magneto-rheological elastomers . ferrites . carbonyl iron . iron oxide core-shell particles . controllable elastic modulus . particle orientation Magnetic elastomer composites based on SBR and NR have been compounded and characterized in order to observe the potential changes in mechanical properties in the presence of an external magnetic field. Micron-sized carbonyl iron and manganese zinc ferrite particles of a variety of sizes as well as nano-sized particles composed of an iron oxide were used as magnetic fillers. Mechanical and rheological experiments show that the modification of the particle surface by silane increases the hardness and tensile strength of the materials significantly. The composites filled with carbonyl iron exhibit a better controllable elastic modulus than samples filled with manganese zinc ferrite or iron oxide.

Magnetorheologische Elastomere mit schaltbaren Eigenschaften Magneto-rheologische Elastomere . Ferrite . Carbonyleisen . Eisenoxid- KernSchalepartikel . Schaltbarkeit . Partikelausrichtung Es wurden magnetische Elastomerkomposite auf Basis von SBR und NR hergestellt und charakterisiert um mögliche Änderungen der mechanischen Eigenschaften in der Gegenwart eines extern angelegten Magnetfeldes zu beobachten. Als Füllstoffe wurden mikroskalige Carbonyleisenpartikel und ManganZink-Ferrit-Partikel unterschiedlicher Teilchengröße sowie Nanopartikel auf Eisenoxidbasis verwendet. Mechanische und rheologische Untersuchungen zeigen, dass durch die Modifizierung der Partikeloberfläche mit Silan die Härte und Zugfestigkeit der Materialien signifikant erhöht werden. Die mit Carbonyl-Eisen gefüllten Materialien weisen eine bessere Schaltbarkeit auf als die Komposite, die Mangan-Zink-Ferrite oder Eisenoxid enthalten.

Figures and Tables: By a kind approval of the authors.

2

KGK ·

Magneto-rheological elastomers with switchable mechanical properties Magneto-rheological elastomers (MREs) belong to the class of smart materials. They consist of an elastomeric matrix and magnetic particles as fillers which have been added to a viscoelastic polymeric material prior to cross-linking [1,2] leading to an alignment of particles into chains along the magnetic field lines [35]. After the matrix curing process, the ferromagnetic particles are fixed in their positions. The viscoelastic or dynamicmechanical properties of the cured composites are altered rapidly and reversibly by an applied magnetic field. The sizes of ferromagnetic particles range from a few to a few hundred micrometers [6]. In the case of a poor dispersion, the chains consist of agglomerated particles. In addition to the particle size diameter and distribution of particles, the magneto-rheological effect is also influenced by temperature and saturation magnetization [5]. Pure iron exhibits the highest saturation magnetization of the known elements and it also has a high permeability and low remanent magnetization, provided high, short-term inter-particle attraction is maintained [7]. As reported in literature, the quantity of particles varies between a few and 50 % by volume [8]. The high iron concentration influences the long-term stability of MREs [9]. The magnetic field induces dipole moments within the particles which tend to achieve minimum energy states. Particle chains with collinear dipole moments are formed and the curing of the polymeric host material fixes the chains in place [3]. The aligned particles can form separate chains, three dimensional simple lattice structures with separate chains or more complex structures with multiple interaction points between the particles [10]. A model to calculate the magnetic interaction energy between neighbouring particles in a chain was developed by Jolly et al. [4]. The energy per unit volume can be derived from this formula. A similar model which highlights the role of particle size has recently been applied by Alshuth et al. [5]. It under-

lines the influence of particle diameter on the shear modulus of the composite when an external magnetic field is applied. A variety of elastomers such as urethane elastomer, silicone rubber, NBR, NR, EPDM, cis-1.4 polybutadiene and particles such as carbonyl iron and barium hexa-ferrite have been explored for MRE preparation [3,5-9,10-13]. The selection of a suitable matrix material is important when considering the possible applications and long-term stability of the MRE. Previous investigations have shown that temperature and magnetic flux density exercise an influence on the particle orientation [5]. Both hard magnetic (strontium ferrite) and soft magnetic (manganese zinc ferrite and carbonyl iron powder) have been used in order to show that elastomer composites with hysteresis-free and ideal soft magnetic behaviour can be compounded [5]. However, there is a dearth of information about the controllable elastic modulus of composites containing iron oxide particles or manganese zinc ferrites. MREs offer many potential applications including variable stiffness components, damping elements, high strain actuators and sensing applications [4,14,18]. In automobile industry circles some research has recently been devoted to this subject [14,15,17]. The goal of this study is to fabricate elastomeric composites which are magnetically active but also exhibit a suf-

Authors C. W. Karl, Hanover, J. McIntyre, Ireland, T. Alshuth, M. Klüppel, Hanover Corresponding author: PD Dr. Manfred Klüppel, Eupener Str. 33, D-30519 Hanover, E-Mail: Manfred.Klüppel@ DIKautschuk.de www.kgk-rubberpoint.de

ficient mechanical strength. With these requirements in mind, the present work is devoted to elastomer composites consisting of natural rubber and SBR containing super-paramagnetic nano-sized iron oxide particles, namely magnetite, micro-sized ferromagnetic ferrite particles of different mean sizes and micro-sized carbonyl iron particles. The iron oxide and iron particles exhibit a thin layer of silica which was functionalized with silane causing a chemical reaction that changed the surface of the particles and allows for a chemical bonding with the rubber matrix. By this procedure, the dispersion was improved significantly. Samples were fabricated from unfilled natural rubber and SBR; filled natural rubber with homogenously dispersed particles and aligned particles. To produce aligned particle chains inside the elastomer, samples with iron particles were subjected to a magnetic field during the curing process.

Free energy density of magneto-rheological elastomers

The stress-strain response of magnetorheological elastomers, in most cases consisting of soft magnetic particles dispersed in a non-magnetic rubber matrix, can be evaluated from the strain dependent part of the free energy density which has two parts: the entropic contribution of the polymer network chains and the potential energy density of the magnetic particles.

( )

( )

( )

F ε µ = Fel ε µ + Fmag ε µ

(1)

with  being the (external) strain of the rubber sample in the spatial direction . In the simple case of a polymer network consisting of linear elastic Gaussian chains it holds:

( )

Fel ε µ

G = 0 2

∑ (λµ 3

µ =1

2

−3

)

(2)

Here G0 = kB T is the shear modulus of the polymer network with  being the chain density, kB is the Boltzmann constant and T is temperature. = 1 + X is the (internal) strain ratio of the polymer chains in the spatial direction  and X is the hydrodynamic strain amplification factor. The local strain of the chains is enhanced due to the presence of rigid magnetic particles. The application of a strain amplification www.kgk-rubberpoint.de

1

Figure 1: Schematic view characterizing the interaction of magnetic dipole  moments mi (red arrows) pointing into the direction ofan external field H . The  separation vector rij and the orientation angle θ ij between two arbitrary particles i and j are indicated.

factor is normally restricted to infinitesimally small strains. An extension of this concept to larger deformations is possible by taking into account that both the volume of the polymer matrix and the whole rubber sample remains constant. This is demonstrated in the work of Lorenz [19] where a more general tube model approach for the description of rubber elasticity is described. For a more sophisticated treatment of the stress-strain response of filler-reinforced elastomers, a more complex free energy density is required which exhibits characteristic stress softening and hysteresis effects, instead of equ. (2). This is realized by the Dynamic Flocculation Model [19-24] and it has also been implemented into an FEM algorithm [25,26]. The strain dependent free energy contribution of the magnetic particles is given by the sum of the potential energy of all particles:

( )

Fmag ε µ =

1 V

N

∑U j (rij (ε µ )) 

(3)

j =1



 Here, rij (ε µ ) is the strain dependent vector separating the centres of particle i and j, N is the particle number and V is the sample volume. Uj is the potential energy of the j-th particle in the field of all other particles [27,28]:  µ µ U j rij ε µ = − 0 r 4π

( ( ))

(

)(

) (

)

       2 3 m  i ⋅ rij m j ⋅ rij − mi ⋅ m j rij    5 i =1  rij   N

∑



(4)

where 0 is the permeability of the vacuum and r is the relative permeability of the rubber matrix. For the non-magnetic

rubber matrix it holds that r = 1. Equ. (3) for the free energy density can be simplified by assuming identical magnetic particles with dipole moments pointing  into the direction of an external field H . Then Uj is independent of j and   mi = m j ≡ m = V p M with Vp = /6dp3 being the volume and M the magnetization of the particles. This yields: 2

c µ0 Vp M 2  Fmag (ε µ ) = c ⋅U j (rij (ε µ )) = 4π

N

i =1

(

 1 − 3 cos 2 θ  ij  3 rij 

∑

) (5)  



where c is the number density of the particles and θ ij is the orientation angle between the particles with respect to the direction of the external magnetic field (compare Fig. 1). Equ. (5) indicates that the potential energy of the j-th particle increases with the square of the particle volume and its magnetization. It varies with the orientation angle θ ij of the j-th particle to all other particles i. The potential energy exhibits a minimum if θ ij = 0, π for all i = 1, N, i.e. when all particles are aligned  in the direction of the external field H . It exhibits a maximum if θ ij = ± π / 2 for all i = 1.... N, i.e. if all particles are arranged in a plane perpendicular to the direction of the external field. Accordingly, a magnetic dipole force results, for example, between randomly distributed particles which tend to form chains by minimizing the free energy of the system. However, the migration of the particles leads to a local deformation of the rubber matrix which increases the elastic free energy part. This results in an equilibrium deformation,eq of the sample which increases with the magnitude KGK ·

3

1 Table 1: Compositions of samples filled with manganese zinc ferrites, nanoparticles (Magsilica 50-85) and carbonyl iron CC particles; The following abbreviations are used: ref.: reference sample (unfilled), MZF: man-ganese zinc ferrite, FeOx: iron oxide (Magsilica 50-85), CC: carbonyl iron type CC, s: silanised, o: plasticizer (oil), MF: magnetic field during vulcanization sample

polymer

SI 69

oil

NR ref. NR-MZF-2105

NR CV50 100 100

Buna VSL 2525 -

particle type FeOx ferrite CC 50 -

-

-

NR-MZF-2077 NR-MZF-0994 NR-MZF-5193 NR-FeOx-50 NR-FeOx-100 NR-FeOx-200 NR-FeOx-200-oil SBR-no MF SBR- MF SBR - s, no MF SBR - s, MF

100 100 100 100 100 100 100 -

100 100 100 100

50 50 50 -

1.63 3.25 6.5 6.5 13 13 13 13

30 -

50 100 200 200 -

400 400 400 400

(

rij ,µ (ε ) = rij ,µ (0 ) 1 + ε µ

2 Magneto-rheological measuring cell (Physica MCR 501)

Rheometer Plate Shear Gap (2,4 mm) Magnetic Flux Guide Non-Magnetic Steel Current Flow Coil Ferro-Magnetic Core Magnetic Field Lines

(a)

(b)

Figure 2: Plate-plate geometry of the measuring system MCR 501(a); schematic representation of the measuring system with magnetic field lines (b).

of the magnetization. This deformation which may be a change in the dimensions of the sample is denoted magnetostriction, and is of interest for sensor applications. It requires a high saturation magnetization of soft magnetic particles, which lose their magnetization almost completely if the field is switched off. The magnetization behaviour of soft magnetic particles can be approximated well by the empirical FröhlichKennely equation [29]:

M =

(

)

MS µ p −1 H

(

)

MS + µ p −1 H

(6)

where MS is the saturation magnetizati-

4

KGK ·

on, H is the external magnetic field strength and p is the magnetic permeability of the particles. The stress-strain response induced by the magnetic dipole moment of the particles and the entropy of the elastic rubber matrix can be evaluated from the free energy density given by equ. (1), (2) and (5) by differentiation with respect to the strain:

( )

σ µ εµ =

dF dε µ

(7) M = const

To perform this derivative it is necessary to estimate the strain dependence of the  distance vector rij ε µ . For sufficiently

( )

low concentrations it can be assumed that the particle displacement is affine and the distance vector transforms in accordance with the external deformation of the sample:

)

(8)

For high filler concentrations this equation fails due to steric hindrance effects and clustering of particles. These clusters will deform elastically in the stress field of the rubber and will finally break with increasing strain delivering the wellknown stress softening and hysteresis behaviour of highly filled rubbers [19-26]. Recently, equ. (7) has been solved for an arrangement of particles on a regular rectangular lattice with different spatial distributions (chainlike, isotropic and planar) at low concentrations by referring to the affinity condition equ. (8) [30]. In particular, the equilibrium strain ,eq has been estimated numerically. For any lattice structure it was found that the sample is uniaxially compressed along the direction of the external field. This compressive magnetostriction effect increases with the magnetization and is more pronounced for an anisotropic sample (containing particles that were aligned into chains) than an isotropic sample. In addition, the shear modulus and the Young‘s modulus of magneto-rheological elastomers have been estimated. The shear modulus increases with increasing magnetization in all considered cases, and is again more pronounced for the anisotropic samples than for the isotropic ones. However, the tensile strength may decrease with increasing magnetization for aligned and homogenous arrangements of particles but increases for a planar distribution of particles.

Experimental Sample preparation

A series of samples consisting of a natural rubber matrix filled with micro-sized manganese zinc ferrite particles of a variety of sizes (supplied from Tridelta Hartferrite GmbH), as well as nano-sized particles (supplied from Evonik Industries, grade Magsilica 50-85) were fabricated. The samples contained 50 phr of ferrite particles and 100, 200 or 400 phr of iron oxide particles. For one sample containing 200 phr of iron oxide particles, the plasticizer WM 450 was added to the mixture. In the Magsilica 50-85 samwww.kgk-rubberpoint.de

ples, the amount of silane added was adjusted to the amount of filler particles. 400 phr of silica coated carbonyl iron particles (BASF grade „CC“) were used as magnetic filler in SBR (Buna S-SBR 2525) with silane Si69 as a coupling agent. The vulcanization system used was a conventional sulphur curing system. The recipe was: 100 phr rubber, 3 phr zinc oxide (ZnO), 1 phr stearic acid, 1.7 phr sulphur, 2.5 phr CBS and 1.5 phr N-isopropyl-N‘phenyl-paraphenylenediamine (IPPD) as an antioxidant. The compositions of the composites are summarized in Table 1. The magnetic particles were mixed into the rubber in a Brabender internal mixer. The vul-canization system was added in a final mixing procedure on a roller mill. Samples were vulcanized at 160 °C for 10 minutes under a pressure of approximately 300 bar. Vulcanized samples containing aligned particles were fabricated according to a procedure described in [5]. Specimens were stamped out from sheets of cured material for use in stressstrain (Zwick Z 010 universal test sys3

4

www.kgk-rubberpoint.de

tem), and Dynamic Mechanical Analyzer (DMA) experiments. The morphological characteristics of the particles were taken from the technical specifications given in the product information literature of Evonik industries and Tridelta industries. Scanning Electron Microscope (SEM, model EVO® MA10) imaging and digital light microscopic investigations were carried out in order to check dispersion of particles and filler and alignment of the carbonyl iron particles.

Rheometric investigations

The samples were subjected to an oscillatory shear strain of 0.5 % deformation, at a frequency of 10 Hz, in a plate-plate rheometer (Anton Paar Physica MCR 501) as shown in Fig. 2. Measurements were carried out at 65 °C. The current controlling the magnetic field was switched on and off at intervals, and was increased in steps of 1 A (0.2 T until a maximum of 5 A (1 T) was reached. An inductor generates the magnetic field which circulated through the Figure 3: SEM micrograph of a cross section of a NR sample containing 200 phr of aligned carbonyl iron particles

shear gap vertically from the ferro-magnetic core (see Fig. 2 b).

Particle analysis and dispersion of the composites Characterization of composites containing carbonyl iron particles

The carbonyl iron particles (type CC) used in this study were mixed with S-SBR (VSL 2525-0). The size distribution of these particles obtained from BASF and is relatively small with a mean particle size of 3.5 m. The particles are coated with a thin silica layer allowing for the application of a coupling agent (silane). Rubbers such as NR and SBR are non-polar and hydrophobic. The surfaces of the micro-sized iron particles are cast with OH groups. Silanisation of the particles modified the particle surface so that it became non-polar. This leads to an im-proved dispersion and a chemical coupling of the particles with the rubber matrix. Fig. 3 shows the alignment of carbonyl iron particles (type SM) in NR obtained by applying a magnetic field during the curing process.

Characterization of manganese zinc ferrite composites

For the magneto-rheological investigations of melt samples, manganese zinc ferrite particles with particle sizes ranging from 5 to approximately 150 µm and having the same chemical composition were employed (see Table 2). Ferrite particles were selected because they exhibit a higher initial permeability than carbonyl iron particles. Fig. 4 shows cross section pictures of the NR composites containing manganese zinc ferrite particles which were obtained by digital light microscopy. Sample “MZF-2077” contained relatively small particles which were well-dispersed in Figure 4: Cross section of NR composites obtained by digital light microscopy containing manganese zinc ferrite particles: sample “MZF-2077“ with small particles (a) and sample „ MZF5193“ with particles having a broad particle size distribution (b).

KGK ·

5

noparticles. The outer silica shell (see Fig 5 b) of the particles was 3 nm thick and enclosed the iron oxide core. Most of the particles consist of magnetite and maghemite which have a similar spinel structure. Both of them are ferrimagnetic.

the elastomer matrix (see Fig. 4 a). An elastomer composite with particles of a very broad size distribution is shown for comparison (see Fig 4 b).

Characterization of nanoparticles (Magsilica 50-85)

Results and discussion

The nanoparticles had a diameter ranging from 5 to 90 nm; more than 80 % of them showed a size between 10 und 25 nm (see Table 3). Fig. 5 a shows the particle size distribution of the Magsilica na-

Characterization of mechanical properties

Concerning the Shore A values of the elastomer composites, an increase was

2 Table 2: Mean particle sizes of manganese zinc ferrite particles sample MZF-2105 [µm] MZF-2077 [µm] MZF-0994 [µm] MZF-5193 [µm] mean particle size

5

8

28

150

5

observed with rising amount of Magsilica (see Table 4). The hardness of the manganese zinc ferrites were lower than 50 Shore A due to the very low magnetic filler loading. Comparing the Magsilica with the manganese zinc ferrite sample, the former exhibits a higher Shore A value. This can be related to the smaller particle size, a better dispersion of particles and chemical bonding of particles to the polymer matrix by silane. The stress-strain curves for each of the NR samples containing MagSilica and manganese zinc ferrite particles can be seen in Figs. 6a and 6b, respectively. Ultimate strain values decrease from 600 % in the unfilled NR to 200 % in the 200 phr filled samples as depicted in Fig. 6a. All samples, except from those with the highest magnetic particle loading of 200 phr, show high tensile strength values between 450 % and 600 %. The tensile strength for these samples is similar with about 25 MPa. The sample with 100 phr exhibits the best ultimate properties, with maximum stress values of about 28 MPa at 450 % strain. Concerning the composites with manganese zinc ferrite particles, strain  as well as stress  increases with decreasing particle size. Samp3 Table 3: Composition of iron oxide particles (MagSilica 50-85) phase amount [%] grain size [nm]

Figure 5: Size distribution of the MagSilica nanoparticles (a); TEM micrograph of Magsilica particles with an iron core and a silica shell (b).

amorphous hematite magnetite

10-25 < 15 55 ±5

40 - 90 15 - 50

maghemite

35 ±5

10 - 40

6

Figure 6: Stress-strain curves of NR/FeOx composites containing MagSilica nanoparticles (a) and NR/MZF com-posites with 50 phr manganese zinc ferrite particles (b).

4 Table 4: Shore-A values of NR samples filled with Magsilica and ferrite particles NR ref. NR-FeOx-50 NR-FeOx-100 NR-FeOx-200 NR-MZF-0994 NR-MZF- 2077

NR-MZF- 5193

NR-MZF- 2105

42,2±0,1

46,9±0,2

50,4±0,1

6

KGK ·

55,2±0,1

70,1±0,2

81,1±0,1

46,5±0,1

47,9±0,2

www.kgk-rubberpoint.de

le „NR-MZF- 5193“ (Table 2), having a very broad particle size distribution, exhibits the lowest ultimate tensile stress and elongation at break values. The best ultimate properties are found for the sample NR-MZF-2105 with the smallest particle size. Stress train curves of the SBR composites filled with 400 phr of carbonyl iron CC particles are shown in Fig. 7. Both silanised and non-silanised samples vulcanized with and without magnetic field are compared, as indicated. As expected, silanised samples reveal significantly higher stress values, because the particles are better dispersed and bonded to the rubber matrix. However, there is no clear trend that particles aligned in the magnetic field show different stress values in comparison with isotropic samples. The strong scattering of data for the silanised samples indicates a heterogeneous morphology probably due to an insufficient particle distribution. The low stress values of the non-silanised samples can be related to an detachment of the polymer matrix from the filler surface due to an insufficient polymer-filler coupling.

7

Figure 7: Stressstrain curves of SBR composites containing 400 phr carbonyl iron CC particles.

8

Dynamic mechanical investigations

The composites filled with the MacSilica nano-particles have been investigated by dynamic mechanical temperature sweeps. The results are depicted in Fig. 8. Fig. 8a shows the storage modulus measured under shear, being proportional to the elastically stored energy, and the loss modulus, which is proportional to the dissipated energy during the deformation cycles. Both quantities increase as the amount of magnetic particles is increased. This behaviour can be related to the formation of a reinforcing filler network [20,21]. It is typical for highly reinforcing fillers and does not differ significantly from other nano-fillers, such as carbon black or commer-cial precipitated silica. The increase of the storage modulus with magnetic filler level can be correlated with the rising hardness values (see Table 4). The ratio between the storage and loss modulus, termed loss factor or loss tangent tan , is depicted in Fig. 8b. It expresses a relative quantity for the energy dissipation during deformation in comparison with the stored energy of the strained rubber. The addition of filler to the rubber results in a decline of the loss factor in the glassy www.kgk-rubberpoint.de

Figure 8: Storage modulus G` and loss modulus G`` of the NR-Magsilica samples vs. temperature (a); variation of the complex modulus G* and tan  with temperature (b).

regime around the glass transition temperature TG -55 °C (maximum of G‘‘). At higher temperatures the value of tan  increases with filler loading. In addition, the complex modulus G* is shown in Fig. 8b which is equal to the square root of the squared sum of G‘ and G‘‘.

Magneto-rheological investigations

In order to verify the effects of alignment and anisotropy respectively, magnetic switching effects of the composites containing Magsilica, ferrite particles and carbonyl iron particles were investigated under oscillatory shear in a plate-plate rheometer (Anton Paar MCR501). The results for various uncross-linked samples are shown in Fig. 9. The time dependent behaviour of the shear storage modulus G‘ for an alternating switched magnetic flux density B of the uncross-linked NR/MagSilica composites is shown in Fig. 9a. As expected, G‘ increases as the volume fraction of particles increases, while the addition of 30 phr softener leads to a significant-

ly lower value of G‘ which is only slightly higher than the value for 50 phr. It is found that G‘ is almost independent of time, indicating that almost no flocculation takes place at the chosen moderate temperatures of 65 °C. In addition, the storage modulus G‘ of the composites with nanoparticles do not exhibit any significant switching effects. This is in clear contrast to the manganese zinc ferrite samples depicted in Fig. 9b which show pronounced switching effects. Obviously, the largest particles show the strongest switching effects. The envelope of G‘ shows a pronounced increase with time, especially for the samples „NR-MZF-5193“ and “NR-MZF-0994“. This indicates that a magnetically supported flocculation takes place due to the strong attractive dipole forces even at moderate temperatures. With increasing time the G‘ values of the smaller particles decline slightly which is probably related to the observed thermal heating of the samples by the magnetic field which reduces the viscosity of the samples. KGK ·

7

5 Table 5: Relative changes (G‘/G0‘) of MRE storage modulus of manganese zinc ferrite and carbonyl iron sam-ples, G0‘: minimum of storage modulus when the external flux density reaches 1 T, G1‘: maximum storage modu-lus at 1 T G0' G1' G' G'/G0' sample NR-MZF-2105 NR-MZF-2077

[MPa] 0,319 0,313

[MPa] 0,32 0,316

[MPa] 0,001 0,003

[%] 0,31 1,04

NR-MZF-0994 NR-MZF-5193 S-SBR, no MF S-SBR, MF S-SBR/silane, no MF S-SBR/silane, MF

0,316 0,32 1,58 2,72 1,19 2,20

0,32 0,327 1,62 3,07 1,29 2,39

0,004 0,007 0,04 0,35 0,10 0,19

1,19 2,19 2,40 12,83 7,98 8,64

9

Figure 9: Effect of switching of magnetic flux density B on the time dependent storage modulus G’ at 65 °C for uncured NR samples filled with different amounts of Magsilica (a) and 50 phr manganese zinc ferrite particles of different size, as indicated (b).

10

Figure 10: Effect of switching of magnetic flux density B on the time dependent storage modulus G’ at 25 °C for cured S-SBR samples containing 400 phr carbonyl iron CC particles.

The reason for the different switching behaviour of the NR/MagSilicaand manganese zinc ferrite samples is mainly due to a size effect as predicted by the pre-factor of equ. (5). In the theoretical section it was shown that the

8

KGK ·

magnetic dipole strength increases linearly with the particle volume implying that the magnetic interaction energy increases with the squared particle volume Vp = /6 d3 with d being the particle diameter. Furthermore, the

average interaction energy decreases with the third power of the mean distance between adjacent particles, i.e. ~ Vp2 / 3. Since the mean particle distance varies with the volume fraction of filler  and the diameter d as = d (/6 -1/3- 1) one obtains for the average interaction energy: ~ d3 (/6 -1/3- 1)-3. This indicates that the strong dependence of the magnetic interaction energy between adjacent particles on particle size is not compensated by the larger distance between the particles if the same volume fraction of small and large particles is compared. Instead, for a fixed volume fraction of filler  the average interaction energy increases with the third power of the particle diameter d. Therefore the magnetic dipole interaction between the micronsized particles is much stronger compared to the nano-particles. This ex-plains the different switching behaviour of the NR/MagSilica- and manganese zinc ferrite samples shown in Figs. 9a and 9b. Since the dipole strength of soft magnetic manganese zinc ferrite particles can be switched on and off by the external magnetic field (equ. (6)), the attraction of the particles changes substantially leading to the observed switching behaviour of G‘ in Fig. 9b. The switching sensitivity of the storage modulus G‘ and the relative changes (G‘/G0‘) of the micron-sized manganese zinc ferrite and carbonyl iron samples is summarized in Tab. 5. The effect of switching of the magnetic flux density B = 0 r H on the time dependent storage modulus G’ for cross-linked S-SBR samples containing 400 phr carbonyl iron CC particles is depicted in Fig. 10. The current direction has been changed as described in [12]. This type of magneto-rheological elastomer composite has shown promising switching effects in previous investigations [5, 12]. They are tested here to analyse the responses of hard and soft magnetic materials. Both silanised and unsilanised samples are investigated. It is well known that the modification of the particle surfaces by silane improves dispersion significantly and supports the bonding at the interface between particle and matrix. This decreases the modulus of the composites and implies improved ultimate properties. However, the switching ability may be reduced because the mobility of the particles declines upon silanisation (Fig. 10). In www.kgk-rubberpoint.de

particular, for the samples without silane alignment of particles in the presence of a magnetic field leads to a six-fold increase in switching ability as expressed by the ratio G‘/G0‘ (see Table 5). For the silanized samples, one observes even a small reduction in switching ability if the systems cured with and without magnetic field are compared. A clear trend is observed regarding alignment of silanised and non-silanised samples. In both cases, the modulus increases by almost a factor two and a more pronounced switching of the modulus G‘ is observed.

Conclusions

Magneto-rheological materials with various magnetic filler particles have been compounded. Both the mechanical properties and the switching ability of elastomer composites containing iron particles reveal promising effects with respect to practical applications. Concerning the alignment of particles in an electric field, there are some important items which have to be taken into account with respect to filling level, viscosity and functionalisation of particles.

Acknowledgements

This work was supported by the Zentrales Innovationsprogramm Mittelstand (ZIM-grand KF2351101HM9). Tridelta, BASF and Evonik industries are gratefully acknowledged for the provision of particle batches.

www.kgk-rubberpoint.de

References

[1] J. de Vicente , G. Bossis, S. Lacis, M. Guyot, J. Magn. Magn. Mater. 251 (2002) 100. [2] D. Wang, J-S. Chen, L. Sun, Finite Elem. Anal. Des. 39 (2003) 765. [3] G. Y. Zhou, Smart Mater. Struct. 12 (2003) 139. [4] M.R. Jolly, J.D. Carlson, B.C. Munoz, T. A. Bullions, J. Intell. Mater. Syst. Struct. 7 (1996) 613. [5] T. Alshuth, M. Ramspeck, R. H. Schuster, B. Halbedel, F. Zschunke, Kautschuk Gummi Kunstoffe 60/9 (2007) 448. [6] M. Lokander, B. Stenberg, Polym. Test. 22 (2003) 677. [7] Y. An, M. T. Shaw, Smart Mater. Struct. 12 (2003) 157. [8] M. Farshad, M. Le Roux, Polym. Test. 23 (2004) 855. [9] M. Lokander, T. Reitberger, B. Stenberg, Polym. Degrad. Stab. 86 (2004) 467. [10] M. Lokander, B. Stenberg, Polym. Test. 22 (2003) 245. [11] A. Boczkowska1, S. F. Awietjan, R. Wroblewski, Smart Mater. Struct. 16 (2007) 1924. [12] J. McIntyre, T. Steinke, A. Maslak, P. Wagner, M. Möwes, T. Alshuth, R. H. Schuster, S. Jerrams, 7th European Conference on Constitutive Models for Rubber (ECCMR) (2011). [13] D. Bellusová, T. Alshuth, R. H. Schuster, M. Myndyk, V. Sepelák, Kautschuk Gummi Kunstoffe 64/11-12 (2011) 22. [14] W. Niedermeier, J. Fröhlich, H.D. Luginsland, Kautschuk Gummi Kunststoffe 55 (2002) 7. [15] S. Bedmarek, Applied Physics A 68 (1999) 63. [16] M. Farshad, A. Benine, Polym. Test. 23 (2004) 347. [17] J. D. Carlson, M. R. Jolly, Mechatronics 10 (2000) 555.

[18] M. Khoo, C. Liu, Sensors Actuators A 89 (2001) 259. [19] H. Lorenz, J. Meier, M. Klüppel, Chap 2 in D. Besdo, B. Heimann, M. Klüppel, M. Kröger, P. Wriggers and U. Nackenhorst (Eds.), „Elastomere Friction: Theory, Experiment and Simulation“, Lecture Notes in Applied and Computational Mechanics, Vol. 51, Springer, Berlin, Heidelberg, New York, ISBN: 978-3642-10656-9 (2010). [20] M. Klüppel, Adv. Polym. Sci. 164 (2003) 1. [21] T. A. Vilgis, G. Heinrich and M. Klüppel, „Reinforcement of Polymer Nano-Composites“, Cambridge University Press, Cambridge, New York, ISBN 978-0-521-87480-9 (2009). [22] H. Lorenz, M. Freund, D. Juhre, J. Ihlemann, M. Klüppel, Macromol. Theory Simul. 20 (2011) 110. [23] H. Lorenz, M. Klüppel, J. Mech. Phys. Solids 60, No. 11 (2012) 1842. [24] H. Lorenz, M. Klüppel, G. Heinrich, Z. Angew. Math. Mech. 92, No. 8 (2012) 608. [25] M. Freund, H. Lorenz, D. Juhre, J. Ihlemann, M. Klüppel, Int. J. Plasticity 27 (2011) 902. [26] M. Freund, J. Ihlemann, H. Lorenz, D. Juhre, M. Klüppel, Kautschuk Gummi Kunststoffe 64/4 (2011) 50. [27] L. D. Landau, E. M. Fifshitz, „Lehrbuch der theoretischen Physik“, Band 2: „Klassische Feldtheorie“, p. 119, Akademie Verlag Berlin (1977). [28] J. D. Jackson, „Classical Electrodynamics“, 3rd Ed., p. 832, John Wiley and Sons, New York, (1998). [29] J . L. Arias, V. Gallardo, F. Linares-Molinero, A. V. Delgado, J. Col. Interface Sci. 299 (2006) 599. [30] D. Ivaneyko, V. P. Toshchevikov, M. Saphiannikova, G. Heinrich, Macromol. Theory Simul. 20 (2011) 411.

KGK ·

9