LABORATORY TEST OF THE NEW SPRAY DOSE ADJUSTMENT SYSTEM FOR FIELD SPRAYERS

Karol GARBIAK, Jan JURGA, Macej TOMCZEWSKI Zachodnopomorsk Unwersytet Technologczny w Szczecne, Wydzał Kształtowana Środowska Rolnctwa Katedra Budowy Użytkowana Urządzeń Techncznych ul. Papeża Pawła VI/3; 71-459 Szczecn, Poland e-mal: karol.garbak@zut.edu.pl LABORATORY TEST OF THE NEW SPRAY DOSE ADJUSTMENT SYSTEM FOR FIELD SPRAYERS Summary The artcle presents an ntal assessment of the new regulatory system of a spray dose adjustment system for feld sprayers. In contrast to tradtonal sprayers, the proposed soluton ncorporates trple nterlocked wth dversfed lqud expense nozzles controlled based on the dose adjustment process utlzng the forward and angular veloctes of the machne. The adjustment system was tested usng data gathered from montorng the movement of a sprayer performng chemcal plant preservaton. The results prove that the system enables a substantal mprovement n the surface dstrbuton of the spray dose. Compared to the aforementoned tradtonal sprayer, where only 57.2% of measurng ponts have shown the spray dose beng wthn 80-120% of nomnal range, usng the proposed soluton the proper dose consttuted 93.6% of all results. Key words: feld sprayer, dose adjustment, momentary dose, forward velocty, angular velocty BADANIA LABORATORYJNE NOWEGO SYSTEMU REGULACJI DAWKI CIECZY APLIKOWANEJ PRZEZ OPRYSKIWACZ POLOWY Streszczene W artykule przedstawono wstępną ocenę nowego systemu regulacj dawk ceczy aplkowanej przez opryskwacz polowy. W porównanu z tradycyjnym opryskwaczem, w prezentowanym rozwązanu zamast pojedynczych rozpylaczy zastosowano zblokowane rozpylacze trójelementowe o zróżncowanym wydatku ceczy, sterowane w oparcu o proces regulacj dawk wykorzystujący wynk dotyczące prędkośc postępowej kątowej maszyny. System regulacj przetestowano wykorzystując dane zebrane podczas montorowana ruchu opryskwacza realzującego chemczną ochronę rośln. Uzyskane wynk dowodzą, że system umożlwa znaczną poprawę rozkładu powerzchnowego dawk ceczy. W porównanu z pracą montorowanego na polu opryskwacza, gdze jedyne 57,2% punktów pomarowych wykazywało dawkę ceczy należącą do zakresu 80-120% dawk nomnalnej, przy użycu proponowanego rozwązana właścwa dawka ceczy stanowła 93,6% wszystkch wynków. Słowa kluczowe: opryskwacz polowy, regulacja dawk, dawka chwlowa, prędkość postępowa, prędkość kątowa. 1. Introducton Chemcal pestcdes are an mportant factor n the process of ncreasng the yeld and qualty of agrcultural output. The European Unon s sustanable use of pestcdes strategy should lead to the elmnaton of threats assocated wth ther use [10]. The man envronmental threat n agrculture s local contamnaton appearng n places where the pestcdes accumulate [2]. Pestcdes, whle bologcally affectng pests and crops, are not nert regardng humans and other flora and fauna n the ecosystem [6]. Durng feld work, t s advsed to perform chemcal plant preservaton tasks along straght lnes and wth a constant sprayer speed. In practce, there are condtons causng cultvators to change drectons durng the task [1]. Chemcal preservaton tasks are usually performed usng tramlnes. Those should be placed precsely n order to mnmze the area wth double spray coverage or no spray at all. The best results can be acheved whle usng automated drvng systems as well [6]. Numerous authors [1, 8, 9] pont out that the sprayer bar s horzontal movement can greatly nfluence the qualty of the spray. Ths observaton allows us to mprove the solutons used to stablze the horzontal fluctuatons of the bar, unjustly neglected untl now On the bass of montorng the work of the sprayer performng the chemcal plant preservaton tasks on the area of 15.2 hectares, t was found that as a result of changes n speed and drecton of movement, as much as 30.9% of the area was covered wth a dose of lqud devatng from the nomnal dose by more than 15%. There was also urgent need of the nstgate of the research and mplementaton of solutons mprovng the operatng parameters of feld sprayer [3]. In order to keep the amount of lqud at acceptable levels n a wde range of speed and drecton of movement of the sprayer, a new regulatory system has been developed at the West Pomeranan Unversty of Technology n Szczecn [4]. 2. The am of the study The objectve of the research s to evaluate the developed spray dose adjustment system appled by a feld sprayer movng wth varable forward and angular velocty. In order to complete the aforementoned objectve, the followng questons have been formulated: 1. What are the momentary spray doses estmated usng predctve smulaton methods based on montorng a feld sprayer s work data and the performed adjustments? 2. What s the outcome from verfyng the doses calculated usng predctve smulaton methods by comparng them wth doses obtaned durng laboratory research usng weght methods? 21

USER INTERFACE Q, v, r, R 1, R 2, R 3 GPS RECEIVER v m, t, x, y DIRECTION SENSOR α MICROCONTROLER SD MEMORY CARD ω, v m, t, x, y, q, qr SET OF S0LENOID VALVES Fg. 1. Block dagram of the spray dose adjustment system appled by ndvdual nozzles Rys. 1. Schemat blokowy układu regulacj dawk aplkowanej przez poszczególne rozpylacze 2 3. Materals and methods The block dagram of the spray dose adjustment system s shown on fg. 1. The system conssts of a GPS recever, drecton sensor, mcrocontroller, user nterface and a set of solenod valves [4]. The GPS recever transmts a $GPRMC sequence from whch retreves the current coordnate related to velocty and tme data. The speed, tme and angle of rotaton of the sprayer measured by the transducer of the drecton of movement are the varables needed to calculate the momentary lqud dose appled by the prmary set of nozzles. The data related to the coordnates s not needed to realzaton the regulaton. In the case of archvng performed treatments they fulfll an auxlary functon, and n ths study, t makes possble to comple the drawngs of the surface dstrbuton of the spray doses. Contrary to sprayers used today, each nozzle s replaced wth three, of dfferent capacty (fg. 2). Ths soluton allows to acheve a total of 8 dfferent spray doses. One of the doses s used as a nomnal value (Q), whereas the remander can be used to compensate for suboptmal forward or angular veloctes, whch results n exceedng the allowed dose, accepted as 80-120 [% Q] n ths paper. The followng relatonshp s used to calculate the dose (1): Q Q q = r [dm3 ha v vm r m -1 ], (1) t where: q spray dose appled by the prmary nozzle set [dm 3 ha -1 ], Q nomnal dose [dm 3 ha -1 ], v nomnal sprayer velocty accepted for the calculaton [m s -1 ], v m real forward velocty of the sprayer [m s -1 ] at tme t, α angel of drecton of sprayer [rad] at tme t, r coordnates of sprayer poston on workng boom [m], t - tme between consecutve sequences $GPRMC [s]. ω angular velocty of the sprayer [rad s -1 ] at tme t. The system conssts of a GPS recever, drecton sensor, mcrocontroller, SD memory card, user nterface and a set of solenod valves [4]. The GPS recever transmts a $GPRMC sequence to the controller, contanng data about current coordnates, velocty and tme. These data, along wth the angular vector of movement, are used to calculate the dose appled through the prmary set of nozzles. Contrary to sprayers used today, each nozzle s replaced wth three, of dfferent capacty (fg. 2). Ths soluton allows to acheve a total of 8 dfferent spray doses. One of the doses s used as a nomnal value (Q), whereas the remander can be used to compensate for suboptmal forward or angular veloctes, whch results n exceedng the allowed dose, accepted as 80-120 [% Q] n ths paper. The values of the parameters n the numerator (1) are entered to the mcrocontroller s memory by the user and are not changed durng the sprayng task. Fg. 2. Head equpped wth three solenod valve-controlled nozzles Rys. 2. Głowca wyposażona w trzy rozpylacze sterowane za pomocą elektrozaworów 22

The varables v m and α, located n the denomnator, can change dynamcally because of vared velocty and movement angle caused by the feld layout and encounterng obstacles, such as masts, trees or ponds, as well as caused by nsuffcent tranng of the machne operator. It has been assumed that the turnng angle of the sprayer to the left s postve and to the rght - negatve. Lookng along the sprayer s drecton of movement, coordnates of the nozzle located on the longtudnal turnng axs have a null value, nozzles on the rght of the axs have progressvely postve values expressed n meters. The coordnates of the nozzles on the left of the axs have negatve values. If the value calculated for the prmary nozzles s not consstent wth the acceptable range of values, the mcrocontroller fnds and actvates the proper combnaton of nozzles through the solenod valves. Table 1 shows the basc data and combnatons of used nozzles. The baselne nozzle set n ths research was combnaton no. 7, consstng of nozzles R 2 and R 3 for a total expendture of 1.38 [dm 3 mn -1 ] and the nomnal dose Q=237 [dm 3 ha -1 ], for speed v-1.94 [m s -1 ] and an nternal pressure of 0.3 [MPa]. Ths set was beng actvated by the mcrocontroller f the calculated dose q was wthn 80-120 [%] of the nomnal dose Q, or between 190 and 283 [dm 3 ha -1 ]. If the calculated dose q had a value between 280-424 [%Q] for example, or 664-1004 [dm 3 ha -1 ], set no. 2 was actvated, whch would change the calculated dose accordngly to the correcton coeffcent, or tmes 0.283, and because of that the momentary spray dose would stll ft between 80-120 [%Q]. The adjustment propertes for the nozzle sets can be shown as a graph of the relatonshp of the adjusted dose value q r to the calculated value q (fg. 3). The graph shows that all calculated doses q wth values between 62-428 [%Q] fall wthn the 80-120 [%Q] range. Doses hgher than 0 but lower than 62 [%Q] are ncreased tmes 1.283, but do not reach the mnmal correct value of 80 [%Q]. If the calculated dose q s hgher than 424 [%Q], or exceeds the nomnal dose Q set by the user four tmes, t has been agreed that the correct decson s to dsable the nozzles. The same reacton takes place for q<0 [%Q], or doubled spray coverage. Smulatons have been carred out n order to test the adjustment system. Durng the tests, montorng movement data from a sprayer performng chemcal preservaton tasks was used. The selected task fragment s approxmately 180 [m] long, wth notceable velocty and drecton changes (fg. 4). Usng the data saved on the SD card concernng tme, forward and angular velocty of the sprayer, as well as the nomnal dose and the speed set for the sprayng task, the mcrocontroller calculated the momentary dose q and qr for nozzles wth coordnates r and subsequently actvated the dfferent combnatons of nozzles. Dose data q and qr were saved on the SD card and later used to prepare charts showng the area dstrbuton, as well as a momentary dose hstogram. Table 1. Basc data for nozzles R 1, R 2, R 3 and ther combnatons (No.) [7] Tab. 1. Podstawowe dane rozpylaczy R 1, R 2, R 3 oraz ch kombnacj (No.) [7] No. Q for Correcton Expendture for press. 0.3MPa v=1.94m s -1 Calculated dose q Adjusted dose qr coeffcent dm 3 mn -1 dm 3 ha -1 dm 3 ha -1 %Q dm 3 ha -1 %Q R1 R2 R3 R1+R2+R3 1. 0 0 >1004 >424 0.000 0 0 2. 0.39 0.39 67 664-1004 280-424 0.283 188-284 80-120 3. 0.59 0.59 101 495-664 209-280 0.428 212-284 89-120 4. 0.79 0.79 135 399-399 168-209 0.572 228-283 96-120 5. 0.39 0.59 0.98 168 331-399 140-168 0.710 235-283 99-119 6. 0.39 0.79 1.18 202 283-331 120-140 0.855 242-283 103-120 7. 0.59 0.79 1.38 237 190-283 80-120 1.000 190-283 80-120 8. 0.39 0.59 0.79 1.77 303 0-190 0-80 1.283 0-244 0-103 Fg. 3. The relatonshp between dose qr after adjustment and the calculated dose Rys. 3. Zależność dawk qr po regulacj od dawk oblczonej q 23

Fg. 4. Montorng values used: forward velocty (v m) and angular velocty (ω) of a sprayer wth an operatng wdth of b=20 [m], Q =234 [dm 3 ha -1 ], v=1.94 [m s -1 ], p=0.4 [MPa], nozzles TT11003 [7] Rys. 4. Wykorzystane wynk montorowana prędkośc postępowej (v m) kątowej (ω) opryskwacza o szerokośc roboczej b=20 [m], Q =234 [dm 3 ha -1 ], v=1,94 [m s -1 ], p=0,4 [MPa], rozpylacze TT11003 [7] In order to verfy the results regardng the momentary dose qr, a laboratory measurement of the real amount of lqud sprayed by the nozzle sets has been carred out. Lqud volume changes durng a 90 [s] adjustment cycle were performed for nozzles located the furthest from the axs of the sprayer: r =-9.75 and r = 9.75 [m]. A laboratory scale wth a vessel for the lqud was used. The measurements of the scale were recorded by a PC wth a frequency of 5 Hz. The acqured results of momentary nozzle spray amounts and data regardng velocty vm, operatng wdth of the nozzle sets and tme t, were the bass for calculatng the adjusted doses (qrw). In order to test f the doses calculated wth smulaton methods (qr) correspond to the values acqured wth weght methods (qrw), the lnear correlaton coeffcent R and the quotent of average values of both doses Dq were used. The verfcaton s postve f the correlaton coeffcent value s not less than 0.98 and the quotent of average values s wthn 0.95-1.05 [5]. Values of the lnear correlaton coeffcent R were calculated usng the relatonshp below: R n 1 n 1 ( qr qr) ( qrw ( qr qr) 2 n 1 qrw) ( qrw qrw) 2, (2) The quotent Dq was calculated based on the average doses qr and qrw: qr Dq. (3) qrw 4. Results In order to facltate nterpretaton, all results n the later parts of the artcle regardng the montorng dose q [dm 3 ha - 1 ], as well as the adjusted dose qr or qrw [dm 3 ha -1 ], wll be presented as a percentage of the adjustment dose Q, measurng 237 [dm 3 ha -1 ]. Results concernng the momentary spray values have been shown on fg. 5. All data shown were acqured va smulaton based on the results from montorng the sprayer s feld route (fg. 5a) and the dose adjustment (fg. 5b). Momentary spray values q acqured based on montorng are vared to a hgh degree along the entre route. Major devatons from the nomnal dose Q appear not only durng change of drecton, but on straght lnes as well, f the momentary velocty was dfferent than nomnal and was changng often. Talkng wth the operator about feld drvng technques has shown that the man cause of errors les n the necessty of keepng close to the route whle avodng varous feld obstacles. Hghly benefcal results have been observed when usng the adjustments. Momentary doses on the turns and straghts have shown mprovements. Spray doses qr over the entre route dd not exceed 120 [%Q]. Doses smaller than 80 [%Q] were rarely observed around the outsdes of turns, where the montored doses had values smaller than 62 [%Q]. Durng the adjustment, doses were multpled tmes 1.283 but ther fnal values dd not reach the acceptable bottom lne (80-120 [%Q]). It s worth notng that despte large adjustment possbltes, the operator should not allow for the dose to rse above the 62-424 [%Q] range. Fg. 6 shows the hstogram of measurng ponts before and after adjustment. The graph was prepared based on data from 2 x 3600 measurng ponts. The ntervals correspond to the adjustment ranges, the 0-62 [%Q] range was dvded nto two parts: 0-50 and 50-62 [%Q]. Before the adjustment, 57.2% of all measurng ponts were showng dose q wthn the acceptable range of 80-120 [%Q]. 7.9% of measurng ponts were n the 120-140 [%Q] range. The ranges wth ncreasng doses had less ponts each. The range of 424 [%Q] or more had only 2% of measurng ponts. A farly large number of ponts appeared wth a dose less than 80 [%Q]. The defntve cause of ths conssted n excessve forward velocty of the sprayer. As much as 15% of doses were n the 62-80 [%Q] range. Less-than-zero values, meanng doubled spray, were 1.2 [%] of all doses. After 24

performng the adjustment, the amount of doses n the 80-120 [%Q] range ncreased notceably and was at 93.6 [%]. Ths range now contaned all of the doses showng values between 62-424 [%Q] before. The adjustment caused 3.2 [%] of doses to zero completely these were the doses wth values lower than 0 [%Q] or hgher than 424 [%Q]. Fg. 7 shows the data gathered n order to verfy the smulaton research by comparng wth the results acqured wth the weght method. These data apply to the dynamcs of dose changes n the ponts furthest away from the sprayer s turnng axs. The momentary dose wll have the most dfference from nomnal Q values there. Both charts contan two straght, horzontal lnes outlnng the 80-120 [%Q] dose value range. The green curves present the adjusted dose qr calculated by the mcrocontroller based on smulatons for nozzles wth coordnates between r=-9.75 and r=9.75 [m]. Red curves present dose qrw after adjustments and the values used to plot them were estmated based on measurng the contnuous lqud expendture appled by nozzle sets. The carred out result verfcaton proves that there are no dfferences between dose qrw amounts estmated usng the weght scale and the qr dose calculaton results saved on the SD card by the mcrocontroller. For nozzles placed at r=-9.75 and r=9.75 [m], the lnear correlaton coeffcent R values were respectvely 0.995 and 0.992, whereas Dq quotent values were 0.994 and 1.006. Fg. 5. Surface dstrbuton of momentary spray doses: a) montorng results, b) adjustment results Rys. 5. Rozkład powerzchnowy chwlowych dawek oprysku: a) wynk montorngu, b) wynk regulacj Fg. 6. Dose quantty (LD) hstogram: q montorng results, qr adjustment results Rys. 6. Hstogram lczebnośc dawek (LD): q wynk montorngu, qr wynk regulacj 25

Fg. 7. Dose change dynamcs after adjustment calculated based on smulatons (qr) and dose estmated usng the weght method (qrw) for edge nozzles wth coordnates of r =-9.75 and r = 9.75 [m] Rys. 7. Dynamka zman dawk po regulacj oblczonej na podstawe symulacj prognostycznej (qr) oraz dawk oszacowanej metodą wagową (qrw) dla skrajnych rozpylaczy o współrzędnych; r = -9,75; r = 9,75 [m] 5. Conclusons 1. The research shows that the descrbed concept of a dose adjustment system for use n feld n feld sprayng s correct. 2. The descrbed system properly adjusts all doses wth values fallng nto the range of 62-424% of nomnal dose. 3. The adjustments carred out on the montored route of the feld sprayer rased the percentage of doses wth values wthn 80-120% of nomnal from 57.3 to 93.6%. 4. No dscrepances between the regulated dose values found n smulatons and weght methods. 5. Further research on the proposed adjustment system s advsed. 6. References [1] Anthons J., Audenaert J., Ramon H.: Desgn Optmsaton for the Vertcal Suspenson of a Crop Sprayer Boom, Bosystems Engneerng, 2005, 90(2), 153-160. [2] Doruchowsk G., Hołownck R.: Przyczyny zapobegane skażenom wód gleby wynkających ze stosowana środków ochrony rośln. Zeszyty IMUZ, 2003, 9, 96-115. [3] Garbak K., Jurga J.: The prognostc smulaton of momentary dose of lqud appled by feld sprayer wthout an automatc pressure control. Journal of Research and Applcatons n Agrcultural Engneerng, 2015, 60 (1), 5-10. [4] Garbak K., Jurga J.: Sposób regulacj dawk ceczy w opryskwaczach polowych podczas zmany kerunku ruchu oraz opryskwacz. Zgłoszene wynalazku P. 407372. Buletyn Urzędu Patentowego, Warszawa, 2015, 19(1088), Rok XLIII, 2. [5] Gawęck J., Wagner W.: Podstawy dośwadczalnctwa w nauce o żywenu żywnośc. AR, Poznań, 1988. [6] Hołownck R., Doruchowsk G., Godyń A., Śwechowsk W.: Technka ochrony rośln w dyrektywach UE. Inżynera Rolncza, 2011, 4(129), 74-84. [7] Katalog produktów TeeJet Technologes. 2011. 51-PL. [8] Lardoux Y., Snfort C., Enfalt P., Mralles A., Sevla F.: Test Method for Boom Suspenson Influence on Spray Dstrbuton, Part II: Valdaton and Use of a Spray Dstrbuton Model. Bosystems Engneerng, 2007, 96(2), 161-168. [9] Lardoux Y., Snfort C., Enfalt P., Sevla F.: Test Method for Boom Suspenson Influence on Spray Dstrbuton, Part I: Expermental Study of Pestcde Applcaton under a Movng Boom. Bosystems Engneerng, 2007, 96(1), 29-39. [10] Makles Z., Domańsk W.: Ślady pestycydów nebezpeczne dla człoweka środowska. Bezpeczeństwo Pracy, 2008, I, 5-9. 26