No abstract

No abstract

Within braided pneumatic Muscle Actuators (pMA) the braid structure is vital to the actuator's performance, preventing over-inflation, converting radial expansion into axial contraction and setting limits for both dilation and contraction. This paper seeks to explore the nature of the contractile limit and the hysteresis observed by researchers during the actuation cycle. Maximum actuator dilation occurs when adjacent braid strands are forced against one another. Within this work this is analyzed mathematically and it is shown that by halving the number of strands used to create the braided shell the actuator's contractile range can be increased by approximately 7%. This also results in a simultaneous peak contractile force increases of over 16%. These results are verified experimentally. Hysteresis due to friction between braid strands during muscle operation is also explored. The paper will show how consideration of the deformation of the strands allows the contact area and therefore friction to be calculated without the need for experimentally obtained data as in previous research. A mathematical model is produced and verified experimentally.

Pneumatic technology has been successfully applied for over two millennia. Even today, pneumatic cylinder based technology forms the keystone of many manufacturing processes where there is a need for simple, high-speed, low-cost, reliable motion. But when the system requires accurate control of position, velocity or acceleration profiles, these actuators form a far from satisfactory solution. Braided pneumatic muscle actuators (pMAs) form an interesting development of the pneumatic principle offering even higher power/weight performance, operation in a wide range of environments and accurate control of position, motion and force. This technology provides an interesting and potentially very successful alternative actuation source for robots as well as other applications. However, there are difficulties with this approach due to the following. (i) Modeling errors. Models of the force response are still nonoptimal and for good results these models are highly complex, which makes accurate design difficult. (ii) Low bandwidth-the bandwidth of the actuator-link assemblies are often considered to be too low for practical success in many applications, particularly robotics.

No abstract