© Strojni{ki vestnik 50(2004)5,277-290 © Journal of Mechanical Engineering 50(2004)5,277-290 ISSN 0039-2480 ISSN 0039-2480 UDK 658.513:004.94 UDC 658.513:004.94 Strokovni ~lanek (1.04) Speciality paper (1.04) Relativne pre~ne vibracije valjev tiskarskega stroja, ki pritiskajo drug ob drugega preko gumijaste obloge The Relative Transversal Vibrations of Printing-Press Cylinders that are Pressed Against Each Other via an Elastic Blanket Vytautas Kazimieras Augustaitis - Nikolaj [e{ok V prispevku predstavljamo rezultate analitične, računalniško podprte raziskave relativnih prečnih vibracij valjev v rotacijskem ofsetnem stroju za obojestransko tiskanje. Predstavljamo postopek izvirne metode, s katero smo izvedli raziskavo in ki temelji na računalniški simulaciji. Tiskarski stroj sestoji iz naslednjih valjev, ki se, v času delovanja tiskarskega stroja, vrtijo in pritiskajo drug ob drugega prek tanke (približno 2 mm) elastične, gumirane tkanine (gumijasta obloga): dveh valjev za ploščo, na katerih sta pritrjeni tiskarski šabloni, in dveh valjev z gumijasto oblogo, med katerima drsi papirni trak, potreben za tiskanje. V poteku raziskave smo ocenili upogibe valjev, pa tudi elastičnost in dušenje gumijaste obloge ter lezajnih enot valjev. Ugotovili smo, da te vibracije povzročajo spremembe v pritisku med valji, s čimer zmanjšujejo tudi kakovost tiska. Potem ko smo raziskali značilnosti prostih in vsiljenih relativnih vibracij valjev, smo se odločili, da bomo intenzivnost vibracij zmanjšali z uporabo valjev z identično dinamično odzivnostjo. Predstavljamo podatke, pridobljene z ekperimentalnim raziskovanjem, ki kažejo vpliv vibracij pritiska med valji na kakovost tiska. © 2004 Strojniški vestnik. Vse pravice pridržane. (Ključne besede: tiskanje, ofset, stroji tiskarski, vibracije) In this paper we present the results of a computer-aided analytical investigation of the relative transversal vibrations of the cylinders in a web-offset double-sided printing press. An original method for carrying out the investigation, using a computer simulation, is presented. The printing press consists of the following cylinders that rotate during its operation and are pressed against each other via a thin (about 2 mm) elastic, rubberized cloth (called a blanket): two plate cylinders, with printing forms attached to them, and two blanket cylinders, with a paper tape moving between them for the printing. During the investigation, the deflections of the cylinders as well as the elasticity and damping of the blanket and the cylinders’ bearing units were assessed. These vibrations caused changes in the pressure between the cylinders, thus reducing the quality of the prints. After investigating the features of the free and forced relative vibrations of the cylinders, it was decided to reduce the intensity of the vibrations by using cylinders with identical dynamic responses. Data from the experimental investigation that show the impact of the variations of the pressure between the cylinders on the quality of the prints are presented. © 2004 Journal of Mechanical Engineering. All rights reserved. (Keywords: printing, offset printing, printing press, vibrations) 0 UVOD V pričujočem prispevku predstavljamo raziskavo relativnih prečnih vibracij med valjema za ploščo in valjema z gumijasto oblogo v rotacijskem ofsetnem stroju za obojestransko tiskanje, ki uporablja metodo računalniške simulacije za zmanjšanje intenzivnosti teh vibracij. Med postopkom delovanja tiskarskega stroja se odtisi prenesejo s površine valjev, prevlečenih s tanko plastjo (približno 2 mm) gumirane tkanine (gumijasta obloga), na papirni trak, ki se pomika skozi stroj. Odtisi se prenesejo iz INTRODUCTION In this paper we present an investigation into the relative transversal vibrations between the plate and blanket cylinders in a web-offset double-sided printing press using a computer-simulation method in order to reduce the intensity of these vibrations. During the operation of the printing press the prints are transferred onto a paper tape, which is moving through the printing press, from the surface of the cylinders that are covered with a thin (approximately 2 mm) rubberized cloth (blanket). The gfin^OtJJIMISCSD 04-5 stran 277 |^BSSITIMIGC Augustaitis V.K., [e{ok N.: Relativne pre~ne vibracije - The Relative Transversal Vibrations tiskarskih šablon (navlaženih s tiskarskim črnilom), ki sta pritrjeni na valja za ploščo, na gumijasto oblogo. Vsi valji se vrtijo in pritiskajo drug ob drugega po celotni dolžini. Valji so nameščeni tako, da se vsi vrtijo in pritiskajo drug ob drugega prek elastične gumijaste obloge: na sredini tiskarskega stroja sta dva valja z gumijasto oblogo, ki pritiskata drug proti drugemu (med njima drsi trak papirja), medtem pa valja za ploščo pritiskata proti drugi strani valjev z gumijasto oblogo. Ta tip tiskarskega stroja je zelo pogost in v nadaljevanju so navedene podrobnosti njegovega delovanja. Pri tem načinu tiskanja je treba določiti pritisk med vrtečimi se valji, ki pritiskajo drug ob drugega prek gumijaste obloge ([1] do [3]). Če valji vibrirajo, se spreminja tudi njihov pritisk. Posledica vibracij je poslabšanje kakovosti odtisov in s tem tudi zmanjšanje produktivnosti tiskarskega stroja. To pomeni, da je vibriranje, ki spremlja postopek tiskanja, resen problem; da bi ga razrešili, smo uporabili metodo računalniške simulacije. Doslej so bile opravljene že mnoge raziskave vibracij rotorjev (gredi, valjev itn.), ki se vrtijo v ležajih [4]; vendar pa te raziskave niso vključevale proučevanja vibracij rotorjev, ki pritiskajo drug ob drugega prek tanke elastične tkanine. Opravljene so bile, na primer, študije valjčne opreme, ki vsebujejo kovinske liste (tu tanek kovinski list drsi med vrtečima se valjema), vendar pa rezulatov analiz teh pojavov ne moremo prenesti na tiskarsko panogo [5]. Pričujoča študija poleg raziskav vibracij valjev prikazuje tudi to, kako lahko spremenjen pritisk med valji učinkuje na kakovost tiskarskih odtisov. Rezultati naše raziskave vibracij valjev so vsesplošno uporabni. Naše ugotovitve lahko prenesemo tudi na primere valjev rotacijskih ofsetnih tiskarskih strojev, ki se po številu valjev in njihovi namestitvi sicer razlikujejo od primerov, opisanih v pričujoči študiji. 1 VPLIV PRITISKA VALJEV NA KAKOVOST TISKANJA Ni nam uspelo najti kakršnih koli zanesljivih in podrobnih eksperimentalnih podatkov o vplivu pritiska valjev na kakovost ofsetnega tiskanja, kar je preprečevalo, da bi ocenili vpliv relativnih prečnih pomikov valjev stroja, ki jih povzroča vibriranje, na kakovost tiskanja. Zato smo se odločili izvesti eksperimentalno raziskavo vpliva pritiska na kakovost ofsetnega tiskanja. S strojem Heidelberg Speedmaster 52-2-P [1] smo, pri različnih pritiskih, tiskali dvobarvne slike (rumena in črna), skupaj s stopenjskim nadzorom sive lestvice in usmerjenimi črtami. Pritisk smo spreminjali s sočasnim spreminjanjem razdalje med osema valja za ploščo in valja z gumijasto oblogo ter osema tiskovnega valja in valja z gumijasto oblogo. Ta razdalja, l, je ^BSfiTTMlliC | stran 278 prints are transferred onto the blanket from the printing forms (moistened with printing ink) that are attached to the plate cylinders. All the cylinders are in rotation and are pressed against each other along their generatrices. The cylinders are positioned in such a way that all of them rotate and press against each other via an elastic blanket: there are two blanket cylinders pressed against each other in the middle of the press (the paper tape moves between them), and the plate cylinders are pressed against the other sides of the blanket cylinders. This type of printing press is very common, and more details are provided below. With this type of printing the pressure between the rotating cylinders that are pressed against each other via the blanket must be fixed ([1] to [3]). If the cylinders tend to vibrate, this pressure varies as well. The result of vibration is a deterioration in the quality of the prints and a reduction in the productivity of the printing press. This means that the problem of vibration during printing is a serious one, and so we have used a computersimulation method in an attempt to solve it. There are many investigations of the vibrations of rotors (shafts, cylinders, etc.) rotating in bearings [4]; however, the vibrations of rotors that are pressed against each other via a thin elastic cloth were not studied during these investigations. For example, there are some studies on metal-sheet rolling equipment (a thin metal sheet moves between rotating rollers); however, the problems discussed are not related to the printing trade [5]. In this study, in addition to investigating the vibrations of cylinders, we have looked at how varying the pressure between the cylinders can have an effect on the quality of the prints. The results of our investigation into the vibrations in cylinders can be universally applied. Our findings can be applied to the cylinders of weboffset printing presses where the number of cylinders and their layout differ from those described here. 1 THE INFLUENCE OF PRESSURE ON THE PRINTING QUALITY We were not able to find any reliable and detailed experimental data on the influence of pressure on the quality of offset printing. This makes it difficult to evaluate the influence of the relative transversal shifts of the cylinders in the press, which are caused by vibration, on the quality of the printing. Therefore, an experimental investigation of the influence of pressure on the offset-printing quality was carried out. Two color images (yellow and black) with a control-step gray scale and narrow lines were printed on a Heidelberg Speedmaster 52-2-P press [1] at various pressures. The pressure was changed by altering the distance between the axles of the plate-and-blanket and the blanket-and-press cylinders simultaneously. This distance, l, is expressed in mm: Augustaitis V.K., [e{ok N.: Relativne pre~ne vibracije - The Relative Transversal Vibrations izražena v mm: = 0mm in pomeni, da se površine valjev le rahlo dotikajo (med njimi ni pritiska); povečani X pa označuje povečanje relativnega pritiska. Normalen pritisk tiskanja je dosežen, ko je X = 35mm. Čeprav ta metoda ne omogoča, da bi določili absolutne vrednosti pritiska, pa omogoča, da razumemo povezavo med prečnimi pomiki valjev in kakovostjo tiskanja. Za določitev kakovosti tiskanja smo uporabili naslednja kriterija: (a) natančnost poltonske reprodukcije (odvisnost optične gostote odtisa D od optične gostote izvirnika D , izmerjene na stopenjski sivi lestvici; (b) spremembo širine usmerjene črte in velikosti zaslonskih točk. Optično gostoto smo merili z merilnikom gostote Macbeth T-297 (z modrim filtrom za rumeni tisk) z natančnostjo ± 0,02B. Velikost elementov smo izmerili z mikroskopom (povečava: 24x , točnost: ±0,01 mm). Rezultati tega preizkusa kažejo na znatne spremembe optične gostote tiska, ki jih povzročijo različne razdalje med valjema v raziskovanem obsegu 8 do 55 mm, in ki posledično vodijo do znatno spremenjenega pritiska. Znatno se spreminjata tudi velikost točk na rastru in širina usmerjene črte. Dobljeni rezultati so prikazani na diagramih 1, 2 in 3. Dp , B 1,50 0 0,15 0,30 0,45 0,60 0,75 0,90 1,05 1,20 1,35 1,50 Do , B Sl. 1. Spreminjanje optične gostote rumene nadzorne skale v odnosu do sprememb X, v razdalji med valjema z gumijasto oblogo in valjema za ploščo: 1- razdalja = 8 mm; 2- razdalja =10 mm; 3- razdalja = 19 mm; 4- razdalja = 35 mm; 5- razdalja = 50 mm; 6- razdalja = 55 mm; 7- idealna reprodukcija poltonov; Do in D sta optični gostoti izvirnika in odtisa p Fig. 1. The variation of the optical density of the yellow control scale versus changes, l, in the distance between the blanket and the plate cylinders: 1, distance = 8 mm; 2, distance =10 mm; 3, distance = 19 mm; 4, distance = 35 mm; 5, distance = 50 mm; 6, distance = 55 mm; 7, the ideal reproduction of halftones; Do and D are the optical densities of the original and the print X = 0mm means that the cylinders’ surfaces are only touching (no pressure); and increasing X means increasing the relative pressure. The normal printing pressure is achieved when X = 35mm. Although this method does not allow us to know the absolute values of pressure, it does allow us to relate the transversal shifts of the cylinders to the printing quality. The printing quality was characterized as follows: (a) by the accuracy of the halftone reproduction (the dependence of the optical density of the print D on the optical density of the original D , measured p on the step gray scale; and (b) by the change in the width of the narrow line and the size of the screen dots. The optical density was measured with a Macbeth T-297 densitometer (with a blue filter for the yellow print) with an accuracy of ± 0.02B. The size of the elements was measured using a microscope (magnification: 24x , accuracy: ±0.01 mm). The results of the experiment show that the changes in the optical density of a print, which are caused by changes in the distance between cylinders in the investigated range 8 to 55 mm, and consequently changes in the pressure, are considerable. The size of the raster points and the width of a line vary considerably as well. Examples of diagrams showing the obtained results are presented in Figs. 1, 2 and 3. Dp, B 10 20 30 40 50 l, mm Sl. 2. Odvisnost optične gostote (rumena nadzorna skala) od sprememb X, v razdalji med valjema z gumijasto oblogo in valjema za ploščo: 1 - polje N1 odtisa nadzorne skale (D = 0,00); 2 - N2 (0,15); 3 - N3 (0,30); 4 - N4(0,45); 5 - N5(0,60); 6 - N6(0,75); 7 - N7(0,90); 8 - N8(1,05); 9 - N9(1,20); 10 - N10(1,35); 11 -N11(1,50) Fig. 2. The dependence of the optical density (yellow control scale) on the changes, l, in the distance between the blanket and plate cylinders: 1 - field N1 of the control scale print (D = 0.00); 2 - N2 (0.15); 3 - N3 (0.30); 4 - N4(0.45); 5 - N5(0.60); 6 - N6(0.75); 7 - N7(0.90); 8 - N8(1.05); 9 - N9(1.20); 10 - N10(1.35); 11 - N11(1.50) | lgfinHi(s)bJ][M]lfi[j;?n 04-5_____ stran 279 I^BSSIfTMlGC Augustaitis V.K., [e{ok N.: Relativne pre~ne vibracije - The Relative Transversal Vibrations 0,100 0,095 0,090 0,085 0,080 0,075 10 20 30 40 50 l, mm Sl. 3. Odvisnost širine posamezne črte od spremembe l, v razdalji med valjema z gumijasto oblogo in valjema za ploščo Fig. 3. The dependence of the width of a separate line on the change, A, in the distance between the blanket and the plate cylinders 2 PREDMET RAZISKAVE Raziskovali smo absolutne in relativne prečne vibracije valjev z gumijasto oblogo in valjev za ploščo pri rotacijskem ofsetnem tiskarskem stroju in se pri tem posebej osredotočili na relativne vibracije med valji, z namenom, da bi zmanjšali njihovo intenzivnost. Naše proučevanje je vključevalo naslednje sklope: - preučevanje prostih prečnih vibracij valjev; - preučevanje vsiljenih vibracij, sproženih kinematično z netočnostjo krogljičnih ležajev v valjih in z vibracijami okvira tiskarskega stroja ter z netočnostjo delovnih površin valjev (ovalnost itn.); - preučevanje vibracij, ki jih povzročata zaponki gumijaste obloge na vrtečih se valjih, ki pritiskajo drug ob drugega. Da bi raziskali ta problem, smo razvili splošno metodo analogno-digitalnega proučevanja prečnih vibracij valjev z gumijasto oblogo in valjev za ploščo pri rotacijskem ofsetnem tiskarskem stroju. Da bi bili rezultati raziskave čimbolj pomembni, smo preučevanje opravili na pogosto rabljenem ofsetnem tiskarskem stroju za obojestransko tiskanje s štirimi valji, od katerih imata dva valja gumijasto oblogo in med njima drsi papirni trak, druga dva pa sta valja za ploščo (tiskarski stroji s šestimi ali tremi valji se prav tako uporabljajo, čeprav bolj redko; in metoda ter računalniški program, uporabljena v naši raziskavi, se lahko prenašata tudi na te stroje). Shema tiskarskega stroja je prikazana na sliki 4. Stroj sestoji iz valjev z gumijasto oblogo, (2) in (3), ki po celotni dolžini pritiskata drug ob drugega in imata delovno površino obloženo z plastjo elastičnega gumijastega materiala (6). Papirni trak (5) drsi med dvema valjema z gumijasto oblogo. Valja za ploščo sta prikazana z (1) in (4). Premeri delovnih površin vseh valjev so enaki in valji se vrtijo prek zobnikov z enako vrtilno hitrostjo. Valji se vrtijo na krogličnih ležajih (7). 2 OBJECT OF THE INVESTIGATION The problem we examined was the absolute and relative transversal vibrations of the blanket and plate cylinders of a web printing press, paying particular attention to the relative vibrations between cylinders in order to reduce their intensity. This study included: - an examination of the free transversal vibrations of the cylinders; - an examination of the forced vibrations, excited in a kinematic way by the inaccuracy of the ball bearings in the cylinders as well as the vibrations of the frame of the press, the inaccuracies of the working surfaces of the cylinders (ovality, etc); - an examination of the vibrations caused by the blanket locks of the blanket cylinders on the rotation of cylinders that are pressed against each other. In order to investigate the problem a universal method for an analog-digital examination of the transversal vibrations of the blanket and plate cylinders of web printing presses has been developed. To make the problems under discussion directly relevant, a widely used double-sided offset printing press with four cylinders, consisting of two blanket cylinders with the paper tape running between them, and two plate cylinders, was chosen as the subject (printing presses with six and three cylinders are also used, although more rarely; however, the method and the software used in this work, also apply to them). A schematic diagram of the printing press is shown in Fig. 4. It consists of the blanket cylinders (2) and (3), pressed against each other, with their working surfaces coated with a layer of elastic rubber-like blanket material (6) along the generatrices of the cylinders. The paper tape (5) runs between the blanket cylinders. The plate cylinders are (1) and (4). The diameters of the working surfaces of all the cylinders are the same, and the cylinders are rotated with the same rotational speed via gears. The cylinders rotate on ball bearings (7). VH^tTPsDDIK stran 280 Augustaitis V.K., [e{ok N.: Relativne pre~ne vibracije - The Relative Transversal Vibrations a) b) Sl. 4. Tiskarski stroj (a - skupina valjev, b - razgrnitev pogleda v smeri A) Fig. 4. The printing press (a - cylinders’ group, b - unfolded view in direction A) 3 DINAMIČNI IN MATEMATIČNI MODEL TISKARSKEGA STROJA Za potrebe raziskovanja prečnih vibracij valjev smo izdelali dinamični model preučevanega sistema. Shema tega modela je prikazana v diagramu 5. Z uporabo modela lahko ocenimo prečne krivine valjev in odstopanja, ki jih le-te povzročajo, elastičnost krogličnih ležajev v valjih (upoštevajoč dušenje), nenatančnosti ležajev ter elastičnost in dušenje gumijastih oblog. Delovne površine valjev (z napakami ali brez njih) so razdeljene v valjaste končne elemente in mesta, na katerih so montirani ležaji, imajo obliko prisekanega stožca. Gumijasti oblogi sta simulirani s pomočjo ločenih, elastičnih elementov, skupaj z dušenjem; ti elementi povezujejo zaključke končnih elementov dveh sosednjih valjev. Elastičnost gumijastih oblog je nelinearna. Za to raziskavo vibracij je elastičnost linearizirana v materialih svoje statične deformacije, ob tem ko valji pritiskajo drug ob drugega. Togost in dušenje ločenih elastičnih elementov, ki simulirajo gumijasto oblogo in ležajne enote, sta izračunani na podlagi literature [6] do [9]. Kroglični ležaji in njihova nenatančnost so simulirani na poenostavljen način, tako da so njihove značilnosti v približku enake značilnostim elastičnih in dušilnih elementov (koeficienta togosti in upora sta k in h z ustreznimima indeksoma). Nenatančnost ležajev je simulirana s kinematično vzbujevanimi elementi (ti so povezani v zaporedje z elastičnimi in dušilnimi elementi); njihova velikost SG,k se spreminja glede na vnaprej določeno zakonitost (slika 5). Kinematično vzbujanje sistema prek dna okrova je simulirano enako kakor nenatančnost ležajev. Razlikujejo se le specifične oblike njunega kinematičnega vzbujanja &k. Netočnosti delovnih površin valjev (odstopanja od idealne površine valja) so simulirane s kinematično vzbujevanimi elementi, 3 A DYNAMICAL AND MATHEMATICAL MODEL OF THE PRINTING PRESS For the investigation of the transversal vibrations of the cylinders a dynamic model of the system under examination was set up. The scheme of the model is presented in Fig. 5. The model evaluates the transversal bends of the cylinders and the deviations caused by them, the elastic shifts of the ball bearings in the cylinders (including damping), the inaccuracies of the bearings, and the elasticity and damping of the blankets. The working surfaces of the cylinders (with or without holes) are divided into cylindrical finite elements, and the places where the bearings are fitted have the shape of a truncated cone. The blankets are simulated by discrete springtype elastic elements with damping; these elements connect the ends of the finite elements of two adjacent cylinders. The elasticity of the blankets is non-linear. For this investigation of the vibrations the elasticity is linearized in the media of its static deformation pressing of the cylinders against each other. The stiffness and damping of the discrete elastic elements that simulate the blanket and the bearing units are calculated on the basis of references [6] to [9]. The ball bearings and their inaccuracy are simulated in a simplified way by approximating them with elastic and damping elements (their coefficients of stiffness and resistance are k and h, respectively, with the corresponding indexes). The inaccuracy of the bearings is simulated by kinematic excitation elements (they are connected in series with the elastic and damping elements), and their size dG,k varies according to the set regularity (Fig. 5). The kinematic excitation of the system via the foundations is simulated in the same way as the inaccuracy of the bearings. Only the specific expressions of the kinematic excitation dG,k differ. The inaccuracies of the working surfaces of the cylinders (the deviations from the ideal surface of a cylinder) are simulated by stran 281 Augustaitis V.K., [e{ok N.: Relativne pre~ne vibracije - The Relative Transversal Vibrations 1./ ' I,J *l..i*L.. Sj,7""-\ **.j '*i.J Sl. 5. Shematični prikaz splošnega dinamičnega modela tiskarskega stroja; y1 do y4, x2, x3 so posplošene koordinate, ki določajo elastične linearne pomike srednjih delov valjev pa tudi smeri vibracij valjev, izbranih za namen raziskave Fig.5. Schematic diagram of the general dynamic model of the printing press; y1 to y4, x2, x3 are the generalized coordinates defining the elastic linear shifts of the middle parts of the cylinders, which also show the directions of the vibrations of these cylinders, chosen for the research spreminjajo se glede na vnaprej določene zakonitosti 3,j, Tovrstna simulacija netočnosti med valji je prikazana na skliki 5. Vsak kinematično vzbujevan element simulira netočnost površin dveh valjev, ki se na določeni točki dotikata prek gumijaste obloge. Na sliki 5 so prikazane le linearne posplošene koordinate y1 do y4, x2 in x3, ki določajo pomike osrednjih delov valjev (druge koordinate, ki določajo elastične pomike končnih elementov, niso prikazane). V dinamičnem modelu so koordinate, ki opisujejo absolutne vibracije valjev in so uporabljene v postopku našega izračuna, zamenjane s posplošenimi koordinatami, ki opisujejo relativne vibracije med valji. Zaradi zahtevnosti dinamičnega modela (dinamični model, prikazan v diagramu 5, ima 168 prostostnih stopenj) je matematični model (enačbe vibracij), ki je nastal na podlagi dinamičnega modela, oblikovan tako, da uporablja poseben algoritem in posebno programsko opremo. Za potrebe poenostavitve in avtomatizacije oblikovanja enačb preučevani sistem vibracij umetno razdelimo na preproste, neodvisne podsisteme in izdelamo pomožne dinamične modele teh podsistemov. To dosežemo z umetno “prekinitvijo” dotika robov elementov, ki simulirajo gumijasto oblogo, z enim od dveh valjev; oba valja sta sicer prek teh elementov vključena v celoten dinamični model. V začetni fazi se enačbe vibracij oblikujejo posebej za vsak posamezen podsistem, kasneje pa se enačbe vibracij, s posebnimi povezovalnimi enačbami, tudi za celoten sistem. Gumijasto oblogo po vsej dol D ini pritrdimo na ustrezni valj. Za njeno pritrditev uporabljamo kinematic excitation elements, varying according to the set regularities dx,j, dy3,j, dh,j. Such a simulation of the inaccuracies between the cylinders is shown in Fig. 5. Each kinematic excitation element simulates the inaccuracies of the surfaces of both cylinders, contacting via the blanket, at a specific point. In Fig. 5 only the linear generalized coordinates y1 to y4, x2 and x3, identifying the shifts of the middle parts of the cylinders, are indicated (other coordinates, identifying the elastic shifts of the finite elements, are not shown). In the dynamic model, the coordinates describing the absolute vibrations of the cylinders that are used in the course of the calculation are replaced with generalized coordinates describing the relative vibrations between the cylinders. Because of the complexity of the dynamic model (the dynamic model shown in Fig. 5 has 168 degrees of freedom), the mathematical model (the equations of the vibrations), which is developed from it, is formed using a special algorithm and software. In order to simplify and computerize the formation of the equations the vibration system under examination is artificially divided into simpler, independent sub-systems, and auxiliary dynamic models of the sub-systems are developed. This is achieved by an artificial “disconnection” of the ends of the blanket-simulating elements from one of two cylinders, which are connected by the elements in the overall dynamic model. Initially, the equations of the vibrations are formed separately for each sub-system, and then, using specially developed link equations, the equations of the vibrations for the overall system are formed. The blanket is fitted to a blanket cylinder along its generatrix. Special locks in the shape of thin metal VBgfFMK stran 282 Augustaitis V.K., [e{ok N.: Relativne pre~ne vibracije - The Relative Transversal Vibrations posebni zaponki v obliki tankih kovinskih ploščic. Med vrtenjem valjev delovne površine ti zaponki pritiskata druga ob drugo prek gumijaste obloge in povzročata utrip udarcev. Posledica tega je, da valja pričneta vibrirati. Kota vrtenja valjev z gumijasto oblogo, (2) in (3), sta naravnana tako, da se zaponki obeh valjev dotakneta v istem trenutku vrtenja. V tem trenutku pride do udarca, ki povzroča vibracije valjev. V pričujoči študiji izračunavamo le udarec med valjema z gumijasto oblogo (2) in (3), (lahko pa bi izračunali tudi udarca med valji (1) in (2) ali (3) in (4)). Udarec simuliramo z utripnimi silami S2k in S3k, ki delujejo vzdolž koordinat y2i in y3i (sl. 5). Oblika pulziranj je polsinusoidna (sl. 6), to je Sj,k=Aj,k.sinvt, kadar je 0aHOB, B. II., Tropun A. A., Enp6paep E. E, IIItomkob B. H. (1999) nenaraoe oGop^OBanue. Msd- eoMTVn. MocKBa. [4] Bn6pamiH b TexuHKe: CnpaBo™. Tom 3. Kojie6aHIM Maimin, kohctpvkuhö h hx MeMenroB. (1980) MamuHocmpoeHue. MocKBa. [5] Allwood, J. M., G.F.Bryant ( 1997) Fast modelling of roll stack behaviour for control applications. Journal of Mechanical Engineering Science. Proceedings Part C, Vol 211, No C5. 373-386. [6] HexMau K H., CeHKycB B. T, Enp6paep E. E. (1987) HenaTHBie Man™. Knusa MocKBa. [7] HexMau *., EenoKpticeHKO B., KpaBHVK H., IHycTBiKeBHH A., IHycTBiKeBHH M. (2000) O^ceTHBie peauHOTKaueBBie nnacnmn. Kojmwoapm. Xypnan dun nomispatpucmoe u usdamejieu. jfc 1. c.24-30. [8] JleBuna 3. M., PemeTOB fl. H. (1971) KoHxaKTHaa ^kčctkoctb Maniuu. MaiuuHocmpoenue. MocKBa. [9] AyrycTauracB. K (1994) Ananua h CHHTea Kone6aHHO Mauiun b npuMeueuHH k npuBOfly MeTannope^Kymux ctzhkob. TexnuKa. Bkjilhioc. [10] EyjiraKOB, E. B. (1954) KoJie6aHIM.Focmexusdam. MocKBa. Avtorjev naslov: profdr Vytautas K. Augustaitis profdr. Nikolaj Šešok Vilnius Gediminas Tech. University Faculty of Mechanics J. Basanavičiaus 28 LT-03224 Vilnius, Lithuania pgses@me.vtu.lt Authors’ address: Prof.Dr. Vytautas K. Augustaitis ProfDr. Nikolaj Šešok Vilnius Gediminas Tech. University Faculty of Mechanics J. Basanavičiaus 28 LT-03224 Vilnius, Lithuania pgses@me.vtu.lt Prejeto: Received: 30.4.2003 Sprejeto: Accepted: 8.4.2004 Odprto za diskusijo: 1 leto Open for discussion: 1 year VH^tTPsDDIK gfin^(5ül[Fi!]DaGC] I stran 290