Numerical simulation has become essential in the design cycle of complex mechanical parts subjected to static, dynamic or thermal loading. To increase their predictability, the numerical prototypes representing these parts need to integrate the maximum number of physical phenomena on increasingly complex geometries with increasingly fine modeling in order to capture local effects. Very often, to select the best solutions, multiple parametric analyses are necessary around nominal designs. Thus, an "extra layer" of optimization or experimental design is to be considered. All of these requirements lead to complex simulations that require high-performance calculation tools in order to efficiently support the design cycle. To validate its products, Valeo implements high-performance solutions to accelerate the simulation process. Three examples of industrial applications are presented : Vibration analysis of a battery cooler The (current) particularity of a battery cooler in an automobile is its size. To accelerate the process of meshing the parts, the automation of this step is considered. It is based on the principle that a fine mesh allows to fit the shape of the parts without too many manual alterations: gain in implementation time but significant increase of information (degrees of freedom) in the final model. Consequently, the finite element models generated can be of significant size (around a hundred million degrees of freedom) on this type of structure. The implementation of a high performance simulation chain (mesh, dynamic solver and post-processing) for the treatment of a battery cooler has allowed a gain in terms of reactivity and time in the design. Crimping of a cooling module manifold The heat exchangers in a car (radiator, CAC ...) contain 2 main sub-assemblies, the bundle by which the heat exchange between the external fluid and the internal fluid takes place, and the inlet/outlet tanks through which the cooling liquid passes. The junction between these two sub-assemblies requires a crimping operation of the teeth of the bundle on the tanks, while taking into account the presence of a seal between the 2 parts. The integration of these non-linearities on a precise and large finite element model requires fast and robust resolution tools (quasi-static approach) to respond in reasonable time (ideally less than a week) to the project teams. Optimizing an HVAC system Due to its location and functionality, an HVAC (Heating, Ventilation and Air Conditioning system) is a major source of noise in an automobile. The fan wheel is the main source of aerodynamic noise and unbalance, with a significant acoustic radiation favored by the large surface of the complete HVAC box. Several design parameters have a significant influence on the radiated noise, such as the shape and position of the decoupling studs between the fan and the air conditioner, as well as the material thickness of the plastic parts. Thanks to high performance resolution tools, the optimization process is made possible in reasonable restitution times (less than a dozen hours, which makes overnight resolution possible). The three analyses presented highlight the high performance numerical simulations deployed industrially to address complex problems, with different physical phenomena to predict on precisely described multi-component products. Structural optimization is also a major asset during the design cycle of a part, which makes it necessary to interface with fast simulation tools. The challenge is to get the most out of advanced simulation and optimization techniques in a short time, compatible with the rapid development cycles observed in the automotive industry.