The “Construction of Machines and Vehicles” research group of Politecnico di Milano is looking for a highly skilled and motivated person to fill in the team. The person shall be involved in a number of tasks, with main focus on automotive.
Main part of the job will be related to an Italian Ministry of University and Research's funded project. The project aims to develop a tool to design high efficiency gears for electric vehicles, with the purposes of increasing efficiency and reducing noise and vibrations compared to the current state-of-art. At the bottom of this paragraph the abstract of the project is reported. The candidate will be, in particular, involved in the experimental campaign necessary to validate the models developed by other universities, partner of the project.
As complementary activities, the candidates will be involved in the current research activities and projects of the team, ranging from optimization to NVH, from design of vehicle’s components to company’s consultancy.
Workhours are expected to be flexible, and some of the work can be carried out remotely.
The candidate shall possess good English skills (B2/C1 level) and prove published papers in international journals.
Here the abstract of the main project the candidate will work on.
The proposed project focuses on the development of innovative models, methods and tools to improve the design process of high-performance gears aiming to increase efficiency and reduce vibrations produced and noise radiated by power transmissions. Notwithstanding the large body of existing scientific literature, the topic remains of primary importance for the industry: for example, in the automotive sector, gear efficiency and transmission noise has become an aspect of considerable importance because, in electric vehicles, the gearbox system is the main mechanical source of losses and noise.
From scientific literature, the most commonly adopted modelling strategies can be classified in analytical, Finite Element, and multibody. This research project is focused on the development of a new methodology for the simulation of high-performance gears based on the concept of flexible multibody model. In particular, the main goal is to identify those parameters and constructive elements of gear pairs that are most relevant for the generation of vibration and noise. This goal is achieved through the development of simulation models to predict vibrations phenomena and evaluate the performance of the gears in complex scenarios. In order to use such models as design tools, they shall necessary be, at the same time, highly accurate and computationally efficient. Two different approaches will be followed throughout the project: the first is based on the concept of the pseudo-rigid body, the latter on combinations of rigid and flexible bodies in which pseudo-rigid and flexible elements coexist.
When dealing with nonlinear dynamics of reduced-order models, the main challenge is the identification of the parameters and their dependency on operative conditions. This task will be performed through model updating based on optimization techniques in an innovative numerical scheme based on fuzzy arithmetic, to assess the uncertainties related to the parameters’ identification and to the simplification of the model itself. Coupled with the multibody model, an innovative acoustic model will be also proposed, converting the vibration output of the multibody models into an acoustic measure. Finally, the models will be verified against experimental results.
Scientific literature lacks a significant body of experimental results to be used as reference, thus an important experimental campaign will be carried out to obtain the necessary reference data, which will be then made available to the public.
The outcomes of this research project will make it possible to develop a simulation model capable of identifying, and therefore predicting, complex operating phenomena in gears such as transients, gear shifts, non-linear dynamics phenomena, noise, and backlash. At the current state of the art, these kinds of phenomena cannot be evaluated within a single simulation environment and are generally not suited for design optimization.