A A NEI.N APPROACH FOR THE AUT01MATIC CONTROL OF PRESSURE DROP (WEIGHT) I.N FILTER ROD MAKING Thomas Leutner~ Christoph Greir.=_- RHONE-POULENC RHODIA AG Deoar`tment Filter Tow 79013 Freiburg/Germany 1. SUMMARY Pressure Drop (Weight) and dla.7=-'Zer ct: filter rods are the most important quality parameters during their manufac:'_re. On-147e diameter control aready is today's technology. For pressure drCD (V.S'ght) *.oday's c.-mmon practice is to test 111iiter off- line on testing stations and do tl~=_ neceessa,-,/ adjustments by manually altering the speed ratio between blooming unil and -'~Iter rod maker (delivery ratio). There have been various approa::-,es to automate this process by introducing 'on- line controls (e.g. density) where-c.., 'the adjustments are still made via the delivery ratio of the blooming unit. In prax's 'this has often lead to "over control" because of the irertmass of the mechanical dr:ves L:s-3d in traditional blooming units. 'This paper intends to describe a :-,evi approach to autcmate quality c--nlrol all the 4 fl~ter rod maker. The system uses :,~ie pressure on the rion-driven pretension rollers in common two stage blooming u-..*'s to 2UtCr_1-atiCaIIy alter and ad,;ust 'the amount of filter tow entering the filter rod maznine. he system is suitable ',-*or mcder,-.'s:r;c tra&:cnai fliter rod makers but is also appli- cable fur the future generation of ..-ter rcd maklers using the "double rod" technology. 2. INTRODUCTION In today's filter rod producticn the.-a- is ~-,ardly any on line c--ntrol fcr pressure drop or weight. (Figure 2) There are several re-,:z-cns W this: First of all there is still a problem to measure +he weight of a filter rod on line but it is even more difficult to obtain an c.-, line signal for pressure drop. Weight mainly is controlled via the density of the f1';:=_:-rcc bit 'there are many more influences an that signal such as product moisture. *.iiac-etin ccntent, yield etc. For The main target value of a ';Iter, pressure drop, -:-,ere has only recently a commercially available system been developed by KOREER and FILTRONA. Another reason still is that when .ne a%d , ;ustrnents are made via the delivery ratio of the blooming unit vs. the rod ma'-.=-r 'Ll-ere is a hich risk of over control due to the inertness of the mechanical drives. And, finally, the product quality cf tc:ca-:,'s fliter to%vs did not reaily justify the huge L?4 efforts necessary into either the rreasur~ng cr L~7,e cont.-ol systems. 1) C:) It needs to be said that the qual,,i ol a filter ircd is determined by the spinning and Lr4 crim , .-,ticn of C71- ping process and that there ~5 cnly ~.:-le cznri-1 'he rod maker. C;N Xh. -10 BATCo document for Legal Services : Health Canada 21 October 1999 Dai.'y pracbce, hcv;ez.-sr. sh-cv,,s that the 'n fi,"er rcd . prcd_:::cn very often (m2ybe too often) ad!just the setting c` Ziocm:ng ~.,nit to corn-e--sate for very minor &,anoes in rct , quality This cerlta - y eads *,:; an unwanted ove.- control. This, as well as the need to further prcd-ictivity and prcd~_::' quality, de- mands 'he installa*~:cn of a re~iable auc-__'_- ;ual:,y for filter rc--4 producticn. 3. THE CURRENT CONTROL SYS7 El',1 Usuatly the first s-,ec, vjill be to adjus', -,7e .2--4e7~3:on pressure to a =nstant value v.,mccn is either Given by the mac,~Iine c- -7,ar.,_fact~;rer cr v.,hich "as been deier- n-,:ned aernpirically. (Figure 3) The vaiue -:,- -.-e p, =_,=_ns;cn p. essure ::_zpends on the tc:al cen~er oi the Fi'.ze, tov., used. 2) Th,en the ratio te- , ..'am I n the stretchin~: ro::e-s v,,:'.1 be set to achiet.-e an optimunn biccming cf the iitre-s. Again values dezen::'-~g cn the filter tow spe--:.,ication will be recomrn,ended by the tow manufacturers After having set the width of the tow L-and ;~ the -plastlicizer applicat:-.n unit (visual control) the amount of tow fed into the r::d -. Eking par, will then or.:y be altered by chanmno the delivery ratio between blc::.-,-,irz unit andrcd maker. This deiivery rat;o is then the only var,:-=:~e lte"e pressure drcp is controlled once it has drifted frcm its target value. A;I otl-,e., settincs, in --articular the pre+=-:7B:_-1- are kept conq,- ,=; it. fr. today's blooming units the non driven prs:snsic7i rcilers play an a-x'remeiy impor- tant rcle: they lor example compensa:e s11!:::nt variaticris in cr;r-,p intensity or the iniluence of increasMg distance bet.,.-n --=;e s-~;rfzce and entrarce into the rcd maker by automatically adjusting the'.- sps-3d. (Figure 4) Thus t~-_y maintain a ean = tanszn rcliers constar~t *Lens;,on onto the tow in the fir-s- zcres i,be".v pre and leed' rollers) The knowledge about this imporltant rc;e prcmpted the idea to investigate whether the preterision rc!lers could not be usnst_ fc,, takirq over the whcle ~-=nlrol over the filter rcd parameters rather than just s~:g"viariaflicns. It therefore needed to be determined ir. the pret-nnsion pressure can be altered without riskina underbloomi,-.= (yiS,J) 117Cr lc%v pretension pressures and overbloming (fly dust) for high pretensicn pr=--;zsu,,=-s. 4. RESULTS The influence of' 4he Pretension pressure cn ,-Ie fs:--ar rcc parameters -.vas determined for varicus tow spec'aications and var~c---s p-zcesz-'ng sceeds but kee-Ping the speed ra~ics within the biccrning unit as well as 44"s -_~elivs-ry r2t;o constant. The results have that in averaz-= VI-,e tow cou!d be a;tered by +/- 5% L^J when the pretens;on pressure was med`.:ed :n a rance of +/- 50%. (Fgure 5) C:D A change in 1mv ,veicht of +/- 51% is eq,_~--alent to a criance in Pressure drop of 2pprcx. +1- 121% is absolutely tc zzver '.1he usual rance nec- 1 essari, %vithin the -rccess,;ng of a t-ale :.- tc for '~,e aver-age 'Cale to bale variat;cn C11. BATCo document for Legal Services : Health Canada 21 October 1999 A C!"ange n. ore'ens=, cressure aikle~s 1,~= friction be!-een '"e --- =,s ZI . - - il 6. -1 --e speec the roile.-s. :Fic,-,,e 6) As ,X'-e speed of the prete-7,sic- -olterS consequently als-- de:e-mix- -.ne _e amcur~z of tow e-e.-na zhe machine I --- - - I I th I . - 1: ._ - Z; -a rod and ta =rrespcriding pressire drcp. Vary.;ng t~-= pretens,::,-, .:,-essura by +/- carltainly -ot ;s againsz :u-rent pr2clice. It cc--'-- hov.,evar be sho,.-.,n that starting from a for the oretensicn force th;s rance has no sionificant effect y's j 2rz.::.- dust. Fic~re 7 :Z APPLICATION'FOR QUALITY CONTROL The syst-enn of vary.nz pr-etensicn force is suitable to .:.-_ssu--_ and .veight aLl,:)MlatlCally;n =-,ibnaticn with a on line measurern=nt -,- mithe.- cne --r boh of these parameters. The procedure woul-- --_ as lollo%vs~ (Ficure 8) 1P.1hen pircduction is s:a.,ted or the specificaticn is changed, the- ;nltial se, uP of the -iooming ~.-na is car,,;-=:: cut traditionally by adjusting pretension -_-essures. rol!er ratio ana tow in the application zone. Also the initia'. adjus-mant of the Pressure drop or Ime ,::J %ve~gght will be dcne via the delivery rati.- tetween IC ooming -~jnit and r:,,d, maker, Then the =utomatic pressure control is T-_ inpL~' s:c:-,-:Is for -I can be c"-~ined by an on line density control cr t, his ccnt,,,- ~-2 newiv d-=,.aloped cressure drop contrcl' as earlier or any other suitable '11~,,en a t:aia chance is carried out it will be necessary to check. tne aC"j2l lavel cf '4he pretension presSur-= 'o nake sure the' it is still in an accepts'_-:9 r2n.=q. If this is not the c2se. a tradit.:cnai resetting of the machine must be carrie~- out. k - lu unyhow a continucs of tt,,e actual pretension pressure va~.;e '-e ~;ser 1 to assure that it is within the possible working range. T he simp.9 design cf -,ne systerr; will me'Ne it pcssible to instaii! it z.-. ex-._:::-z z:~:,ter iodd T,ankers. Cn y a few '-rd , ._,.vare changes at the rod maker are necassarj~ tire manual ~~ressure controller f1c, 'he pretension pressure needs to be exc:-27~ge-4 foCr an elec- ,rcnic pressure conturc,ler connected to a control unit. (Figu, e 9) in combiration with s~.:J'table software it will therefore also be pos-z'ble to ccnnect the system to traditional ~="cratory measuring equipment which suc=',es, fli-e- ir.Put data :he 6. APPLICATION IN COMBINATION VVITH "DOUBLE ROD" TECHNOLOGY RHONE-POULEENC deveic-ped 2 nevv generation of Filter Tow, V"e so called T1jViNT0%,%/. TWINT~-i'.' is a tcw that consists of two exac,:y idendcal each U4 having the denier p- fi , and total d3 .. 1. , I , iiament - nier required for the ---:ters, which car, be CD J cn ccnver.-:cnal b1corning units and which only aft processer ~er 1~e delivany rollers of LA tz ;nto two parts vihi-h are then fed ir,,o a *,',,at uses ,~,e blocmirc unit '"a '*dct.:`e s nct require i - rod" tec7.-CiCrv Th-? blooming unit for this type of dca- 014 ccntrc:_z .c adiust 'he sett;--7gs or the tow celrc- All ac- are mad-3 lc%v, i.9 the two halves, BATCo document for Legal Services : Health Canada 21 October 1999 p-sz,nilit, to use the rod" technolco-v in be the processing of two s=-p-aratte bales of the s,_-_ l.'.Iifh the prccedtures to control the iiiiter rod quality this wc- z ',"e use cf. either _z se - parate blooming units or one bicoming unit Whm,-e a C==rS, dr:ves etc. =.- present ir, double. T-;s of course would eliminate all the advantages 'the use TV.'!NTCW on a single ::ming ur.;,, oflefs. This problem can be solved using techrco'.-gy de- ---:ced ;n U-is pa-er If a bloom:nc unit sh2H be mo,:fie~ -.,c-,E?ss:r- of two :-:arate batess the only modifica?'cr needed is to intrz::-:=- ;-_:ension :z:.:ers, i.e. 'to split the exis*,ing roller pair xto t~,vo. (Figure 1.,,:a ,he pretensicri pressure It is thezn -cssib;-- to adjust prsssLre drop of --a filters made from each bale separat-sly so that id-entic-a. ::,c:=,,,;es for filters f--- the t\vo tows can be achieverd. Moreover it is poss~z:e t:: =n1brol the overall -~=a:~very of and therefore the pressure drop Of the filt-E-S C-- -,--tlh tLO'.VS cgether, lo adj, ust t,he ceneral level of the filter rod weight and ,, s-s-s-L!"S d,-cp. 0 CONCLUSIONS cculd be _s,".own that the variau;cr, of the pretension Pra___~. s a ::~;;fa'.a tcol for au,,cma,;c -r semi-autcrnatic quality control at a fliter rz::: m-:.Ner, Within the usual -~ge in vwhic~,, adjustments have to be made while a be!= - fi '- ::I er ,c-.-,, is pr1ccessed ---e are no -.ecative effects on ',he quality of blooming ~-ser.,ed, ararnete systerns 2,;Icv.,s a very quick response to changes in f:,*er -zd p- - --s as it !ccm:r-,-., un its. ~s -nuc~-i =sfer and more precise than the gear bcxes :- sxis;~ng, t- tern can eilher come from any z-.~!taz cn ~ine ...a input signal 'or the syst rn=nt or from off fine laboratory equipment via the respective c-!c::7nirc unit a =-n ks can e2Sily lba mcdif';ed with 'Lh SVS'_ T',-,-a control tethtno!cgy descrited is also needed whe- ".vo separate ~--Blres are .::-:--essed cn a single blooming unit using the "double r::z" Va a split cre-rension rc:!er the adjustments can be made individually `z.- ea-_h tow band-4. 0 %,nristoph Greiner, RHbNE-POULENC RHODIA: "Filts- ':,r4 rnak:rc vv;z"zut ccn- paper prasented at the Schiuchsee-Kolloquiurn In V:::2 2 3: RH6NE-F0,ULENC, Depanme-, F;.--=r Tc,.-.,, Technii:2,' E _- 1 le,, i - - 0 - pr c- es_~ IL4 s,no with K!-.F2!AF2" ON Dr Ra , ~~er Vlendebicirn, RHCNE-POULENC RHCCI,-* =4 using (_4 ~nolcgy- paper presenled a, CORESTA Cc-::-s-- n, Octcbe.- 1-9-4 rcd BATCo document for Legal Services : Health Canada 21 October 1999 A NEW APPROACH FOR THE AUTOMATIC CONTROL OF PRESSURE DROP (WEIGHT) IN FILTER ROD MAKING Thomas Leutner, Christoph Greiner RHONE-POULENC RHODIA AG Department Filter Tow On Line Control is not Standard ù On Line Measurement of Pressure Drop and VVeight is still difficult and not very reliable ù High Risk of Over-Control via Mechanical Drives n Hiah Product Quality does not justify the Installation of expensive Equipment ~_4 CD L4 L14 BATCo document for Legal Services : Health Canada 21 October 1999 Traditional AdjUstment of a Blooming Unit mAdjust Pretension Pressure mAdjust Blooming Ratio m Adjust Tow Band Width m Control Pressure Drop and Weight via Delivery Ratio fcz;-e 3 Behaviour of Pretension Roller Speed P -1 S L' e =-,-q (MIM i n) 440 . c i 420 400 - il", -W.,, 360 loc 80 60 40 20 0 CD gate height (cm. ) 01% BATCo document for Legal Services : Health Canada 21 October 1999 Rod Weight vs. Pretension Pressure .:-'cd Weight k-g) Rod Weight 7-731 1,5 2 2,5 3 Pretension Pressure (bar) Pretension Speed vs. Pretension Pressure P--:ension (rnmmin) Pretension Speed 4 d,7 r~ 5 1,5 2 2.5 3 Pretension Pressure (bar) CN (-.4 CIN L7 t_r BATCo document for Legal Services : Health.Canada 21 October 1999 Yield and Fly Dust vs. Pretension Pressure Yield usable range 0 0,5 1 1,5 2 2,5 3 3,5 Pretens~on Pressure %Ure Adjustment Procedure a traditional adjustment of pretension pressure, blooming ratio, tow width and delivery ratio n activation of automatic (semi-automatic) . 5_~ , control n alteration of pretension pressure to maintain filter parameters ON BATCo document for Legal Services : Health Canada 21 October 1999 System Design Input through On-Line or Off-Line (externai) sianal 1k, J On-Line Extr----rnal AppliCation for "Double Rod" Technology ammarata wwS:..::: M: Q*bdGS ; I , -, m ,0--- WrArols for:.:. ro r i CD FF]U 1 1,71 ~mh:: PT -. BATCo document for Legal Services : Health Canada 21 October 1999 Conclusions is It is possible to control filter parameters via the pretension pressure in the blooming unit ù The system allows a quick and precise response to changes in weight and PD ù It can be combined with on-line or off-line measuring systems ù Existing blooming units can easily be modified ù It is essential for the processing of two bales on one bloomirng unit using the "double rod" technology 0 a% U-4 U_. co BATCo document for Legal Services : Health Canada 21 October 1999