Water sorption in polyamide 6.6 has been characterized for a wide range of temperature (25°C to 80°C) and various water activities using a Dynamic Vapor Sorption testing machine. Complex sorption mechanisms govern the water uptake in the material. The competition between two main temperature dependant mechanisms has been observed: a Henry's sorption mechanism that mainly governs the sorption curve at low water activities, and a second mechanism at high water activities that could be related to the formation of water clusters. It is observed that the temperature dependency can mainly be attributed to the Henry's contribution. Four physically based models are then used and identified thanks to the extended experimental database. It is shown that a simple Flory-Huggins model is not able to capture the experimental observations at very high water activities for all the temperatures tested. The ENSIC model is a better choice, but good prediction for very high water activity cannot be obtained. Both modified Park and GAB models can accurately predict the volume fraction of water for the whole ranges of water activity and temperature, although the modified Park model should be preferred considering the number of parameters and the mathematical simplicity.

Diffusion of water in polyamide 6.6 has been characterized for a wide range of temperatures (from 25 to 80 °C) and various humidities using dynamic vapor sorption machine. The decrease in glass transition temperature (Tg) has also been measured using DMA tests. As usually observed, PA66 absorbs a large amount of water (up to 5% at 90%RH) with a Fickian behavior with a diffusion coefficient that depends on water activity for all temperatures. Moreover, it appears that the diffusion coefficient for tests performed below Tg is almost independent of the water activity whereas a strong dependency is observed above Tg. This behavior is to be compared to a large decrease of Tg with the absorption of water. The increase of the water diffusion can therefore be related to a change of the amorphous phase (the crystalline phase is supposed to absorb no water) from the glassy to the rubbery states. A model based on the free volume theory is used successfully to describe the wide experimental database. It is therefore possible to describe the dependency of the water diffusion kinetics on both temperature and water uptake using the approach described in this paper. Please note that this is an author-produced PDF of an article accepted for publication following peer review. The definitive publisher-authenticated version is available on the publisher Web site.

is an open access repository that collects the work of Arts et Métiers ParisTech researchers and makes it freely available over the web where possible. a b s t r a c tTensile properties and crack propagation properties, especially critical strain energy release rate in mode I, G IC , have been used to investigate fracture properties of elastomers and their relationships with microstructure. These investigations were mainly based on a series of comparisons: first, the behaviour of polychloroprene rubber (CR), undergoing stress hardening due to strain induced crystallization (SIC) and oxidative crosslinking (OCL) was compared with that of chlorinated polyethylene (CPE), which undergoes SIC but not OCL, and with a polyurethane based on hydroxyl terminated polybutadiene (PU) which undergoes OCL but not SIC. Comparisons were also made on CR between fracture behaviour at ambient temperature, where SIC occurs and at 100 C where there is no SIC. Finally, oxidative crosslinking was used to vary in a continuous way the crosslink density in CR and PU, in order to evaluate the role of crosslinking in fracture behaviour. The results reveal the strong contribution of SIC to fracture strength. Crosslinking, even at low conversion, inhibits SIC which explains the sharp decrease of CR toughness in the early period of exposure to oxidation. When SIC has disappeared, it is possible to appreciate the effect of crosslinking on fracture behaviour. This effect, as evaluated from the density of deformation energy at rupture in tension or from G IC value, is almost negligible while the sample modulus increases regularly as a consequence of crosslinking. It appears that the toughness remains almost constant because it is under the influence of two contradictory phenomena: the negative effect of a reduction of ultimate elongation and the positive effect of a modulus increase. Such behaviour can be explained in terms of heterogeneous distribution of the lengths of elastically active chains. After long exposure, the sample behaviour becomes brittle, very high modulus values indicate that the samples approach, presumably in a heterogeneous way, the glassy state.

No abstract