The use of uncoupled damage models has been widely used over the years for the prediction of ductile fracture in engineering applications. Nevertheless, its applicability in the prediction of failure has been shown to be limited in the wide range of loading conditions encountered in different manufacturing processes. In order to enhance the formulation of former damage models, the Lode angle has been recently used to characterize the stress states along with the stress triaxiality. This new family of damage models has been demonstrated to give excellent results when proportional loading paths are considered, but its efficiency in more complex applications still need further analysis. To this end, a comparative study of former and recently developed uncoupled damage models is performed in this work. The identification of material parameters is done considering simple mechanical tests under different conditions. Then, the models are used to predict the onset and propagation of cracks during blanking, where numerical predictions are compared with experimental results. Finally, the selected damage models presented a remarkable overall performance in the range of clearances under study.

One of the most widely used cutting techniques in sheet metal forming processes for mass production is the blanking process. In this process, a metallic sheet is placed between a die and a blankholder, and is then cut by the action of a punch which moves downward. The quality of the final product is directly linked to the resulting shape of the cut edge. Due to the complexity of the separation step, the set-up of the blanking process in practice is often driven by empirical knowledge. Thus, an accurate numerical tool is extremely desirable to optimize the setting parameters of this technique and will lead to a better understanding of the entire process. The numerical approach must be able to deal with three main issues involved in blanking: large and localized deformation, friction and contact, and ductile fracture. Furthermore, due to requirements of mass production the punch velocity is normally high and the effects of the strain rate must also be considered. Several approaches have been developed in order to model this cutting process but its accuracy still presents some numerical challenges.

The fatigue and crack study of HSS S690QL steel welded pieces is first experimentally analyzed by fatigue tests on small samples of Base Material (BM), Heat Affected Zone (HAZ) and Weld Metal (WM) separately for several stress ratios, stress levels, on smooth and notched samples. The Lemaître-Chaboche fatigue model combined with the stress gradient method and the three sets of parameters are found to accurately describe the behavior of each material. Another experimental campaign with larger pieces (closer to the scale of the bridge connections) called "small scale samples", with/without welded stiffeners, with several geometries and with/without post-treatment has shown each effect separately: the scale and machining effects, residual stresses due to the welding, geometrical and post-treatments effects. Numerical simulations of these experiments are used to improve and validate the material data, with a study in real context and size. Then, the characterized fatigue damage model is coupled with the element deletion method to model the propagation of cracks along the welded samples and the numerical predictions are compared with experimental results in order to validate the approach.

The use of the blanking process has been widely spread in mass production industries. In this technique, the quality of the final product is directly related to the setting parameters of the process and the material response of the sheet. In the present work, a general framework based on the finite element method for the simulation of the sheet metal blanking process is presented. The proposed approach properly addresses all the numerical challenges related to blanking. First, an extension of elasto-viscoplastic constitutive equations for the large strain regime is used to take into account the material strain-rate sensitivity. Then, the inertial effects coming from high velocity operations are considered by means of an implicit time integration scheme. Moreover, the frictional contact interactions are simulated with the classical Coulomb law and an energetically consistent formulation of area regularization. Finally, ductile fracture is modeled thanks to the element deletion method coupled with a fracture criterion. The blanking process is then simulated for different setting parameters. The accuracy of this approach is evaluated by comparing the numerical predictions to experimental results for both quasi-static and dynamic conditions. Good agreement is found between experimental and numerical results for all cases.

Viscoelastic problems deserve great interest in Computational Mechanics literature. In the last years different approaches have been proposed in order to model viscoelastic problems, as in the case of the generalized Maxwell model and its numerical implementation. In particular Kaliske and Rothert (M. Kaliske and H. Rothert, Comput. Mech., 19(3): 228-239 (1997)) discussed basic reological models and the formulation of a generalized Maxwell model and the corresponding implementation of three dimensional viscoelastic model both for small and large strain cases. The numerical implementation addressed by Kaliske and Rothert is quite simple for small strain case and can be extended to a large strain format amenable to be included in finite element codes SOGDE and Metafor. The implementation of the discussed model in a 1D constitutive model, written in Matlab, is addressed. The well known relaxation and creep tests are simulated and compared with analytical results. Furthermore, the influence of constitutive parameters on the viscoelastic response is discussed. In addition, the model is implemented in Finite Element codes and the obtained results are compared with the 1D ones.

Method used to characterise and numerically study the fatigue behaviour of HSS welded plates

A large strain elasto-viscoplastic Perzyna model was proposed by Ponthot. The resultant numerical model computes the elastic response using an hypoelastic model. A viscoplastic multiplier was proposed, then both inviscid (elastoplastic) and elastic limiting cases can be easily recovered. More recently the authors have included this viscoplastic model in a large strain constitutive model based on hyperelasticity and multiplicative decomposition of deformation gradient tensor. In this work a brief summary of both, hypoelastic and hyperelastic based large strain models of viscoplasticity is provided. Large strain examples are simulated in order to test the discussed models. Different parameters of the constitutive model are tested in order to study the sensitivity of the resultant algorithm. From the obtained results can be said that both models show a very good agree and represent very well the characteristic of the viscoplastic constitutive model.

The numerical simulation of crack propagation in solids is of main importance in fracture mechanics and has been extensively studied over the years. Several approaches have been developed in order to describe the evolving geometry of a crack [1{6], but despite the research e orts some challenges are still present. A commonly used technique in Finite Element codes is the element deletion method due to the simplicity of its numerical implementation and possible extension to 3D. Furthermore, it is possible to couple this method with any failure criterion or damage model without additional considerations. This advantages are extremely desirable for numerical approaches involving high computational costs, e.g. the multi-scale computational homogenization [7, 8], where the element deletion method can be used at the micro-scale to simulate the nucleation, growth and coalescence of micro-voids [9].

Over the years, the simulation of manufacturing processes has introduced several numerical challenges for researchers in computational mechanics. In particular, the numerical modeling of sheet metal blanking process involves different numerical issues that must be carefully treated: a large and highly localized deformation in the shearing zone prior to fracture, complex contact interactions between the tools and the metallic sheet and finally, the ductile failure phenomenon. Despite that this process is one of the most widely used cutting techniques for mass production, the process parameters are normally set by empirical evidence due to the physical complexity resulting from the extreme amount of shearing involved. In addition, the strain-rate dependent behavior of the material must be taken into account due to high punch velocities encountered in practice. Thus, an accurate numerical tool is extremely desirable to optimize the setting parameters of this technique and will lead to a better understanding of the process.

Fracture prediction in blanking has gained great attention due to increasing requirements of high-quality products. In this work, the predictive capabilities of different uncoupled and coupled damage models, recently implemented in a fully implicit homemade FE code, for blanking process are compared. Some advanced models considered here include damage sensitivity to both triaxiality and Lode angle in their formulation. The material characterization is based on the history of internal variables during loading, where different mechanical tests are considered. Finally, numerical and experimental results are compared for a wide range of different geometries of blanking.

presentation of genesis of the project, his goal and objectives, the partnership, the task flowchart and main achievements

The use of high strength steel in bridges is studied through the welding, the fatigue and the stability for three designs and also in term of life cycle impacts.

For the case ofmetalswith large viscoplastic strains, it is necessary to define appropriate constitutive models in order to obtain reliable results from the simulations. In this paper, two large strain viscoplastic Perzyna type models are considered. The first constitutive model has been proposed by Ponthot, and the elastic response is based on hypoelasticity. In this case, the kinematics of the constitutive model is based on the additive decomposition of the rate deformation tensor. The second constitutive model has been proposed by García Garino et al., and the elastic response is based on hyperelasticity. In this case, the kinematics of the constitutive model is based on the multiplicative decomposition of the deformation gradient tensor. In both cases, the resultant numerical models have been implemented in updated Lagrangian formulation. In this work, global and local numerical results of the mechanical response of both constitutive models are analyzed and discussed. To this end, numerical experiments are performed and different parameters of the constitutive models are tested in order to study the sensitivity of the resultant algorithms. In particular, the evolution of the reaction forces, the effective plastic strain, the deformed shapes and the sensitivity of the numerical results to the finite element mesh discretization have been compared and analyzed. The obtained results show that both models have a very good agreement and represent very well the characteristic of the viscoplastic constitutive model.

numerical model based on the finite element method has been developed to simulate the blanking process. Thanks to this model we analyse the influence of punch velocities during blanking on the quality of the sheared edge and the characteristic parameters governing the process (maximum punch force and displacement, temperature increase). In this model, inertia, viscous and thermal effects are properly considered by means of a unified thermomechanical framework. A full remeshing approach is adopted to overcome the high distortion of elements due to large deformations, prior to fracture. The material strain-rate sensitivity is introduced by means of an extension of elasto-viscoplastic constitutive equations for the large strain regime. The inertial effects are considered thanks to an implicit time integration scheme. Crack propagation during the process is tracked using the element deletion method driven by an uncoupled damage criterion. Finally, the coupled thermomechanical problem is solved by an isothermal staggered scheme. Experimental and numerical results are compared for the entire range of punch velocities under consideration. Good agreement between both results has been found.