Optimal Dimensioning of Elastic-Link Manipulators regarding Lifetime Estimation

TL;DR

The work tackles durable, lightweight elastic-link robotic manipulators by embedding fatigue-life estimation into the design process. It introduces a fatigue-analysis pipeline that couples plane-stress transformation with the cutting-plane method and Tresca equivalent stress to identify load cycles via rainflow counting, then applies Haigh/Wöhler data and Miner accumulation to predict life, expressed as $t_{\text{life}} = t_{\text{task}}/D$. This lifetime estimate feeds a multi-criteria Pareto optimization over weight and vibration to select geometries for a 3DOF elastic-link arm performing pick-and-place tasks. The approach enables principled design choices that trade mass reduction against durability and vibration performance, with demonstrated guidance on selecting configurations from the Pareto front that achieve meaningful lifetime improvements for industrial automation.

Abstract

Resourceful operation and design of robots is key for sustainable industrial automation. This will be enabled by lightweight design along with time and energy optimal control of robotic manipulators. Design and control of such systems is intertwined as the control must take into account inherent mechanical compliance while the design must accommodate the dynamic requirements demanded by the control. As basis for such design optimization, a method for estimating the lifetime of elastic link robotic manipulators is presented. This is applied to the geometry optimization of flexible serial manipulators performing pick-and-place operations, where the optimization objective is a combination of overall weight and vibration amplitudes. The lifetime estimation draws from a fatigue analysis combining the rainflow counting algorithm and the method of critical cutting plane. Tresca hypothesis is used to formulate an equivalent stress, and linear damage accumulation is assumed. The final robot geometry is selected from a Pareto front as a tradeoff of lifetime and vibration characteristic. The method is illustrated for a three degrees of freedom articulated robotic manipulator.

Figures (9)

• Figure 1: Elastic manipulator ELLA, Institute of Robotics @ JKU
• Figure 2: Elastic link (beam) of an articulated robot, with elastic displacements and critical point for the subsequent lifetime estimation
• Figure 3: Stress state in a thin walled beam
• Figure 4: Definition of mean and amplitude stress (left) and exemplary load history (right)
• Figure 5: Pagoda roof method for compressive valleys (left) and tensile peaks (right)