Design, Fabrication and Characterization of Layered Jamming Bistable Composite Structures for Assistive Robotics 55 inherent load-displacement behavior for the biased bistable composite plate can be seen in Figure 3 (seen in wide black in the figure, we only tested the biased plate, since it maintains contact during the entire compressive displacement range that we tested through the bistable response region. The response of the two jammed bistable composite plates can also be seen in Figure 3 with vacuum pressures as low as -3 kPa all the way (designated in wide red) to a full vacuum. It can be seen that at -3 kPa, the response represents the combination of the two bistable composite plates. As the vacuum pressure increases, the hysteresis loop expands as the jammed bistable composite plates stiffen by increasing the energy level associated with the instability to make it more difficult to initiate the plate buckling response. Fig. 3 Experimental results obtained in three-point bending for two jammed layered bistable composite plates starting at -3 kPa of vacuum pressure (wide red) all the way to a full vacuum, as compared to the biased bistable composite plate. The increase in hysteresis with vacuum pressure can be attributed to the maximum shear force that was expected at the interface. It can be seen that the minimum load after the buckling instability is initiated upon loading goes from near the response of the biased bistable composite plate at -2.9 N when there is -3 kPa of vacuum pressure all the way to -7.4 N at full vacuum. From the model, the expected shear load resistance at full vacuum pressure would be increasing from 0 N up to 4.5 N for a coefficient of friction of 0.24 at the interface between the bistable composite plates. Thus, it appears the vacuum pressure is controlling that dip in load based on the vacuum pressure. Thus, it is possible to control the bistable response to adjust to the desired needs of the patients. At higher pressures, there is a much lower negative stiffness, which increases the interaction distance with patients. It was also noted that as the vacuum pressure increased, the load approached 0 before the buckling instability initiated when the jammed plates were unloaded. Initially, the adhesion of the two plates induced by layer jamming is sufficient to treat the structures as one continuous structure with stiffness associated with the combined thicknesses. This “tightening up” of the structure can also be seen as the initial displacement for three-point bend loading decreases with increasing vacuum pressure, which decreases to -2.19 mm. That stiffened response was maintained until a level of -6 N was reached. At that point, the response separates with increasing take-off stiffness with increasing vacuum pressure. The same tightening effect can also be seen as the plates are unloaded, until they approach 0 N at approximately -12 mm, at which point the take-off response upon buckling becomes more compliant with increasing vacuum pressure until the specimen is fully unloaded. Clearly, this indicates that the specimen is starting to approach a state where there is residual energy left to initiate reverse buckling and allow the plates to deform. In fact, the plates tended to stay deformed after 10 cycles at a displacement of
RkJQdWJsaXNoZXIy MTMzNzEzMQ==