Experimental identification of fracture toughness of a carbon black-filled styrene butadiene rubber undergoing energy dissipation by Mullins softening
Introduction
The strength of elastomers can be affected by the presence of cracks initiated during use by fatigue or accidentally by sharp objects. The growth of pre-existing cracks has been studied for several decades, following the seminal work of Rivlin and Thomas (1953). These authors proposed an energy fracture criterion defined as an extension of the linear elastic fracture mechanics Griffith criterion (Griffith, 1921). The latter applies to non-dissipative elastic materials only and was extended to rubbers under the assumption that any source of energy dissipation is confined in the crack tip vicinity, and can therefore be neglected in the global energy balance. This assumption was experimentally verified by Rivlin and Thomas (1953) on non-filled natural rubbers. It allows calculating the energy release rate, commonly noted , which designates the decrease in total potential energy per increase of crack surface area. In a structure, when reaches a material critical energy release rate , the crack propagates creating new surfaces. Therefore, when G can be determined, the growth of a crack can be predicted independently of the specimen shape or size.
The energy criterion such as defined by Rivlin and Thomas (1953) has been commonly used for filled rubbers (Glucklich and Landel, 1976, Medalia, 1987, De and Gent, 1996, Hamed and Park, 1999, Gherib et al., 2010 among others), and is often asserted on a pure shear geometry, which provides a simple analysis and reliable values (Roucou et al., 2019). However, carbon-black filled elastomers undergo significant softening when first submitted to a level of strain never applied before, due to the well known Mullins effect (Mullins, 1969, Diani et al., 2009). The phenomenon is rate independent and induces energy dissipation due to the degradation of the bonded layer around the carbon-black particles (Diaz et al., 2014). Therefore, when considering a notched pure shear sample of such a material, submitted to a monotonic loading, the energy dissipation cannot be considered localized to the vicinity of the crack tip only. Recently, Qi et al. (2018) have proposed a theoretical framework to account for bulk energy dissipation for mode-I plane stress cracks under steady state propagation in a model rubber undergoing Mullins softening. The strain energy release rate is defined as the sum of an intrinsic material parameter , and a dissipative part depending on the amount of energy consumed by Mullins softening upon loading and crack propagation. Such a decomposition is inspired by the same decomposition adopted in order to account for the viscoelasticity of soft materials (see the topical review from Persson et al., 2005). Their theory is developed for an incompressible Neo-Hookean material modified by the Ogden and Roxburgh (1999) model to account for the Mullins softening. Their computational approach aims to estimate the value of . Capitalizing on this remarkable theoretical work, the present study intends to investigate experimentally the resistance to failure of an actual elastomer undergoing Mullins softening, in order to estimate the dissipative part of the energy release rate and address the possible existence of an intrinsic parameter .
For this purpose, thin plates of carbon black filled SBR were manufactured by the French tire manufacturer and cut into pure shear test geometries. Fracture tests were run on virgin and preloaded notched samples in order to estimate the contribution of the Mullins energy dissipation on the critical energy release rate. On one side, a simple global energy balance analysis using the experimental macroscopic stress–strain curves provides access to two quantities, the classical global critical strain energy release rate as defined by Rivlin and Thomas (1953), which characterizes the global material resistance to fracture, and the original critical energy release rate characterizing the change of stored energy during the crack growth. On the other side, monitoring the local strain using digital image analysis on the free surfaces of the pure shear specimens, allows estimating the data required to calculate and by following the theoretical analysis of Qi et al. (2018). Moreover, the calculations carried out give access to quantitative separation of the Mullins energy dissipation during the loading driving to the crack propagation, and of the Mullins energy dissipation during crack propagation. This work provides a direct comparison between the theory proposed by Qi et al. (2018) and actual experiments.
Section snippets
Material
Rectangular plates of styrene butadiene rubber (SBR) filled with 50 phr of N347 carbon-black were manufactured by Michelin. The star-branched solution SBR presents a molar mass of kg/mol with a styrene content of 15%. Some plates were cut into pure shear geometry samples of 147 mm width, 20 mm height and 2.4 mm thickness without any preloading. Other plates were first submitted to uniaxial tensile tests in order to soften the material by Mullins effect, before cutting the same pure shear
Theory
The theory developed by Rivlin and Thomas (1953) driving to Eq. (1) has been established for non-dissipative hyperelastic materials based on the classic fracture energy balance made through the virtual steps:
Loading the specimen up to a global displacement without crack propagation allows characterizing the energy put into the system.
While the global displacement is maintained constant at , the crack propagates of a small length and stops, the crack new length becomes . Part of the
Theory
The framework introduced by Qi et al. (2018) considers the existence of an intrinsic material parameter , which characterizes the rate of energy consumed exclusively by the creation of new crack surfaces. By definition, such a parameter is independent of any other bulk energy dissipation process such as the Mullins effect, and the total material toughness decomposes as , where represents the additional contributions to fracture resistance caused by other dissipation processes.
Conclusion
The Mullins effect occurring in carbon-black filled rubbers has been known and studied for years, nonetheless, it has long been disregarded within the study of fracture of such materials. Recently, a theoretical approach based on thermodynamics has been proposed by Qi et al. (2018), which also provided numerically predicted results. In order to provide an experimental insight on this subject, an extensive study has been carried out to evaluate the impact of the Mullins softening on the
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
Authors are grateful to Julien Caillard for helpful discussions.
References (19)
- et al.
A review on the mullins effect
Eur. Polym. J.
(2009) - et al.
Physical interpretation of the mullins softening in a carbon-black filled SBR
Polymer
(2014) - et al.
Experimental characterization and modelling of the cyclic softening of carbon-black filled rubbers
Mater. Sci. Eng. A
(2011) - et al.
Fracture toughness of soft materials with rate-independent hysteresis
J. Mech. Phys. Solids
(2018) - et al.
Mapping the nonlinear crack tip deformation field in soft elastomer with a particle tracking method
J. Mech. Phys. Solids
(2019) - et al.
Tear strength of carbon-black-filled compounds
Rubber Chem. Technol.
(1996) - et al.
Influence of the filler type on the rupture behavior of filled elastomers
J. Appl. Polym. Sci.
(2010) - et al.
The tearing test as a means for estimating the ultimate properties of rubber
J. Appl. Polym. Sci.
(1976) VI. the phenomena of rupture and flow in solids
Philos. Trans. R. Soc. Lond. A Math. Phys. Eng. Sci.
(1921)