Various experiments have been conducted to investigate the influence of fluorescent whitening agent (FWA) on aspects of colour evaluation. Large amount of samples were prepared by direct dyeing with various concentration of colorant dyes as well as FWA. These dyed samples, together with high-FWA-content samples provided by third party, were used as stimuli in instrumental measurement by spectrophotometer and tele-spectroradiometer as well as in psychophysical experiment. The influence of FWA on instrumental illumination variation, calibration methods' performance, calibration standards' performance are studied. It was found that there is no significant relationship between the material of measured target and calibration standard. It is also found that spectrophotometers agrees well if the UV content in illumination is properly calibrated. A practical design of viewing cabinet with improved quality in UV range was proposed. The additional calibrated UV lamp significantly improved the fluorescent daylight lamps performance in term of metamerism index. Such improvement in UV enabled common viewing cabinets be upgraded to cater the need of visual assessment involving FWA. Three pairs of metameric samples were developed which matched under illumination with certain UV component while mismatched under illumination with other UV conditions. These metameric pairs were developed as UV Visual Aid which could be used as practical checker to visualize the UV quality in light conditions. Psychophysical experiment with category-judgment method was conducted to investigate the preferred white of human. Mean-category-value and ellipsoid fitting were used to find the resulting preferred white colour. There were significant amount of samples which were not within the Tint limit of CIE Whiteness index but were categorized by observers as acceptable white. This indicated the possible need of expanding the Tint limit of CIE Whiteness. Psychophysical experiment with magnitude estimation method was conducted to evaluate the prediction of visual whiteness by various whiteness formulae. CIE Whiteness index had the best agreement between instrumental prediction and visual perception, outperforming Uchida and C/V Whiteness.

The primary objective of this thesis was to develop a novel adsorption calorimeter to assess the equilibrium and dynamic behaviour of adsorbent-bond- metal sub-system, a constituent of an adsorption chiller. In this research, a novel calorimeter consisting of an adsorber and an evaporator section has been built. Thermoelectric modules were used to control the temperature and measure the instantaneous heat flow at the adsorber and evaporator sections. This thermoelectric module system was tested against an inert aluminium sample. The aluminium sample temperatures were controlled to constant, sinusoidal and triangular set point to within 0.1 K (1 standard deviation). The effective Seebeck coefficient and thermal conductance of the module were extracted from the calibration experiment. The inferred heat flow had a typical accuracy of 10 %. Further tests with a sinusoidal temperature variation yielded the root mean square error of cumulative heat flow as - 8 %. Three and one mm nominal diameter Type A silica gel beads were chosen as the adsorbent, and water as the refrigerant. In the adsorption-evaporation experiment, the plot of heat of adsorption versus time was well fitted to exponential recovery (r2 > 99%). Cumulative net heat flow at the adsorber at time equal to infinity gave the adsorptive capacity, while the time to reach 63.2 % of this value gave the apparent Linear Driving Force constant (Kldf). Adsorptive capacities of the silica gels were fitted to Henry's law and were within 20 % of previously published data. Specific heats of adsorption (obtained by Arrhenius plots of the Henry's law constants) were 2495 kJ/kg and 2634 kJ/kg for 3 and 1 mm diameter gel beads respectively, which were within 9 % of the published data (2710 kJ/kg). Specific heat of adsorption for both samples, calculated from energy balance, were within 8.4 % of values inferred from Arrhenius plot of the Henry Law coefficient. The apparent Kldf for 1 mm diameter gel-beads was 3 times higher compared to the one for 3 mm diameter beads. Activation energy (1261 kJ/kg for 3 mm diameter beads and 1537 kJ/kg for 1 mm diameter beads) was obtained from a further Arrhenius plot, and was between 50 and 58% of the specific heat of adsorption. This suggests the surface diffusion is the governing mechanism for the water adsorption onto silica gel. Tests with de sorption- condensation and temperature varied sinusoidally versus time revealed the ineffectiveness of the water-containing section when acting as a condenser. The adsorption behaviour of coated fins was studied experimentally. The apparent K1df for silica gel coated to a stainless steel fin bonded to an isothermal aluminium plate was ~ 50 % less than that of silica gel bonded directly to flat plate. For an aluminium fin that was cut from a single piece of metal, the apparent Kldf was reduced by 10 %. A numerical model predicted the apparent K1df of the coated fin with an accuracy better than 22 %. The model was developed using the rate constant for silica gel bonded onto a flat plate, the thermal contact resistance at the. root of the fin (if applicable) and the temperature gradient along the fin. Other fin configurations that were not tested experimentally were simulated in the model. In an adsorption chiller, it is estimated that a specific cooling power as high as 275 W/kg and 750 W/kg could be obtained for the 3 and 1 mm diameter gel beads bonded to the flat plate heat exchange surface. For the silica gels coated onto 20 mm high aluminium fins (1 mm thickness), the specific cooling power could be reduced by 30 and 33 % for 3 and 1 mm diameter gel beads respectively. The specific cooling power is sensitive to the adsorber configuration and cycle time.

La détection de très faibles quantités de Matériaux Energétiques (ME) est un challenge important dans la lutte contre le terrorisme. En plus des méthodes de détection des ME par affinité chimique, il est aussi intéressant d'utiliser les variations enthalpiques dues à la décomposition des ME pour les détecter par analyse thermique. Cependant, la sensibilité des methodes classiques est insuffisante pour la détection des particules dont la masse se situe dans le domaine des nanogrammes. En revanche, la nanocalorimétrie est parfaitement adaptée pour la caractérisation de très faibles quantités d'échantillons et est de ce fait adaptée aux exigences de la détection. Afin d'explorer la possibilité de détecter et d'identifier des micro-particules solides de ME à l'aide de l'analyse thermique, nous avons élaboré des protocols optimisés pour la détection et l'identification de particules pures unitaires de quelques nanogrammes de ME ainsi que leurs mélanges. Les résultats montrent que la limite de détection se situe environ à quelques centaines de picrogrammes. Les expériences ont été complétées par de l'analyse structurale in-situ en utilisant sa combinaison avec de la DRX par faisceau nanofocus synchrotron. / Being able to sense the minuscule amounts of energetic materials is crucial in the context of the fight against terrorism. Apart from the methods of detection of EM, which are specific to the chemical structure, one could use the enthalpy variations of the EM decomposition process for their detection by means of thermal analysis. However, the sensitivity of classical methods would be still insufficient to sense particles in the nanogram range. By contrast, the recently developed technique of chip calorimetry is perfectly suited for characterizing small amounts of samples and is therefore fully adequate for this task.In order to explore the possibilities of detection and identification of solid micro-particles of EM with thermal analysis, we discuss on the protocols optimized for the detection and identification of nanogram-size particles of EM and its mixtures with the chip calorimeter accessory. The results obtained on pure EM and its mixtures show that the detection threshold can be put at approximately several hundred picograms. The experiments were completed by the in-situ structural analysis using a combination with nanofocus synchrotron XRD.

Ces travaux exposent et analysent les performances d’un composite à matrice thermoplastique semi-cristallin au passage et au-delà de la transition vitreuse. Il est nécessaire de donner des éléments objectifs afin d’évaluer et discuter dans quelles mesures ce matériau peut être utilisé de manière innovante. L’objectif visé est de donner des perspectives en termes de plage de températures d’utilisation des matériaux composites à matrice thermoplastique et plus spécifiquement pour des applications à hautes températures. Le composite à fibres continues de carbone et matrice polyétheréthercétone PEEK a ainsi été étudié sur une large gamme de température, avant et après sa température de transition vitreuse (Tg = 143°C).La phase de caractérisation quasi-statique a mis en évidence l’importance du renfort et le bon transfert de charge de la matrice au-delà de la Tg pour les différentes sollicitations. En particulier, la résistance à la rupture en compression s’est avérée intéressante dans l’optique d’une application structurelle. De plus, les observations fractographiques ont mis en évidence des comportements radicalement différents au passage de la transition vitreuse. Le caractère plus ductile de la matrice permet de limiter la propagation de fissures au travers des plis par dissipation de l’énergie : la plastification de la matrice augmente la capacité du composite à dissiper de l’énergie en limitant ainsi la fissuration. Cependant pour des chargements où la matrice pilote la réponse mécanique du composite tels que des efforts de cisaillement, les comportements non linéaires sont fortement accentués. Des mécanismes de déformations dépendant du temps ont été observés à travers des essais de charge-décharge et de fluage au-delà de la Tg : le caractère visqueux de la matrice joue un rôle prépondérant.Ces mécanismes non linéaires étant identifiables sur des temps longs, il était intéressant de proposer des méthodes de modélisation pour prédire le comportement du composite. C’est pourquoi des modèles à l’échelle du pli ont été adaptés en fonction de la température et de la prépondérance des caractères viscoélastique et viscoplastique. Différents essais de fluage-recouvrance en torsion rectangulaire menés sur un rhéomètre ont permis d’évaluer les composantes viscoélastiques et viscoplastiques de la déformation à des températures inférieures et supérieures à la Tg. / The present study shows and analyses the specifications of a semi-crystalline thermoplastic composite as function of temperature, below and above the glass transition. In order to assess and discuss about what extent this material could be innovately use, objective facts must be necessary exposed: the main target is to give the outlooks about the temperature range, in particular the high temperatures. The studied material is a continuous carbon fibre composite with a polyetheretherketone (PEEK) matrix. Its glass transition temperature is around 143°C. It has been characterized throughout a wide temperature range.For several kinds of quasi-static loadings, the load transfer from the matrix to the fibre reinforcement is good even above the glass transition temperature. The compression strength is indeed very interesting for an aeronautical application. In addition, the fracture surface analysis have significantly revealed a different behaviour above the glass transition temperature: the matrix is more ductile and thus the crack propagation is limited thanks to the energy dissipation. However when the mechanical response is driven by the matrix behaviour such as shear loadings, the nonlinear mechanical behaviour of the composite are highly increased. Therefore the time-dependent behaviours have been characterized by using creep experiments and loading-unloading tensile tests as function of the temperature.In order to predict those non-linear behaviours, meso-models have been developed as function of the temperature. Thus viscoelasticity and viscoplasticity have been taken into account to model the nonlinear mechanical behaviour of the composite material, thanks to creep-recovery tests which have been carried out with a torsion rheometer.