No abstract

A hybrid method is introduced for the calibration of the spring constants of atomic force microscopy cantilevers. It is based on the minimization of the difference between the modelled and experimentally determined full-field displacement maps of the surface of the cantilever in motion at several resonant frequencies. The dynamic mechanical response of the cantilever to periodic motion is measured in a vacuum by means of a scanning vibrometer. Given the dimensions of the cantilever, the obtained surface displacements together with analytical or numerical models are used to resolve the physical unknowns of the probe. These are the elastic properties of the cantilever, and the residual stress state built up during the deposition of the reflective coating on the backside of the cantilever. The scanning vibrometry experiment allows the precise determination of the first ten resonant frequencies and the modes associated. After optimization of the elastic properties and the surface stress, the relative agreement between all resonant frequencies is better than 1% with the finite element model and 2% with the Timoshenko beam equation. The agreement between surface displacements is also excellent when the damping constant of the system has been determined, except for the first lateral mode, which exhibits strong coupling to a reflection of the first torsional mode. Because all the displacements at resonance are known, it is possible to decouple these modes, and the result is shown to compare well with the model. The cantilever being fully characterized (geometry, materials, residual stress state and boundary conditions), it is straightforward to deduce all its spring constants, in the linear and nonlinear elastic regimes.

The development and adoption of lab-on-a-chip and micro-TAS (total analysis system) techniques requires not only the solving of design and manufacturing issues, but also the introduction of reliable and quantitative methods of analysis. In this work, two complementary tools are applied to the study of thermal and solutal transport in liquids. The experimental determination of the concentration of water in a water-methanol mixture and of the temperature of water in a microfluidic T-mixer are achieved by means of fluorescence lifetime imaging microscopy (FLIM). The results are compared to those of finite volume simulations based on tabulated properties and well-established correlations for the fluid properties. The good correlation between experimental and modelled results demonstrate without ambiguity that (1) the T-mixer is an adiabatic system within the conditions, fluids and flow rates used in this study, (2) buoyancy effects influence the mixing of liquids of different densities at moderate flow rates (Reynolds number Re ( 10 -2 ), and (3) the combination of FLIM and computational fluid dynamics has the potential to be used to measure the thermal and solutal diffusion coefficients of fluids for a range of temperatures and concentrations in one single experiment. As such, it represents a first step towards the full-field monitoring of both the extent and the kinetics of a chemical reaction.

We have explored a new technology based on chemically induced phase separation that yields porous epoxies and cyanurates with a closed cell morphology and micrometer sized pores with a narrow pore size distribution. When the precursor monomers are cured in the presence of a low molecular weight liquid, the desired morphology results from a phase separation and a chemical quench. After phase separation, the porosity is achieved by thermal removal of the secondary liquid phase, specifically by diffusion through the crosslinked matrix. In respect to the thermodynamics and kinetics, the origin of the phase separation process can be identified as nucleation and growth. The influence of internal and external reaction parameters, such as chemical nature of the low molecular weight liquid, its concentration and the curing temperature on the final morphology are presented. Thus, the morphology can be controlled ranging from a monomodal to bimodal pore size distribution with pore sizes inbetween 1 to 10 μm. These porous thermosets are characterized by a significantly lower density, without any loss in thermal stability compared to the neat matrix. Such new materials demonstrate great interest for lowering the dielectric constant and for improving the fundamental understanding of the role of voids in stress relaxation and toughening.

A model for elastic stress transfer across the interface in composites has been developed in particular to overcome the known problem of the matrix effective radius. Stresses were determined using a two dimensional method by introducing a novel hypothesis of structural response of the matrix. This is in contrast to previous approaches, which mainly consider that the axial derivative of axial stress does not depend upon the radius. This leads to a new form of the stress transfer equation, free of an adjustable parameter. This analysis was applied to linear elastic single fibre and single platelet composites. The comparison of the two reinforcements allowed the identification of a structural parameter for which the two geometries are shown to be equivalent. The model was validated using literature results obtained by Raman spectroscopy means or photo-elastic methods. The agreement between measured data and the model showed the relevance of the matrix structural response approach to stress transfer.