Relationship between surface shape and interaction force
The exerted shear force shown in Fig. 14 and Table 2 suggests that there is a relationship between the pattern of unevenness and the interaction force. First, the interaction forces among the different surface shapes did not differ until around \(y=-8\) mm, although the shape under the robotic cuff differed, as shown in Fig. 12. This was probably because the effect of the variance of the surface shape was absorbed by the deformation of the sponge of the robotic cuff. The shear modulus may also have contributed to this easing. Then, the interaction force started to show the difference among gradient values, as stated above.
When the front edge of the robotic cuff reached approximately \(y=5\) mm, the increase rate of the interaction force started to decline. In this phase, the partial contact of the robotic cuff, a part of which separated from the surface of the device, was observed. Because the normal force of the robotic cuff is controlled to a constant value, the robotic cuff moved upward when getting over the convex. Finally, the interaction force of the baseline shape and large gradient shape started to decrease after \(y=10\) mm. Although the shear force did not increase, the partial contact probably causes stress concentration, which could not be observed even using this measurement system.
Because the target of this study was to identify the increase of the wounds risk caused by the variance of surface shape when using the wearable robot, some features of the contact state between the robot and human, such as the shape and stiffness of the robotic cuff, viscoelasticity of the wearer’s tissue, and the loading pattern of the robotic cuff on the surface, were simplified. Thus, more precise experimentation and analysis is required for practical validation testing of the wearable robot. However, the necessity of considering the surface shape and its variance based on the human were expounded in this study.
Effect of surface shape variance from the viewpoint of contact safety
When evaluating the risk that appears after several exposures to the hazard, the method using the threshold curve is commonly used. By using the threshold curve, the endurance time, which is the threshold time of acceptable risk, can be estimated under specific loading conditions. Thus, using the curve that evaluates the wounds risk caused by continuous rubbing , the endurance time under various surface shapes was estimated and compared.
The shear force exerted in rubbing motions increases with the normal force. According to previous study , the normal force exerted when using a wearable robot seldom exceeded 50 N. Thus, in this section, considering the severe condition when using the wearable robot, a rubbing motion under 50 N of normal force was assumed. In addition, because of the limitation of the specification of the device, it was difficult to keep the surface shape under such a large compression force. Thus, the shear force under the normal force of 50 N was estimated by increasing the pattern of the interaction force observed in the experiment mentioned above proportionally.
The estimated maximum shear stress and the endurance time under such a condition area displayed in Fig. 15 and Table 3. The risk curve displayed here was obtained from the rubbing test using porcine skin and human skin . These results suggest that the endurance time significantly changes even when the surface shape changes only slightly. In addition, the risk curve used in this study did not include the effect of local stress concentration. As mentioned above, it suggests that the stress concentration occurs over the large convex. Moreover, the existence of stress concentration was verified by visualizing the pressure distribution on the surface of the device. A tactile sensor sheet (I-SCAN, Nitta Corporation, Japan), which could measure the normal force exerted on square cells lined on the surface, was used. By putting this flexible sheet on the uneven surface, the concentration of normal force was roughly observed via a rubbing test. The observed force concentration made the endurance time short, as shown in the risk curve.
For the safety design of the cuff of the wearable robot, the increase and concentration of shear force around the convex will cause a problem. Thus, it seems better to use maximum pressure instead of mean value when evaluating the contact force under the robotic cuff. Furthermore, the shape of robotic cuff should probably be designed to decrease the pressure exerted around the body convex. One option is to support the interaction force with a flat area of the body surface. In addition, the contact between flat cuff and body convex such as the situation of this study was not suitable from the view of contact safety. Thus, the area of robotic cuff, which contacts the body convex, should have sufficient indentation to avoid force concentration. However, it should be noted that the index to evaluate the body convex, which could be used to estimate the range of individuality, was not determined yet.