Abstract. Cranes used in factories and ports are frequently exposed to potential risks arising from complex operating conditions. To comprehensively assess operational reliability, digital technologies are being increasingly employed for dynamic monitoring and optimization during crane operations. This study focuses on the dynamic stress variations in the main truss of cranes during operations. First, a digital mapping model between a scaled physical crane and its virtual counterpart was developed using a parametric design approach based on a real-world engineering crane. Using this model, a digital-twin representation of the crane's dynamic stress field was constructed. Nodal stresses of the twin crane were obtained using radial basis function (RBF) interpolation in conjunction with finite-element stress field calculations. Subsequently, the K-nearest-neighbor algorithm was used to select relevant nodes for training an interpolation-based surrogate model, enabling end-to-end stress prediction at the crane's nodes. Finally, dynamic stress rendering using the HSV (hue, saturation, value)-color-model-enabled synchronized visualization of the stress field within the digital twin, supporting real-time monitoring, simulation-based optimization, and dynamic life cycle management of the crane. Experimental comparisons of three different lifting conditions show that the average error between finite-element analysis and stress-rendering results is 8.29 %, while the average error between measured data and stress-rendering results is 9.98 %, verifying the predictive reliability of the interpolation model. These findings guide the application of dynamic digital twins of stress fields in industries such as construction, manufacturing, and energy.
Yan et al. (Tue,) studied this question.