sensors Communication

An Optical-Fiber-Based Airborne Particle Sensor Yifan Wang and John F. Muth * Department of Electrical and Computer Engineering, North Carolina State University, Raleigh, NC 27695, USA; [email protected] * Correspondence: [email protected]; Tel.: +1-919-513-2982 Received: 1 July 2017; Accepted: 6 September 2017; Published: 14 September 2017

Abstract: A new optical-fiber-based airborne particle counter is reported. Unlike traditional light-scatter-based techniques, the particle is detected through the drop in optical fiber coupling efficiency as the particle disrupts the electromagnetic mode of the optical beam. The system is simple, substantially smaller than traditional systems, and does not require high power laser input. This makes it attractive for wearable air quality monitors where size is a premium. There is close agreement between theoretical model and experimental results for solid and liquid particles in the 1 to 10 µm range. Keywords: particle sensor; fiber optic sensor; aerosol detection; scattering measurements

1. Introduction Airborne particulate matter (PM) is a complex mixture of extremely small particles and liquid droplets that is usually generated by different forms of combustion or chemical processes, or through mechanical wear [1–3]. The sizes of particulates range widely from submicron to ~100 µm, with >10 µm particles settling rapidly on the scale of minutes in still air, while smaller particles remain suspended for longer periods of time. Particulate exposure can be harmful to human health [1]. Several health problems, such as aggravated asthma, irregular heartbeats, and lung cancer have been linked to exposure to particulate matters [4]. Particulates can also act as abrasives or contaminates, and are detrimental to the cleanness and performance of machinery. Generally, regulatory agencies have divided airborne particulates into 4 classifications. PM10—particles less than 10 µm in diameter, PM2.5—particles less than 2.5 µm in diameter, PM1.0—particles less than 1 µm in diameter, and PM0.1—100 nm or less in diameter, and known as ultra-fine particles or nanoparticles. Particles with sizes between PM10 and PM2.5 are sometimes referred to as the coarse fraction and PM2.5 particles and smaller are called the fine fraction. Traditionally, an optical particle counter (OPC) relies on a focused light beam and ambient/forward particle scattering [5,6]. Using a long optical path, the scattered light can be separated from the main optical beam, detected and counted. Since the energy of the scattered light is very small compared to the original light field, this kind of OPC usually needs a high power laser input, a set of optics to focus the light field, and sensitive light detecting equipment to capture the signal. This makes traditional OPCs relatively large and expensive. Recently, researchers have been interested in lower-cost and compact devices for atmospheric measurements and personal particle monitoring. Our goal was to investigate methods where long optical path lengths and high laser powers are not required. This would reduce the size and cost of the detection system. In 2013, R.S. Gao et al. reported a high-sensitivity optical particle counter using a single drilled lens [7]. The particle sensing was based upon detecting the forward laser scattering signal instead of perpendicular scattering signal, which resulted in a higher signal amplitude. The lowest detectable size limit could be as small as 100 nm, as suggested by theoretical calculation. In Zhu and Ozdemir’s work [8], particles are driven towards the tapered region of a single mode fiber, as particles adhere to the surface of the fiber, the Sensors 2017, 17, 2110; doi:10.3390/s17092110

www.mdpi.com/journal/sensors

Sensors 2017, 17, 2110

2 of 10

Sensors 2017, 17, 2110

2 of 10

driven towards the tapered region of a single mode fiber, as particles adhere to the surface of the fiber, the cladding mode is altered, resulting in signal drops at the signal receiving end. By measuringmode the number andresulting magnitude thesedrops drops,atthe information revealed, but cladding is altered, in of signal theparticles’ signal receiving end.can Bybe measuring the the accumulation of particles on thethe fiber limitsinformation long-term can usebeasrevealed, a sensor. example number and magnitude of these drops, particles’ butAnother the accumulation involving the effort towards scale sensing byexample L. Cui, T.involving Zhang and H. Morgan [9], of particlesoptical on the fiber fiber is limits long-term use chip as a sensor. Another optical fiber is the in which they chip tookscale advantage fibers and and H. microfluidic detect 10 µm effort towards sensingofbymultimode L. Cui, T. Zhang Morgan [9],channels in whichtothey tookboth advantage and 45 µm bioparticles a fluid. of multimode fibers andinmicrofluidic channels to detect both 10 µm and 45 µm bioparticles in a fluid. paper, we we describe describe an optical optical fiber particle particle counter where the air flow with with particles is In this paper, coupling region region between between two two single single mode mode fibers. fibers. We avoided using particle particle guided through the coupling scattering as a method, since the desire was to minimize the volume of the sensor package and the amount of of energy energyconsumed consumedby bythe thelight lightsource. source.The Theresulting resultingsensor sensor measures drop intensity amount measures thethe drop in in intensity as as the particle disrupts electromagnetic mode of light the light beam. The implementation described the particle disrupts the the electromagnetic mode of the beam. The implementation described here here best suited for measuring the coarse fraction it has best performance for particles in is bestissuited for measuring the coarse fraction since itsince has the bestthe performance for particles in the 1 to theµm 1 to µm size range.since However, there interest is increased in theofmeasurement of 10 size10range. However, there issince increased in the interest measurement submicron and submicron particles, and nano-sized particles, point that, with the usecell, of awhere condensation cell, where nano-sized we point out that,we with the out use of a condensation smaller particles act smaller act as nucleiliquid to form larger-sized liquidmay particles, device maysmaller also beparticles. used to as nucleiparticles to form larger-sized particles, the device also bethe used to count countcondensation smaller particles. This condensation has been usedcounting in commercial particle counting This technique has been usedtechnique in commercial particle systems to increase the systems to increase the detection of submicron particles. detection of submicron particles. 2. Materials and Methods The principle behind this method is simple. However, rather than simple geometric occlusion of light passing particles disrupt thethe electromagnetic mode, changing the passing through throughthe thebeam, beam,the thepassing passing particles disrupt electromagnetic mode, changing coupling between the the twotwo fibers. ThisThis disruption causes a drop in transmission thatthat cancan be be used to the coupling between fibers. disruption causes a drop in transmission used count particles. As shown in Figure 1, two optical fiber probes are placed head to head, with some to count particles. As shown in Figure 1, two optical fiber probes are placed head to head, with distance The distance between particle and the fiber probe L1 and distance in inthe themiddle. middle. The distance between particle andemitter the emitter fiber isprobe is the L1 distance and the between particle and the receiver probe is Lprobe . distance between particle and thefiber receiver fiber is L 2 . 2

Figure 1. 1. The is aa single and the the receiving fiber is is also also single single Figure The light light from from the the emitting emitting fiber fiber is single mode mode beam, beam, and receiving fiber mode. When the particle through the the beam beam the the resulting resulting signal signal loss loss in in mode mode coupling coupling is is larger larger mode. When the particle passes passes through than just just the the geometric geometric occlusion occlusion of of the the size size of of the the particle particle blocking blocking the the beam. beam. than

Each fiber probe is made by fusion splicing a quarter pitch graded index (GRIN) fiber cap onto Each fiber probe is made by fusion splicing a quarter pitch graded index (GRIN) fiber cap onto the end of a single mode fiber. The GRIN caps serve as collimating optics for the output beam from the end of a single mode fiber. The GRIN caps serve as collimating optics for the output beam from the single mode fibers. The output from a single mode fiber can be roughly seen as a Gaussian beam, the single mode fibers. The output from a single mode fiber can be roughly seen as a Gaussian beam, without collimating optics, the width of the Gaussian wavefront would enlarge very quickly after without collimating optics, the width of the Gaussian wavefront would enlarge very quickly after leaving the fiber. As a result, only a small fraction of energy from the emitting fiber would reach the leaving the fiber. As a result, only a small fraction of energy from the emitting fiber would reach receiving fiber, which limits the overall sensitivity of the system. A comparison between the two the receiving fiber, which limits the overall sensitivity of the system. A comparison between the

Sensors 2017, 17, 2110

3 of 10

Sensors 2017, 17, 2110

3 of 10

two probes’ coupling coefficients with and without quarterpitch pitchGRIN GRINcaps capsisis as as shown fiberfiber probes’ coupling coefficients with and without thethe quarter shown in in Figure 2. The introduction of GRIN caps greatly increases the coupling coefficients, making it possibleit Figure 2. The introduction of GRIN caps greatly increases the coupling coefficients, making to maintain relatively high couplinghigh efficiency whileefficiency having a comparatively separation between possible toa maintain a relatively coupling while havinglarge a comparatively large the fiber probes, which is important for the particles to efficiently pass through the coupling region separation between the fiber probes, which is important for the particles to efficiently pass through and the light coupling the separated fiberbetween probes the interact couplingwith region and beam. interactImproving with the the light beam. efficiency Improvingbetween the coupling efficiency by a graded of appropriate act as lens, as shown in Figure 2, ahas been thefusing separated fiber index probesfiber by fusing a gradedlength indexto fiber of aappropriate length to act as lens, as well established [10,11]. The fiber probe was made by first splicing a single mode fiber and a GRIN shown in Figure 2, has been well established [10,11]. The fiber probe was made by first splicing a fiber and and thenacutting GRIN fiber and to a desired lengththe under a magnifier. In making the singletogether, mode fiber GRIN the fiber together, then cutting GRIN fiber to a desired length fiber, it is important to ensure that there is no air bubble at the splicing junction, and that the GRIN under a magnifier. In making the fiber, it is important to ensure that there is no air bubble at the cap’s cross-section afterthat thethe cutGRIN is flat cap’s and clean. splicing junction, and cross-section after the cut is flat and clean.

Figure 2. Coupling coefficient with and without GRIN caps spliced to the ends of single mode fibers. Figure 2. Coupling coefficient with and without GRIN caps spliced to the ends of single mode fibers. The single single mode mode fiber fiber and and GRIN GRIN fiber’s fiber’s parameters parameters are are from from Newport Newport F-SF F-SF single single mode mode fiber fiber and and The Thorlabs GIF-625 graded index fiber. Thorlabs GIF-625 graded index fiber.

At the the entrance entrance window window of of the the receiver, receiver, according according to to scattering scattering theory, theory,the thetotal totalelectric electric field fieldEEtt At equals the the summation summation of of the the incident incident field field without without perturbation perturbation EEii and and scattering scattering field field Es Es [12]. [12]. The The equals incident light field from the emitter fiber can be regarded approximately as a transverse Gaussian incident light field from the emitter fiber can be regarded approximately as a transverse Gaussian wave. Assume Assume the the incident incident light light is is x-polarized, x-polarized, then then EEii == E Exx,, and wave. and 2   ws 2y 2 2x x + y2 ]exp[i(k ( z  z0 )  z  xy )] w    E E exp[ s 0 [−   x Ew0 exp ] exp[−i (k(z − z0 ) − φz − φxy )] Ex =  w  2  w w  √   λ// AA   s w =  s ws  √ √   2   λ √ cos 2 2( + Al g )   πw0 nw00 n0 sin  sin(2 ( Al cos ( Algg ))  Al 2 g) πw A 2 n0 0  w n A



(1) (1)

0 0

where where EE00 is is the the amplitude amplitude at at the the center center of of the the beam beamwaist, waist,EExx and and EEyy are are the the electric electric field field amplitudes amplitudes along x and y axes, respectively. w is the local beam width as the beam propagates. k is the wavenumber. along x and y axes, respectively. w is the local beam width as the beam propagates. k is the zwavenumber. z axis. ws is the beam emitting [13–15]. 0 is the position z0 of is the the emitter positionprobe of thealong emitter probe along z axis.width ws is at thethe beam widthwindow at the emitting λwindow is the wavelength ofthe thewavelength light. 2w0 and mode (MFD) core refractive index of 0 are [13–15]. λ is of n the light. 2w0field anddiameter n0 are mode fieldand diameter (MFD) and core the single mode fiber. A is called the GRIN constant of the GRIN lens. lg is the length of the GRIN cap. refractive index of the single mode fiber. A is called the GRIN constant of the GRIN lens. lg is the φlength φxythe areGRIN functions that describe and transverse profiles of thetransverse beam. Onprofiles the other z and of cap. ϕ z and ϕxy arelongitudinal functions that describe longitudinal and of

the beam. On the other hand, the scattering field could be obtained by applying the generalized Lorenz-Mie theory (GLMT) and far-field approximation, in which

Sensors 2017, 17, 2110

4 of 10

hand, the scattering field could be obtained by applying the generalized Lorenz-Mie theory (GLMT) and far-field approximation, in which  ∞ E0 e−ikr 2n + 1   cos( ϕ) ∑ gn ( an τn + bn πn )  Esθ = −ikr n =1 n ( n + 1 ) ∞  2n + 1 E e−ikr   Esϕ = 0 sin( ϕ) ∑ gn ( an πn + bn τn ) ikr n n =1 ( n + 1 )

(2)

where Esθ and Esϕ are electric fields in azimuthal and polar directions, respectively. r is the radial distance from the scattering center. The shape factor gn describes the form of the incident field [16]. an and bn are the Mie coefficients for external scattering amplitude calculation using the below expressions:      an =     bn =

m2 jn (mx )[ xjn ( x )]0 − µ1 jn ( x )[mxjn (mx )]0 (1)

(1)

m2 jn (mx )[ xhn ( x )]0 − µ1 hn ( x )[mxjn (mx )]0 µ1 jn (mx )[ xjn ( x )]0 − jn ( x )[mxjn (mx )]0 (1)

(3)

(1)

µ1 jn (mx )[ xhn ( x )]0 − hn ( x )[mxjn (mx )]0

where m is the relative refractive index of the sphere, x is the size parameter, µ1 is the relative magnetic (1)

permeability of the sphere, jn is spherical Bessel function of order n, and hn is first order Hankel function. The primes denote derivatives with respect to the argument. Transfer expressions in Equation (2) from spherical coordinates to Cartesian coordinates, and sum the incident and scatter fields up, the electric fields after scattering could be expressed as (

Ex = Ei + Esθ cos(θ ) cos( ϕ) + Esϕ sin( ϕ) Ey = Esθ cos(θ ) sin( ϕ) + Esϕ cos( ϕ)

(4)

The perturbation in the wavefront, together with energy extinction caused by the presence of the particle, result in a mode mismatch between the emitter and the receiver fiber probes. This mismatch causes coupling efficiency loss that is detected at the other end of the receiver fiber. This loss is larger than the geometric occlusion of the particle in the beam. The normalized energy drop is accounted for by taking the integration of the fiber mode and the distorted wavefront. T=[

∞ ∞ 2 w w Ex ( x, y, x )|z0 =0 × Ex∗0 ( x 0 , y0 , z0 )|z0 =0 dx 0 dy0 ]2 πw02 −∞ −∞

(5)

where T is the power transmission rate between the two fiber probes with particle scattering. o and o’ are the origins of two Cartesian coordinate systems centered at the emitting GRIN lens surface and the receiving GRIN lens surface, respectively. Ex ’ is the fundamental fiber mode which is extrapolated to the end of the receiving GRIN lens. Figure 3a shows simulated wavefront at the receiving window with both on and off axis particle scattering. Assuming the particle is at the center of the laser’s Gaussian beam profile yields the relationship between particle size and coupling energy drop as shown in Figure 3b. As can be seen in Figure 3b, with >2.5 µm particles, the coupling energy drop is larger than 10%, which could be easily detected by a common photodiode. Figure 3c depicts the relationship between energy drop and distance between particle and z axis.

Sensors 2017, 17, 2110 Sensors 2017, 17, 2110

5 of 10 5 of 10

Figure 3. (a) Some simulated total forward intensity distribution on the xy plane (the wave Figure 3. (a) Some simulated total forward intensity distribution on the xy plane (the wave propagation propagation is along the z axis) after scattering by on-axis and off-axis particles (here, on-axis means is along the z axis) after scattering by on-axis and off-axis particles (here, on-axis means the particle is the particle is on the z axis, off-axis means the particle is not on the z axis). OG specifies the position on the z axis, off-axis means the particle is not on the z axis). OG specifies the position of the particle, of the particle, the unit of the coordinates is µm, the particle being simulated has 2 µm radius and a the unit of the coordinates is µm, the particle being simulated has 2 µm radius and a real refractive real refractive index of m = 1.52, laser beam width at beam waist is 14.2 µm, (b) calculated coupling index of m = 1.52, laser beam width at beam waist is 14.2 µm, (b) calculated coupling energy drop for energy drop for particles with different refractive indices, the distance between the two fiber probes particles with different refractive indices, the distance between the twoenergy fiber probes is on-axis 200 µm,and laser is 200 µm, laser beam width at waist is 14.2 µm, (c) theoretical coupling drop for beam width at waist is 14.2 µm, (c) theoretical coupling energy drop for on-axis and off-axis particles off-axis particles with 5 µm, 4 µm and 2 µm diameters, the refractive index m = 1.52. with 5 µm, 4 µm and 2 µm diameters, the refractive index m = 1.52.

3. Results and Discussion 3. Results and Discussion The transmitting and receiving fiber probes were made by fusion splicing Newport F-SF single mode with Thorlabs graded index fibersmade using by a Sumitomo Type 36 fiber splicer. Thefibers transmitting and receiving fiber GIF-625 probes were fusion splicing Newport F-SF The single graded index fiber was thengraded trimmedindex to ¼ pitch (246.5 µm length) act as a focusing collection mode fibers with Thorlabs GIF-625 fibers using atoSumitomo Type and 36 fiber splicer. lens [17]. The lightfiber source used wastrimmed a 4 mW Newport 633(246.5 nm He-Ne laser. The were The graded index was then to 14 pitch µm length) totwo actfiber as aprobes focusing and placed head-to-head in a V-groove carved on the surface of a brass plate, covered by glass plates. collection lens [17]. The light source used was a 4 mW Newport 633 nm He-Ne laser. The two fiber Between theplaced two fiber probes, there 400 µm diameter hole allowsofair flow to pass covered through. by probes were head-to-head in is a aV-groove carved on thethat surface a brass plate, Thus, the distance emitter and the isreceiver probes is alsohole 400that µm.allows During glass plates. Betweenbetween the twothe fiber probes, there a 400 µm diameter airthe flow experiments, a Newport optical fiber scrambler was used to ensure that there was only oneµm. to pass through. Thus, the distance between the emitter and the receiver probes is also 400 fundamental mode in the emitting fiber probe. A picture of fiber probes in the V-groove is as shown During the experiments, a Newport optical fiber scrambler was used to ensure that there was only one in Figure 4b. fundamental mode in the emitting fiber probe. A picture of fiber probes in the V-groove is as shown in Three kinds of particles with different refractive indices: water droplets, sucrose solution Figure 4b. droplets and soda lime glass beads, were tested. As depicted in Figure 4a, a Mabis CompMist Three kinds of particles with different refractive indices: water droplets, sucrose solution droplets compressor nebulizer kit was used to produce droplets from water and sucrose solutions. Soda lime and soda lime glass beads, were tested. As depicted in Figure 4a, a Mabis CompMist compressor glass beads were introduced into the air flow through a sealed glass jar which was connected to the nebulizer kit was used to produce droplets from water sucroseusing solutions. lime glassand beads system (not shown in the graph). The particle sizes wereand controlled an SKCSoda plastic cyclone were introduced into the air flow through a sealed glass jar which was connected to the system (not

Sensors 2017, 17, 2110

6 of 10

shown in the Sensors 2017,graph). 17, 2110 The particle sizes were controlled using an SKC plastic cyclone and a6regulator. of 10 With a certain flow rate, particles larger than the cut-off size of the cyclone would be filtered out of the a regulator. With certain flow rate, particles larger than the cut-off size of the cyclone would be air flow. When the aira flow rate increases, the cyclone’s cut-off size decreases. Thus, by controlling filtered out of the air flow. When the air flow rate increases, the cyclone’s cut-off size decreases. the flow rate, the largest particle size remaining in the air flow can be determined. In Figure 5, to be Thus, by controlling the flow rate, the largest particle size remaining in the air flow can be conservative, we assume that the largest particle left in the air flow after passing through the cyclone determined. In Figure 5, to be conservative, we assume that the largest particle left in the air flow is 0.5 µm larger than the cyclone’s 50% cut-off size. after passing through the cyclone is 0.5 µm larger than the cyclone’s 50% cut-off size.

Figure 4. (a) Experimental setup allowing generation and size segregation of particles, (b) fiber

Figure 4. (a) Experimental setup allowing generation and size segregation of particles, (b) fiber probes probes in the V-groove holder with 400 µm separation, (c) picture taken during experiment, where a in thelow V-groove holder withparticles 400 µmare separation, (c) picture taken light during experiment, a low flow flow rate and large used to enable the scattered from the particleswhere to be seen. rate and large particles are used to enable the scattered light from the particles to be seen.

The signal from the receiving probe was detected by a Thorlabs PDA36A photodiode. A digitizing was connected thedetected photodiode show the PDA36A waveformphotodiode. captured by it. The signaloscilloscope from the receiving probeto was by atoThorlabs A Since digitizing the air flow contains a variety of particles with sizes below the cut-off size, for each flow rate, the oscilloscope was connected to the photodiode to show the waveform captured by it. Since the air flow largest excursion was caused by particles of the maximum size being near the center between the contains a variety of particles with sizes below the cut-off size, for each flow rate, the largest excursion fibers (position 0,0). The percentage of coupling energy drop for a certain sized particle was was caused by particles of the maximum size being near the center between the fibers (position 0,0). determined by dividing the amplitude of the largest downward spike by the average amplitude of The percentage coupling energy dropwaveform. for a certain sized particle was determined by dividing the the signal onofthe oscilloscope’s output amplitudeThus, of the largest downward spike by rate, the average of the on the by controlling the system’s flow we couldamplitude easily measure thesignal percentage of oscilloscope’s coupling output waveform. energy drops caused by particles with different sizes. For higher flow rates where smaller particles disruptby thecontrolling coupling, the in transmission the baseline noise level. For the smallest Thus, thedrops system’s flow rate,are wecloser couldtoeasily measure the percentage of coupling particles counted the quantization of the digitizing oscilloscope was limiting. energy drops caused by particles with different sizes. For higher flow rates where smaller particles Experiments performed with water droplets, sugar solution droplets soda lime disrupt the coupling, were the drops in transmission are closer to the baseline noiseand level. For theglass smallest beads. It was found, as shown in Figure 5, that for water particles and soda lime glass particles counted the quantization of the digitizing oscilloscope was limiting. particles—where the index and sizes are known—there was good agreement with the theoretical Experiments were performed with water droplets, sugar solution droplets and soda lime glass model. The results of the sucrose solution were about 10% higher than expected, which may be the beads. It was as shown in Figure that forinwater particlesthan and soda lime or glass particles—where result of found, a higher concentration of 5, sucrose the droplets expected, possibly water

the index and sizes are known—there was good agreement with the theoretical model. The results of the sucrose solution were about 10% higher than expected, which may be the result of a higher

Sensors 2017, 17, 2110

7 of 10

evaporation during transport. However, it was clear that the sucrose solution particles followed7 the of 10 same trend, and could also be measured. The percentage energy drop is not strictly monotonic along the Sensors increase size, which means that within a certain interval, 2 particles with 7different 2017,in 17, particle 2110 of 10 concentration of sucrose in the droplets than If expected, orwere possibly water evaporation during transport. sizes might cause an identical energy drop. the desire to precisely determine the particle size, evaporation during transport. However, itrequire was particles clear thatinvestigation. the sucrosethe solution followed However, it was clear the sucrose solution followed same particles trend, and couldthe also be rather than count the that particles, this would more same trend, could also be measured. energy drop is notthe strictly monotonic along size, measured. The and percentage energy drop isThe notpercentage strictly monotonic along increase in particle the increase in particle size, which means that within a certain interval, 2 particles with different which means that within a certain interval, 2 particles with different sizes might cause an identical sizes might cause an identical energy drop. If the desire were to precisely determine the particle size, energy drop. If the desire were to precisely determine the particle size, rather than count the particles, rather than count the particles, this would require more investigation. this would require more investigation. Sensors 2017, 17, 2110

Figure 5. Solid lines are calculations of the loss as a function of particle given refractive index and particle dotted lines connectof experimental whereofthe particlegiven size is determined by and the Figure 5. size. SolidThe lines are calculations the loss as adata function particle refractive index Figure 5. Solid lines are calculations of the rate loss and as a function of particle given refractive index andto be cut-off point of the cyclone at a given flow the largest signal excursions are assumed particle size. The dotted lines connect experimental data where the particle size is determined by the particle size. The dotted lines connect experimental data where the particle size is determined by the causedpoint by particles with sizes toflow the cut-off plus 0.5largest µm. signal excursions are assumed to be cut-off of the cyclone at aequal given rate and the cut-off point of the cyclone at a given flow rate and the largest signal excursions are assumed to be

caused by particles with sizes equal 0.5µm. µm. caused by particles with sizes equaltotothe thecut-off cut-off plus plus 0.5

In addition to revealing the largest particle size in an air flow, the system may also be used to In addition to revealing the size in anair airflow, flow, the system may be used to determine the size of the particle particlessize inin the air flow. Figure 6a shows the count In addition todistribution revealing thelargest largest particle an the system may alsoalso be used to of percentage energy drops with the sensing volume directly connected to the output of the nebulizer, determine the size distribution of the particles in the air flow. Figure 6a shows the count of percentage determine the size distribution of the particles in the air flow. Figure 6a shows the count of without cyclone in the thedrops system restrain the size connected ofdirectly particles. arethe water energy drops with sensing directly toThe theparticles output of without percentage energy withtovolume the sensing volume connected to the output ofnebulizer, thedroplets. nebulizer, Figure 6b is the particle distribution aThe nebulizer’s output measured with Malvern’s without cyclone in to the systemsize to size in of particles. The particles are water droplets. cyclone in the system restrain therestrain size ofthe particles. particles are water droplets. Figure isparticle the particle size device. distribution in a nebulizer’s output measured with Malvern’s Spraytec particle size measurement The particles areoutput water droplets, the measurements were Figure 6b is6b the size distribution in a nebulizer’s measured with Malvern’s Spraytec Spraytec particle size measurement device. The particles are water droplets, the measurements were taken with a 2 lpm air flow rate.The Particles with µm diameters aremeasurements excluded fromwere this taken statistical particle size measurement device. particles are

An Optical-Fiber-Based Airborne Particle Sensor.

A new optical-fiber-based airborne particle counter is reported. Unlike traditional light-scatter-based techniques, the particle is detected through t...
3MB Sizes 0 Downloads 12 Views