ESTIMATION OF PARAMETERS OF THE TORQUE CONVERTER OF AN AUTOMATIC TRANSMISSION OF A PASSENGER VEHICLE Elias Dias Rossi Lopes1, André Flora Alves Pinto1 and Caio César do Prado Dorea Reis1 1Instituto Militar de Engenharia E-mails: [email protected] , [email protected] , [email protected] ABSTRACT The aut omatic transmission has been becoming more usual in the modern vehicles, even those considered low cost ones. The torque converter is the device responsible for the coupling between the engine and the gearbox in automatic transmission vehicles. In this con text, this research aims the simulation and estimation of parameters of the automatic transmission of a passenger vehicle. In that purpose, it is implemented the Particle Swarm Optimization Method, which assists in obtaining parameters of the torque conver ter’s curve from the engine full-load curve and the car real traction diagram. The torque converter’s curves have fundamental importance, and only manufacturers have possession of that information. Beyond that, through those curves, it is possible: to rais e the profit of the company, to develop an automobile that is economic in terms of fuel consumption and offers a better adaptation of the engine characteristic curve to the ideal traction curve. INTRODUCTION The methodology used in this article aims to obtain parameters of the characteristic curves of a TRILOK torque converter through the Particle Swarm Optimization (PSO) Method. To this end, this introductory section is subdivided into three subsections, in order to detail the process as a whole. The first subsection details the operating principle of a TRILOK torque converter, as well as it presents its characteristics curves. At its end, the parametrization adopted for these curves and the parameters that the method seeks to estimate are specified. In the second subsection, the matching process is approaches, which is the process of coupling between the full -load engine curve and the transmission system. The matching equating is presented and specifies as the process is performed for vehicle equipped with a torque converter. The matching process is also known as the direct problem and is used as the basis for formulating the inverse method. Finally, in the last subsection, is presented the methodology of inverse problems in general. The concept of ob jective function is defined and the formulation adopted is presented. Focus is given to the stochastic method known as Particle Swarm Optimization, which is detailed and equated. 1. TORQUE CONVERTER The basis of most of the types of automatic and semi -autom atic transmissions used in vehicles engines is hydrodynamic coupling, so called because it operates with kinetic energy of the circulating fluid. This type of coupling involves a hydrodynamic clutch or a torque converter. Despite being similar in their con structions, the torque converter is capable of multiplying torque while the hydrodynamic clutch is not [1]. The hydrodynamic couplings use the inertia of the flow of a fluid. The individual components of a transmission of this type are the fluid flow devi ces. A rotatory pump plays the role of source of energy and the turbine of a primary engine. The mechanical energy applied through the transmission shaft is converted into hydraulic energy from the fluid and then back into the mechanical energy in the turb ine, which is available (except losses) on the output shaft. The basic difference between a hydrodynamic clutch and a torque converter is that the torque has, in addition to the pump and turbine, a stator [2]. The advantages of a hydrodynamic clutch and a torque converter can be combined to avoid a decaying section in the efficiency curve of the torque converter. This type of device is known as a TRILOK torque converter and its simple construction and high level of efficiency make this convert