RICERCA ITALIANA     

Research program

THE ENHANCEMENT OF MECHANICAL TRANSMISSIONS FOR THE REDUCTION OF THE ENERGY CONSUMPTION

University Co-ordinator

Politecnico di BARI - INGEGNERIA MECCANICA E GESTIONALE - BARI(BA)

Research Unit Leader

Luigi MANGIALARDI

Description

The well known potential advantages of the CVT transmissions, as for example the possibility to improve the passenger comfort, and reduce the vehicle fuel consumption and emissions, has lead to an increased interest of the car manufacturer Companies in this kind of transmissions. Almost all car Companies have, among their several models, some vehicles equipped with the pushing V-belt CVT transmission (e.g. Fiat, Bmw, Audi, Rover) or with the toroidal traction drive (e.g. Nissan with the NSK half-toroidal). Also some gear production Companies as Getrag-Bari is intereseted in the full-toroidal traction drive by Torotrak UK. Moreover, in the field of motorcycles the rubber V-belt CVT has become the most used variator, thanks to its simplicity and low cost. Thus, it is clear that a better and a more detailed comprehension of the dynamical behavior of CVTs would allow to single out the geometric, kinematical, and dynamical parameters, that has to be adjusted and modified in order to optimize the performances of the CVT variator and in particular its mechanical efficiency. This, in turn, will allow for a better power management, a fuel saving improvement, a emission reduction, and a better comfort.
The results achieved by the Research Unit of Bari in the last two years, pointed out the necessity of performing a much more detailed experimental investigation of the CVT variator, and also the necessity of a better comprehension of the sliding friction phenomenon between the belt and the pulley in the case of the V-belt CVT, or between the rollers and discs in the toroidal traction drives. In fact, the power and torque transmission in these typologies of variators is allowed by the tangential stresses that are produced at the interface between the contacting surfaces. Since the tangential stresses and the sliding motion largely influence the energy dissipation of the variators, it is clear that a insight into the sliding friction phenomenon is indispensable to improve the CVT performances. In the case of toroidal traction drives, it is known that the tangential actions are caused by the strain motion of the traction oil, that is almost "trapped" between the two sliding surfaces. But it is not yet well-quantified what is the effect of the local temperature increment on the mechanical efficiency of the variator and its traction capabilities. The situation is less clear for the metal belt CVTs. In this case it is not clear what is the actual lubrication regime. Different models can be found in the literature. Some of them describe the lubrication as a EHL regime, some other are inclined to consider a boundary or mixed lubrication regime. On the other hand, as regards the rubber V-belt CVT variator, the simple assumption of a constant friction coefficient is not well-suited to describe the belt-pulley interaction and the traction capabilities of the variator. Some of the models proposed in the literature, indeed, lead to a set of equations describing the belt motion, of which the solution is highly influenced by any small change of the initial conditions (in some sense these models appears to be not well-posed since any small variation of the initial data produce a large variation of the solution of equations, this is not physically consistent since the measure of those initial conditions cannot be done with the accuracy requested by those model). Thus, the Research Unit of Bari intends to follow two parallel directions. The first one consists in performing a campaign of experimental investigations on the metal belt CVT, needed to validate and improve the theoretical dynamical model of the variator, already obtained during the last four years of research (thanks also to the financial support of COFIN 2000 and COFIN 2002). After the validation, the theoretical model will be given to the other Research Units as a simulation tool to be used during the design of the control strategy of the CVT variator. The second direction that this Research Unit will undertake, consist in a deep theoretical study of the interaction at the interface between the sliding surfaces of the variator. The analysis will allow for a much more detailed comprehension of the sliding friction and wear, that, so much, influence the mechanical behaviour of the variators. In particular, the rubber belt-pulley interaction will be investigated in order to single out the physical properties of the material, and the morphological features of the surfaces that affect, more than others, the traction capabilities of the system belt-pulley.
The research project is split in the following five steps:
1) Experimental activity: Completion and tuning up of a Van-Doorne type CVT test rig. Theoretical activity: study of the sliding contact between the pulley and the rubber belt: the influence of the rubber viscoelasticity and of the hysteresis of adhesion will be taken into account in order to valuate the extension of the actual contact area, and the magnitude of the friction forces (6 months, 20000 Euro).
2) Experimental activity: the test rig will be used to measure the mechanical efficiency of CVT variators equipped with Van Doorne metal and rubber V-Belts respectively and their traction properties (transmitted torque, and power flow) during the steady-state running of the variator. Moreover, the CVT dynamical response will be measured during the transitory phases of shifting (4 months 10000 Euro).
3) Interaction with the other research units (2 months, 5000 Euro)
4) Comparison of the experimental results, obtained at step 2, with those predicted by the existing theoretical model: the experimental results will be used to validate and improve the theoretical model, once the causes of the differences between theory and experiments will be understood. The theoretical model, so validated, will be given to the other Research Units, and will be used as a simulation tool during the design of the control strategy of the CVT variator (6 months, 10000 Euro).
5) Theoretical activity: the influence of the roughness and temperature of the contacting surfaces on the adhesive properties of the belt will be studied in order to single out what are the physical and morphological properties of the material and surfaces capable to increase the friction forces, reduce the slip at the interface, and improve the mechanical efficiency of the variator. Experimental activity: the power spectra of the surface roughness will be measured by means of appropriate tribo-measurement systems (6 months, 19000 Euro).

Step 1 will be dedicated to the completion of the CVT test rig already available in the laboratories of this Research Unit. The test rig, that has been yet utilized to measure the performances of some car transmissions, will be properly modified in order to allow the CVT Van-Doorne belt-box to be mounted on it. This belt-box represents the core of the variator, made up of the drive and driven pulley and the belt. This transmission has been provided to the Research Unit of Bari by the Advanced Development CVT-System Department of Bosch-Van Doorne situated in Tilburg (NL). The tuning up of the belt-box will be performed on the basis of the Know-How that the Van-Doorne Transmissie is transferring to this Research Unit. The theoretical and experimental results that will be obtained during step 2 (see below) will be used to validate the theoretical models of metal V-belt CVT dynamics already developed by the Research of Bari, and these ones will be given to the Advanced Development CVT-System Department of Bosh-Van Doorne and to the Technical University of Eindhoven, that, in turn, will use these results in order to improve the design of the variator control strategy. Beside the experimental work, the Research Unit will study the contact mechanics between the belt and the pulleys in the case of rubber belt CVT of common use in scooters. In particular the influence of the material viscoelasticity and interfacial forces (energy of adhesion, hysteresis of adhesion) on the actual contact area will be analysed. The extension of the actual contact ares is, indeed, of crucial importance as regards the magnitude of the friction forces. The polymeric material will be described by a linear viscoelastic model. The model will allow for the calculation of the energy losses caused by the cyclic deformation produced by the substrate rough profile.
During the step 2 the Research Unit of Bari will focus on the experimental investigations of the variator. The techniques of Design of Experiment will be used to plan the tests, whereas Data Analysis techniques will be used to single out the geometrical, kinematical and dynamical parameters that influence more the mechanical efficiency, the traction properties of the variator and the magnitude of the belt slip. As regards the behaviour of the variator during the shifting, a significant number of tests will be concerned with the analysis of the relation between the shifting speed, the rotating velocity of the pulleys, and the transmitted torque. A similar investigation will be also performed on the rubber V-belt CVT, and in this case the time evolution of energy losses per unit time will be measured during shifting phases.
All these activities will enable the Research Unit to map the dynamical response of both type of variators.
During the step 3 the Research Unit of Bari will have an intense interaction with the other Units and especially with the Research Unit of Palermo that, during the last two-three years, have been studying the influence of the viscoelastic behaviour of the belt on the mechanical behaviour of the variator. In particular the theoretical results of Palermo and Bari will be analysed in order to interpret the physical consequences and point out if the proposed models are susceptible to be further improved. The results of the experimental investigation, performed by the Research Unit of Bari, will be made available to the other Research Units. Palermo will use these results in order to analyse the performance of a more complex transmission, i.e. the Split-Path transmission. Moreover the map of the CVT dynamical response, will be given to the research Units of Torino and Firenze, that will use this map in a complex vehicle simulation model. The Research Unit of Torino will use these model to perform a HIL simulation of the dynamic behaviour of vehicles equipped with CVT transmission, whereas the Research Unit of Firenze will use the same variator models to simulate by means an HIL approach the performance of vehicles equipped with the continuously variable transmission in terms of fuel consumption and polluting emissions. Regarding the interaction with the Research Unit of Roma, it is foreseen that it will be provided with the Mathematica® or Fortran codes, already developed by the Research Unit of Bari to study the non conformal EHL contacts in toroidal traction drives. Because of the similarity between the cam-follower contact and the roller-disk contact in toroidal traction drive, these codes, once slightly modified, will be used to study contact and friction problems in cam-follower mechanism which is the research topic of the Roma Research Unit.
During step 4 the Research Unit of Bari will compare the experimental results with the theoretical ones, already available and obtained by the same Unit thanks to the COFIN 2000 and COFIN 2002 grants. The comparison will allow to point out the weak points of these theoretical models, to establish their field of applicability, and to clarify how to improve the models. After the validation, these theoretical models will be used, in place of the experimentally calculated map, to predict the CVT behaviour under different running conditions, and for different geometries. After an additional interaction with the other Units of the program, the last step 5 will be dedicated to the theoretical analysis of the influence of the surface roughness and temperature on the adhesive properties of the belt. The dependence of the friction coefficient on the sliding velocity will be estimated. This friction model will be used to developed a theoretical model of the rubber-belt variator, that may enable to find the optimal CVT geometry and material properties, able to increase the friction forces, reduce the relative slip and improve the mechanical efficiency of the variator. At the same time some experimental measurement of the rough surface power spectra will be performed in order to use these data as input for the theoretical model of friction.