Riassunto analitico
The aim of the following thesis work is to develop a well-structured methodology to fully characterize the front axle suspension system of an agricultural tractor CCM 36 LWB in order to extract the optimal passive damping for best comfort performance.
The Tractor under study features a double acting cylinder adaptive stiffness and semiactive continuous variable damping.
The front axle suspension subsystems have been tested through extensive experimental quasi-static and dynamic tests for the full characterization.
Both suspension stiffness for different tractor setups (Soft- Normal- Hard) and damping characteristics have been extracted for a broad range of cylinder velocities at each supplied current command , the global friction hysteresis of the hydraulic cylinder and Panhard mechanism has been characterized in compression and rebound using an instrumented test bench ad hoc.
The simultaneous pressures in piston-side and rod-side for both the cylinder and accumulators have been recorded using pressure transducers, a 100KN loadcell was positioned between the test cylinder and the tractor cylinder in order to record the total force exerted.
The total hysteretic behaviour of the cylinder was obtained by the subtraction between the load cell force and gas force.
The mathematical model of the suspension model was built in Simulink environment and is composed mainly by three subsystems (Accumulators Gas Model – Damping characteristics Model - Friction Model).
To capture the dynamic pressures in both piston-side and rod-side accumulators ,Ideal and empirical Benedict Webber Rubin (BWR) models have been implemented and compared to simulate the dynamic gas pressure during extension and compression of the hydraulic cylinder , both the gas models results have been compared.
In order to capture the frictional hysteresis behavior in the hydraulic cylinder observed during the tests, the decision was taken to use the asymmetric Hyperbolic tangent function-based ,and the total hysteresis of the cylinder was given by the summation of damping characteristic curves obtained by nonlinear least square fitting method and the hysteresis cycles at each current setup.
The Parameters of the model were identified at each combination of frequency, amplitude and supplied current for the numerous tests , a sensitivity analysis was performed in order to correlate the dependency of the model to the dynamic amplitude ,frequency and supplied current , at the end a fitness function was created to combine the singular effect of each of them.
The validation of the Hyperbolic tangent friction-based model was done by comparing the friction -velocity characteristic curves of the experimental data to those produced by the model.
The validation of the gas model was made by comparing the P-V gas cycles for each piston-side and rod-side and the total gas force at different amplitudes and frequencies , a good correlation between the results was found.
The steady state frictional force of both cylinder and Panhard was modelled using the Lugre Model, and the parameters were identified using the Tringular wave tests data.
To explore the capability of the model to predict the gas forces and the total suspension forces on track ,a series of bump tests at different vehicle speed were used, the damping setting of the vehicle was varied from soft to hard setting passively by changing the current supplied to the electrohydraulic valve , the front axle relative position signal has been used as an input to the single degree of freedom model, piston-side and rod-side accumulator pressures have been compared , the total predicted force on the track was also shown , and the model showed great fidelity with some highlighted limitations.
Finally , Iso-Smooth 5008 test was performed at 3 different vehicle speed on soft mode, the power spectrum density (PSD) of the chassis acceleration and RMS values of the test were compared to the one predicted by the model.
|
