| Back |
Development of Effective Force Testing: Test Results |
Details |
|
|Home|Pilot study|Development|Validation|References|Site map|Contact us| |
| Static Loading Test |
|
 |
|
The springs were designed to lose contact with the mass at displacements exceeding 1 in., resulting in a reduced stiffness. Thus, the structure was a linear elastic structure with an initial stiffness of 4 kip/in. when the displacement response was within the 1-in. precompression, while it acted as a nonlinear elastic structure when the displacement response exceeded the precompression (the stiffness reduced to 2.0 kip/in.).
|
| Free Vibration Test |
|
|
|
|
Free vibration test with an initial displacement of 0.75 in. was conducted, and parametric simulations were made to determine a combination of viscous damping and friction force which would minimize the error between measured displacements and simulation displacements based on a least square technique.
|
| Direct Implementation of EFT |
|
|
|
|
Tests with 0.5 kip sine sweep excitation are conducted to investigate the effect of the linear velocity feedback compensation. The above graphs show the significant amplitude reduction of applied forces near the natural frequency of the structure, which confirms the existing of natural velocity feedback problem. Simulation results follows the force measurements. |
| EFT with Linear Velocity Feedback Compensation |
|
 |
|
Great improvements in applying forces near the natural frequencywas achieved with the linear velocity feedback compensation. The discrepancy near the natural frequency was believed due to the fact that the load pressure affected the servovalve flow property as shown in equation (0.6) while this nonlinearity was not considered in the linear velocity feedback compensation in both simulation and physical implementation. |
| EFT with Linear Velocity Feedback Compensation under Large Excitation |
|
 |
|
Tests with larger amplitude excitations (2 kip), which caused displacement greater than 1 inch (yield displacement at the current setup), are conducted to investigate the feasibility of linear velocity feedback compensation scheme. Comparison of experimental results show that the amplitude drop at the natural frequency is obvious. Simulations with linear flow model with a constant flow gain and the influence of load pressure (Equation (0.6)) cannot explain this large amplitude drop at the natural frequency where the flow demand reaches its maximum value. Further studies find that the nonlinear flow properties should be considered in addition to the influence of load pressure. A typical flow-spool opening relation is shown below. As can be seen, the flow gain (the slope of the curve) decreases as an increase of spool opening. Further simulations with a nonlinear flow model matches experimental results well. This indicates that more advanced control algorithms are necessary when large amount of flow is required. |
| EFT with Nonlinear Velocity Feedback Compensation |
|
 |
|
The FFT of the measured force does not show any obvious drop across the whole frequency range, indicating that the actuator was able to apply forces correctly at all frequencies. Compared to the test with the linear velocity feedback compensation, the ability of the actuator to follow force commands was greatly improved. Small discrepancies can be seen in the frequency domain from 1.5 Hz to 2.5 Hz, which corresponds to 4s to 8s in the time domain, where the force output was noisy. The structural velocity response between 4s to 8s was large, which indicates large required spool openings (55% maximum) during the test. At large spool openings, the system uncertainties become significant. Hence, the velocity feedback compensation based on the piece-wise linear flow curve might instantly incomplete. |
|Home|Pilot study|Development|Validation|References|Site map|Contact us|