No abstract

In order to build upon the exceptional interest for flexible sensors based on carbon nanotube networks (CNNs), the field requires high device-to-device reproducibility. Inkjet printing has provided outstanding results for flexible ohmic sensors in terms of reproducibility of their resistance. However, the reproducibility of the sensitivity, the most critical parameter for sensing application, has been only marginally assessed. In the present paper, CNN based resistive strain sensors fabricated by inkjet-printing on flexible Ethylene Tetrafluoroethylene (EFTE) sheets are presented. The variability on the device initial resistance is studied for 5 different batches of sensors from 3 to 72 devices each. The variability ranges between 8.4% and 43% depending on the size of the batches, with a 20% average. An 8-device batch with 15% variability on initial resistance is further studied for variability on the strain and thermal sensitivity. Standard deviation values are found to be as low as 16% on the strain sensitivity and 8% on the temperature sensitivity. Moreover, the devices are hysteresis free, a rare achievement for CNT strain sensors on plastics

International audienc

The stability and post-bifurcation of a non-linear magnetoelastic film/substrate block is experimentally exploited to obtain active control of surface roughness. The non-intuitive interplay between magnetic field and elastic deformation owes to material and geometry selection, namely a ferromagnetic particle composite film bonded on a compliant passive foundation. Cooperation of the two otherwise independent loading mechanisms-mechanical pre-compression and magnetic field-allows one to bring the structure near a marginally stable state and then destabilize it with either magnetic or mechanical fields. We demonstrate for the first time that the critical magnetic field is a decreasing function of pre-compression and vice versa. The experimental results are then probed successfully with full-field finite element simulations at large strains and magnetic fields. The magnetoelastic coupling allows for reversible on/off control of surface wrinkling under adjustable critical magnetic and mechanical fields, thus this study constitutes a first step towards realistic active haptic and morphing devices.

To cite this version:Dipayan Mukherjee, Laurence Bodelot, Kostas Danas. Microstructurally-guided explicit continuum models for isotropic magnetorheological elastomers with iron particles. AbstractThis work provides a family of explicit phenomenological models both in the F − H and F − B variable space. These models are derived directly from an analytical implicit homogenization model for isotropic magnetorheological elastomers (MREs), which, in turn, is assessed via full-field numerical simulations. The proposed phenomenological models are constructed so that they recover the same purely mechanical, initial and saturation magnetization and initial magnetostriction response of the analytical homogenization model for all sets of material parameters, such as the particle volume fraction and the material properties of the constituents (e.g., the matrix shear modulus, the magnetic susceptibility and magnetization saturation of the particles). The functional form of the proposed phenomenological models is based on simple energy functions with small number of calibration parameters thus allowing for the description of magnetoelastic solids more generally such as anisotropic (with particle-chains) ones, polymers comprising ferrofluid particles or particle clusters. This, in turn, makes them suitable to probe a large set of experimental or numerical results. The models of the present study show that in isotropic MREs, the entire magnetization response is insensitive to the shear modulus of the matrix material even when the latter ranges between 0.003-0.3MPa, while the magnetostriction response is extremely sensitive to the mechanical properties of the matrix material.two-dimensional MREs. Moreover, in an effort to resolve some of the surrounding air and specimen effects, Kalina et al. (2016) have modeled directly the specimen, the surrounding air and the microstructure at the same scale. While this study has led to satisfactory qualitative agreement with experiments, it did not resolve the different length scales as one goes from specimen to microstructure, since that would require an untractable mesh size. Along this effort, Keip and Rambausek (2015) proposed a two-scale finite element approach in order to solve simultaneously the magneto-mechanical boundary value problem and the microstructural problem by properly resolving the separation of the very different length scales. While this last approach is the more complete one, it still remains numerically demanding, especially if complex unit cells with large number of particles are considered. Moreover, in all these approaches, it is very hard to decouple from the estimated response the relative effect of the specimen geometry and that of the microstructure.In this regard, the recent study of Lefèvre et al. (2017) proposes an alternative view to the problem by first solving the homogenization problem at the RVE scale analytically and then using these estimates at the macroscopic scale to analyze the specimen shape effects. In that effort, the authors obtained a very usef...

This work studies experimentally and numerically the post-bifurcation response of a magnetorheological elastomer (MRE) film bonded to a soft non-magnetic (passive) substrate. The film-substrate system is subjected to a combination of an axial mechanical pre-compression and a transverse magnetic field. The non-trivial interaction of the two fields leads to a decrease of the critical magnetic field with applied precompression, while the observed wrinkling patterns evolve into crinkles, a bifurcation mode that is defined by the accompanied curvature localization and strong shearing of the side faces of the wrinkled geometry. Using a magneto-elastic variational formulation in a two-dimensional finite element numerical setting, we find that the crinkling is an intrinsic feature of magnetoelasticity and its presence is directly associated with the repulsive magnetic forces of the neighboring wrinkled-crinkled faces. As a result, the presence of the magnetic field prohibits the formation of creases and folds. In an effort to obtain a good quantitative agreement between the numerical and the experimental results, we also introduce an approximate way to model the friction of the lateral film-substrate faces. This analysis reveals the strong effects of friction upon the magneto-mechanical wrinkling modes.

In this paper, we investigate and quantify the thermal effects induced by plastic deformation at the level of the microstructure of a polycrystalline metallic sample. For the first time, this investigation is conducted on a specimen containing hundred of grains. We use a unique experimental setup to access-simultaneously in-situ and in real time-strain and temperature fields of an austenitic stainless steel under tensile loading. We show that strain fields are directly linked to the expression of plasticity at the grain scale. We show, on the other hand, that thermal fields at the last increment of deformation are linked to the microstructural expression of plasticity on a larger lengthscale corresponding, instead, to grain clusters. Hence strain fields exhibit stronger localization features than the temperature fields in terms of both values and space. For a mean temperature rise of 0.75°C and a global deformation of 2.4% in the fastest quasi-static regime investigated in this paper, the maximum local temperature rise is measured to be 0.88°C even though local strain in grains can reach up to 6.7%. These fully-coupled measurements also provide the first experimental evidence that an instantaneous coupling takes place within grains between strain gradients and thermal dissipation. Finally, an estimation of a grain-scale field of the fraction of plastic work converted into heat is conducted and shown to be not only heterogeneous but also to be related to the microstructural features of deformation at the surface of the material, namely to the absence or presence of slip bands. The results obtained support the relevance of establishing energy balances and acquiring stored energy data at the microstructural scale where damage localization takes place.

Vertically aligned carbon nanotube (VA CNT) arrays are considered as promising thermal interface materials (TIMs) due to their superior out-of-plane thermal conductivities. However the air gaps between adjacent CNTs within the CNT array hinder the in-plane heat transfer, thus significantly degrading the thermal performance of VACNT-based TIMs. To improve the inter-tube in-plane thermal conduction within of VACNT arrays, we propose a novel three dimensional CNT (3D CNT) network structure, where the CNTs in a VACNT array are cross-linked by randomly-oriented secondary CNTs. Three different catalyst preparation methods for the secondary CNT growth are compared in terms of their ability to produce a dense network of secondary CNTs. Among the tested methods, the chemical impregnation method shows a denser 3D CNT network structure. The 3D CNT network grown using this method and is thus chosen for further thermal characterization via a framework especially developed for the evaluation of in-plane thermal properties of such devices. The temperature fields of the corresponding 3D CNT network under different heating powers are recorded using a 15 μm-resolution infrared thermal imaging system. The in-plane thermal conductivity is then derived from these fields using numerical fitting with a 3D heat diffusion model. We find that the in-plane thermal conductivity of the 3D CNT network is 5.40±0.92 W/mK, at least 30 times higher than the thermal conductivity of the primary VACNT array used to grow the 3D CNT network.