Conception d'un effecteur terminal immergé à deux degrés de liberté

Details

Supervisors

Ronsse, Renaud

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil électromécanicien, à finalité spécialisée: mécatronique

Abstract

enThe Louvain School of Engineering is currently developing and assembling a robot with four degrees of freedom to test swimming robots or vibrating cylinders. The user of the installation would also like to simulate free flight with an air foil. For such test, the robot needs to extra degrees of freedom : pitch and roll. The aim of this master thesis is to design an end-effector with these two degrees of freedom to equip the robot. Accuracy is the most severe requirement~: the difference between the actual pose of the air foil and its theoretical pose cannot exceed 0.5 mm and 0.5°. These values include the deformations, the tolerances and the errors in the transducers. As the contribution of the robot is still unknown, we have distributed the errors equally between the six degrees of freedom, which means that the accuracy for the end-effector must be 0.17 mm. There are two important constraints. First, the end-effector is immersed in a towing tank, which causes drag forces on the structure. Secondly, the robot and the walls of the towing tank limit the outer dimensions of the end-effector. To prevent coupling between the translation and the rotation (one of the design objectives), we have considered mechanism with a remote center of motion (RCM) corresponding the center of mass of the air foil. A comparative study of several solutions showed that the parallelogram was probably the most suitable RCM mechanism. It consists in two main parts : an immersed block with a motor to generate the roll of the air foil, and an articulated parallelogram structure to transmit the pitch movement to the block and to the air foil. The structure is immersed when the robot is in its lowest position. Regarding the block roll, we could not reconcile the objectives of accuracy and of easy disassembling. To obtain accuracy, clearances between the bearings, the housing and the shaft must be avoided, and interference is required. But tolerances to the hundredth of a millimetre (micron is excluded because of the cost) could generate interference between the parts that prevent them from being disassembled without damage. The structure of the end-effector has been designed with SolidWorks. By increasing the cross section of the tubes up to the maximum dimensions allowed by the tank walls and by the robot, the deformations of the structure still cause a displacement of the air foil of 0.29 mm. And the manipulator weights 40 kg. An analyse of the deformations revealed that the drag force due to the water is responsible for about 90% of the displacement of the air foil. Therefore, we think that the enquiry and the givens of the problem should be discussed. First of all, would it be possible to perform the test in a wind tunnel instead of in the towing tank ? A another solution would be to accept the deformations of the end-effector and to measure or estimate the actual position of the air foil. But to measure or estimate the actual position of the air foil will be difficult, respectively because of the proportions of the tank and because of the computational cost