Modeling Soft Material Adhesion for In vivo Robotic Locomotion

In a continued effort to develop an active Robotic Capsule Endoscope (RCE), research is being conducted on the motion of the RCE through the small and large bowel. One of the factors in uencing the motion of the RCE is the adhesion of the RCE's micropatterned treated wheels to the inner lumen of bowel tissue. While adhesion is needed to generate traction, too much adhesion can also cause malfunctioning of the device, reduce autonomy and damage the tissue. This study is focused on the characterization of adhesion as a function of three critical RCE design parameters (Pre-load, Dwell Time and Separation Rate). In this work, a method for adhesion characterization is detailed and a complete model for the adhesion between a cylindrical probe with a smooth polydimethlysiloxane (PDMS) surface and synthetic tissue is presented. An explicit nondimensional model for the adhesive response is constructed with an R2 value of 0.9996 and a maximum relative error less than 5.6%. In addition, physical meaning of the proposed mathematical model is experimentally verified. Accurate models for the maximum stress supported by the tissue (R2 = 0:9895, maximum error = 2:04%), effective adhesion energy consumed in the separation (R2 = 0:9936, maximum error = 3:23%) and the total probe displacement from the beginning of the adhesion region to the point where complete separation occurs (R2 = 0:9964, maximum error = 1:47%) as a function of critical design parameters are obtained. Finally, a qualitative approach to the extrapolation of these models when varying the substrate stiffness and probe area is presented. Outgoing