Printer Friendly

Dynamic behavior of short aramid fiber-filled elastomer composites.


Short fiber-reinforced elastomers and, especially, those reinforced with aramid, are used for a broad range of applications such as tires, belts, hoses, friction products, and seals [1-3]. In general, the improvement in composite properties in terms of dynamic mechanical behavior [4, 5], heat build-up (HBU) [6, 7], and wear properties [8-11] is the reason for this broad application range.

Aramid is often chosen, because it better withstands the high shearing forces occurring in the mixing process of elastomers compared to other man-made fibers such as carbon or glass [7, 11-14], A challenge of applying aramid fibers is their poor adhesion to the elastomeric matrix. To promote adhesion often a resorcinol formaldehyde latex (RFL) dip is applied, consisting of an epoxy layer and a RFL layer [4, 15-18]. That such RFL dip can improve the fiber-elastomer interaction was shown for sulfur cured tread compounds as well as for peroxide cured hose compounds like those investigated in this study [16].

In general it is important to separate the different contributions to the final composite properties coming from the interfering parameters like short fiber content, aspect ratio, orientation, and fiber-elastomer interaction. These aspects were investigated by the authors in previous studies by using suitable model elastomer compounds [18-20]. It was found that the fiber breakage behavior was dominating the physical properties of the composites, which itself was influenced by the applied fiber surface treatment [18], the fiber type [19] or the corresponding processing conditions [20], In this study, the so gained knowledge was used to discuss some of these influence factors on the dynamic behavior of typical carbon black-filled compounds for which the fiber morphology was not directly accessible.



The two different base rubbers were ethylene-propylenediene rubber (EPDM) and natural rubber (NR). The EPDM was the base elastomer in a typical hose compound cured by a peroxide curing system. The sulfur cured truck tread formulation was based on NR. Table 1 shows the compound formulations used in this study. As usual in rubber technology the components were given in (weight) parts per hundred parts of rubber (phr). The EPDM type used was Keltan[R] 8340A from DSM Elastomers Europe B.V. with a Mooney viscosity of 80 MU (ML 1 + 4) 125[degrees]C, an ethylene content of 55%, a propylene content of 39.5%, and ethylidene norbomene content of 5.5%. The NR applied for the tread compound was a Standard Malaysian Rubber (SMR CV60) type with a Mooney viscosity of 60 MU (ML 1 + 4) 100[degrees]C. The reinforcing fillers carbon black N220 and N550 were obtained from Evonik (Germany).

The compounds were also consisting of plasticizers. The oil type added to EPDM compounds was the paraffinic oil Sunpar[R] 2280 from Sunoco and in case of NR it was the naphthenic oil Nytex[R] 840 from Nynas (Sweden). Several other antioxidants, antiozonants, and vulcanization additives were added to the compounds. ZnO with Red Seal quality was coming from Union Miniere (Belgium), and stearic acid was a technical quality grade. Polyethyleneglycol (PEG) with a molecular weight of 1800 g/mol was obtained from Merck (Darmstadt, Germany). As antioxidant-polymerized 1,2-dihydro-2,2,4-trimethylquinoline (TMQ) was supplied from Flexsys (Belgium). The peroxide curing system consisting of trimethylolpropane-trimethacrylate (TRIM) and Perkadox[R] 14/40 peroxide was obtained from AkzoNobel (Deventer, the Netherlands).

Sulfur and the accelerator tertiary-butyl-benzothiazol sulfenamide (TBBS) for the curing system of the truck tread formulation were provided by Rhein Chemie (Germany).

Both aramid fiber types applied in this study had approximately 3 mm fiber length and 12 pm diameter and were referred as Twaron[R] [poly(para-phenylene terephtalamide)] and Technora[R] [co-poly-(paraphenylene/3,4'-oxydiphenylene terephthalamide)]. The two fiber types had two different fiber treatments.

One treatment was an oily standard finish, the other one, a RFL dip, was consisting of an epoxy layer and the RFL layer. In the following the treatments will be abbreviated with "stan" and "RFL." The effective fiber volume content for 5 phr fibers was 1.4 or 2.3 vol% for EPDM or NR compounds, respectively.


The two master batches were mixed in a 150-L industrial internal mixer without curatives and short fibers. Then the curatives were added and the short fibers were oriented on a laboratory two roll mill, because this method results in higher final fiber lengths in the composites compared to incorporation in an internal mixer as shown in a previous study [20]. That the fibers were sufficient oriented was shown in another work [18] comparing the tensile testing results of a transparent EPDM model system with those of the EPDM hose compound applied in this study.

Vulcanization was carried out by compression molding with a hot press from Fortijne Grotnes BV (AC Vlaardingen, the Netherlands) at a pressure of 10 MPa and an optimum cure time [t.sub.90] + 2 min. The temperature was 170[degrees]C for EPDM compounds and 140[degrees]C for NR tread compounds. For dynamic mechanical analysis 2 mm thick rectangular specimens with the dimensions 35 mm x 10 mm were punched out in the direction of assumed fiber orientation. For HBU measurements, cylindrical samples were compression molded in cavities with a diameter of 18 mm and a height of 25 mm by piling up slices punched out of the milled sheets. The temperature and pressure were kept constant for 1 h for EPDM or 1.5 h for NR samples to secure completed vulcanization. The fiber orientation was assumed in a plane parallel to the circular surface. Specimens for measuring the fatigue growth resistance were cured in a special mold for [t.sub.90] + 2 min.


Curing Behavior

The curing behavior in terms of minimum and maximum torque, and specific curing times [t.sub.10] and [t.sub.90] was measured with a rubber process analyzer type Scarabaeus SIS V50 (Scarabaeus Mess-und Produktionstechnik GmbH, Langgons, Germany). The specimens were tested at an amplitude of 0.2[degrees] and a frequency of 0.83Hz, with temperatures set to 170[degrees]C for EPDM and 140[degrees]C for NR, respectively.

Dynamic Mechanical Thermal Analysis

The frequency-temperature dependent viscoelastic mechanical properties of the vulcanizate, i.e., storage modulus E', loss modulus E", and loss factor tan [delta] were estimated in a temperature range from -80[degrees]C to 80[degrees]C at a heating rate of 2 K/min and constant frequency. For the measurements a dynamic mechanical analyzer, EPLEXOR 2000N (Gabo Qualimeter Testanlagen GmbH, Ahlden, Germany) a 2-mm thick rectangular rubber vulcanizate strip of 35 mm x 10 mm was applied. The measurements were executed in tensile mode with the fibers oriented in load direction with strain amplitude and frequency being set to 0.5% and 10 Hz, respectively. One sample for each compound was tested. The static prestrain was set to 1%, corresponding to a low elongation where the fibers were still stretched but not pulled out.

Heat Build-Up

DIN 53533 specifies the flexometer test for the determination of the temperature rise and resistance to fatigue of vulcanized rubbers. Many rubber products, such as tires and belts, are tested by subjecting them to an oscillating load with either a constant peak stress amplitude or strain amplitude. In order to obtain good correlation between accelerated tests and in-service exposure of these products measurement of HBU and failure behavior should be carried out.

Samples were prepared according to DIN 53533-3 while the fibers in the test specimen were hereby oriented perpendicular to the direction of applied stress. One sample was tested for each material. Before every test the geometry of the specimens and their Shore A hardness were measured. Tests were executed for 25 min in compression mode and started at room temperature for the condition of constant dynamic amplitude (4.45 mm) or constant dynamic stress (0.8 MPa), respectively. When the flexometer test was finished, the final compression set and internal temperature change in every sample were measured. The permanent compression set was hereby the change in height before and after 1 h of testing in percent and was determined with a digital caliper. The final internal temperature was recorded immediately after the test by sticking a thermal sensor in the middle of the cylindrical specimen. The temperature increase during testing was the calculated as the difference between the initial and final internal temperature.

Fatigue Crack Growth Resistance

The service lifetime of rubber articles can by characterized by special testing procedures. However, such methods as the DeMattia tester obtain no proper correlation between the measured behavior and the real fatigue crack growth (FCG) resistance of the studied compounds [21].

Therefore, the influence of different orientations of the short reinforcing fibers on FCG behavior was studied by an adequate testing device, a tear analyzer (TA). Fatigue tests were carried out with notched specimens under cyclic tension. The used TA (Co. Coesfeld, Germany) was modified in order to simultaneously test notched tensile and pure shear mode specimens, the most common specimen geometries for such an analysis. However, in the current experiments only pure shear mode specimens were used due to its better reproducibility [21, 22]. Figure 1 shows the arrangement of the tensile and pure shear mode specimen in the TA device and the geometry of the pure shear mode specimen used.

EPDM hose compounds without and with 5 phr Twaron[R] standard fibers oriented parallel and perpendicular to the crack propagation direction were tested. The pure shear test specimen had the following geometrical dimensions: length L = 15 mm, thickness B = 1.5 mm, width D = 120 mm, and initial one side cut [a.sub.o] = 30 mm, according to the definition of minimal notch length [a.sub.0mm] in dependence on geometry ratio L/D [23].

The test conditions for the specimen were set to a preforce of 1 N, sinusoidal loading condition and a frequency of 10 Hz. The external nominal strain with respect to the height of specimens amplitudes was ranging from 4.9% till 19.68%.

Characterization of crack propagation in elastomers is based on a global energy balance in which the tearing energy T was introduced by Rivlin and Thomas [24] as the energy required for the creation of an unit area of new crack surfaceA. Considering that the external work is zero, i.e., the clamp distance l is constant during the process of crack propagation (fixed grip conditions), the tearing energy T can be determined from the loss of elastic strain energy W

T = dW/dA[|.sub.l=const.] = - dW/Bda[|.sub.l=const.. (1)

Considering the special case of crack propagation in a pure shear specimen without crack deflection, the tearing energy can be estimated by

T = [W.sup.*] L (2)

with [W.sup.*] being the elastic energy density ahead the crack front where the strain fields are not disturbed by the presence of the crack. This quantity can be determined by a simultaneous test of an identical unnotched pure shear specimen [22],


Considering a cyclic tensile solicitation, the crack propagation per cycle is usually plotted as a function of the maximum tearing energy attained within the cycle in a double logarithmic plot. In most cases, an extended regime with a nearly constant slope can be identified and, therefore, the relation between both quantities can be described by

da/dn =C[T.sup.m] (3)

similar to the proposed law by Paris and Erdogan [25] for stable crack propagation under cyclic loading. Here, n is the cycle number and C and m being parameters describing the crack propagation behavior within this regime.


Curing Behavior

In Table 2 the curing behavior of the investigated compounds is displayed regarding the minimum ([S'.sub.min]) and maximum torque ([S'.sup.max]), curing times [t.sub.10] and [t.sub.90]. By adding fibers generally the torque levels were increasing while the optimum curing times values were decreasing. These effects were caused by the better heat transfer through the fiber-matrix interfaces and were therefore more pronounced for the NR compounds due to the higher effective fiber volume content. Similar findings for the curing behavior of short fiber-reinforced elastomers were reported elsewhere [9, 10, 20]. Between the different fiber types and treatments no significant differences could be observed. The exception was the specimen for EPDM filled with Technora[R] fibers with standard finish, which showed unexpected high [t.sub.90], even exceeding that of the fiber-free compound.

Dynamic Mechanical Thermal Analysis

In fig 2 storage modulus E', loss modulus E", and mechanical loss factor tan [delta] versus temperature for EPDM hose compounds are displayed regarding the used fiber types and treatments. In general, E' and E" levels were rising with the addition of short fibers as reported by Ashida et al. [4], The degree of reinforcement was hereby related to the remaining fiber lengths after processing also found in transparent model compounds in a previous study [18]. The lowest residual fiber length was obtained for Twaron[R] with standard finish, followed by Twaron[R] with RFL dip where the dip obviously prevented the fiber breakage partly. Technora[R] fibers showed negligible fiber breakage leading to the highest reinforcing effect, independent of the surface treatment. However, also the RFL dip itself improved the reinforcing effect by increasing the fiber-elastomer interaction due to chemical bonds and higher surface roughness which was reported elsewhere [16, 18]. The reduction of the tan [delta] maxima around [T.sub.g] followed the same trends. Especially in the case of Technora[R]-reinforced composites, the peak was reduces most and additionally broadening and shifted to lower temperatures.


In the following, the temperature dependence of the NR tread samples was tested with respect to different applied fiber types and treatments. The results are shown in fig 3. Again the incorporation of fibers increased the levels of storage and loss modulus, but the differences between the used fiber types and fiber treatments in terms of E', E", and tan [delta] were moderate. The tan [delta] decrease was slightly higher for the samples reinforced with Technora[R] compared to the specimens reinforced with Twaron[R]. The curves for Technora[R] samples showed maxima shifted to a lower temperature and a shoulder around 25[degrees]C. The difference between the results for EPDM and NR compounds can be attributed to a possible lower fiber orientation in the NR matrix due to the higher elasticity of the matrix. Therefore, the other influence factors like different fiber breakage or fiber-elastomer interaction were less dominant.


Heat Build-Up

Flexometer tests were executed according to DIN 53533, which assumes tests under the condition of constant amplitude. This condition is valid when thinking of applications like seals. By adding fibers to these compounds the hardness was increased. This rise in hardness required a higher force to compress the fiber-reinforced specimen in comparison to the fiber-free reference, which led to a general increase in HBU and permanent compression set. This effect can be seen in fig 4. The temperature increase inside the samples during the test, the permanent compression set after testing, and their initial Shore A hardness values are displayed.

For the EPDM hose compound, the Shore A hardness increases from 57 for the reference to 69 for compounds filled with Twaron[R] or Technora[R] fibers dipped with RFL. With growing hardness also the temperature increase jumped from 98 K for the reference to around 130 K for the specimen with highest hardness. The permanent compression set after testing nearly doubled from 2.55% to over 4%. The hardness rise due to fiber addition was increased for the NR compared to EPDM compounds, because of the higher effective fiber volume content in NR. While the NR reference specimen had a Shore A hardness of 62, those filled with Technora[R] fibers obtained values close to 80. Although the sample with Twaron[R] had just a hardness of 73, it displayed the highest increase in temperature of 130 K and permanent compression set of 10%, respectively. The high permanent compression set was caused by the exceeding of a critical temperature which led to the failure of the fiber-elastomer interphase. Since the fibers were oriented perpendicular to the axe of the cylindrical specimen this led to significant height reduction, which was responsible for the higher permanent compression set.


To discuss situations for operating tire treads or timing belts, some tests were also carried out under the condition of constant stress. This led to a decrease in HBU level and permanent compression set. The harder fiber-reinforced samples achieved a lowered elongation for a constant stress, which caused generally a lower HBU in these compounds. Figure 5 shows that there is minor drop of temperature increase and permanent compression set for the fiber-filled compounds.

In case of EPDM the temperature increase in the fiber-filled samples was always lower than that of the reference. The permanent compression set, however, was on the same level or slightly above it. For NR the same trend was observed. The permanent compression set and temperature increase stayed below the level of the reference, except for the sample with filled with Technora[R] with standard finish. Beside this deviation it could be shown that the fibers had less effect on permanent compression set and temperature increase in the composites when tested in the condition of constant stress.

FCG Resistance

To predict service lifetimes of the vulcanizates under cyclic loadings FCG tests were carried out. Due to strain induced crystallization NR has itself excellent tear fatigue preventing a pronounced visible effect of short fiber reinforcement. Therefore, a short fiber-reinforced EPDM hose compound was chosen for the test in order to better demonstrate the effect of short fiber reinforcement on the fatigue crack propagation. Specimens of the EPDM hose compound with Twaron[R] fibers with standard finish in fiber direction and perpendicular to the crack propagation direction were tested for five different external strain amplitudes as well as the fiber-free reference.


Figure 6 shows data points obtained at the different strain amplitudes (amplitude values are displayed near the data points) and the linear fits in plot of the logarithmic crack propagation per cycle versus the logarithmic maximum tearing energy. It can be easily seen that crack propagation is hindered in the case of fibers oriented perpendicular to the crack propagation direction in the considered range. This is due to the local crack stopping at fibers following a fiber pull-out due to the low fiber-matrix adhesion. In contrast, in case with fibers oriented in crack direction, the crack propagation per cycle was even increased in comparison to the fiber-free reference. This was caused by the local propagation of cracks at the weak fiber-matrix interphases for Twaron[R] fibers with standard finish reported elsewhere [16, 18].



In this work, dynamic behavior of short fiber-filled elastomer composites was characterized in terms of curing behavior, dynamic mechanical behavior, HBU, and FCG resistance under cyclic loading conditions. The addition of short fiber generally increased the vulcameter torque values and decreased the curing times regardless of the fiber type and treatment. The two applied aramid fiber types showed different reinforcement behavior for the viscoelastic properties of the composites. The dynamic mechanical thermal analysis (DMTA) results of the EPDM composites reinforced with Technora[R] fibers showed the highest increase in storage modulus and drop of tan [delta] peak, while EPDM composites with Twaron[R] fibers obtained a lower reinforcement level. For the NR compounds the results were not that clear because of less oriented specimen. HBU of the composites depended significantly on the testing conditions and the Shore A hardness changes due to fiber incorporation. For testing under condition of constant strain, the hardness increase due to the fiber reinforcement led to a higher temperature increase and therefore higher permanent compression set due to failure in the fiber-matrix interphase. However, for testing under the condition of constant stress the addition of fibers showed less effect on the HBU behavior of the composites. TA measurements showed a significant influence of the fiber orientation on the crack propagation behavior. If the fibers were oriented perpendicular to the crack direction obviously crack stopping mechanisms arose. Further research will focus on the influence of fiber volume content and fiber-elastomer interface degradation on the dynamic behavior of short aramid fiber-reinforced elastomer composites.


Authors gratefully acknowledge Teijin Aramid B.V. Arnhem and DSM Elastomers B.V. Geleen, the Netherlands, for their support.


(1.) T.F. Schuler, Kautsch. Gummi. Kunstst., 7, 548 (1992).

(2.) J.F. Van der Pol, Rubber World, 32, 210 (1994).

(3.) J. Clarke and J. Harris, Plast. Rubber Compos., 30, 406 (2001).

(4.) M. Ashida, T. Noguchi, and S. Mashimo. J. Appl. Polym. Sci., 29, 611 (1984).

(5.) C. Vajrasthira, T. Amornsakchai. and S. Bualek-Limcharoen, J. Appl. Polym. Sci., 87, 1059 (2003).

(6.) B. Rijpkema, Kautsch. Gummi Kunstst., 47, 748 (1994).

(7.) F.J. Cataldo, Macromol. Sci. Phys., 47, 818 (2008).

(8.) N. Wada and Y. Uchiyama, Int. Polym. Sci. Technol., 21, 1 (1994).

(9.) R.P. Kumar and S. Thomas, Polym. Int., 38, 173 (1995).

(10.) S. Joseph, P.A. Sreekumar, J.M. Kenny, D. Puglia, S. Thomas, and K. Joseph, Polym. Eng. Sci., 50, 1853 (2010).

(11.) D. Roy, A.K. Bhowmick, and S.K. De, Polym. Eng. Sci., 32, 448 (1992).

(12.) J.E. O'Connor, Rubber Chem. Technol., 50, 945 (1977).

(13.) A.P. Foldi, Rubber Chem. Technol., 49, 379 (1976).

(14.) P.B. Jana, S. Chaudhuri, A.K. Pal, and S.K. De., Polym. Eng. Sci., 32, 448 (1992).

(15.) P.J. de Lange, P.G. Akker, S. Willemsen, and R.N. Datta, J. Adh. Sci. Technol., 27, 139 (2009).

(16.) M. Shirazi, A.G. Talma, and J.W.M. Noordermeer, J. Adh. Sci. Technol., 23, 1084 (2013).

(17.) T. Gotze, R. Hickmann, R. Fleischhauer, K.W. Stockelhuber, G. Heinrich, O. Diestel, C. Cherif, and M. Kaliske, Kautsch. Gummi Kunstst., 44, 1141 (2013).

(18.) C. Hintze, M. Shirazi, S. Wiessner, A.G. Talma, G. Heinrich, and J.W.M. Noordermeer, Rubber Chem. Technol., 86. 579 (2013).

(19.) C. Hintze, G.Heinrich. S. Wiessner, and U. Wagenknecht, Gummi Fasern Kunstst., 9, 562 (2012).

(20.) C. Hintze, R. Boldt, S. Wiessner, and G. Heinrich, J. Appl. Polym. Sci., 130, 1682 (2013).

(21.) R. Stocek, Ph.D. thesis. Dynamische Rissausbreitung in Elastomerwerkstoffen. Technische Universitat Chemnitz, Germany (2012).

(22.) R. Stocek, G. Heinrich, M. Gehde, and R. Kipscholl, Fracture Mechanics and Statistical Mechanics of Reinforced Elastomeric Blends, Vol. 70, Springer Berlin Heidelberg, Germany, 269 (2013).

(23.) R. Stocek, G. Heinrich, M. Gehde, and A. Rauschenbach, J. Plast. Technol., 1, 2 (2012).

(24.) R.S. Rivlin and A.G. Thomas, J. Polym. Sci., 10, 291 (1953).

(25.) P.C. Paris and F. Erdogan, J. Basic Eng., 85, 528, (1963).

Christian Hintze, (1,2,3) Radek Stocek, (4,5) Thomas Horst, (1,2) Rene Jurk, (2) Sven Wiessner, (1,2) Gert Heinrich (1,2)

(1) Institut fur Werkstoffwissenschaft, TU Dresden, D-01069 Dresden, Germany

(2) Leibniz-Institut fur Polymerforschung Dresden e.V., D-01069 Dresden, Germany

(3) Dutch Polymer Institute DPI, P.O.Box 902, 5600 AX Eindhoven, The Netherlands

(4) PRL Polymer Research Lab, s.r.o. 760 01, Zlin, Czech Republic

(5) Centre of Polymer System, Tomas Bata University, 760 01 Zlin, Czech Republic

Correspondence to: Christian Hintze; e-mail:

This study is part of the research program of the Dutch Polymer Institute (DPI), under project # 664.

DOI 10.1002/pen.23854

Published online in Wiley Online Library (
TABLE 1. Compound formulation in phr.


EPDM             100        100           100
N550             105        105           105
Stearic acid      1          1             1
Sunpar 2280       60        60            60
PEG              2.5        2.5           2.5
TMQ              1.25      1.25          1.25
Perkadox 14/40   7.5        7.5           7.5
TRIM             4.0        4.0           4.0

Twaron            0     5 (1.4) (a)        0
Technora          0          0        5 (1.4) (a)


NR               100        100           100
N220              55        55            55
Stearic acid      2          2             2
Nytex 840         8          8             8
ZnO               5          5             5
TMQ              1.5        1.5           1.5
6PPD              2          2             2
Wax               2          2             2
Sulfur           1.5        1.5           1.5
PVI              0.1        0.1           0.1
T waron           0     5 (2.3) (a)        0
Technora          0          0        5 (2.3) (a)

(a) Values in italic brackets are the corresponding fiber vol.%.

TABLE 2. Curing behavior of the compounds.

           [S'.sub.min]   [S'.sub.max]   [t.sub.10]   [t.sub.90]
              (dNm)          (dNm)         (min)        (min)

EPDM           0.50           5.34          0.99        10.56
Stan           0.61           5.98          0.95         8.96
RFL            0.52           6.15          0.98         8.85
Stan           0.46           6.10          1.10        11.12
RFL            0.44           5.96          0.99         8.53

           [S'.sub.min]   [S'.sub.max]   [t.sub.10]   [t.sub.90]
              (dNm)          (dNm)         (min)        (min)

NR             0.80           6.89          3.70        25.25
T waron
Stan           0.58           8.24          1.77        23.46
Stan           0.31           8.29          3.47        18.94
RFL            0.35           7.99          2.81        19.58
COPYRIGHT 2014 Society of Plastics Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2014 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hintze, Christian; Stocek, Radek; Horst, Thomas; Jurk, Rene; Wiessner, Sven; Heinrich, Gert
Publication:Polymer Engineering and Science
Article Type:Report
Geographic Code:4EUGE
Date:Dec 1, 2014
Previous Article:A novel thermoplastic polyurethane scaffold fabrication method based on injection foaming with water and supercritical carbon dioxide as coblowing...
Next Article:Preparation and characterization of porous PDMS beads for oil and organic solvent sorption.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters