Pascal Doumalin

Experimental and numerical characterisation of in-plane deformation in two-phase materials

The aim of the present work consists in the comparison of in-plane strain fields with out-of-plane displacements in micro-areas of an Ag/Ni-composite after a macroscopic compressive deformation of 8.6%. The in-plane deformations in an Ag/Ni-composite have been analysed experimentally with a high resolution object grating technique and numerically using the finite element method. The out-of-plane displacements were measured with an atomic force microscope (AFM). The development of local strain fields in micro-areas at the surface of an Ag/Ni-composite was simulated numerically using the FE-method in plane strain condition. A real cut-out of the microstructure served as input for the calculation. The out-of-plane displacements determined by AFM measurements were used further to correct the in-plane values of strains evaluated by the object grating technique. The roughness on the surface of the sample was characterised by fractal dimensions and compared with the in-plane strains in the same micro-region.

Micromechanical Applications of Digital Image Correlation Techniques

Digital image correlation (DIC) is applied to images obtained with a scanning electron microscope (SEM) in order to determine the local strain field in a heterogeneous material on a representative domain of the microstructure, millimetric in size, the local contrast being provided by a microgrid deposited on the surface of the sample, with a typical 5μm step. The specific difficulties encountered when applying this classical technique for CCD images to SEM images are discussed. They are related to geometric images defects, high noise levels or imperfect scanning or to the non optimal local contrast, which are all potential sources of loss of accuracy. Several procedures to correct such imperfections are presented. The investigation of the residual strain field in a ferritic-martensitic steel after 1% total strain under uniaxial traction is then shortly presented.

Refinement of digital image correlation technique to investigate the fracture behaviour of refractory materials

Refractory materials exhibit a heterogeneous microstructure consisting in coarse aggregates surrounded by fine grains that form an aggregate/matrix composite. This heterogeneous microstructure often leads to a complex mechanical behaviour during loading. This paper is devoted to the study, thanks to an optical method, Digital Image Correlation (DIC), of the fracture behaviour of two industrial refractory materials in relation with their microstructure resulting from both the chosen constituents and the sintering process. The aim is here, specifically, to highlight and to characterize the evolution of kinematic fields (displacement and strain) observed at the surface of sample during a wedge splitting test typically used to quantify the work of fracture. DIC is indeed a helpful and effective tool, in the topic of experimental mechanics, for the measurement of deformation in a planar sample surface. This non-contact optical method directly provides full-field displacements by comparing the digital images of the sample surface obtained before and during loading. In the present study, DIC has been improved to take into account the occurrence of cracks and performed so as to better identify the early stage of the cracking behaviour. The material transformation, usually assumed homogeneous inside each DIC subset, is thus more complex and a discontinuity of displacement should be taken into account. Then each subset which crosses a crack can be cut in two parts with different kinematics. By this way, it is possible to automatically find the fracture paths and follow the crack geometries (length, opening).

Verification of a spherical plain bearing finite-element model using scattered light photoelasticity tests

Spherical plain bearings are used on many components of Airbus aircrafts like engine-to-pylon and pylon-to-wing links. Their design is based on contact pressure distribution on spherical surfaces. Until now, empirical laws are applied to determine it, and the aim of this work is to develop a numerical modelling in order to validate the use of these laws. The assumptions on the numerical model and their significance have been verified and meas-ured by scattered light photoelasticity (SLP) on an epoxy replica of the bearing. The analysis of this type of bearing is difficult due to its three-dimensional behaviour and its spherical confined contact. The SLP technique allows getting information on stress fields inside an epoxy specimen with the analysis of photoelastic fringes. Comparing photoelastic fringes with simulated results, the finite-element model has been improved to be more representative of the real life bearing. In addition, some critical factors on the bearing behaviour have been obtained.

Characterization of Local Strain Distribution in Zircaloy-4 and M5 ® Alloys

Zirconium alloys with low alloying content are mainly used in the nuclear industry as structural materials because of their superior properties in terms of neutron transparency, mechanical strength, and corrosion resistance. In order to further improve the corrosion resistance as well as the integrity of Zr based cladding tubes under severe thermomechanical loading, the M5® alloy was developed to replace stress-relieved Zircaloy-4. An experimental study conducted at the macroscopic scale between 20 and 500°C shows that the mechanical behavior of the studied Zr based alloys depends on the metallurgical state (stress-relieved or recrystallized) rather than on the chemical composition. To try to understand these mechanical differences, an experimental multiscale investigation was devised at ambient temperature (20°C) in order to characterize the strain distribution at the scale of the grains and at that of the representative volume element. Local strain fields were measured by means of a microscale digital image correlation technique, based on microgrid deposits and scanning electronic microscopy (SEM). Tensile tests were performed inside the SEM chamber. Here, the original method of strain distribution quantification based on statistical strain field analysis is used. First, this analysis reveals a particular strain distribution consisting of bands with an orientation greater that 45° with regards to the direction of macroscopic tension, and second, shows that these interaction lengths are much greater than the average size of the grains, which clearly demonstrates that local investigations cannot be limited to a few grains. Therefore, the macroscopic mechanical response of these materials is not only governed by intragranular heterogeneities but by the local deformations which become organized between the grains in a pattern of bands at a mesoscale, which is determined by medium to long-range interactions. The difference of values in the band characteristics could partly explain the anisotropic global behavior of these materials linked with their microstructure.

Detection of cracks in refractory materials by an enhanced digital image correlation technique

This paper is devoted to the study of the fracture behaviour of two industrial refractory materials thanks to the development of a new technique of digital image correlation (DIC). DIC, already known as a helpful and effective tool for the measurement of displacement and deformation fields in materials, has been adapted to take into account displacement discontinuities as cracks. The material transformation, usually assumed homogeneous inside each DIC subset, is thus more complex, while each subset can be cut in two parts with different kinematics. By this way, it is possible to automatically find the fracture paths and follow the crack geometries (length, opening) during the loading with a higher spatial resolution than the one obtained by standard DIC. After having presented the principle of the new technique, its metrological performances are assessed from synthetic images and the choice of crack detection criterion is discussed. The capacity of this new technique is shown through a comparative study with standard DIC. Its application is led on magnesia-spinel refractory materials, specifically to highlight and to characterize the evolution of kinematic fields (displacement and strain) observed at the surface of sample during a wedge splitting test typically used to quantify the work of fracture. We show that refractories with aggregates of iron aluminate spinel present a fracture mechanism with crack branching and can dissipate more energy thanks to a longer crack network.

In vitro production and biomechanical experimental analysis of thoracolumbar burst fractures

Spinal traumatisms constitute a current pathology in traumatology (10%) and thoracolumbar burst fractures are the most frequent and generally concern young patients. Burst fractures are characterised by compression loadings involved lesion of the vertebrae endplate between the anterior and the posterior walls (Magerl et al. 1994). Determinations of instability and treatment indications remain particular points of questioning. Various treatment solutions exist: immobilisation by an aircast, surgical treatment by posterior or anterior instrumentation and, recently, kyphoplasty with cement injection after vertebral reduction. Choice of the treatment can be delicate and few biomechanical studies have been performed to establish the mechanical response of each option. For that, a cadaveric model of the burst fracture is helpful. This study concerns the development of experiments to generate burst fractures.

Biomechanical analysis of the thoracolumbar spine under physiological loadings: Experimental motion data corridors for validation of finite element models

Biomechanical studies that involve normal, injured or stabilized human spines are sometimes difficult to perform on large samples due to limited access to cadaveric human spines and biological variability. Finite element models alleviate these limitations due to the possibility of reusing the same model, whereas cadaveric spines can be damaged during testing, or have their mechanicals behaviour modified by fatigue, permanent deformation or structural failure. Finite element models need to be validated with experimental data to make sure that they represent the complex mechanical and physiological behaviour of normal, injured and stabilized spinal segments. The purpose of this study is to characterize the mechanical response of thoracolumbar spine segments with an analytical approach drawn from experimental measurements. A total of 24 normal and fresh cadaveric thoracolumbar spine segments (T11–L3), aged between 53 and 91 years, were tested in pure flexion/extension, lateral bending and axial torsion using a specific experimental setup. Measurements of global and intervertebral angle variations were performed using three-dimensional mark tracking methods. Load/angle curves for each loading were fitted by a logarithmic approach with two coefficients. The coefficients for the functions describing the response of the spinal segments are given and constitute predictive models from experimental data. This work provides data corridors of human thoracolumbar spine motion segments subjected to pure bending in the three physiological planes. These data could be very useful to validate finite element models of the human spine.

Development of an experimental model of burst fracture with damage characterization of the vertebral bodies under dynamic conditions

Background: Burst fractures represent a significant proportion of fractures of the thoracolumbar junction. The recent advent of minimally invasive techniques has revolutionized the surgical treatment of this type of fracture. However mechanical behaviour and primary stability offered by these solutions have to be proved from experimental validation tests on cadaveric specimens. Therefore, the aim of this study was to develop an original and reproducible model of burst fracture under dynamic impact. Methods: Experimental tests were performed on 24 cadaveric spine segments (T11‐L3). A system of dynamic loading was developed using a modified Charpy pendulum. The mechanical response of the segments (strain measurement on vertebrae and discs) was obtained during the impact by using an optical method with a high‐speed camera. The production of burst fracture was validated by an analysis of the segments by X‐ray tomography. Findings: Burst fracture was systematically produced on L1 for each specimen. Strain analysis during impact highlighted the large deformation of L1 due to the fracture and small strains in adjacent vertebrae. The mean reduction of the vertebral body of L1 assessed for all the specimens was around 15%. No damage was observed in adjacent discs or vertebrae. Interpretation: With this new, reliable and replicable procedure for production and biomechanical analysis of burst fractures, comparison of different types of stabilization systems can be envisaged. The loading system was designed so as to be able to produce loads leading to other types of fractures and to provide data to validate finite element modelling. HIGHLIGHTSA specific experimental dynamic setup for spinal loading has been developed.A biomechanical analysis was performed during impact on spines from a dynamic optical method.Mechanical effects of a dynamic impact on spine were characterized by X‐ray imaging and optical analysis.A reproducible model of burst fracture on human cadaveric specimens was developed.

Metrological Analysis of the DIC Ultimate Error Regime

In DIC, the “ultimate error regime” corresponds to situations for which the shape function used to describe the material transformation perfectly matches the actual one. We propose to confront results obtained from numerically-shifted images with the predictions of theoretical models developed in the literature to describe bias and random error evolutions with respect to the imposed displacement. Results show the overall good predictions of these models but small deviations arise, mainly around integer values of imposed displacements for noisy images. These deviations are interpreted as the unrepresentativeness of the underlying hypotheses of the theoretical models in these particular cases.