Search

Direct methods in continuum mechanics start from the thermomechanical characteristics and by solving the equilibrium or motion equations, satisfying boundary and critical conditions, result in stress, strain and/or displacement fields. Inverse methods are starting from the obtained solutions in order to compute a priori unknown characteristics or parameters. Direct experimental measurements of some characteristics or parameters are sometimes very difficult, if not impossible, to perform. Special applications of inverse methods combining some experimental global results with computed results are known as mixed numerical-experimental techniques or hybrid methods. Those MNET can be applied in order to obtain thermomechanical characteristics of material systems with elastic, viscoelastic, elastoplastic or viscoplastic behaviour. They can also be applied for damage, interphase, damping and processing characterisation. The MNET-approach can also be used in many other fields out of mechanics if a good theoretical model exist and a sensitive “experimental” method is available.

The basic principle of an inverse method for material stiffness identification is to compare measured values of a test specimen with computed values from a numerical model of the test specimen. The parameters in the numerical model are the unknown material stiffness properties and they are iteratively updated starting from an initial set of parameters until the computed output matches the measured values. This paper first outlines the general procedure used for material stiffness identification with inverse methods. The procedure is next illustrated with an example on the identification of complex engineering constants of composite materials. In this example the vibration behaviour of a freely suspended test plate is used as the information source. The measured quantities are the resonance frequencies of the test plates and the damping ratios of their associated mode shapes. The resonance frequencies are used to tune the real part of the complex moduli, while the damping ratios are used to identify the imaginary part.

Complex moduli are used to characterize the dynamic material behaviour of fibre reinforced polymers (FRP). This paper presents an efficient method to identify the in-plane complex moduli of thin orthotropic FRP plates, based on a mixed numerical experimental approach. The required experimental data are the modal parameters of a set of specific eigenmodes of a plate specimen. In combination with a numerical model for the calculation of the modal parameters of the specimen, all the independent complex moduli of the material are derived.

The mixed numerical/experimental techniques (MNET), in which the numerical results are correlated with experimental data in an iterative fashion, has been developing for more than two decades. An overview of the various estimation algorithms in the MNET is given. Emphasis is placed on the different philosophies which may lead to a deterministic optimisation procedure, or a statistical estimator, or a combination of the two. Formulations, limitations and internal relations of the algorithms are also presented. All the discussions are oriented to the researchers and engineers who are interested in the design optimisation, parameter identification and finite element model updating.

The paper uses a mixed numerical-experimental technique (MNET) to determine material properties of cement matrix composites. The MNET is based on modal analysis of free vibrating plates, and is applied for the case of microconcrete and polymer modified cement matrix. The obtained values for the complex moduli are discussed, as well as the influencing parameters. It is concluded that both stiffness and damping appear to be sensitive parameters for the characterization of concrete materials. It is also noted that the methodology and exploitation of the MNET is somewhat different for concrete than for polymer composite materials, for which the MNET was originally developed.

A method, based on the finite element technique, is proposed in order to study the behaviour of an adhesive single lap joint, together with stochastic material properties, taking into account the uncertainties of design parameters. The technique is applicable to linear and non linear problems, with more or less complicated structures. The needs are experimental determination of the components properties and numerical tools.

This paper presents a finite element analysis of coupled and semi-coupled thermal structural response of composite laminated plates. Our purpose is to use the results in order to quantify and’ calibrate’ the SPATE ( Stress Pattern Analysis by Thermal Emission ) experimental procedure. SPATE uses the thermoelastic effect, first quantified by Lord Kelvin [1], in order to obtain the stress distribution in a structure. Unfortunately the application of SPATE as NDT method on composite materials has so far been limited. The main problem has been identified to be associated with the existence of true adiabatic conditions, therefore the valid application of Kelvin’s law and the specimen motion due to the cyclic loading, known as’ motion effect’ [2-4].

Static torsion tests and fatigue torsion tests are carried out on quasi-isotropic carbon epoxy laminates consisting of 16 plies with different stacking sequences in order to determine the influence of the stacking sequence on the value of G12. Therefore, an automated torsion setup developed by COSARGUB, was used. The results are compared to those obtained with CLT (Classical Laminate Theory). Based on the conclusions in [2], the axial displacement was also monitored during the fatigue tests because this parameter seems to be more sensitive than the evolution of G12 in order to describe the torsional fatigue behaviour.

The paper presents a finite element formulation of the problem of moisture ab-/desorption in composite materials structural elements (e.g. laminated structures), and in which the materials show deviations from classical Fickian behaviour, principally through the characteristics of the matrix. The phenomena are considered as time-dependent and - through the previous consideration - nonlinear. The finite element formulation is analysed, a time integration scheme described and the selection of the time-step for time integration discussed. Some experimental data are verified.

The author presents techniques of structural properties identification which were developed during the past ten years at Brussels University (VUB) by the Composite System and Adhesion Research Group. The paper is concerned with the identification of stiffness in structural elements and elaborates quite extensively on a method developed by the author and H. Sol.

As far as experiments are concerned, life time prediction of adhesively bonded structures implies an extensive test program. The properties of the joint components, i.e. adhesive and composite adherend are experimentally determined as they serve as data input for finite element model of the bonded joint. The adhesive characterization includes shear tests performed on bulk adhesive material and on thin film form (rail shear and the Arcan ring test), tensile test, torsion test and resonalyser test, based on eigenfrequencies measurement of a free plate. An elastoplastic and a viscoelastic model are fitted to the experimental data. The unidirectional carbon-epoxy composite is also characterized, and single lap structures are fabricated and tested. Destructive techniques give information about (the evolution of) the residual strength of the tested joints. The evolution of physical properties are monitored with non destructive techniques like ultrasonic scanning. Experimental results on single lap joints are further used to assess numerical model of bonded structures.

In a composite continuum we have at least two distinct continua with thermomechanical characteristics available on macroscopic scale. Generally one of the continua has an outer surface with some measurable surface characteristics. The second phase, due to the presence of the first one, cannot fully develop his bulk properties during the processing in the region close to the geometrical interface. This “constraint”, together with some surface changes of the first phase, results in the existence of a third phase, the transition region between the two original continua or interphase region.

In this chapter a method is presented which determines the elastic properties of a composite material plate using experimentally measured resonant frequencies.

In this paper, a method is presented which determines anisotropic plate rigidities Dij using vibration data (eigenfreauencies, modeshapes).