CRITT Materiaux Poitou Charentes

Rochefort, France

CRITT Materiaux Poitou Charentes

Rochefort, France
Time filter
Source Type

Leger R.,Ecole des Mines d'Ales | Roy A.,CRITT Materiaux Poitou Charentes | Grandidier J.C.,CNRS Pprime Institute
International Journal of Adhesion and Adhesives | Year: 2013

This paper aims at investigating the aging of a single lap joint (SLJ) in water at 70 °C. Experimental observations show an important decrease of deformability from 1 day of aging, while initial behaviour is not impacted. Further investigations reveal that the initial crack occurs at the interface near extremities of overlaps at a load depending from aging time. To predict the behaviour of the aged SLJ submitted to tensile loading, it is necessary to know precisely the state of stress in the adhesive layer and at the interface. The authors develop a finite element model to simulate the diffusion of water in the adhesive layer of the SLJ and the subsequent residual strength testing. The mechanical behaviour of the adhesive is predicted by a coupled elasto-plastic law which is deduced from an experimental work on bulk adhesive aged in water. It can model simultaneously its intrinsic behaviour as a function of temperature and water concentration and allow predicting the tensile tests on aged adhesive specimens and the volume change due to temperature and diffusion of water. Using the finite element software ABAQUS, the authors show that during diffusion of water, stresses can be important in the adhesive layer and can generate small damage. Then, for a load that led without aging to crack initiation and final fracture, stress field in the aged adhesive layer is not that critical. Such results tend to confirm that interface has been mostly impacted and weakened by water. © 2013 Elsevier Ltd.

Denay A.-G.,CRITT Materiaux Poitou Charentes | Denay A.-G.,Institute p | Castagnet S.,Institute p | Roy A.,CRITT Materiaux Poitou Charentes | And 3 more authors.
Journal of Cellular Plastics | Year: 2012

Size effects on the monotonic and creep compression behavior of closed-cell polyurethane foams is investigated experimentally. The aim is to determine a minimum representative volume of mechanical behavior for both mechanical tests. Cubic specimens of nonreinforced and glass fiber-reinforced foams are tested in compression and creep. In the first series of tests, various sizes of samples are tested, cube with edge between 10 and 50 mm. In this case, edge effects have a variable influence depending on the sample size and the minimal size is about 40×40×40 mm 3 for the nonreinforced foam and 50×50×50 mm3 for the glass fiber-reinforced one. In the second series, the mechanical behavior of more or less extended domains is analyzed from the surface of a single 50×50×50 mm3, that is, at given edge effect. The minimal size was smaller in the latter case. Influence of the loading path was investigated comparing monotonic tension and short-term creep kinetics. Unlike in monotonic compression, and in the initial creep strain, no influence of the sample size was evidenced on the creep kinetics. Surface analysis in cubic samples upon monotonic compression and creep lead to a minimum representative surface between 4 and 13 mm edge. © The Author(s) 2012 Reprints and permissions:

Denay A.-G.,CRITT Materiaux Poitou Charentes | Denay A.-G.,Institute p | Castagnet S.,Institute p | Roy A.,CRITT Materiaux Poitou Charentes | And 2 more authors.
Journal of Cellular Plastics | Year: 2013

This work experimentally addresses the effects of negative temperature on compression behavior of non-reinforced and glass-fiber-reinforced polyurethane foams. Monotonic, creep and cyclic compression tests are performed down to cryogenic temperatures. In this range, temperature is shown to influence both the Young's modulus and the yield stress, which increase with decreasing temperature. The influence is nearly the same on either reinforced or unreinforced foams, mostly due to the preferential orientation of fibers perpendicular to the compression direction. Among contributions that could explain low temperature deformation, viscoelasticity and damage are focused on. The former contribution was investigated by creep and recovery tests, which showed that a non-negligible part of the compression strain could be recovered. The latter contribution was studied from cyclic tests and scanning electron microscopy observations. Damage analyses showed that a non-negligible initial damage issued from cooling down itself prior to mechanical loading. During further compression, a decrease of the cyclic modulus was observed, significantly before the yield stress at low temperature. Generalized brittle fracture of cavity walls was suggested from micrographs. As a conclusion, damage and viscous phenomena seem to coexist down to cryogenic temperatures. © 2013 The Author(s).

Denay A.G.,CRITT Materiaux Poitou Charentes | Denay A.G.,CNRS Pprime Institute | Castagnet S.,CNRS Pprime Institute | Roy A.,CRITT Materiaux Poitou Charentes | Alise G.,CRITT Materiaux Poitou Charentes
EPJ Web of Conferences | Year: 2010

The objective of this work was to predict the long term behavior of Polyurethane foam (PU) at very low-temperature, by applying the time-temperature superposition principle (TTSP). The experimental background of the TTSP was based on a Dynamical Mechanical Analysis technique. Two issues arise from this experimental approach: the relevance of the temperature range to apply the TTSP, and the possible size-effect associated to the small DMA samples. Firstly, on the studied temperature range (-170°C; +180°C) many transitions have been observed, particularly from-20°C. Thus to apply the TTSP, it would be necessary to limit the temperature range (between temperature of molecular transitions, i.e. from-20°C up to 80°C). At very low temperatures, DMA spectra did not evidence any viscoelastic domain. However a deformation has been measured during creep tests in the same temperature range. So it would be necessary to determine which micro-mechanism is responsible for the observed deformation. Secondly, it was important to determine if the volume of DMA sample was representative. Several techniques have shown that a representative volume would be reached between 8 and 12mm3. © 2010 Owned by the authors, published by EDP Sciences.

Loading CRITT Materiaux Poitou Charentes collaborators
Loading CRITT Materiaux Poitou Charentes collaborators