Biologically inspired robots are by now the object of several engineering programs because of a vast set of useful applications. Among these projects, very few are tackling autonomous aerial locomotion. This is indeed a very challenging problem. Aeromodelism has already produced impressive small scale rotorcrafts or airplanes, but a well controlled flapping wing vehicle does not exist. Industry is demanding for such a type of vehicle which could mimic the fascinating capacity of insects to manage adverse ambient conditions. As an example, stability of a hovering dragonfly subjected to gusts or to turbulence is really something technology would want to mimic. This is why investigation of flapping wing animals is particularly interesting. With approaches and tools from physics, biology and engineering we propose here to explore the biomimetic path from experimental observation on real flying insects to the physical models explaining the physics of thrust and lift generation. We aim at the final target of providing a set of sounded basic elements or 'rules' which can be useful in the design of efficient future flapping-wing devices. The present project is thus about investigating insect flight from a fundamental and multidisciplinary point of view.

This project aims to improve the fatigue crack initiation and propagation understanding in polycristalline materials. When the crack length in in the same order than the grain size (typically when the crack evolves within a dozen of grains), the geometry and orientation of the grains crossed by the crack front highly influence the propagation. Existing surface analysis are not sufficient to assess the real mechanisms at work and models in use are by far too empirical leading to overly conservative design of structures. The idea behind this project is to use state of the art 3D synchrotron imaging techniques as well as the new capabilities in terms of numerical simulation to tackle the short crack problem. On the experimental side, X-ray micro-tomography makes available the 3D shape of the crack in its classical form and the local crystallography using diffraction contrast. Being a non-destructive technique, the crack shape can be probed during in-situ testing as the damage progress in a network of grains of known shapes and orientations. This key aspect relies on the experience of the MATEIS team which co-developed these techniques in the past ten years. On the modelling side, crystal plasticity computations will be conducted at Centre des Matériaux on numerical avatars of both the real (obtained by synchrotron imaging) and synthetic microstructures. Crack initiation and propagation using material constitutive behaviour modified to account for damage in the local approach to fracture framework. To ensure a good description of the stress strain field within each grain, high density meshes coupled to parallel computing will be used. Regularisation tools are to be used and improved to reduce mesh dependency often observed within the local approach to fracture framework. Automated remeshing procedures will be necessary to make the crack propagate in the grain network. This large numerical effort will rely on a tight cooperation between ARMINES Centre des Matériaux and DMSM/CEMN at l'Onéra who have very complementary experience and are bond by the same development platform: The Z-Set/ZeBuLoN software. In modelling, for the very first time, crack initiation and propagation in a grain network of known shapes and orientations and comparing with the experimental growth, it is expected that the prevalent physical mechanisms at work can be precisely determined. Accounting for those in the material constitutive behaviour represents the final goal to achieve much more realistic predictions of short fatigue crack propagation.

Most cases of severe visual loss in developed countries are due to retinal diseases affecting a specialized class of neurons, the photoreceptors. Currently available systems for retinal imaging in humans do not allow neuronal imaging at the cellular level, which is crucial to understand, diagnose and monitor retinal diseases. In recent years, adaptive optics (AO) fundus imaging has proven its capability to image individual photoreceptor cells in the human retina. This technology is now reaching technical maturation. A prototypic AO system (manufactured by Imagine Eyes) is currently in operation in a clinical setting (Clinical Investigation Center 503) and has proven its reliability to monitor single photoreceptors over time. Yet, the clinical evaluation of AO imaging is still in its infancy, and biomarkers issued from AO imaging have not been validated. The goal of the iPhot project is thus to optimize AO imaging process (from the implementation of novel technical solution to image processing and data analysis) in order to obtain morphological, quantitative and longitudinal information at the level of single photoreceptors in humans. We will aim at detecting early photoreceptor damage during genetically and phenotypically defined inheritable retinal dystrophies. This industrial research project comprises four workpackages: 1- transversal and longitudinal evaluation of photoreceptor imaging in normal and diseased retinas (Leader: CIC 503); 2-adaptation of the existing AO prototype to several technical needs that have been identified in preliminary clinical studies (Imagine Eyes, with L2TI and ONERA); 3-development of data management softwares (Telecom ParisTech); 4- data analysis, comprising correlation with functional data (U INSERM 707). We expect that AO imaging will detect cell loss/dysfunction earlier than any other technique currently available. To our knowledge, it is the first project aimed at detecting neuronal loss at the cellular level in humans. Beyond hereditary retinal degeneration, we expect that this project, if successful, will have broad application in other retinal diseases.