelength changes for Experiments #4 and #5 with adding water followed by decane and adding propylene followed by decane. Figure 8b shows that a bigger difference in refractive index as water propylene followed by decane. Figure 8b shows that a bigger difference in refractive index as water

Sensors 2018, 18, 3094

9 of 14

Sensors 2018, 18, x FOR PEER REVIEW

9 of 14

wavelength changes for Experiments #4 and #5 with adding water followed by decane and adding propylene by decane. Figure 8b shows that a abigger difference refractive index as water (n = 1.33)followed and decane (n = 1.41) would introduce bigger resonantinwavelength change when (ncompared = 1.33) and = 1.41) would introduce a bigger resonant wavelength change when compared to decane smaller(ndifference in refractive index as PG (n = 1.43) and decane (n = 1.41). to smaller difference in refractive index as PG (n = 1.43) and decane (n = 1.41). 0

Intensity (dBm)

-5 -10 -15 -20 -25 1520

Propylene glycol: n=1.43

Decane: n=1.41 1528

1536

Wavelength (nm) (a)

1544

Center Wavelngth Changes (nm)

4

Water: n=1.33

2 0

Water & Decane Propylene & Decane

-2 -4 -6

Propylene glycol: n=1.43

-8 -10 -12

Decane: n=1.41

-14 -16 -18

0

5

10

15

20

25

30

35

Liquid Level (mm) (b)

Figure Figure8.8.(a) (a)Spectrum Spectrumchanges changesvs. vssimultaneous simultaneous liquid liquidinterface interfaceand andlevel levelchanges changes(Experiment (Experiment#5); #5); (b) (b)Resonant Resonantwavelength wavelengthchanges changesofofLP07 LP07vs.vs simultaneous simultaneous liquid liquid interface interface and and level level changes changes (Experiments (Experiments#4#4and and#5). #5).

5.5.Device DeviceSensitivity SensitivityStudy Study 5.1. Influence from Rod Materials (Experiment #4 and #6) 5.1. Influence from Rod Materials (Experiment #4 and #6) To investigate the influence of different holding rod materials on the sensitivity of the developed To investigate the influence of different holding rod materials on the sensitivity of the developed sensor device, two different rod materials were tested in lab, including metal and glass material sensor device, two different rod materials were tested in lab, including metal and glass material as as shown in Figure 9a,b for the photos of the two tested rods. Both rods had same outer diameter shown in Figure 9a,b for the photos of the two tested rods. Both rods had same outer diameter of 3 mm. of 3 mm. Figure 9c shows the experimental results for measuring level changes and interface of Figure 9c shows the experimental results for measuring level changes and interface of layered liquids, layered liquids, water and decane. The liquid level sensitivity of the LPFG sensing unit did not water and decane. The liquid level sensitivity of the LPFG sensing unit did not change significantly change significantly with the use of the two different materials of the holding rod. On the glass with the use of the two different materials of the holding rod. On the glass rod, the sensor had a level rod, the sensor had a level sensitivity for water of −0.222 nm/mm and −0.505 nm/mm for decane. sensitivity for water of −0.222 nm/mm and −0.505 nm/mm for decane. On metal rod, the sensor had On metal rod, the sensor had a level sensitivity of −0.205 nm/mm for water and −0.504 nm/mm a level sensitivity of −0.205 nm/mm for water and −0.504 nm/mm for decane, respectively. The for decane, respectively. The materials did affect a little on the actual value change of the interface materials did affect a little on the actual value change of the interface detection. Glass rod showed detection. Glass rod showed 6.66 nm changes between the interface of water and decane, and metal 6.66 nm changes between the interface of water and decane, and metal rod had 6.76 nm changes. The rod had 6.76 nm changes. The 0.1 nm difference between the two different materials of the holding 0.1 nm difference between the two different materials of the holding rods can be attributed to the rods can be attributed to the variance of interface locations when performing experiments two different variance of interface locations when performing experiments two different times. Thus, these times. Thus, these experiments provide confidence in ignoring the influence from materials of the experiments provide confidence in ignoring the influence from materials of the holding rods. For holding rods. For practical applications, best available and most cost-effective holding tube material practical applications, best available and most cost-effective holding tube material can be used. can be used. However, laboratory calibration is recommended before any field applications of the However, laboratory calibration is recommended before any field applications of the sensor device. sensor device.

Sensors 2018, 18, x FOR PEER REVIEW

10 of 14

Sensors Sensors2018, 2018,18, 18,3094 x FOR PEER REVIEW

10of of14 14 10

(a) (a)

Resonant Wavelength Change (nm) (nm) Resonant Wavelength Change

2 20

Glass Metal Glass Metal

Water

0-2 -2-4

Water

-4-6 -6-8

Decane

-10 -8

Decane

-10 -12 -12 0 0

5

10

15

20

25

30

35

Liquid Level Change (mm)

5

10

(b)

15

20

25

30

35

40 40

(c)Change (mm) Liquid Level

(b) Figure 9. (a) Metal holding rod of 3 mm; (b) Glass holding rod of 3 mm; (c) (c) Resonant wavelength changes of the LP 07 vs layered liquid changes for two holding rods with different materials: metal and Figure 9. (a) Metal holding rod of 3 mm; (b) Glass holding rod of 3 mm; (c) wavelength Figure 9. (a) Metal holding rod of 3 mm; holding rod of 3 mm; (c) Resonant Resonant wavelength glass in layered liquids of water and decane. changesofofthe theLP LP0707 vs. vs layered changes layeredliquid liquidchanges changesfor fortwo twoholding holdingrods rodswith withdifferent differentmaterials: materials:metal metaland and glassininlayered layeredliquids liquidsof ofwater water and and decane. decane. glass

(a) (a)

Resonant wavelength changes (nm) Resonant wavelength changes (nm)

5.2. Influence from Rod Diameters (Experiment #4 and #7) 5.2. 5.2.Influence Influencefrom fromRod RodDiameters Diameters (Experiment (Experiment #4 #4 and and #7) #7) The influence of rod diameter on the sensitivity of the developed sensor device was also tested The of on sensitivity of two the developed developed sensordevice device was alsoholding tested in laboratory by comparing the sensing different diameters of the glass Theinfluence influence ofrod roddiameter diameter on the the responses sensitivity of of the sensor was also tested in laboratory by comparing the sensing responses of two different diameters of the glass holding rods. Figure 10a,b show the photos of theresponses two testedofglass with diameters different diameters of 2holding mm and in laboratory by comparing the sensing two rods different of the glass rods. 10a,b showthe thephotos photosofofthe the two tested glass rods with different oflayered 2and mm Figure show two tested rods with different diameters of 2 mm 3rods. mm.Figure Figure10a,b 10c shows the experimental results for glass measuring level changes anddiameters interface of and 3 mm. Figure 10c shows the experimental results for measuring level changes and interface of 3 mm. Figure 10c shows the experimental results for measuring changes and interface of layered liquids for water and decane. For the two different diameterslevel of the holding rod, the sensor device layered for water and decane. For two different diameters of the holding rod, the sensor liquids liquids for well water and decane. For the twothe different diameters of the holding rod, the sensor performed for liquid interface measurements with similar center wavelength changes ofdevice 5.68 nm device performed well for liquid interface measurements with similar center wavelength changes of performed well for liquid interface measurements with similar center wavelength changes of 5.68 nm and 5.82 nm, respectively. However, Figure 10c clearly shows that the diameter of the holding rod 5.68 nm and 5.82 nm, respectively. However, Figure 10c clearly shows that the diameter of the holding and 5.82 nm, However, Figure 10c clearly shows that theThe diameter the holding rod influences therespectively. accuracy of the liquid level measurement significantly. sensorof device with smaller rod influences the holding accuracy ofliquid the2 liquid leveldiameter measurement The sensor device with influences accuracy of the levelouter measurement significantly. The sensor device with smaller diameter inthe glass rod of mm failed significantly. to measure the level changes in water smaller diameter glass ofouter 2 mm outer diameter to measure level changes diameter in glassinholding rod of 2rod mm diameter failed tofailed measure the changes ineffect. waterinA although it behaved wellholding in decane. The phenomenon can be explained bylevel thethe capillarity water although it behaved well in decane. The phenomenon can be explained by the capillarity effect. althoughdiameter it behaved in decane. The phenomenon can becapillarity explainedeffect, by thewhich capillarity effect. Aof smaller of awell holding rod may introduce significant the thickness diameter ofofcannot a aholding rod introduce capillarity effect, which the ofof Aasmaller smaller diameter holding rodmay may introducesignificant significant capillarity effect, which thethickness thickness small epoxy layer overcome. Therefore, for practical applications, the diameter of a holding asmall small epoxy layercannot cannot overcome. diameter of arod layer Therefore, practical applications, the diameter ofaaholding holding of epoxy the sensor deviceovercome. is recommended tofor bepractical at leastapplications, larger thanthe 3 mm according to our rod of the sensor device is recommended to be at least larger than 3 mm according to ourto rod of the sensor device is recommended to becalibration at least larger than 3 mm according to ourapplications experimental experimental data. In addition, laboratory is required before any field experimental data. In addition, laboratory calibration isdevice required beforeitsany applications to it data. In an addition, laboratory calibration is required before any field applications toofensure an effective ensure effective size of the holding rod. If the sensor changes sizefield the hosting rod, ensure an effective size of the holding rod. If the sensor device changes its size of the hosting rod, itto size of the holding rod. If the sensor device changes its size of the hosting rod, it cannot be assumed cannot be assumed to have the same sensor sensitivity without laboratory calibration. cannot be assumed to have the same sensor sensitivity without laboratory calibration. have the same sensor sensitivity without laboratory calibration. 2 2

0 0 -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 0

(b) (b)

Larger Diameter Larger Diameter Smaller Diameter Smaller Diameter

Water Water

Decane Decane

0

5

5

10 15 20 25 30 35 40 10Liquid 15 Level 20 Changes 25 30(mm)35 40 Liquid Level Changes (mm)

(c) (c)

Figure 10. 10. (a) rod with outer diameter of 2 mm; Glass rod withrod outer diameter Figure (a)Glass Glassholding holding rod with outer diameter of 2 (b) mm; (b) holding Glass holding with outer Figure 10. (a) Glass holding rod with outer diameter of 2 mm; (b) Glass holding rod with outer diameter diameter mm; (c) Resonant wavelength LP07 liquid vs. layered liquid changes for two vsthe layered changes for two holding rods of 3 mm; of (c)3Resonant wavelength changes ofchanges the LP07of of 3 mm; (c) Resonant wavelength changes of the LP07 vs layered liquid changes for two holding rods holding rods with two different sizes in layered liquids water and decane. with two different sizes in layered liquids of water andofdecane. with two different sizes in layered liquids of water and decane.

Sensors 2018, 18, 3094

11 of 14

5.3. Influence from LPFG Cladding Modes (Experiment #4, #8 and #9) In Section 2.1, Equation (2) predicts that a higher cladding mode of LPFG sensor can result in a larger Γs , leading to a higher sensitivity to measure the surrounding medium changes such as liquid interface or level changes. From previous study [18], although the resonant wavelength changes with liquid nonlinearly, linear approximation can be used to estimate the liquid level changes. More details about the nonlinearity of the liquid level to the resonant wavelength change refer to [18]. In addition, if the cladding mode is too high, it may show a turning point in spectrum, which is undesired for this application [28]. To investigate the influence of LPFG cladding modes on sensitivity of the developed sensor device, two other LPFG cladding modes, LP06 and LP08 , in addition to the LP07 as shown in the previous sections from Figures 6–8, have been investigated here. All the three cladding modes (LP06 , LP07 , and LP08 ) are visible in a 400 nm wavelength range. For these three cladding modes, no dual resonance is expected to present with the surrounding refractive index changing from 1.0 to 1.46 [28], which is the common range for liquids. Figure 11a shows the transmission spectrum changes of LP06 with adding water followed by decane and Figure 11b shows its resonant wavelength changes. A similar phenomenon as LP07 (shown in Figure 7) was observed for LP06 ; however, lower sensitivity is observed for LP06 for both liquid interface and level measurements. For water, the resonant wavelength of the LP06 was gradually shifting to the left with a sloop of −0.041 nm/mm. For decane, the level sensitivity LP06 was −0.059 nm/mm. In between water and decane, a sudden resonant wavelength shift occurred with a change of 2.5 nm. The boundary between water and decane was clearly detected by the LP06 cladding mode also at 12.5 mm away from the beginning of the LPFG starting point. Comparing Figure 7b to Figure 11b, it can be clearly seen that the sensor sensitivity of the cladding mode LP06 was five times smaller than the cladding mode of LP07 for both liquid interface and level measurements. For the cladding mode LP08 , the spectrum changes with the liquid level and interface changes could be found in Figure 11c and the corresponding resonant wavelength changes could be found in Figure 11d. From the experimental results, it can be seen that a big change of resonant wavelength occurred for LP08 when decane was added into water. For water, the LP08 had a level sensitivity of −0.643 nm/mm and for decane, it has a level sensitivity of −1.154 nm/mm. The center wavelength shift for LP08 at the interface in between water and the decane is 16.2 nm. Comparing Figure 7b to Figure 11d, the cladding mode of LP08 is much more sensitive with at least two times higher than that of the cladding mode LP07 . No peak split occurred along the experiments. The interface of the water and decane was also located to be at 12.5 mm away from the LPFG starting point. The comparison between Figures 6 and 9 clearly indicates that a higher cladding mode of a LPFG sensing unit will have a higher sensitivity for detecting the layered liquids. The LP08 is two times more sensitive than the LP07 , and the LP07 is five times more sensitive than the LP06 . For practical applications, it is recommended to use the highest cladding modes with no peak split in the range of liquids to be measured to achieve the highest sensitivity of measurements.

Sensors Sensors2018, 2018,18, 18,3094 x FOR PEER REVIEW

1212ofof1414

Resonant Wavelength Changes (nm)

-2.5

Intensity (dB)

-5.0 -7.5 -10.0 -12.5 -15.0 -17.5 1306

Water: n=1.3

Decane: n=1.41 1308

1310

1312

1314

0.5 0.0 Water

-0.5 -1.0 -1.5 -2.0 -2.5

Decane

-3.0 -3.5 0

1316

5

Wavelength (nm)

10 15 20 25 Liquid Level Changes (mm)

(a)

30

35

(b)

Intensity (dBm)

-8 -16 -24 -32 Decane: n=1.41 -40 1500

1510

1520

Water: n=1.33 Air: n=1.0 1530

1540

Wavelength (nm)

(c)

1550

1560

Resonant Wavelength Change (nm)

5 0

0 -5 -10

Water

-15 -20 -25 Decane

-30 -35 -40 0

5

10 15 20 25 Liquid Level Change (mm)

30

35

(d)

Figure11. 11.(a) (a)Spectrum Spectrumchanges changesofofLP LP (b)Resonant Resonantwavelength wavelengthchanges changesofofthe theLP LP (c)Spectrum Spectrum Figure 0606; ;(b) 0606; ;(c) changes (d)Resonant Resonantwavelength wavelengthchanges changesofofthe theLP LP0808with withlayered layeredliquid liquidinterface interfaceand andlevel level changesofofLP LP0808;;(d) changes changes(water (waterand anddecane). decane).

6.6.Conclusions Conclusionsand andFuture FutureWork Work In Inthis thispaper, paper,aasensor sensordevice devicebased basedon onLPFG LPFGwas wasdeveloped developedtotodetect detectsimultaneous simultaneouschanges changesinin the and level of insoluble layered liquids. Based Based on the findings from thisfrom paper,this the paper, following theinterface interface and level of insoluble layered liquids. on the findings the conclusions can be drawn: following conclusions can be drawn: (a) (a) Operational Operational principles principles of of the the sensor sensor device device show show that thataaconstant constantslope slopewill willbe bepresent presentfor forlevel level changes change in the the level measurements is expectedisfor the occurrence changesand anda sudden a sudden change inslope the of slope of the level measurements expected for the of boundaries layered insoluble liquids. occurrence of in boundaries in layered insoluble liquids. (b) experiments clearly clearlyvalidated validatedthat thatthe thedeveloped developed sensor device could detec (b) The The laboratory experiments sensor device could detec the the interface changes of layered liquid example of water, decane, interface and and levellevel changes of layered liquid usingusing the the example of water, decane, and and propylene glycol. propylene glycol. (c) thethe materials of of thethe holding rods of the sensor on the (c) Experimental Experimentalstudy studyalso alsoshowed showedthat that materials holding rods of LPFG the LPFG sensor on sensor device had had veryvery littlelittle influence on the sensitivity; however, the diameter of theof the sensor device influence onsensor’s the sensor’s sensitivity; however, the diameter holding rod did the accuracy of the which will require a calibration before the the holding rod influence did influence the accuracy ofdevice the device which will require a calibration before sensor’s practical application. the sensor’s practical application. (d) The The selection selection of cladding significantly thethe sensitivity of the (d) cladding modes modesof ofthe theLPFG LPFGwill willalso alsoaffect affect significantly sensitivity of sensor device. A cladding mode of ofLPLP 08 08 was more sensitive sensitive when when the sensor device. A cladding mode wastested testedtotobe be two two times more comparedto toLP LP07 07, and more than ten times times more more sensitive sensitive than than LP LP06 06. compared

Inthe thefuture, future,more more investigation will performed to identify the liquid chemical components In investigation will be be performed to identify the liquid chemical components and and apply the developed sensorfordevice for field applications. With theofadvantages of high apply the developed sensor device field applications. With the advantages high measurement measurement accuracy and great developed optical fiber sensing couldfor be accuracy and great durability, the durability, developedthe optical fiber sensing device could device be applied applied for layered liquid detection in various applications such as fuel storage systems, chemical layered liquid detection in various applications such as fuel storage systems, chemical processing, processing, and aerospace engineering. However, since the sensitive range of the developed device

Sensors 2018, 18, 3094

13 of 14

and aerospace engineering. However, since the sensitive range of the developed device is limited by the length of the LPFG unit, if the boundary of the layered liquids is not located within the length of the LPFG sensor unit, it may not be able to detect the boundary. For practical applications which require a larger detection range, multiple LPFG units as a series or long-gauged LPFG sensor unit is suggested to obtain a potential larger measurement range, which will be studied in future. Author Contributions: Conceptualization and Methodology, Y.H. and H.X.; Experimental Validation, Investigation, and Data Curation, Z.P. and Y.H.; Writing-Original Draft Preparation, Z.P. and Y.H., Writing-Review and Editing, Y.H.; and Supervision, H.X. Funding: This research was partially funded by the U.S. DOT PHMSA under Agreement DTPH56-15-H-CAAP06. The findings and opinions expressed in the paper are those of the authors only and do not necessarily reflect the views of the sponsors. 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.

16.

Beutel, T.; Ferreira, N.; Leester-Schädel, M.; Büttgenbach, S. Robust pressure sensor for measurements in boundary layers of liquid fluids with medium total pressures. In Proceedings of the International Society for Optical Engineering, Prague, Czech Republic, 5 May 2011. Infineon Technologies AG. Liquid Level Sensing Measuring Liquid Levels Using Hall Effect Sensors. Application Note, Rev. 1.0. 2009. Available online: https://www.infineon.com/dgdl/AppNote_Liquid_ Level_Sensing_Rev.1.0.pdf?fileId=db3a30432313ff5e0123a385f3b2262d (accessed on 12 February 2009). Makanjuola, N.T.; Shoewu, O.O.; Akinyemi, L.A.; Ajasa, A.A. Design and development of microcontroller based liquid level detector with graphical output. Pac. J. Sci. Technol. 2015, 16, 173–181. Kumar, B.; Rajita, G.; Mandal, N. A review on capacitive-type sensor for measurement of height of liquid level. Meas. Control 2014, 47, 219–226. [CrossRef] Meribout, M.; Habli, M.; Al-Naamany, A.; Al-Busaidi, K. A new ultrasonic-based device for accurate measurement of oil, emulsion, and water levels in oil tanks. In Proceedings of the 21st IEEE Instrumentation and Measurement Technology Conference, Como, Italy, 18–20 May 2004; pp. 1942–1947. Wang, A.; Gunther, M.F.; Murphy, K.A.; Claus, R.O. Fiber-optic liquid-level sensor. Sens. Actuators A 1992, 35, 161–164. [CrossRef] Guo, T.; Zhao, Q.; Duo, Q.; Zhang, H.; Xue, L.; Huang, G.; Dong, X. Temperature-insensitive fiber bragg grating liquid-level sensor based on bending cantilever beam. IEEE Photonics Technol. Lett. 2005, 17, 2400–2402. Wang, D.; Cao, M.; Li, C.; Li, D.; Chen, Y.; Xu, X.; Xu, J.; Li, Y.; Wan, Z.; Wang, B. Fiber bragg grating liquid level sensor with double pressure and temperature sensitivities. Procedia Eng. 2011, 15, 704–709. [CrossRef] Bhatia, V. Optical fiber long period grating sensors. Opt. Lett. 1996, 21, 692–694. [CrossRef] [PubMed] Ye, X.W.; Su, Y.H.; Han, J.P. Structural health monitoring of civil infrastructure using optical fiber sensing technology: A comprehensive review. Sci. World J. 2014, 652329. [CrossRef] [PubMed] Zheng, S.; Zhu, Y.; Krishnaswamy, S. Fiber humidity sensors with high sensitivity and selectivity based on interior nanofilm-coated photonic crystal fiber long-period gratings. Sens. Actuators B 2013, 176, 264–274. [CrossRef] Tian, F.; He, Z.; Du, H. Numerical and experimental investigation of long-period gratings in photonic crystal fiber for refractive index sensing of gas media. Opt. Lett. 2012, 37, 380–382. [CrossRef] [PubMed] Khaliq, S.; James, S.W.; Tatam, R.P. Fiber-optic liquid-level sensor using a long-period grating. Opt. Lett. 2001, 2, 1224–1226. [CrossRef] Grice, S.; Zhang, W.; Sugden, K.; Bennion, I. Liquid level sensor utilizing a long period fiber grating. In Proceedings of the International Society for Optical Engineering, San Diego, CA, USA, 3–4 August 2009. Fu, H.; Shu, X.; Zhang, A.; Liu, W.; Lin, Z.; He, S.; Bennion, I. Implementation and characterization of liquid-level sensor based on a long-period fiber grating Mach–Zehnder interferometer. IEEE Sens. J. 2011, 11, 2878–2882. [CrossRef] Xue, H.; Xu, Z.; Chen, H.; Yang, Y.; You, J.; Yan, J.; Fu, H.; Zhang, D. Continuous liquid level sensor based on a reflective long period fiber grating interferometer. Meas. Sci. Technol. 2015, 26, 037001. [CrossRef]

Sensors 2018, 18, 3094

17. 18. 19.

20. 21.

22.

23.

24. 25. 26. 27. 28.

14 of 14

Wang, J.; Luo, C. Long-period fiber grating sensors for the measurement of liquid level and fluid-flow velocity. Sensors 2012, 12, 4578–4593. [CrossRef] [PubMed] Huang, Y.; Chen, B.; Chen, G.; Xiao, H.; Khan, S.U. Simultaneous detection of liquid level and refractive index with a long-period fiber grating based sensor device. Meas. Sci. Technol. 2013, 24, 095303. [CrossRef] Yu, Y.; Hung, H.; Liaw, S.; Shih, M.; Kishikawa, H.; Goto, N. Simultaneously two-parameter measurement using tilted fiber grating and long period fiber grating. Microw. Opt. Technol. Lett. 2017, 59, 1122–1125. [CrossRef] Jiang, M.; Zhao, Z.; Li, K.; Yang, F. Long-period fiber grating cascaded to thin-core fiber for simultaneous measurement of liquid refractive-index and temperature. Sens. Rev. 2018, 38, 79–83. [CrossRef] First-Sensor.com. Measuring Transition Points in Liquids with Different Densities with CLC and CLW Capacitive Level Sensors. Application Note, E-11160. Available online: https://www.yumpu.com/ en/document/view/43889645/measuring-transition-points-in-liquids-with-different-sensortechnics (accessed on 11 November 2015). Huang, Y.; Xiao, H. Detecting layered liquids using long period fiber grating sensors. In Proceedings of the 2nd International Electronic Conference on Sensors and Applications, Basel, Switzerland, 15–30 November 2015; pp. 1–8. Li, Y.J.; Wei, T.; Montoya, J.A.; Saini, S.V.; Lan, X.W.; Tang, X.L.; Dong, J.H.; Xiao, H. Measurement of CO2 laser irradiation induced refractive index modulation in single mode fiber toward long-period fiber grating design and fabrication. Appl. Opt. 2008, 47, 5296–5304. [CrossRef] [PubMed] Shu, X.W.; Zhang, L.; Bennion, I. Sensitivity characteristics of long period fiber gratings. J. Lightw. Technol. 2002, 20, 255–266. Yaws, C.L. Chemical Properties Handbook; McGraw-Hill: New York, NY, USA, 1999; pp. 159–179. O’Neil, M.J. The Merck Index: An Encyclopedia of Chemicals, Drugs, and Biologicals; Merck Research Laboratories: San Francisco, CA, USA, 2006. Ivanov, O.V.; Nikitov, S.A.; Gulyaev, Y.V. Cladding modes of optical fibers: Properties and applications. Physics 2006, 49, 167–191. Patrick, H.J.; Kersey, A.D.; Bucholtz, F. Analysis of the response of long period fiber gratings to external index of refraction. J. Lightwave Technol. 1998, 16, 1606. [CrossRef] © 2018 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/).