PSI - Issue 13

Jacopo Schieppati et al. / Procedia Structural Integrity 13 (2018) 642–647 Schieppati et al. / Structural Integrity Procedia 00 (2018) 000 – 000

643

2

1. Introduction

The unique mechanical properties of rubbers make them suitable for applications in which cyclic loadings are involved. In this loading condition, failure is mainly related to fatigue phenomena, Gent (2012), and therefore the study of the fatigue behaviour of materials is of great practical importance in the rubber industry. The ultimate task of fatigue analysis is to estimate the lifetime of components, hence the prediction of fatigue life of materials is of critical importance in this frame. Mars and Fatemi (2002) reported that in the field of elastomers, two main approaches are followed for fatigue life prediction: (i) crack nucleation and (ii) crack growth. The first one deals with the nucleation and growth of cracks up to a certain limit and is based on a continuum mechanics approach. The second one is based on the study of the growth of pre-existing cracks up to end of service life using fracture mechanical approaches. The fatigue behaviour of rubbers is influenced by a large number of parameters, which can be related to the mechanical history, environmental conditions and rubber formulation as pointed out by Mars and Fatemi (2004). Among them, temperature has a relevant effect on the fatigue properties of rubbers: Lake and Lindley (1964) reported a drop of 4 order of magnitude of fatigue life passing from 0 °C to 100 °C. Further analysis on the effect of temperature can be found in Young (1986), Young & Danik (1994), Legorju-Jago & Bathias (2002). Even though an increase in temperature is considered to cause a decrease in the fatigue properties of rubbers, this effect has not been yet proved. The aim of this work is to have a better comprehension of the impact of temperature on the fatigue properties of elastomers. In order to do this, crack growth measurements have been established measuring the surface temperature and studying the effect of frequency on the crack growth rate and on temperature variation. Moreover, the impact of temperature has been assessed through thermal conductivity measurements. The measurements were carried out at different temperatures, from 30 to 130°C, in order to have a better description of the impact of temperature on the material properties. This study represents a basis to extend the fracture mechanics approach, by implementing the temperature effect in common used models for fatigue life prediction.

2. Experimental part

2.1. Materials

The material used for this research is a commercial acrylonitrile butadiene rubber (NBR) filled with 42 phr of carbon black. Due to confidentially, no additional information about the precise formulation can be given.

2.2. Crack growth measurements

For the implementation of the crack growth measurements, a pure shear specimen geometry was used. The height of the specimen was 16 mm, the width 200 mm (width to height ratio 1/12.5) and the thickness 4 mm. The samples were mounted on special clamps, preloaded to 20 N and notched. The initial notch was introduced using a razor blade, mounted on a customized tool guided on the clamping system, for an initial notch of about 25 mm. The tests were carried out using an MTS 858 Table Top System testing machine and the crack length was monitored through a camera system CV-5701P by Keyence. In order to avoid light reflections the specimens were sprayed with a white powder coating. The tests were utilized in force control mode with a load ratio of 0.1 and a maximum load of 1300 N. The surface temperature was monitored using an IR sensor placed around 10 mm in front of the initial crack tip. From the position of the crack tip, evaluated through the pictures recorded with the camera system, the crack growth rate was calculated. The hysteresis were monitored during the cyclic loading through the MTS test device and the tearing energy G was calculated dynamically, i.e. during testing on the same specimen. For pure shear geometry, G can be calculated with equation 1, Rivlin and Thomas (1953): = 0 ∙ ℎ 0 = ⁄ (1)