Adaptive magnetorheological sliding seat system for ground vehicles.

摘要

Magnetorheological (MR) fluids (MRFs) are smart fluids that have reversible field dependent rheological properties that can change rapidly (typically 5--10 ms time constant). Such an MRF can be changed from a free flowing fluid into a semi-solid when exposed to a magnetic field. The rapid, reversible, and continuous field dependent variation in rheological properties can be exploited in an MRF-based damper or energy absorber to provide adaptive vibration and shock mitigation capabilities to varying payloads, vibration spectra, and shock pulses, as well as other environmental factors. Electronically controlled electromagnetic coils are typically used to activate the MR effect and tune the damping force so that feedback control implementation is practical and realizable. MR devices have been demonstrated as successful solutions in semi-active systems combining advantages of both passive and active systems for applications where piston velocities are relatively low (typically 1 m/s), such as seismic mitigation, or vibration isolation. Recently strong interests have focused on employing magnetorheological energy absorbers (MREAs) for high speed impact loads, such as in helicopter cockpit seats for occupant protection in a vertical crash landing. This work presents another novel application of MREAs in this new trend---an adaptive magnetorheological sliding seat (AMSS) system utilizing controllable MREAs to mitigate impact load imparted to the occupant for a ground vehicle in the event of a low speed frontal impact (up to 15 mph).;To accomplish this, a non-linear analytical MREA model based on the Bingham-plastic model and including minor loss effects (denoted as the BPM model) is developed. A design strategy is proposed for MREAs under impact conditions. Using the BPM model, an MREA is designed, fabricated and drop tested up to piston velocities of 5 m/s. The measured data is used to validate the BPM model and the design strategy. The MREA design is then modified for use in the AMSS system and a prototype is built. The prototype MREA is drop tested and its performance, as well as the dynamic behavior in the time domain, is described by the BPM model. Next, theoretical analysis of the AMSS system with two proposed control algorithms is carried out using two modeling approaches: (1) a single-degree-of-freedom (SDOF) rigid occupant (RO) model treating the seat and the occupant as a single rigid mass, and (2) a multi-degree-of-freedom (MDOF) compliant occupant (CO) model interpreting the occupant as three lumped parts---head, torso and pelvis. A general MREA is assumed and characterized by the Bingham-plastic model in the system model. The two control algorithms, named the constant Bingham number or Bic control and the constant stroking force or Fc control, are constructed in such a way that the control objective---to bring the payload to rest while fully utilizing the available stroke---is achieved. Numerical simulations for both rigid and compliant occupant models with assumed system parameter values and a 20 g rectangular crash pulse for initial impact speeds of up to 7 m/s (15.7 mph) show that overall decelerations of the payload are significantly reduced using the AMSS compared to the case of a traditional fixed seat. To experimentally verify the theoretical analysis, a prototype AMSS system is built. The prototype seat system is sled tested in the passive mode (i.e. without control) for initial impact speeds of up to 5.6 m/s and for the 5th percentile female and the 95th percentile male. Using the test data, the CO model is shown to be able to adequately describe the dynamic behavior of the prototype seat system. Utilizing the CO model, the control algorithms for the prototype seat system are developed and a prototype controller is formulated using the DSPACE and SIMULINK real time control environments. The prototype seat system with controller integrated is sled tested for initial impact speeds of up to 5.6 m/s for the 5th female and 95th male (only the 95th male is tested for the Bi c control). The results show that the controllers of both control algorithms successfully bring the seat to rest while fully utilizing the available stroke and the decelerations measured at the seat are substantially mitigated. The CO model is shown to be effective and a useful tool to predict the control inputs of the control algorithms. Thus, the feasibility and effectiveness of the proposed adaptive sliding seat system is theoretically and experimentally verified.