The relation between thermo-oxidative aging and mechanical fatigue under repeated loading of cured elastomer compounds is investigated by referring to a wide variety of polymer types for special applications. Beside technical elastomers, typical tire tread compounds and special elastomers for high temperature resistant applications are studied. On one side, the lifetime for various loads is evaluated, whereby mechanical relaxation effects and the evolution of surface temperature are investigated. Furthermore, the internal temperature profile is calculated in relation to the measured surface temperature by referring to heat conduction theory in order to get information about the full heat history of the samples during thermo-mechanical aging. On the other side, the effect of thermal aging at various temperatures on mechanical and ultimate properties is investigated and compared to the lifetime characteristics of the samples without thermal aging. Significant differences between the various polymer types and the applied curing systems and filler types have been found. Thermal aging effects during mechanical aging are unlikely to affect the lifetime for most of the samples due to their insensitivity against thermal aging or due to relatively short aging times combined with small temperature increases. By means of the measured stress decrease during mechanical fatigue tests, several aspects influencing the lifetime of rubber compounds have been identified and analyzed.

To avoid higher cost for development of new materials and to accelerate development cycles, industrial partners are interested in more reliable prediction of abrasion properties on laboratory scale. In this work a new method to evaluate abrasion properties of elastomeric materials with the focus on tire tread compounds is presented. It is found that the abrasion behavior of elastomers depends strongly on how the wear test is carried out. Abrasion properties depend not only on material composition, but also on conditions under which the material is tested. The comparison with the standard DIN ISO 4649 method reveals a much better differentiation between materials made of various formulations. Moreover, the rating is strongly varying with the applied severity condition during the test. In this frame different abrasion contributions, e.g. abrasive wear and fatigue wear, can be identified and combined with each other to have an improved guess for real tire performance.

Elastomer blends are of high interest for tailoring materials with specific mechanical behavior. However, while for the pure components the experimental characterization of the temperature dependent viscoelastic properties is usually well feasible, doing this for the blends is often difficult or impossible. One reason is that the blend components often have different glass transition temperatures which makes the resulting behavior quite complex. In most cases when blending elastomers, heterogeneous morphologies are formed consisting of different regions with (nearly) pure components and finite interphases in between. Additional to the pure phases, especially these interphases influence the resulting viscoelastic properties significantly. For such cases, material modeling and numerical simulations can help to better understand the interactions between phases and interphases and to forecast the resulting viscoelastic properties. In this contribution we model and simulate an RVE of a binary blend consisting of natural rubber (NR) and styrene butadiene rubber (SBR). A phase field variable is used to describe the blend morphology within the simulation. The blend morphology is based on microscopic images and the dependency on the field variable is derived from an energy formulation allowing sharp and diffuse interphases between the NR and SBR phases. Both, sharp and different diffuse interphases are numerically investigated and their influences on the mechanical behavior are compared to elaborate experiments.

We consider the problem of stick-slip and noise of lubricated rubber seals that sometimes occurs in master cylinders and clutches of car braking and transmission systems. The problem is investigated for different brake fluids with a special tribological device consisting of a steel ball moving oscillatory on the lubricated EPDM samples of different roughness. The stick-slip-noise appears reproducibly for selected brake fluids exclusively for sufficient roughness of the rubber sample. Chemical analysis on the influence of brake fluids on the EPDM-material resulted in neglible effects. Based on stationary friction measurements in a broad velocity range at different temperatures and loads it is demonstrated that the stick-slip-noise problem is present for selected brake fluids delivering a pronounced negative velocity gradient of the friction force in the mixed lubrication regime, which is known to produce instable sliding. We consider a scaling approach to interrelate the frequency of stick-slipsliding with the characteristic velocity range of mixed lubrication and the maximum roughness of the rubber sample. Obviously, stick-slip-noise can be avoided for all brake fluids if the predicted frequency lies out of the sensitivity range of the human ear. From the scaling relation we conclude that this is the case if the rubber surface is sufficiently smooth

. The effect of thermally induced crystallization of high-cis polybutadiene (BR) on dynamic mechanical spectra is analyzed under different heating and cooling conditions. It is studied how the addition of carbon black and the blending with styrene-butadiene rubber (SBR) and natural rubber (NR) is affecting the crystallization. For unfilled BR an increased heating rate after fast cooling delivers less crystallization demonstrating that crystal formation requires time. With rising carbon black (CB) loading, a pronounced increase of crystallization speed is found indicating that CB supports nucleation and growth of BR crystals. In BR/SBR blends, which are detected as a single phase in dynamic mechanical spectra, crystallization is reduced for low SBR amounts and seems to disappear totally for 50/50 blends. In blends of BR, SBR, and NR the filler distribution is determined by comparing the peak heights in the loss modulus G ″ of unfilled and filled compounds. The filler located in each phase is determined from the increase of the peak height corresponding to the respective phase. It is found that the filler is mainly located in the NR phase.

Tire treads are exposed to rolling and sliding impacts leading to damages, which are known as Chip and Cut (CC) effects. A reliable prediction of the CC behaviour of new designed tire tread compounds is very difficult without field tests. In this paper a first comparison between the real tire tread wear, appearing under high severity conditions, with a fast laboratory testing method is presented. We predict the CC behaviour of a series of truck tire tread compounds when using a laboratory Instrumented Chip & Cut Analyser (ICCA) which operates under realistic practice like conditions. As a result, the carbon black filled NR/BR/SBR-blends used in this study show an identical trend of CC behaviour as observed during the tire field tests.

The mechanical response of filled rubber depends on load history, strain rate and state, temperature and even direction of previous loading. Although there is a plurality of both physical and phenomenological models, only few are able to reproduce this rich spectrum of effects. Moreover, many of them suffer from physical or mathematical inconsistencies. We present a model, which is based on physical ideas and plausible assumptions about the material’s microstructure, while being designed for high efficiency and robustness in finite element applications. It is shown by fits to extensive experimental data that it reproduces almost the full phenomenology of filled rubbers, both at low and high strains, for different deformation states and rates, holding times, and at different temperatures. The main modeling paradigm is the stress-induced breakdown and reorganization of microscopic structures which defines the time-dependent behavior of the material and allows to reproduce logarithmic relaxation effects. Moreover, its nine fit parameters evolve in a physically reasonable way under variation of filler and cross-linker content. A static limiting case of the model is derived, reducing the number of parameters and computational effort wherever necessary. Finally, a FE-implementation using computer-generated subroutines is presented and tested against experimental data of a simplified bushing under torsional, radial, cardanic and axial loading.

To determine the influencing parameters on tire tread properties and in particular the friction behaviour, it is necessary to investigate the contact conditions between rubber and rough substrates. The contact properties between two materials are affected by a large number of extrinsic and intrinsic factors such as substrate roughness, surface energy, temperature, load, sliding speed, lubrication and elastomeric material properties. In this work, the gap between a model substrate (multi-indenter) and contacting tire tread compounds with varying viscoelastic properties as a function of the normal load and sliding speed is investigated. Therefore, a new experimental technique based on laser distance measurements is used to determine the relaxation of a rubber block into cavities of a model substrate and thus the depth of rubber indentation depending on sliding distance. In this way, a validation of the contact theories and FEM simulations concerning crucial parameters such as the indentation depth and contact area could be carried out.

Filled rubber has a complex mechanical response which depends on temperature, load history and deformation rate. More specifically, the material relaxes under load and becomes softer. The degree of softening is a function of maximum load and time spent at the respective load level. Although this effect is understood rather well at low strains in terms of breakdown of the filler network, it remains an open question what happens at high strains. It may be this lack of understanding that only a few models deal with the phenomenon (Carleo, Barbieri, Whear, and Busfield 2018). An extension of a recently published micromechanical model is presented (Plagge and Kluppel 2017), which focuses on the time- and load dependent breakdown of rubberfiller structure to explain softening. The model is based on the assumption of a microscopically heterogeneous material with differently strain-amplified rubber-filler domains. The free energy density of the model is simplified to contain only elementary mathematical functions. The maximum amplification factor of the system is assumed to decrease according to a simple differential equation, whose relaxation time is reduced if local load surpasses a critical value. It is shown that the formalism automatically generates logarithmic stress relaxation as observed in experiments performed at different temperatures and deformation states. Moreover, a similar approach is used to create nonlinear Prony elements which naturally generate Payne-effect like behavior and may be used to represent viscoelastic data obtained at strains outside the linear regime.

In this work, tire tread rubber materials with simple composition based on the green tire compound are studied regarding friction properties. Using two different fillers, namely carbon black N234 and silica, friction experiments for different conditions varying speed, load and water-soap solutions are performed on two granite surfaces with different roughness. Additionally, the extended theoretical concept of Kluppel & Heinrich is used to quantify the different inputs contributing to total friction forces at dry and wet conditions. Friction results show highly nonlinear behavior of the coefficient of friction by changing different influencing factors. Both contributions, adhesion forces and hysteresis friction can be identified using simulations and quantified for experiments with pure water.

The structure and dynamics of confined polymer between adjacent filler particles is addressed, forming glassy-like polymer bridges between adjacent filler particles. They play a key role in understanding the mechanical properties of filler-reinforced elastomers. It is demonstrated that several aspects of linear and non-linear viscoelasticity of filled rubbers can be traced back to the specific rate and temperature dependent properties of these filler-filler bonds. Since they consist of immobilized polymer they are quite stiff, transmitting the stress between adjacent particles of the filler network. Accordingly, the response of the filler network is viscoelastic in nature, which has consequences for the small strain modulus and the construction of viscoelastic master curves. The pronounced non-linear behavior of filled elastomers is related to the rupture of glassy-like polymer bridges, which deform under strain and break if a critical strain is exceeded. The rupture mechanism is modelled analytically in the frame of a microstructure-based model of rubber reinforcement, denoted Dynamic Flocculation Model. It describes the mechanical response due to cyclic breakdown and re-aggregation of tender filler clusters connected by glassy-like polymer bridges. This provides a microscopic understanding of the complex stress-strain properties during repeated, quasi-static loading up to large strains, i.e. the well-known filler induced stress softening and hysteresis effects. Based on evaluated material parameters, various energy dissipation mechanisms are discussed for elastomer systems filled with carbon black (CB) and graphene nano-platelets (GNP), respectively. The latter carbon-based nano-fillers provide a promising performance regarding mechanical strength, hysteresis and tribological properties.

Softening of filled rubber under high loads, often called Mullins effect, is still far from understood. This poses major problems on modeling, simulation, composition and even analyzation of technical elastomers. The present paper shows many experiments characterizing the effect in different ways. We present data on the mechanical relaxation, temperature-induced recovery and load-dependent residual strain. The latter is analyzed under the hypothesis of plastic deformation, allowing the determination of crosslink density. Non-crystallizing (EPDM, SBR, HNBR) as well as strain-crystallizing rubbers (NR), filled with different amounts of carbon black and silica are investigated. It is found that sulfur cross-linked rubbers recover at elevated temperatures, while peroxide cross-linked rubbers do not. Furthermore, residual strain is discussed in terms of different maximum load measures, providing valuable information for modeling purposes.

In this paper improved concepts regarding the prediction of dynamic crack growth in truck tire tread compounds under praxis conditions are presented and optimized with respect to the special application "chip & cut wear" appearing under high severity conditions. The phase morphology of the carbon black filled NR/BR/SBR-blends used in this study is characterized based on DMA measurements. The effect of stress softening around the crack tip on the tearing energy is analysed by the 1 Integral method. In addition, an improved method for measuring the directional crack propagation is presented.

Strain induced crystallization is essential to the physicochemical properties of polymer materials, but is difficult to investigate, as it usually requires X-ray sources in combination with stretching machines. We improve and validate a recently developed method which allows the calculation of the crystallinity index using easily available thermography and stress-strain data. For natural rubber, the method is shown to be reproducible and delivers results quantitatively comparable to spectroscopic methods such as wide angle X-ray scattering. The incorporation of different amounts of carbon black is shown to increase the level of crystallization and to change the shape of the strain-crystallization curves. Additionally, crystallinity during partial retraction is investigated and reveals that crystallization characteristics change at sufficiently high strain.

This article features the current state of research with respect to plasma and varnished elastomer composites, which have been characterized by the sessile drop and the modified Wilhelmy balance technique. The sessile drop measurement is well suited to assess smooth and plasma modified surfaces of elastomers qualitatively. However, rough samples should be analyzed by other methods such as modified Wilhelmy balance technique having several advantages compared to the sessile drop method, e.g. high accuracy and reproducibility. It is shown that contact angle hysteresis values correlate well with roughness parameters that are obtained by white light interferometry.

It is widely believed that the extraordinary mechanical properties of natural rubber (NR) are mainly caused by its ability to crystallize at large strains. While several authors believe crystallites working equivalent to nanoscopic filler particles in terms of amplification and filler networking, recent works have identified a crystal-induced strain regulation process as a possibility to explain its outstanding properties. We present a theory that is able to quantitatively describe crystal formation and melting in stretched NR in dependence of temperature and cross-link density. The theory gives reasons for the constant crystals length observed in NR and answers the question why crystallization onset strain is independent of cross-link density. It is tested on the data set of Trabelsi (2003), Albouy (2005), and Rault (2006) reproducing stress-strain data, degree of crystallinity, and crystal sizes at different temperatures using a physically well-defined set of parameters. Additionally, a scheme for NR crystallization involving linear and folded chain crystals is drawn.

A constitutive model for filled elastomers is developed by combining the framework of the Dynamic Flocculation Model (DFM) (Kluppel 2003) and the continuum damage model (Govindjee & Simo 1991). The model extends the previously proposed micro-mechanical formulation describing both the polymerfiller network damage and the induced filler breakage (Darabi, Itskov, & Kluppel 2016). Deformation induces damage both in the network rubbery matrix and inside the filler aggregates. This leads to the evolution of the probability density function of the number of segments and the filler size, which, in turn, causes the stress softening and the Mullins effect. These effects result in the hydrodynamic strain amplification being the major topic of this work. The model is capable of describing the deformation induced anisotropy as well as permanent set and includes a few number of physically motivated material constants characterizing the average filler cluster dimension, filler-filler and filler-matrix interaction properties.

Whereas unfilled elastomers show nearly ideally hyperelastic behavior, the differences in the stress response by adding filler are connected to the interaction between the polymer and the filler. Filling of the polymer leads not only to a reinforcement but increases also the hysteresis and stress-softening. By evaluating the displacements of an airbrushed pattern on notched Pure-Shear samples with an ARAMIS system, the displacement fields around the crack tip are obtained. Using a physically motivated model of stress softening and hysteresis of filled rubber the energy density and stress distribution can be calculated. We are able to determine the J-Integral J for closed contours around the crack tip. For purely elastic materials the value of J is path-independent but due to energy dissipation the J-Integral depends on the integration path. By variation of the strain amplitude and the integration path the impact of stress softening on the J-Integral is evaluated.

Strain-induced crystallization (SIC) in unfilled and carbon black or silica filled Natural Rubber (NR) with and without silane is investigated. The method introduced in this paper is based on measurements of the surface temperature during tensile test, whereby SIC is quantified by dividing the produced heat into different contributions, namely the dissipative heat, entropy-related reversible heat and crystallization enthalpy. It turns out that there is pronounced SIC in unfilled and carbon black filled NR, while silica/silane systems show less SIC. The degree of crystallinity correlates with the tensile strength of the samples. For silica/silane systems at the same strain level self-reinforcement by SIC is less pronounced possibly due to a lower crosslink density or strain amplification factor. Because of its simplicity, the method developed here is a promising option to investigate SIC on a broad experimental scale and provides an alternative access next to well-established methods like WAXS.