Éditeur(s) : HAL CCSD American Institute of Mathematical Sciences
Résumé : International audience
In this paper, shape optimization tools are applied to coastal engineering. A barred beach protection study is described, using a new type of water wave attenuator device, namely geotextile tube, to reduce the mobilization of sediments. A global optimization algorithm, able to pursue beyond local minima, is used to search for the optimum properties of the geotextile tube. The optimal configuration is then post-validated for generated currents.
ISSN: 1531-3492

hal-00475636
https://hal.archives-ouvertes.fr/hal-00475636
DOI : 10.3934/dcdsb.2010.13.759

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
While vertical motion induced by long-term geological loads is often used to estimate the flexural rigidity of the lithosphere, we intend to evaluate the shear rigidity of the lithosphere using horizontal motion. Our approach considers that the rigidity of the lithosphere may be defined as its resistance to horizontal tectonic lateral forces. In this case, a spatial distribution of an effective shear rigidity can be estimated from the analysis of the interseismic velocity fields. We consider the Western United States zone where weakly strained areas (e.g., the Sierra Nevada) are connected with areas of large strain rate (e.g. San Andreas Fault system). By inverting interseismic strain distribution measured by geodetic methods, we infer the effective shear rigidity of the lithosphere. The forward problem is defined using the equations of linear elasticity. The inversion relies on the minimization of the sum of a quadratic measure of the differences between measured and modelled velocity fields. The functional also includes regularization terms for the parameters of the model. The gradient of the functional with respect to the minimization parameters is computed using an adjoint formulation. This permits the treatment of large dimensional minimization problems. Finally, a measure of the uncertainty of our inversion is illustrated through the covariance matrix of the parameters at the optimum. The optimization chart is validated on two synthetic velocity distributions. Then, the effective shear rigidity variations of the Western United States are estimated using the CMM3 interseismic velocities. The inversion displays low effective rigidities along the San Andreas Fault system, the Mojave Desert and in the Eastern California Shear Zone, while rigid areas are found in the Sierra Nevada and in the South Basin and Range. Finally, we discuss the differences between our strain rate and rigidity maps with previously published results for the Western United States.
ISSN: 0040-1951

Droits : http://hal.archives-ouvertes.fr/licences/copyright/
hal-01505024
https://hal.archives-ouvertes.fr/hal-01505024
https://hal.archives-ouvertes.fr/hal-01505024/document
https://hal.archives-ouvertes.fr/hal-01505024/file/Paper_tectophysics.pdf
DOI : 10.1016/j.tecto.2017.03.015

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
ISSN: 0378-3839

hal-00385746
https://hal.archives-ouvertes.fr/hal-00385746

Éditeur(s) : HAL CCSD Elsevier
Résumé : International audience
We show how to introduce the Value at Risk (VaR) concept in the analysis of the adaptation of the shape of a soft bed to a flow knowing the Probability Density Function (PDF) of the responses of the shape to flow perturbations (bed receptivity). Our aim is to quantify our confidence on simulation scenarios by an available morphodynamic model for the shape. The approach permits to perform this task at low complexity as it does not require any sampling of the bed receptivity parameter space. The paper goes beyond stationarity for the variability of the bed receptivity by linking its dynamics to the bottom morphodynamics through an original transport equation for the local bed receptivity standard deviation. The approach has been applied to the analysis of bed morphodynamics based on minimization principles. The results show the importance of including uncertainty information during the coupling and not only eventually through simple margins on the results.
ISSN: 0045-7930

hal-01054937
https://hal.archives-ouvertes.fr/hal-01054937
https://hal.archives-ouvertes.fr/hal-01054937/document
https://hal.archives-ouvertes.fr/hal-01054937/file/paper_var_litto.pdf
DOI : 10.1016/j.compfluid.2013.10.021

Éditeur(s) : HAL CCSD
Résumé : International audience
Optimizing the resource extraction and limiting the environmental constraint motivate to accurately characterize reservoirs. The extraction of resources induces volume change of the reservoir causing displacements in the surrounding medium up to the surface. The resulting surface deformation is usually significant enough to be measurable by geodetic instruments (GPS, satellites, tiltmeters, strainmeters). However, the data needs to be associated to a reservoir model using an inverse scheme to allow the estimation of physical properties of the reservoir. Complete geodetic data including InSAR, GPS, tilt, strain and gravimetric data are uncommon for conventional, unconventional or mining exploitations. In this study, we work on the one site that gathers a wide range of geodetic data: the salt exploitation of Vauvert, France, operated by KemOne. Our approach aims at estimating the volume changes of the cavities created by the extraction combining all the available geodetic data. To do so, we minimize a global functional including weighting factors for each data type. The weighting factors are optimally estimated for our study and can be adapted depending on the issue. Hence, an optimal distribution of data can be suggested for a specific reservoir characterization.
79th EAGE Conference
Paris, France

Droits : http://hal.archives-ouvertes.fr/licences/copyright/
hal-01575426
https://hal.archives-ouvertes.fr/hal-01575426
https://hal.archives-ouvertes.fr/hal-01575426/document
https://hal.archives-ouvertes.fr/hal-01575426/file/reservoir-characterization-inversion.pdf
DOI : 10.3997/2214-4609.201700987