In this work we present the first implementation of the incremental scheme for coupled cluster linear-response frequency-dependent dipole polarizabilities. The implementation is fully automated and makes use of the domain-specific basis set approach. The accuracy of the approach is determined on the basis of a test suite of 47 molecules and small clusters. The local approximation in the coupled cluster singles and doubles polarizability exhibits a mean error of 0.02% and a standard deviation of 0.32% when using a third-order incremental expansion. With the proposed approach, it is possible to compute polarizabilities with larger basis sets compared to the canonical implementation and thus it is possible to obtain higher total accuracy. The incremental scheme yields the smallest errors for weakly-bound and quasi-linear systems, while two- and three-dimensional (cage-like) structures exhibit somewhat larger errors as compared to the full test set. / Dieser Beitrag ist aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

In this work we present the first implementation of the incremental scheme for coupled cluster linear-response frequency-dependent dipole polarizabilities. The implementation is fully automated and makes use of the domain-specific basis set approach. The accuracy of the approach is determined on the basis of a test suite of 47 molecules and small clusters. The local approximation in the coupled cluster singles and doubles polarizability exhibits a mean error of 0.02% and a standard deviation of 0.32% when using a third-order incremental expansion. With the proposed approach, it is possible to compute polarizabilities with larger basis sets compared to the canonical implementation and thus it is possible to obtain higher total accuracy. The incremental scheme yields the smallest errors for weakly-bound and quasi-linear systems, while two- and three-dimensional (cage-like) structures exhibit somewhat larger errors as compared to the full test set. / Dieser Beitrag ist aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Nous proposons dans cette thèse diverses approches, pour l'amélioration de la modélisation et la simulation multi-échelle du comportement des matériaux composites. La modélisation précise et fiable de la réponse mécanique des matériaux composite demeure un défi majeur. L'objectif de ce travail est de développer des méthodologies simplifiées et basées sur des techniques d'homogénéisation existantes (numériques et analytiques) pour une prédiction efficiente du comportement non-linéaire de ces matériaux. Dans un premier temps un choix à été porté sur les techniques d'homogénéisation par champs moyens pour étudier le comportement élastoplastique et les phénomènes d'endommagement ductile dans les composites. Bien que restrictives, ces techniques demeurent les meilleures en termes de coût de calcul et d'efficacité. Deux méthodes ont été investiguées à cet effet: le Schéma Incrémental Micromécanique (SIM) en modélisation mono-site et le modèle Mori-Tanaka en modélisation multi-site (MTMS). Dans le cas d'étude du comportement élastoplastique, nous avons d'une part montré et validé par la méthode des éléments finis que la technique d'homogénéisation SIM donne un résultat plus précis de la modélisation des composites à fraction volumique élevée que celle de Mori-Tanaka, fréquemment utilisée dans la littérature. D'autre part nous avons étendu le modèle de Mori-Tanaka (M-T) généralement formulé en mono-site à la formulation en multi-site pour l'étude du comportement élastoplastique des composites à microstructure ordonnée. Cette approche montre que la formulation en multi-site produit des résultats concordants avec les solutions éléments finis et expérimentales. Dans la suite de nos travaux, le modèle d'endommagement ductile de Lemaître-Chaboche a été intégré à la modélisation du comportement élastoplastique dans les composites dans une modélisation multi-échelle basée sur le SIM. Cette dernière étude révèle la capacité du modèle SIM à capter les effets d'endommagement dans le matériau. Cependant, la question relative à la perte d'ellipticité n'a pas été abordée. Pour finir nous développons un outil d'homogénéisation numérique basé sur la méthode d'éléments finis multi-échelles (EF2) en 2D et 3D que nous introduisons dans le logiciel conventionnel ABAQUS via sa subroutine UMAT. Cette méthode (EF2) offre de nombreux avantages tels que la prise en compte de la non-linéarité du comportement et de l'évolution de la microstructure soumise à des conditions de chargement complexes. Les cas linéaires et non-linéaires ont été étudiés. L'avantage de cette démarche originale est la possibilité d'utilisation de toutes les ressources fournies par ce logiciel (un panel d'outils d'analyse ainsi qu'une librairie composée de divers comportements mécaniques, thermomécaniques ou électriques etc.) pour l'étude de problèmes multi-physiques. Ce travail a été validé dans le cas linéaire sur un exemple simple de poutre en flexion et comparé à la méthode multi-échelle ANM (Nezamabadi et al. (2009)). Un travail approfondi sera nécessaire ultérieurement avec des applications sur des problèmes non-linaires mettant en évidence la valeur de l'outil ainsi développé / We propose in this thesis several approaches for improving the multiscale modeling and simulation of composites’ behavior. Accurate and reliable modeling of the mechanical response of composite materials remains a major challenge. The objective of this work is to develop simplified methodologies based on existing homogenization techniques (numerical and analytical) for efficient prediction of nonlinear behavior of these materials. First choice has been focused on the Mean-field homogenization methods to study the elasto-plastic behavior and ductile damage phenomena in composites. Although restrictive, these techniques remain the best in terms of computational cost and efficiency. Two methods were investigated for this purpose: the Incremental Scheme Micromechanics (IMS) in One-site modeling and the Mori-Tanaka model in multi-site modeling (MTMS). In the framework of elastoplasticity, we have shown and validated by finite element method that the IMS homogenization results are more accurate, when dealing with high volume fraction composites, than the Mori-Tanaka model, frequently used in the literature. Furthermore, we have extended the Mori-Tanaka's model (MT) generally formulated in One-site to the multi-site formulation for the study of elasto-plastic behavior of composites with ordered microstructure. This approach shows that the multi-site formulation produces consistent results with respect to finite element and experimental solutions. In the continuation of our research, the Lemaître-Chaboche ductile damage model has been included to the study of elasto-plastic behavior in composite through the IMS homogenization. This latest investigation demonstrates the capability of the IMS model to capture damage effects in the material. However, the issue on the loss of ellipticity was not addressed. Finally we develop a numerical homogenization tool based on computational homogenization. This novel numerical tool works with 2D and 3D structure and is fully integrated in the conventional finite element code ABAQUS through its subroutine UMAT. The (FE2) method offers the advantage of being extremely accurate and allows the handling of more complex physics and geometrical nonlinearities. Linear and non-linear cases were studied. In addition, its combination with ABAQUS allows the use of major resources provided by this software (a panel of toolbox for various mechanical, thermomechanical and electrical analysis) for the study of multi-physics problems. This work was validated in the linear case on a two-scale analysis in bending and compared to the multi-scale method ANM (Nezamabadi et al. (2009)). Extensive work will be needed later with applications on non-linear problems to highlight the value of the developed tool

In this thesis, the implementation of the incremental scheme for open-shell systems with unrestricted Hartree-Fock reference wave functions is described. The implemented scheme is tested within robustness and performance with respect to the accuracy in the energy and the computation times. New approaches are discussed to implement a fully automated incremental scheme in combination with the domain-specific basis set approximation. The alpha Domain Partitioning and Template Equalization are presented to handle unrestricted wave functions for the local correlation treatment. Both orbital schemes are analyzed with a test set of structures and reactions. As a further goal, the DSBSenv orbital basis sets and auxiliary basis sets are optimized to be used as environmental basis in the domain-specific basis set approach. The performance with respect to the accuracy and computation times is analyzed with a test set of structures and reactions. In another project, a scheme for the optimization of auxiliary basis sets for uranium is presented. This scheme was used to optimize the MP2Fit auxiliary basis sets for uranium. These auxiliary basis enable density fitting in quantum chemical methods and the application of the incremental scheme for systems containing uranium. Another project was the systematical analysis of the binding energies of four water dodecamers. The incremental scheme in combination with the CCSD(T) and CCSD(T)(F12*) method were used to calculate benchmark energies for these large clusters.

In this thesis, the implementation of the incremental scheme for open-shell systems with unrestricted Hartree-Fock reference wave functions is described. The implemented scheme is tested within robustness and performance with respect to the accuracy in the energy and the computation times. New approaches are discussed to implement a fully automated incremental scheme in combination with the domain-specific basis set approximation. The alpha Domain Partitioning and Template Equalization are presented to handle unrestricted wave functions for the local correlation treatment. Both orbital schemes are analyzed with a test set of structures and reactions. As a further goal, the DSBSenv orbital basis sets and auxiliary basis sets are optimized to be used as environmental basis in the domain-specific basis set approach. The performance with respect to the accuracy and computation times is analyzed with a test set of structures and reactions. In another project, a scheme for the optimization of auxiliary basis sets for uranium is presented. This scheme was used to optimize the MP2Fit auxiliary basis sets for uranium. These auxiliary basis enable density fitting in quantum chemical methods and the application of the incremental scheme for systems containing uranium. Another project was the systematical analysis of the binding energies of four water dodecamers. The incremental scheme in combination with the CCSD(T) and CCSD(T)(F12*) method were used to calculate benchmark energies for these large clusters.

In this thesis, several development steps for the incremental method are presented. At first, the extension of the incremental scheme to other quantities than the energy is advanced in terms of molecular dipole moments. In this context, a revised error correction as well as the template localization for the treatment of aromatic systems are introduced. As a second enhancement, a new implementation of the template localization ensures a higher stability of this algorithm step and, thus, of the incremental scheme. Finally, pair natural orbitals (PNOs) are utilized in the incremental method with the aim of an increased efficiency. The PNO approach is re-assessed in context of the incremental expansion leading to both small incremental and PNO local errors for reaction, intermolecular interaction and cluster binding energies. The higher efficiency due to the twofold reduction of the computational efforts by the PNO and the incremental approaches is demonstrated for molecular clusters. Additionally, the complete basis set (CBS) limit is targetted by using the efficient MP2-based focal-point approach to the incremental scheme (with and without PNOs). Finally, based on these improvements of the performance, the PNO-based incremental scheme is applied to support a computational study regarding the modelling of the reaction mechanism for the base-catalyzed twin polymerization.