1、2010 ASHRAE 581ABSTRACTAlthough the use of a piezoelectric transducer has trans-formed a traditional range hood to a self-contained one andgive it real-time performance on the detection of cooking fumesthat is never achieved by the temperature and the optic sensorscommonly used for continuous ventil
2、ation in the kitchen, theexhaust flow rate controlled precisely without unnecessaryrunning noise and power consumption on the basis of reliableestimation of the amount of plume-like cooking fumes is stilldesired. In this study, the hardware of the self-contained rangehood equipped with twelve piezoe
3、lectric transducers is devel-oped first. It gives reliable estimation about the cooking fumesand can be used as a general material to implement preferablerange hoods with acceptable costs. Then, a cheap but effectivemethod to examine the escape of cooking fumes is presented.This method helps to dete
4、rmine the proper exhaust flow rate forthe released amount of cooking fumes detected by the employedtransducers. Finally, the performance on the detection of cook-ing fumes is inspected with a series of analyses proposed herein order to save the number of transducers objectively andreasonably.With th
5、e implementation in this study, cooking fumes canbe detected immediately and estimated properly within foursampling periods that is about 23.8 ms. Then the self-containedrange hood removes the cooking fumes with proper exhaustflow rates instead of the full one. As a result, the noise reductionor the
6、 energy saving can be reached up to 39.53%.INTRODUCTIONIt is well known that home cooking produces high levelsof contaminants into the air. These contaminants are com-posed of particulate matter, moisture, heat, and odors, whichusually pose adverse effects on comfort and even health (Seeet al. 2006;
7、 Svendsen et al. 2002). Therefore, the ways toremove cooking fumes are always important issues to resi-dents for the air quality in the kitchen. Contemporary rangehoods exhaust cooking fumes based on two principles (Liet al. 1997). Consider the heat at the cook-top, thermal plumesare generated. As a
8、 result, cooking contaminants are broughtupwards by the nature of the thermal plumes. This mechanismfacilitating cooking contaminants to be captured by the rangehood over the cook-top is called buoyancy-capture principle(Figure 1). The range hood continues to collect and removecooking fumes moving t
9、owards to it with powerful mechanicalairflow, which is called velocity-capture principle.Although the range hood exhibits its capability on theelimination of cooking fumes, it is sometimes left off duringthe cooking. That might reflect the annoyance of the acousticnoise generated from the exhaust fa
10、n (Grimwood 1997).Hence, in our early study, a range hood was proposed to switchits exhaust velocity to a proper level according to the detectionof cooking fumes completed by a sensitive piezoelectric trans-ducer in order to reduce the running noise (Liu and Young2002). The average noise level of th
11、e proposed range hood is72 dB while the exhaust velocity is fixed at its maximum(12.55 m/s) measured at the exhaust outlet (10 23 cm2). Instir-frying vegetables, a very common cooking in Orientalkitchens, the average noise level becomes 65.66 dB when theproposed range hood works with the given funct
12、ion.Although the early proposed range hood can work withlower noise, its reliability needs to be further confirmed. First,the detection area of the piezoelectric transducer is not suffi-cient to cover all of the plume-like cooking fumes. Second,with the equivalent capacitance inside (Dally et al. 19
13、93), theThe Novel Use of Piezoelectric Transducers in the Implementation of Reliable Self-Contained Range HoodsTang-Jen Liu, PhD I-Cheng Shen, PhDTang-Jen Liu is an Associate Professor with the Department of Electronic Engineering, Far East University, Hsin-Shih, Taiwan. I-Cheng Shenis a Post-doctor
14、ate researcher with the Department of Electrical Engineering, National Cheng Kung University, Tainan, Taiwan.OR-10-062 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use onl
15、y. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. 582 ASHRAE Transactionsavailable output of the piezoelectric transducer always lagsbehind the detection. Third, although the peak level of thetransduce
16、r output can be an estimate for the amount of thecooking fumes composed of the same constituent, as illus-trated in the detection of water particles (Liu, 2007), it is hardto conclude this connection for all cooking processes andmaterials used. One evidence is that the peak levels induced bythe same
17、 amount of water and oil particles separately on thepiezoelectric transducer are different. Consequently, the peaklevel of the transducer output cannot be a common scale.In this paper, a reliable self-contained range hood ispresented to solve the issues mentioned above. Its reliabilitycomes from the
18、 novel use of the piezoelectric transducersproposed here. From the design of the self-contained rangehood with the use of piezoelectric transducers to the analysisof the detection performance achieved by these transducers, ageneral road map to create a demanded self-contained rangehood is constructe
19、d.IMPLEMENTATIONIn our early study (Liu and Young, 2002), cooking fumeswere detected locally with only one transducer. To plume-likecooking fumes, one transducer is not fully adequate to guar-antee the fidelity of the measurement. To improve the preci-sion of the measurement, global detection of coo
20、king fumes istaken into consideration. In addition, the equivalent circuit ofthe piezoelectric transducer works as a low-pass filter whichintroduces a phase shift in its output. As a result, the availablepeak level of the transducer output always lags behind thedetection. Even the lag can be accepte
21、d, the peak level still cannot be used as a common index for the amount of cookingfumes as mentioned before. Consequently, the temporal vari-ation of the transducer output is used instead of its peak levelon the basis of the translation algorithm proposed in our previ-ous study (Liu 2007). Figure 2
22、is the schematic diagram of areliable self-contained range hood. Its implementation isdescribed as follows.Global DetectionThe majority and the most concerned contaminants incooking fumes are the liquid particles of water and culinarygrease or oil (Wolbrink and Sarnosky 1992). When the buoy-ancy and
23、 velocity capture principles both work on these parti-cles, perceptible kinetic impact occurs on the piezoelectrictransducer (400ET180) placed close to one of the two exhaustinlets of the range hood. The piezoelectric transducer convertsthe impact to a peak level at its output voltage in a sensitivi
24、tyof 61.6 mV/Pa (Liu and Young, 2002). This sensitivity can bemagnified to facilitate the inspection of the change on thedetection.To have immediately a reliable estimation about theamount of plume-like cooking fumes in order to remove themin time with a proper exhaust flow rate, global detection is
25、carried out not only with more transducers but also a transla-tion algorithm to process their output voltages. Figure 3 showsFigure 1 Cooperation of the buoyancy-capture principleand velocity-capture principle on the eliminationof cooking fumes. This figure comes after animage processing to enhance
26、the almost invisiblecooking fumes.Figure 2 A reliable self-contained range hood designedwith twelve piezoelectric transducers. The ane-mometer (LCA 6000) and the PC-based datarecorder are used to monitor the action of themicrocontroller. In addition, the feedback buttonsare used to modify the volume
27、tric exhaust flowrate referred by the microcontroller in order topass the test of escape detection. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional repro
28、duction, distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission. ASHRAE Transactions 583the positions of the transducers employed in Figure 2. For effi-ciency, transducers were only arranged on the way out for thecaptured fumes. In ord
29、er to detect the fumes from everywhere,the way out were surrounded by the transducers.Real-Time Signal ProcessingTo figure out the result of the global detection immedi-ately from the output voltages of all transducers, the transla-tion algorithm developed early is used. To apply this algorithmto al
30、l transducers, the signal conditional module and mappingcircuit (Liu, 2007) developed at the same time are also dupli-cated for each transducer, as indicated in Figure 2. This algo-rithm can figure out the start and end time points about thepresence of cooking fumes with four successive samples onth
31、e temporal variation of the transducer output. Here, thisalgorithm is implemented by a Microchip microcontroller(PIC16C73A) and its on-chip 8-bit A/D converter. In Figure 2,the twelve transducer outputs are polled in sequence by usinga multiplexer. Each transducer output is polled at a frequencyof 1
32、68 Hz. The polled output is first converted to its digitalform by the on-chip A/D converter and then processed with thetranslation algorithm in order to generate a binary signalwhich indicates the presence of fumes with the higher level onit. As a result, twelve binary signals are generated as corre
33、-sponding time series. Between successive sampling timepoints for polling all transducer outputs, the amount of thefumes around the two exhaust inlets can be represented as thesum of the twelve binary signals. As a result, the change on theamount of cooking fumes can be recognized within a samplingp
34、eriod. With the setup shown in Figure 4 to stir-fry spinach, asummation result was obtained, as partially shown in Figure 5.Configuration of the Volumetric Exhaust Flow RateThe level of the summation waveform coming from thecount of the transducers which detect the fumes at the sametime usually sign
35、ifies the extent of the plume-like cookingfumes. Therefore, the summation level can approximate theamount of cooking fumes at each sampling time point. As aresult, the adaptation of the exhaust flow rate can follow thechange of the summation waveform. Figure 6 depicts the rela-Figure 3 Arrangement o
36、f the twelve piezoelectric transduc-ers employed in Figure 2.Figure 4 Setup of the cooking appliances in this study. Themeshes used here are to detect the escapingcooking fumes. Figure 5 Summation of the twelve binary signals at aroundthe (a) beginning, (b) middle, and (c) end of onespinach stir-fry
37、ing with the setup shown in Figure 4. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digi
38、tal form is not permitted without ASHRAEs prior written permission. 584 ASHRAE Transactionstionship between the summation level and the proper volumet-ric exhaust flow rate developed as follows.The self-contained range hood is designed to power theexhaust fan fully in the beginning and then to have
39、it stand byat a base flow rate (181.470 L/s, 384.511 cfm) if cookingfumes have not come yet. During cooking, the range hoodwould speed up its exhaust flow rate from either the base levelor other higher levels to remove the released cooking fumesand would slow down and towards the base level if cooki
40、ngfumes are getting smaller. Consequently, the exhaust fan wasalways speeded up from the base flow rate in the experimentsdetermining the proper volumetric exhaust flow rate to removethe cooking fumes represented as a specific level on thesummation waveform. This experiment was conducted withwater i
41、nside two covered pans heated respectively on the twoburners (Figure 4).When water inside the pans was both boiling, the coverswere opened face to face at a specific angle shortly and thenwere put back again. With the release of fumes, the exhaust fanwas immediately accelerated from the base flow ra
42、te to apreset one which was assigned for the present level of thesummation waveform. If the preset flow rate was insufficientto prevent the released fumes from escaping, the “Up” feed-back button in Figure 2 was used to increase its value for thecorresponding summation level by a half of the increme
43、nt ordecrease in the last modification. The initial modification wasset to 92 L/s (195 cfm). If the preset flow rate was sufficient,the “Down” feedback button would be tried to decrease it in asimilar way, in order to reduce the proportional running noiseas it could be. In Figure 6, each proper volu
44、metric exhaust flowrate is normalized to the full one (288.650 L/s, 611.611 cfm)of the centrifugal exhaust fan used in Figure 2. The adaptationof the exhaust flow rate was implemented by driving the fanwith an inverter (VFD002S11A).Performance InspectionReferring to the configuration of the required
45、 volumetricexhaust flow rates depicted in Figure 6, the time profiles of thenormalized exhaust flow rate responding to the detection resultsof Figure 5 are exhibited in Figure 7. It was measured at theexhaust outlet (10 23 cm2, 3.937 9.055 in.2) of the rangehood with an anemometer (LCA 6000) read by
46、 a PC-based datarecorder, as shown in Figure 2. The recording operation wassynchronized with the system clock of the microcontroller.The acoustic noise pressure level of the exhaust fan isfound proportional to its velocity. As a result, the noise inten-sity (Fahy, 2001) produced and the work down by
47、 the exhaustfan can be rated with the square of the exhaust velocity gener-ated to remove the cooking fumes. Hence, the performance ofthe range hood in this study is investigated by accumulatingthe square of the normalized volumetric exhaust flow rates atall sampling time points, because the normali
48、zed volumetricexhaust flow rate has the same magnitude as the normalizedexhaust velocity. A smaller accumulated sum means morenoise reduction and energy saving having been achieved.Figure 6 Configuration of the volumetric exhaust flow ratefor the released amounts of cooking fumes repre-sented with t
49、he twelve levels of the summationwaveform. In this figure, all the exhaust flowrates are normalized with the maximum one(288.650 L/s, 611.611 cfm).Figure 7 Normalized volumetric exhaust flow rate re-sponding to the detection results of Figure 5according to the configuration of Figure 6. 2010, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2010, Vol. 116, Part 1. For personal use only. Additional reproduction, distribution, or transmission in either print or digital form is not permitted without AS
