论文标题
部分可观测时空混沌系统的无模型预测
Torsional Instabilities in the Delamination of Soft Adhesives
论文作者
论文摘要
储层计算是预测湍流的有力工具,其简单的架构具有处理大型系统的计算效率。然而,其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间,然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用,该系统模拟了湍流的若干特征。我们表明,使用径向基函数作为非线性投影器,即使只有部分观测并且不知道控制方程,也能稳健地捕捉复杂的系统非线性。最后,我们表明,当测量稀疏、不完整且带有噪声,甚至控制方程变得不准确时,我们的网络仍然可以产生相当准确的预测,从而为实际湍流系统的无模型预测铺平了道路。
Soft adhesive contacts are ubiquitous in nature and are increasingly used in synthetic systems, such as flexible electronics and soft robots, due to their advantages over traditional joining techniques. While methods to study the failure of adhesives typically apply tensile loads to the adhesive joint, less is known about the performance of soft adhesives under shear and torsion, which may become important in engineering applications. A major challenge that has hindered the characterization of shear/torsion-induced delamination is imposed by the fact that, even after delamination, contact with the substrate is maintained, thus allowing for frictional sliding and re-adhesion. In this work, we address this gap by studying the controlled delamination of soft cylinders under combined compression and torsion. Our experimental observations expose the nucleation of delamination at an imperfection and its propagation along the circumference of the cylinder. The observed sequence of `stick-slip' events and the sensitivity of the delamination process to material parameters are explained by a theoretical model that captures axisymmetric delamination patterns, along with the subsequent frictional sliding and re-adhesion. By opening up an avenue for improved characterization of adhesive failure, our experimental approach and theoretical framework can guide the design of adhesives in future applications.