TPA-Based Indirect Force Identification Methodology For Agricultural Machines

This dissertation presents a systematic methodology for the indirect estimation of blocked and contact forces applied to a modern agricultural machine. The primary purpose of this study is to enhance vehicle product development by using the final results as input for virtual simulations, troubleshooting, and future informed decision-making. These forces need to be defined at the connective interface, made of mechanical links, between the tractor chassis (to which the engine, driveline and hydraulic ancillaries are rigidly connected), and the operator's cabin, respectively constituting the active side, where the blocked forces belong, and the passive side, where the contact forces belong. For this purpose, the in-situ TPA approach was adopted for estimating the former forces, while the matrix inversion method was adopted for the latter ones. However, the agricultural vehicle's large structural dimensions, the presence of highly isolating non-linear mechanical links, and the difficulty in exciting the connection points with sufficient accuracy and repeatability degrade the Signal-to-Noise Ratio (SNR), which must be maximized to minimize numerical errors. The strategy proposed to counteract the negative effect of these factors, which have been proven by a thorough data analysis, lies in diversifying numerical methods to estimate forces and establishing a criterion for discerning the physically acceptable regions from the incorrectly estimated ones. Further, the results were made scalable by also accounting for the possibility of measuring multiple FRF sets, thereby leading to enhanced results. The entire described process was implemented in MATLAB to make the methodology repeatable within a company application context. In summary, this study offers a comprehensive methodology for identifying forces at the tractor's chassis-cabin interface. This is achieved through the integration of experimental TPA theory, diverse numerical methods, a validation criterion, and a scalable FRF improvement strategy. Furthermore, the experimentally obtained data and the methodology's results were exhaustively analysed.