PRFEN UND MESSENTESTING AND MEASURINGKGK - PDF document

PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.deThe tested machine parameters for mastication are (table Rotor speed:  to  rpm.Mastication duration:  to  min.Fill factor: % of the mixer volume.Initial mixer temperature: °CThree tests at  rpm for  minutes are successively performed, in order to stabilize the mixer under steady state conditions. The tests are conducted with the following operating precautions:Gum cut into small cubes  x  x  cm, for each test.Introduction time in the mixer:  sec.Time  corresponds to the stroke of the piston (Fig. After the dump, temperature of the gum taken by pricking with a thermocouple.The masticated gum is then passed on a roll mill with a . mm gap, reduced to  mm in a second time, for less than one minute, with a friction ratio of /, and a fluid temperature in °C.The masticated rubbers were characterized by dynamic rheometry with a Rubber Process Analyser RPA (Dynisco Alphatechnology®), in Small Amplitude (SAOS) e.g. linear domain, and in large amplitude (LAOS) with Fourier Transform. This rheometer is strain controlled, and the torque is measured by a transducer on the upper plate. SAOS experiments were performed at °C, with a strain of %, and frequency sweep from . Hz to  Hz. LAOS experiments were performed at °C, . Hz and strain sweep from  to %. For LAOS and SAOS, measurements were performed with at least  oscillations and  minutes delay is applied for the sample relaxation before measurements.Mooney viscosities were measured with a Monsanto R- rheometer (Dynisco Alphatechnology®). All the rheological test were performed at °C, with a preheat time of one minute and four minutes of measurements, the rotor speed remaining at  rpm.Results and discussions: rheological measurementsExamples of temperature and power evolution during mixing are given in figure . The mixing times are short in order to be done in industrial factories. The steady state is reached after two minutes that’s the shorter time for our trials.The temperatures obtained at the mastication issue change a little with the rubber viscosity (Fig. ). The gum temperature measured by the mixer stabilizes, at a value which increases linearly with rotor speed (Fig. a), a shift is observed with a different fill factor. Reducing the fill factor from % to efficient for temperature control as re Processing parametersRotor speed Time Fill factor Fig. : (a) Relationship between mixer temperature and rotor speed, with fill factors . (b) Relationship between rubber temperature and power in steady state.  Stabilized Power (kW)Rubber temperature (°C) Rotor speed (RPM)Stabilized temp. (°C)\r    \f\f% VM\f% VMVMVM Fig. : Recording during trails on rubber VM H for extreme parameters: (a) temperature, (b) power.  Time (s)Temperature (°C)-      \r-\f% Time (s)Power (kW)-      \r-\f% PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.deducing the rotor speed from  rpm to  rpm.The temperature is stabilized when a balance is reached between the energy supplied b

y the self-heating of the rubber and the energy extracted by the mixer cooling system (Fig. ). Thus the link can be made with the power output and steady state temperature, regardless of fill factor, initial viscosity of the gum and of rotor speed (Fig. ). Gum temperatures were measured just after discharge, with a thermocouple probe. There may be between °C and °C difference with the one measured on the mixer. This temperature control allows us to increase the Specific Mechanical Energy (SEM), without reaching too high temperature close to rubber thermo-oxidation. Another way to reach high SME with quite low temperature could be to masticate at low rotor speed (below  rpm) for a long time (more than  min). Taking into account that our study is based on an industrial process, the mastication time could not exceed  minutes.The Mooney viscosity decrease versus time is almost linear, regardless rotation speed and type of gum (Fig. ). If we look at the relative decrease of viscosity (Fig. ) we note that under mild conditions ( rpm, and up to  minutes at  rpm) the relative loss is quite the same, whereas the decrease is more important for VM H than VM L under harder mixing conditions ( minute at  rpm or anytime at  rpm). Theses changes of Mooney viscosity with the parameters do not take in account the differences of mechanical energy due to rubber viscosity.Quantification of the mechanical contribution can be made with SME defined by Eq. . Where C is the torque, rotor speed and m the mass of gum. = . ( ) Reducing the decrease of the Mooney viscosity as a function of the Specific Mechanical Energy (SME) (Fig. ), two trends are highlighted:The viscosity reduction is more pronounced when the initial viscosity is higher.Under  kJ/kg the fill factor has no influence on the viscosity decrease, it depends only on the SME. At a fill factor of % the Mooney viscosity fits well with the SME and Mooney viscosity (Eq. ). This empirical relation based on the Mooney viscosity before mastication (VM) is simple but related to the mixer model and the fill factor. = + ( , , × ) × For SME higher than  kJ/kg, we observed different trends as the function of the fill factor, if this factor do not exceed Fig. : VM H and VM L gum at three speeds (,  and  rpm) for a fill factor of % versus: (a) Mooney viscosity ML (+) at °C and (b) Relative decrease of viscosity. Time (min)) at VM LVM LVM LVM HVM HVM H Time (min)Relative viscosity decrease VM HVM HVM HVM LVM LVM L Fig. : Relative viscosity decrease for both rubbers in function of rubber temperature. Rubber temp. (°C)Relative Mooney decrease    Mechanical breakdownThermal oxidation .\f VM . VM.\f VM. VM Fig. : Mooney viscosity for both rubbers in function of SME: comparison between the model and experimental points. EMS (kJ/kg)) at     \f% VM H% VM H\f% VM L% VM LModel VMHModel VM L PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.de%, the viscosity reach a limit value between  and  m.u. (Fig. ). These two trends as a function of fill factor imply that chain scissions do not originate only from mechanical degradation. If the viscosity drop is expressed in function of temperature (Fig. ), it ap

pears that above a certain level, chain ruptures are due to the temperature and time. These results are consistent with the literature [], the mastication efficiency mal effects. Figure schematically represents these two trends, with the U shape of total mastication efficiency. At low temperatures, the mastication is achieved by mechanical shearing if the temperature increases isoprene macromolecule relaxation time increases and mechanical scission are less effective. In the case of a fill factor of %, the temperature stabilizes at around °C, at this temperature the mastication efficiency is very poor. To reduce the viscosity below  MU, polymer chain breakdown must be thermally activated (above °C), with an unavoidable oxidation (Fig. ). However, exposure duration to a given temperature will also affect the viscosity, which explains the poorer correlation Mooney viscosity/material temperature.Dimier & Al. [] plotted the complex viscosity of a masticated rubber with a Carreau-Yassuda model with a shift factor related to the SME, this factor depending on the temperature regulation. In their work the rubber temperature ranged between °C and °C at the end of mixing, so that the viscosity decrease was due to mechanical breakdown. Recently Wortmann & Al. [] developed a model of relative loss of viscosity as a function of mechanical breakdown and thermal oxidation for higher mastication temperatures. This physical model based on experimental results describes the ‘U-Shape’ known for years [] with different SME.If we look at the SAOS measurement (Fig. ), we can observe the same shift of the elastic modulus’s decrease at low frequency according to mastication. We can easily apply a shift factor as Dimier & Al. done in their work. This exploitation could be interesting but doesn’t enhance the different mechanism of chains scission. In conclusion SAOS measurements give the same information as Mooney viscosity measurement, the moduli decreases with the increase of SME and temperature.LAOS tests give access to non-linear domain, were the moduli are not constant with the applied strain, and thus to observe gums disentanglements at large deformations. LAOS experiments are quite useful for polymer with several mesostructures. Studies performed on LAOS showed that in the non-linear domain, the measured torque is not as sinusoidal as the imposed strain but is distorted. Hyun & al. wrote a complete review on LAOS from the ’s to actual state of the art []. These distortions are specific to the architecture of the studied polymers, and depend on the amount of long branches. Fourier Transform can be used to decompose the periodic torque signal into a sum of sinusoidal signals (Eq. -). Only odd harmonics are relevant []. Where : strain ; applied strain ; : pulsation ; t: time ; G’n: n harmonics of the modulus in phase the strain (elastic part) ; G’’n : n harmonics of the modulus in phase the strain rate (viscous part). = 0 ( ) + ( ) = 1 ( ) = 1 0 2 0 ( ( ) ) () ( ) = 1 0 2 0 ( ( ) Fourier Transform rheology gives interesting results for synthetic polymers with controlled and well known structure [-]. In NR, some links between aldehydes function at the end of isoprene macromolecules and the non rubber component like fatty acids or proteins lead to associative structures like branched molecules, gel or physical crosslinks [----]. This links between isoprene molecules could be called

associative structure. In this study we propose to use the Fourier transform LAOS rheology to measure the importance of this associative structure. Leblanc & al. [] and Burhin [-] experimented this technique on NR samples. In a first time Lissajous figures (rectangular projection of periodic signal) are a good way to characterize branched or linear structures. Branched polymer gives thick end of Lissajous figures, more linear structure gives thinner end, and linear polymer shows some secondary loops []. Secondary loops appear for polymer with strong elastic non-linearity, and elastic stress ’= for strain \n  []. By testing some polymers with controlled structure and studying the mathematical conditions to obtain secondary loops, Burhin [] introduced a Long Chain Branching index (LCB index) calculated with the G’ odd harmonic according to Eq. . This index is related to the long branched chains content, we noted that this index is strain depend Fig. : Schematic mastication efficiency curve due to mechanical and thermal actions based on Ohm (). Temperature (°C)Mastication Efficiency \f   \r  \fMechanical BreakdownOxidative cleavageCombined mechanical and oxidative effect Fig. : G’ modulus versus frequency at  °C in SAOS (different mastication treatments. \f Frequence (Hz)G' and G'' Modulus (Pa), ,   VM  VMVMVM  VM  VM \r- VM - VM - PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.deand could be negative. So we use this index in order to compare the associative structure but we are aware that it has no absolute physical meaning and could be adjusted. = + × × The LCB index for both rubbers at % before mastication is around .. It is noted that for the short time mastication, the quantity of structured material (long branches, gel…) is smaller than before mastication (Fig. ). This would mean that the scission in NR occurs preferentially at these associations between isoprene and non rubber component. This destruction of the associative structure is close to the observation made by Shiibashi [] during mastication on roll mill at °C of IR with chemical gel which was destroyed into smaller gel particles or branched molecules and NR with gel due to associative structure which was destroyed into linear isoprene molecules. Ehabe & al. [] also enhance the gel destruction by mastication. The destruction of associative structures during mastication where more recently described by Nimpaiboon et al. [] by mastication of rubbers with different gel contents. Then, for longer mastication durations, a gradual emergence of new associative structures at high rotor speeds and not at low speeds happens. The trend is the same for both VM L and VM H. This increase of the branching index is not only rotor speed dependent. At  rpm, and % fill factor, despite the high level of SME there is no evidence of formation of associative structures. The formation of these structures during mastication is probably thermally activated by double bond oxidation largely described as thermal oxidation. It can also be caused by reactions with degraded non rubber constituents. According to Fig. , the increase of the LCB index at high rotor speed depends on the material temperature. We can conclude that this increase enhances the thermal oxidation during mastication at

high temperatures and we can define two domains, one for mechanical degradation and the other one for thermo-oxidation according to the literature.The observations made on LCB index, could be confirmed with Lissajous figures of three cases (unmasticated rubber, and rubbers masticated under mechanical and thermal conditions). Lissajous figures (Fig. ) show that rubbers masticated in smooth conditions (low speed, e.g. low temperature) are quite linear, as a secondary loop is present. The secondary loops on Lissajous figures for linear polymers, was described by Burhin et al. [-]. For the unmasticated rubbers and rubbers masticated at high temperature, Lissajous figures show thicker end, point out more associative structures. These Lissajous figures confirm the results obtain with the LCB index calculation.Whatever the elasticity at low strain, structured rubbers are more elastic than linear rubbers at high strain. This elasticity at high strain was also enhanced by Buhrin []. On different unmasticated rubbers which present a high level of associative structures keep a low loss factor (e.g. tan delta) at high strain (Fig. The rubbers masticated  minutes at  rpm which stay at low temperature and present Lissajous figure (Fig. ) of a linear polymer show a very high loss factor at high strain (Fig. ). This could be interpreted as a flow due to disentanglement Fig. : LCB index at °C, . Hz and % of strain in function of time with different processing parameters. Rubber Temperature (°C)LCB index \r  \f,,,\r-,\r-,VM \f%VM \f%VM  \f%VM  VM H \r \f%VM H  \f%VM H  \f%VM H  Mechanical breakdowndegradationTime Fig. : Lissajous figure at  % of strain for rubber unmasticated (), and masticated with different processing parameters for (a) VM L rubber and (b) VMH rubber. \t Shear rate (sShear stress (kPa)VMVMVM Shear rate (sShear stress (kPa)VM H VM H VM H  PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.deat high strain. The rubber masticated at  rpm which reach a very high temperature (°C), at this temperature oxidation occurs and Lissajous figures show the presence of associative structures due to thermal oxidation (Fig. ). However, this rubber have a very low viscosity ( MU) and high loss factor at low strain, the loss factor at high strain keep quite low as the unmasticated rubber (with a viscosity of  MU). Consequently the elasticity at high strain enhances associative structures which not appear by frequency sweep measurement in linear domain (Fig. ) or Mooney viscosity measurement.Finally, comparing LCB index with the G’ slopes (power law index in strain sweep: Fig. ), we get a fairly good correlation between them (Fig. ). This correlation is in agreement with the observation on elastic non-linearity made by Ewoldt and McKinley [], in fact secondary loops on Lissajous figures are obtained for visco-elastic fluids with a strong non linear elasticity. The LCB index is based on mathematical condition on the G’ harmonic ratio to obtain secondary loops or thick end of Lissajous curves. J.L. Leblanc [] integrated a strain sensitivity parameter in his model describing the complex modulus as a function of strain. The hypothesis is that the associative structures render disentanglements more difficult at high strain and reduce strain sensitivity. This sensitivity on complexes modulus does not take into account the non linearity of

the elasticity enhanced by Buhrin []. When extending this analysis to the TSR  Ivory Coast rubbers, with different botanical origin (clone) or maturation duration, (Fig. ) we observe that they follow the same trend as masticated rubbers, but with an offset between them. For the same LCB index masticated rubbers are less sensitive to strain, probably due to lower modulus in the linear domain. A point was added to this figure, it relates to a test conducted in LRCCP, with a SMR  CV masticated in an internal mixer type Rheomix HAAKE P, at high temperature with . phr dicumyl peroxide (DCP). Shortly after the addition of DCP a slight increase in torque was noticed indicating an incipient crosslinking. This masticated gum with low Mooney viscosity ( MU) shows high structure and high elasticity due to chemical crosslinking with DCP. The slight crosslinked structure, which could be considered as a high level of branched molecules, was enhanced by LAOS experiment, despite a quite low viscosity of the gum.ConclusionsThis study assesses the influence of various internal mixer parameters on the mastication of natural rubber. The results point out the two different mechanisms which reduce the viscosity of the rubber. From these results the following conclusions can be drawn:The temperature rise is due to a balance between the power loss and the cooling capacity of the mixer. Its limit in steady state will depend on the rotor speed and fill factor. It is difficult to get a viscosity lower than  MU with a fill factor of deed such a fill factor causes temperature stabilization between °C and °C, at this temperature the mastication efficiency is the lowest. In order to reduce the viscosity to  MU in our Fig. : Tan delta in classical strain sweep at °C and . Hz, in all case the fill factor is Strain (%)Tandelta   \r,,VM L VM \rVM  Fig. : G’ modulus in classical strain sweep at °C and . Hz: difference of slope in power law.  Strain (%)G' Modulus (kPa)VM L  R = ,VM-,\r = ,VM  -, = , Fig.  : G’ slope in power law compared to LCB index at %, °C and , Hz for different masticated and unmasticated rubbers from Ivory Coast :  and  GT clone, maturation  days;  PB clone maturation  days,  GT clone maturation  days LCB Index at -, , \r, , \f,,,,,,MasticatedTSRSMR  + , Pce PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.deconfiguration, it is necessary to induce thermo-oxidation.  MU NR still processable and a fill factor of in order to avoid thermo-oxidation, even % is more traditionally used for industrial compounds.A model law was derived and predicts the viscosity reduction as a function of the mastication parameters (speed / duration or energy) and of the initial viscosity of the gum. The fill factor shall be % of the useful volume, and the Specific Mechanical Energy should be around  kJ / kg. This energy can be reached within  minutes at  rpm or  rpm for  minutes. For better th

ermal control it is interesting to masticate the rubber only  minutes at  rpm.The LAOS appears to be a very good tool to distinguish, through the branching index, the thermal degradation from the mechanical breakdown on natural rubber. This test is performed very quickly, and gives additional information than only Mooney viscosity.AcknowledgementsThis study has been done in the framework of the CANAOPT project which aims to develop new grades of natural rubber with reduced and controlled variability, for damping applications. Another goal is to develop new tests to characterize raw natural rubber.The authors would like to thank the partners of the CANAOPT project: Hutchinson, Anvis, Emac, Metaldyne, Michelin, the Maine University (le Mans), and the CIRAD for the very fruitful collaborative work. Thanks to French BPI (Banque Publique d’Investissement), Elastopole, the French counties of Bourgogne (CR), Pays de la Loire (CR), Languedoc Roussillon (FEDER), Loiret (CG), Nièvre (CG) and the Pyrénées Atlantiques (CG) and the Montargis town for their financial support. Special thanks to Marie Cartault from Anvis for providing the samples of TSR  from Ivory Cost. Thanks to Didier Kessab from IFOCA for his help on the mastication of rubber. Referenceses Allen P.W., Bristow G.M. J. Appl. Polym. Sci. . Sci.  Tanaka Y., Tarachiwin L., Rubber Chem. Techech Tarachiwin L., Tanaka Y. Kautsch Gummi Kunst  (unst  ( Amnuaypornsri S., Nimpaiboon A., Sakdapipanich J., Kautsch. Gummi Kunst. ()  Amnuaypornsri S., Tarachiwin L., Jitladda T. Sakdapipanich T., J. Appl. Polym. Sci., ,  ISO - standard Rubber, unvulcanized: Determinations using a shearing-disc viscometer - Part : Determination of Mooney viscosity ().osity (). Pike M., and Watson W.F. “Mastication of Rubber. I. Mechanism of Plasticizing by Cold Mastication” J. Appl. Polym. Sci. () ()  Mullins L., Watson W.F. “Mastication: shear dependence of degradation on hot mastication”. J. Polym. Sci. . Sci.  Fries H., Pandit R.R. “Breakdown of rubbers”. Rubber Chem. Technol. echnol.  Ohm in “the mixing of rubber “; Champmann et Hall., .chapter , London, ()., (). Shiibashi T., “Gel structure characterisation of NR and IR and direct observation of individual polymer molecules by electron microscopy”, (translated by Langsley. G.) Int. Polym. Sci. and technol. . Sci. and technol.  Ehabe E. E., Bonfils F., Sainte-Beuve J., Collet A., Schué F., “Mastication of Raw Natural Rubber: Changes in Macrostructure and Mesostructure”. Polym. Eng. Sci. ()  ISO  standard Rubber, raw natural and raw synthetic Sampling and further preparative procedures. ().. (). Dimier F., Vergnes B., Vincent M., “Relationships between mastication conditions and rheological behavior of a natural rubber”. Rheol. Acta. ta.  Wortmann C., Lindner P., Dettemer F., Steiner F., Scheper T., “Mastication Behavior of cis-,-P

olyisoprene as a Model for Natural Rubber” J. Appl. Polym. Sci. () ()  Hyun K., Wilhelm M., Klein C.O., Cho K.S., Nam J.G., Ahn K. H, Lee S.J., Randy H. Ewoldt G.H. “A review of nonlinear oscillatory shear tests: Analysis and application of large amplitude oscillatory shear (LAOS)”. Prog. Polym. Sci. . Sci.  Giacomin A. J., and Dealy J.M., in Techniques in Rheological Measurement, A. A. Collyer. Chapman & Hall, Chapter , London  Wilhelm M., Maring D. & Spieb H. W. Rheol Acta, ta,  Wilhelm M, Reinheimer P. & Ortseifer M. Rheol Acta, ta,  Wilhelm M., Macromol. Mater. Eng. ()  Leblanc J.L., Pilard J.F., Pianhanuruk E., Campistron I.,Buzare J.Y. “Characterizing gum NR samples through advanced techniques” J. Appl. Polym. Sci. . Sci.  Burhin H. “FT rheology, a tool of quantity long chain branching in NR and its effects on mastication, mixing behaviour and final properties”. Kautsch Gummi Kunst, ()  Burhin H. Viscoélasticité non-linéaire, historique, pertinence, applications et applications potentielles Séminaire scientifique AFICEP, Vitry-sur seine ().-sur seine (). Burhin H. “Linear and non linear viscoelasticity” IISRP, Workshop Brussels ().shop Brussels (). Stadler F.J., Leygue A., Burhin H., Bailly C., “The potential of large amplitude oscillatory shear to gain an insight into the long chain branching structure of polymers”. In: The th ACS national meeting, polymer preprints ACS, , New Orleans, LA, USA  Ewoldt R.H., McKinley G.H., “On secondary loops in LAOS via self-intersection of Lissajous-Bowditch curves“ Rheol. Acta, ()  Nimpaiboon A., Amnuaypornsri S., Sakdapipanich J., “Role of Gel Content on the Structural Changes of Masticated Natural Rubber”. Advanced Materials Research,  PRÜFEN UND MESSENTESTING AND MEASURINGKGK www.kgk-rubberpoint.de Natural Rubber · Mastication · Rheology · LAOS · associative structure · internal mixerMastication experiments were performed on natural rubber in an intermeshing mixer by varying the process parameters. The control of rubber temperature during mixing emerged as the key factor of the process. Two different fill factors and several processes were used to reduce viscosity. LAOS measurements Eigenschaften von mastifiziertem Naturkautschuk – gezeigt durch LAOS Rheologie-Messungen Fourier Transform · Mechanische Energie · Molekulare Struktur · Naturkautschuk · Oxidation · RheologieMastikationsexperimente an Naturkautschuk wurden mit einem Mischer mit ineinandergreifenden Rotoren unter Variation der Prozessparameter durchgeführt. Die Kautschuktempera Behavior of masticated natural Rubber highlighted by LAOS RheologyAuthorsFlorence Bruno, Martin Herblot, Vitry-sur-Seine, Florian Deme, Frontonas, FranceCorresponding author:Florence Bruno LRCCP French Rubber and Plastics Research and Testing Laboratory rue AuberF- Vitry-sur-Seine Cedex, FranceE-Mail: herblot@lrccp.com Discover more interesting articles and news on the subject!        www.kgk-rubberpoint.de Entdecken Sie weitere interessante Artikel und News zum Thema