Search Results

Current vehicle regulations demand for a challenging decrease in the overall exterior noise as a benefit for the health of citizens and road users. New limits have been implemented in UN R51.03 (based on ISO 362-1:2015) to reduce the emitted noise both at constant speed and in full load so as to cover most of the real urban driving conditions. In order to achieve those targets the carmakers have to refine the trim of their vehicles and an experimental approach can take place too late. This paper shows a method for the pass-by noise simulation exploiting the numerical transfer functions and a library of experimentally characterized sources with the aim to reduce the noise and find out a better tradeoff between costs and effectiveness of the modifications. Moreover a simple software tool for the treatment of the data and to ease the workflow has been created and used for the rank assessment of the different paths.

The time to market in the automotive industry is constantly decreasing pushing the carmaker companies to increase the efforts in numerical simulations and to decrease the number of prototypes. In the NVH field, this time constraint reflects in moving the well-established finite element simulations towards the so called “full-vehicle simulations”. Specifically, the CAE techniques should be able to predict the complete behavior of the vehicles in mission conditions, so to reproduce some usual tests, such as the “coast down” test on different roads. The aim of this paper is to present a methodology to improve rolling noise simulations exploiting an integrated full-vehicle approach. An accurate modeling of all the subsystems is needed, with particular attention to the wheels and the suspension systems. Therefore, the paper firstly covers the modeling approach used to obtain the FE models of tires and suspension system.

Within the automotive industry, a typical way to account for tires in a roadnoise mission simulation is to use the “modal model” supplied by tire manufacturers. Even though this kind of models is certified by the suppliers and is very simple to use, it has the drawback to be disconnected from the physical description of the tire. This reflects in limiting the carmaker company to be able only to request certain modal characteristics to the supplier. The aim of this paper is to present an accurate, yet easy to use, methodology to develop an FE model of a tire, to be used in a full-vehicle simulation. The determined model must be connected to the tire physical properties. These properties are not measured directly, but determined by tuning a properly created geometric FE model to the measured point inertances of the inflated tire. This allows creating the model only by using an optimization algorithm to tune such properties.

Methodologies of a vehicle assessment through computer simulation comes to enable every day to preview difficulties in developing models, which also contributes to reducing the time to develop a new model. For initial assessment of the vibroacoustic behavior of a vehicle, in the early months of development, the frequency response functions, known as inertance (a/F), are analyzed, at the points of attachment of the engine and suspension to the body still in the Body-in-White configuration. Usually the finite element simulations are performed up to the limit of 300Hz. In the aim at increasing the range of inertance analysis, enabling a more comprehensive analysis in NVH, the results by elements finites simulation were compared, in this work, with the results obtained in experimental measurements focused on the validation of this simulation methodology until the limit of 800Hz.

In the last years engineers have to deal with multiple, often conflicting targets, where improvement of one quantity leads to deterioration of others, therefore it is impossible to obtain simultaneous structure enhancements without automatic optimizations tools. The so-called trade-offs have to be applied, providing less efficient modifications, nevertheless, for all of the design objectives. The Pareto front is a method that helps to determine a set of equipotential designs. In order to explore entire design space, response surface methodology supplemented by genetic algorithms is often used. In the work presented, the Gaussian Processes Methodology and an Adaptive Range Multi-Objective Genetic Algorithm - ARMOGA were implemented. Basing on the solutions obtained by the design of experiment, response surface methodology is used to predict the values of the measured outputs throughout the full range of interest.