Bernard Douchin

Mechanical Characterization of an Alternative Technique to Embed Sensors in Composite Structures: The Monitoring Patch

Sensor embedding is one of the main operations in dealing with composites in-core instrumentation. In this work, an alternative encapsulation technique called “monitoring patch” is proposed to achieve correct sensor embedding, to facilitate the industrialised instrumentation procedure and to adapt the sensors according to the geometry and material heterogeneities required of the composite structures. The monitoring patch is mainly developed with the aim to reduce the variability effects produced if the sensor alone is placed. In this initial study, a first patch’s configuration is manufactured with CTMI pre-impregnate epoxy–woven glass, hosting two kinds of silicon prism sensors. The monitoring patch is then placed in the thick middle plane of an epoxy-carbon M21 T700GC quasi-isotropic plate. The plates are instrumented with strain gauges and tested using digital image correlation (DIC). The strain field maps are calculated to analyse the over-strain zones and to infer fracture paths. At the same time, a FEM model is developed to compare the numerical and the experimental observations. The results show that the mechanical strength of the instrumented plates is not significantly affected by the presence of the patch. The failure path of the instrumented plates with monitoring patch is found along the patch perimeter; therefore, the sensors can be recovered without damage even after the failure of the instrumented structure. The feasibility of the monitoring patch is discussed with other embedding techniques. In further studies, the monitoring patch will host a streaming sensor with an aim to carry out in-core strain measurements.

Comparison between the Classic Sensor Embedding Method and the Monitoring Patch Embedding Method for Composites Instrumentation

In this paper, the classic embedding technique, with bared sensors, and a recent proposal, the monitoring patch, are compared with the aim to improve the composites in-core instrumentation. The monitoring patch emerges with the need to industrialize sensors integration inside composite structures; thus, a complete evaluation of its mechanical performance has to be done. Numerical and experimental campaigns are carried out on elementary carbon-epoxy coupons to evaluate the benefits and disadvantages of this procedure compared with the typical interlayer sensor embedding. The results show that the use of monitoring patch does not affect significantly the mechanical performance of instrumented coupons. An instrumentation transfer function (ITF) is proposed to link the information that electronic devices can detect, the mechanical phenomena around these electronic devices and the measurements data acquired by global or local techniques (DIC, FEM, gauges). A good correlation between the strain data acquired and the strain values calculated by FEM confirms the approach of the ITF to evaluate the influence of the monitoring patch on the measured signal.

Spatial Evolution of the Thickness Variations over a CFRP Laminated Structure

Ply thickness is one of the main drivers of the structural performance of a composite part. For stress analysis calculations (e.g., finite element analysis), composite plies are commonly considered to have a constant thickness compared to the reality (coefficients of variation up to 9% of the mean ply thickness). Unless this variability is taken into account reliable property predictions cannot be made. A modelling approach of such variations is proposed using parameters obtained from a 16-ply quasi-isotropic CFRP plate cured in an autoclave. A discrete Fourier transform algorithm is used to analyse the frequency response of the observed ply and plate thickness profiles. The model inputs, obtained by a mathematical representation of the ply thickness profiles, permit the generation of a representative stratification considering the spatial continuity of the thickness variations that are in good agreement with the real ply profiles spread over the composite part. A residual deformation FE model of the composite plate is used to illustrate the feasibility of the approach.

Development of Instrumented Composite Cylinders

This paper describes a technique to fabricate instrumented composite cylinders made by the filament winding process and based on the insertion of Bragg grating sensors contained in an optical fiber, so that real time monitoring of the temperature and strain response can be made. The main advantage of this method is that, for the first time, the strain response of the wound layers as a function of time and temperature can be monitored during fabrication without alteration of the composite structure, since the sensor is a filament itself. The sensor data reveals the existence of several phenomena related to the manufacturing process and material response. The results show that during the curing stage of the fabrication of the composite cylinders there is a dimensional contraction without change in temperature which indicates that a material transformation occurs, and also that at the last stage there is a residual compressive strain which may affect the in-service behavior of the composite cylinder.

Identification and Modelling of the In-Plane Reinforcement Orientation Variations in a CFRP Laminate Produced by Manual Lay-Up

Reinforcement angle orientation has a significant effect on the mechanical properties of composite materials. This work presents a methodology to introduce variable reinforcement angles into finite element (FE) models of composite structures. The study of reinforcement orientation variations uses meta-models to identify and control a continuous variation across the composite ply. First, the reinforcement angle is measured through image analysis techniques of the composite plies during the lay-up phase. Image analysis results show that variations in the mean ply orientations are between −0.5 and 0.5° with standard deviations ranging between 0.34 and 0.41°. An automatic post-treatment of the images determines the global and local angle variations yielding good agreements visually and numerically between the analysed images and the identified parameters. A composite plate analysed at the end of the cooling phase is presented as a case of study. Here, the variation in residual strains induced by the variability in the reinforcement orientation are up to 28% of the strain field of the homogeneous FE model. The proposed methodology has shown its capabilities to introduce material and geometrical variability into FE analysis of layered composite structures.

Benefit from Embedded Sensors to Study Polymeric Composite Structures

The present paper deals with the use of optical fibers with Bragg gratings for the identification of process-properties relations for composite structures. This study relates a 7 year-experience in the field of process and mechanical characterization of composite parts. It highlights some major topics in terms of technological issues such as sensor integration, information discrimination and measurement accessibility. Results of process monitoring concerning autoclave cure and filament winding are shown through which initial states are identified. Finally, structural tests of instrumented parts are presented, showing the efficiency of the embedded sensors to allow test-calculation dialogues and thus properties identification.

Variability in Monolithic Composite Parts: From Data Collection to FE Analysis

A main challenge facing the structural applications of composite materials is related to the uncertainty in the material performance due to their inherent variabilities. Structural properties of composites are not only dependent on the manufacturing steps, but also on the constituent materials, reinforcement architecture and design choices. By introducing geometrical variabilities into finite element (FE) model through meta-models, the effect of variations in the composite structure can be studied at different scales. To assure that the FE results are in accordance with the real composite structure, the input parameters of the models must be in agreement with the actual material and its configuration in the structure. In this chapter, a methodology to study and introduce variabilities into a composite structure FE analysis is presented.

Assessments on the mechanical behaviour of a monolithic composite structure instrumented with a monitoring patch

In this paper, the monitoring patch is evaluated as an alternative instrumentation technique for aircraft-type composite structures, by means of the Multi-Instrumented Technological Evaluator. In this case, the goal is to evaluate the strength and failure modes of a carbon-epoxy composite plate with two drop-offs instrumented with a monitoring patch. With the aid of finite element models, the testing of the plate under combined loads is analysed to have a first numerical approach of its behaviour. Then, the experimental campaign is accomplished by testing the plate with multi-instrumentation devices and techniques such as strain gauges and digital image correlation. A correct calculation/test correlation is achieved by comparing the strain values calculated by the finite element model and the experimental strain data acquired by gauges and digital image correlation. The results confronted provide a first evidence to quantify the influence of the monitoring patch on the mechanical performance of the composite plate. Therefore, it could be employed in the near future as instrumentation technique on large composite structures.

10 – Repairing composites

This chapter deals with the issues and main challenges concerning large repair of primary composite structures. Feedback about use of composite parts, during a million cumulated flight hours, demonstrate the higher damage tolerance of composite solution but also the necessity to define optimized repair solutions. A global vision of repair is proposed through a case study, with a focus on two elementary bricks: the MITE toolbox and low-cost tools adapted to a case-by-case basis. Finally, a new method is highlighted leading to reduce the patch size and the removed material for a given damaged volume.

Identification of a macroscopic anisotropic damage model using digital image correlation and the equilibrium gap method

The aim of the work is to demonstrate how an anisotropic damage model may be identified from full field measurements retrieved during a heterogeneous test. The example of a biaxial test performed on a 3D C / C composite is used. In a first step, the displacement fields measured by classical Digital Image Correlation are used as input data of a finite difference version of the Equilibrium Gap Method. A benefit from unloadings (assumed to be elastic) is shown to retrieve a damage law. In a second step, inelastic strains can be assessed from the total measured strain and the elastic estimated strains. The constitutive parameters relative to the inelastic part of the model are then identified.