Netzsch_Gabo

Bildquelle: alle Gabo Netzsch

Sealing elements are used in technical applications to prevent mass transfer between two components or auxiliary chambers [1]. The desired property profile is achieved primarily through a variety of design options. Besides the polymer and necessary additives, the filler used also plays a crucial role in establishing a sealing element’s characteristics such as compressive strength, thermal and chemical resistance. Depending on the mounting conditions for the different sealing systems, creep and relaxation processes lead to premature mechanical aging and premature wear of the seal. If identical sealing elements are used in different applications, the aging processes in the respective application can vary considerably. Removing the seals and checking their mechanical properties is not a solution. As a rule, the seals cannot be used after disassembly. Ideal in this context would be a so-called „marker“, which „reports“ the mechanical state of the mounted seal and enables continuous „monitoring“. Ideally, as “marker” a combination of simultaneous dynamic mechanical and dielectric analysis is used.
Elastomer seal materials are mostly reinforced with electrically conductive fillers like carbon black. Under static load, the „virgin“ carbon black structures present in the rubber matrix change. The dielectric conductivity changes accordingly due to breakdown or even collapse of the carbon black network. The dielectric conductivity becomes load-dependent [2]. After the installation of the „virgin“ or new seal relaxation processes take place. The static contact pressure, corresponding to mounting conditions, relaxes into an equilibrium state. The force and thus the load on the seal decreases.
The next step is obvious. Measuring the mechanical loads at mounted seals are quite impossible. However, recording a dielectric spectrum corresponding to the steady-state stress state is quite possible at any time. Online sensors for dielectric analysis undertake the „monitoring“ task. First, a correlation between the static load and the dielectric sample response must be determined. This is done best with simultaneous dynamic mechanical and dielectric analysis. Each static load level corresponds to a conductivity spectrum. If the failure limit of the seal element is known, the associated dielectric spectrum too. The work presented here deals with the load dependence of the mechanical stress and the associated dielectric spectra.

Dielectric analysis: theoretical basics

The dielectric relaxation spectroscopy enables to determine the behavior of rubber materials subjected to an oscillating electric field. This technique identifies the microstructure dynamics of the

Fig. 1: Simplified measuring setup for dielectric relaxation spectroscopy. The electrodes are displayed in yellow. The sample is displayed in grey.

Fig. 1: Simplified measuring setup for dielectric relaxation spectroscopy. The electrodes are displayed in yellow. The sample is displayed in grey.

test samples at the molecular level through dynamics of chain segments, orientation of already existing as well as induced dipoles and transport of electrical charge carriers [2-4]. The test sample is placed between 2 electrodes connected to an electrical voltage source U(t) generating an alternating electric current I(t) as shown in Fig. 1.

Fig. 2: Oscillating electrical field E(t) (black curve) and the 
associated phase shifted electric displacement field D(t) (red 
curve).

Fig. 2: Oscillating electrical field E(t) (black curve) and the
associated phase shifted electric displacement field D(t) (red
curve).

The electric current charges the plate capacitor, whereby the electrodes are charged in opposite sense. An electrical potential is build up within the plate capacitor whose field strength E(ω) is proportional to the applied voltage, the electrode gap and the electrode area.
The response of a dielectric placed into a plate capacitor with an electric field E(ω) is the polarization P(ω). The following relationship applies:

(1)

where ε0 = 8.854 ∙ 10-12 As/(Vm) is the vacuum permittivity and ε*(ω) is the complex permittivity.
The polarization is carried out by several frequency-dependent mechanisms. At measurement frequencies ranging between 10-1 Hz and 107 Hz, a transport of free charge carriers like electrons, holes and ions may occur as well as the orientation of already existing permanent dipoles along the electric field lines. Induced dipoles can additionally be formed through heterogeneous charge distribution within one atom or simply between single atoms. However, this happens at frequencies up to 1010 Hz outside our measurement frequency range.
The application of an oscillating electrical voltage induces in rubber materials a phase shifted current because the response of the sample does not instantly occur. An oscillating electric field of the form

(2)

where E0 is the electric field amplitude and ω = 2πf is the angular frequency, generates a electric displacement field D(ω) of the form

(3)

where δ is the phase shift. Fig. 2 shows the curve progression of both the exciting electrical field E(t) and the phase shifted dielectric displacement field D(t). A proportionality between E(ω) and D(ω) in the linear regime is given according to

(4)

where ε*(ω) – the complex permittivity – can be expressed as

(5)

ε‘ (ω) and ε“ (ω) represent the real and the imaginary part respectively and are related to phase angle δ according to

(6)

The complex permittivity ε* (ω) is calculated by dividing the capacitance of a capacitor with a test sample C* by the idle capacity C0 (capacitance of a capacitor without sample):

(7)

C* is derived from the complex impedance Z* according to

(8)

Experimental Data

https://www.kgk-rubberpoint.de/24616/static-seals-provide-information-about-their-own-wear/

https://www.kgk-rubberpoint.de/24616/static-seals-provide-information-about-their-own-wear/

Sample preparation
A styrene butadiene rubber (SBR) filled with 70 phr carbon black (N 234) was prepared. SBR contains 23.5% styrene groups and is extended with 37.5 phr of highly aromatic oil. The furnace carbon black N 234 was used as reinforcing filler. The compound recipe is listed in Table 1.
The rubber compound was mixed for 8 minutes at an initial temperature of 30 °C and a maximum rotor speed of 40 rpm with an internal mixer of type Werner & Pfleiderer GK5E. SBR was mixed first for 1 minute. Carbon black N 234, zinc oxide (ZnO), stearic acid, the antioxidants 6PPD and TMQ as well as the protective wax Antilux 500 were then added. The crosslinking system – sulfur, CBS and DPG – was mixed at a reduced rotor speed of 25 rpm in order to avoid a pre-crosslinking. The mixing process was finished by homogenizing the rubber compound on a roll mill.
2 mm thick sample sheets were cured in an electrical heating press at a vulcanization temperature of 160°C. The vulcanization time was derived from the TC90-time, determined by the rheometer MDR 2000 E from Alpha Technologies. TC90-time is the time at which the real part of the torque reached 90% of its maximum.

Measurement System

The simultaneous DEA and DMA measurements were performed with the dynamic-mechanical analyzer DMA Gabo Eplexor 500 N from Netzsch Gabo Instruments, equipped with special sample holders for compression

Fig. 5: Thickness variation of the SBR sample filled with 70 phr N 234 due to increasing static load from 5 N to 45 N.

Fig. 5: Thickness variation of the SBR sample filled with 70 phr N 234 due to increasing static load from 5 N to 45 N.

measurement and a broadband dielectric spectrometer BDS from Novocontrol, shown in Fig. 3. In this combination the device is also called Diplexor 500 N. The compression clamps serve as electrodes. They are electrically isolated from the rest of the instrument in order to ensure that the dielectric properties of the SBR sample are the only aspect being measured.
Dielectric measurements under several static mechanical loads were performed in compression mode at room temperature. The dynamic voltage was 1 V. The 2 mm thick samples had a diameter of 10 mm. A good surface contact between the samples and the electrodes was ensured by coating the samples with a very thin silver layer as shown in Fig. 4.
The SBR samples were subjected to varying static forces in 10 N steps from 5 N to 45 N. For each static load, dielectric spectrum was recorded in a frequency range between 100 Hz and 107 Hz.

Results and discussions

Fig. 6: Real (left) and imaginary part (right) of the complex conductivity as a function of the electrical frequency by varying the static load from 5 N to 45 N.

Fig. 6: Real (left) and imaginary part (right) of the complex conductivity as a function of the electrical frequency by varying the static load from 5 N to 45 N.

(see above) Fig. 7: Real (left) and imaginary part (right) of the complex permittivity as a function of the electrical frequency by varying the static load from 5 N to 45 N.

(see above) Fig. 7: Real (left) and imaginary part (right) of the complex permittivity as a function of the electrical frequency by varying the static load from 5 N to 45 N.

To illustrate the simultaneous mechanical and dielectric behavior of a sealing material and how the progression of mechanical damage can be characterized at the same time, dielectric spectra were recorded for the SBR sample filled with 70 phr N 234 while it was under mechanical load. Consciously, just one sample was used for the total measurement in order to avoid material inhomogeneities and different equipment settings.
The rubber matrix is electrically insulating. The carbon black N 234 is electrically conductive because its surface area has a graphitic nano-crystallite structure. It is also important to note that the carbon black amount of 70 phr is above the dielectric percolation threshold. This is an absolute prerequisite for building up a closed filler network providing the necessary electrically conductive paths.

 

Mechanical analysis

The SBR sample was subjected to a static load in compression mode. Increasing the static load is associated to a change of the sample thickness accordingly. Fig. 5 shows the correlation between the static load amplitude and the sample thickness. Increasing the static load amplitude is accompanied by a reduction of the sample thickness. A change in thickness of 10% due to mechanical loading is observed. This value correlates quite well with seal performance in real applications.

Fig. 8: Variation of the real part of the dielectric conductivity σ‘ as a function of the static force at electric frequencies fel of 100 Hz, 1 MHz and 10 MHz.

Fig. 8: Variation of the real part of the dielectric conductivity σ‘ as a function of the static force at electric frequencies fel of 100 Hz, 1 MHz and 10 MHz.

The reason lies in the fact that a continuous increase of the static load induces large deformation of the SBR samples. As a consequence, the internal friction within the SBR sample increases due to diffusion processes. The filler clusters start to break up. The filler network cannot withstand the mechanical load and it is progressively destroyed. The sample stiffness decreases and its thickness also [5-8].
It can be concluded that, the damage progression is associated with a gradual destruction of the filler network. This lead to serious consequences for the density of the electrically conduction paths within the sample.

Dielectric properties under static load

Carbon black filled elastomer composites above the dielectric percolation threshold are physically approximated by a binary system with conductive and non-conductive components. The polymer matrix is the dielectric while the filler network represents the electrical conductor. The conduction paths go through the filler clusters and the gaps between. The gaps are mostly filled with nanosized mobilized rubber layer – also known as bounded rubber – which represents a potential barrier that the electrical charge carries must pass through [2].Dielectric measurements under static mechanical loads were carried out in compression mode at room temperature. Both real and imaginary part of the complex conductivity as a function of the electrical frequency by varying the static load from 5 N to 45 N are shown in Fig. 6.
Considering the general curve shape of the real and imaginary part of the complex conductivity, two different frequency ranges can be distinguished.
Until a frequency of 104 Hz, σ‘ – the real part of the complex conductivity σ*- remains constant independently of the electrical frequency and reaches a plateau value known as DC-conductivity. This value has mainly to do with the normal diffusion or transport of the electrical charge carriers along the filler network. The only change is that σ‘ continuously decreases with increasing the mechanical static load. The reason for this is that the uninterrupted breakage of the filler network reduces the density of available conduction paths throughout the entire SBR sample and consequently the electric current density.
At higher frequencies, σ‘ becomes frequency-dependent. This area is called dielectric dispersion because the variation in the electric field E(ω) is not followed with an instantaneous change in the sample polarization P(ω). Additional relaxation processes come into play. The orientation of the already existing permanent dipoles, interface effects like Maxwell-Wagner polarization or electrode-sample polarization cannot be ruled out. All those processes are now active and contribute additionally to conductivity.
It cannot be overlooked at this point that, the frequency dependence of conductivity at higher frequencies does not correspond to a power law according to Jonscher as seen in Fig. 6. Jonscher coined the term “Universal dielectric response” which states that the real part of the conductivity σ‘ (ω) can be written as the sum of a frequency-independent value that matches with the DC-conductivity and a term varying as a power of the frequency [9]. σ‘ (ω) can be expressed as follows:

(9)

where σ0 is a constant and χ is an exponent smaller than 1.
To clarify this behavior, it is useful to consider the complex permittivity ε*(ω). It serves to quantify weak superimposed effects of electric field within dielectric materials. The complex permittivity as a function of the electrical frequency by varying the static load from 5 N to 45 N is shown in Fig. 7.
The dielectric loss ε“(ω) decreases when the electrical frequency increases. At low frequencies, a high proportion of free-moving components is available. Even the styrene and vinyl side groups of the SBR chains are able to fluctuate and to rotate. Therefore, significant heat losses are generated in the sample. This behavior supports the transport of electrical charge carriers on the filler network and the consequent conduction effect, DC-conductivity. At high frequencies, the movement of these different subcomponents is very restricted. A large number of chain segments remain unaffected and do not contribute to the sample polarization P(ω). Consequently, ε“(ω) decreases.
At frequencies above 105 Hz, the curve progression of ε“(ω) shows a hill. This is a characteristic feature of occurring relaxation process and not a conduction effect. This could refer to relaxation of one component like polymer chains or filler network or even more.
To conclude, it can therefore be stated that the variation in the real part of the conductivity σ‘ due to varying static load during the operational life of an elastomeric sealing material can be used as a smart way of monitoring the actual damage state.
Fig. 8 illustrates the frequency-dependent relationship between increasing the static loading Fstat and decrease of the real part of the complex conductivity σ‘. This relationship – attributed to the density decrease in the conduction paths within the SBR sample – allows monitoring the actual state of damage of the filler network and thus the static seal.

 

Conclusions

Mechanical analysis is the main quality control system for technical products under mechanical load. Dielectric analysis (DEA) further supports the development process due to an available very large frequency range (as compared
to DMA), which permits an in-depth molecular understanding of the inner dynamics of test samples. This valuable insight into a material’s microstructure allows conclusions to be drawn – with minimal effort – about the actual state of damage of a finished technical product during active operation, when electrically conductive fillers are used.
It was shown that the current changes in dielectric conductivity are in accordance with the state of its filler network, and hence the damage in the sealing element.
The Diplexor 500 N offers a good advantage. It permits characterization
of the dielectric properties of sealing
elements under high mechanical load, in order to determine first their properties and later their actual performance during operation.

Literature

[1] W. Tietze, „Handbuch Dichtungspraxis“, Vulkan-Verlag, Essen, 2003.
[2] J. G. Meier, J. W. Mani and M. Klüppel, “Analysis of carbon black networking in elastomers by dielectric spectroscopy,” Physical Review B, vol. 75, p. 054202, 2007.
[3] J. G. Meier, J. Fritsche, L. Guy, Y. Bomal and M. Klüppel, “Relaxation Dynamics of Hydration Water at Activated Silica Interfaces in High-Performance Elastomer Composites,” Macromolecules, vol. 42, no. 6, pp. 2127-2134, 2009.
[4] J. Fritzsche and M. Klüppel, “Structural dynamics and interfacial properties of filled-reinforced elastomers,” Journal of Physics: Condensed Matter, vol. 23, no. 3, p. 035104, 2011.
[5] H. Kawamoto, Carbon black polymer composites, New York: E.K. Sichel, M. Dekker (Eds), 1982.
[6] H. Batzer, Polymere Werkstoffe, vol. III, Stuttgart, New York: Georg Thieme Verlag, 1984.
[7] G. Kraus, “Mechanical losses in carbon-black filled rubbers,” Journal of Applied Polymer Science, vol. 39, pp. 75-92, 1984.
[8] J. Ulmer, “Strain dependence of dynamic mechanical properties of carbon black-filled rubber compounds,” Rubber chemistry and technology, vol. 69, p. 15, 1996.
[9] A. K. Jonscher, “The ‘universal’ dielectric response”, Nature, vol. 267, pp. 673–679, 1977.

About the authors

Horst Deckmann

Netzsch-Gerätebau GmbH
Schulstraße 6
29693 Ahlden, Germany

Sahbi Aloui

Netzsch Gerätebau, Ahlden