Abstract
This paper presents a novel approach for implementation of digital twin (DT) based Structural Health Monitoring (SHM) of an Palfinger offshore knuckle boom crane. Contrary to most harbor gantry cranes, knuckle boom cranes are highly nonlinear mechanisms that cannot be represented by static reduced order twin models. Such cranes need to be solved by non-linear finite element solvers. The digital twin representation of the Palfinger crane is modelled and simulated real-time in a nonlinear finite element (FE) program, driven by inputs from physical sensors. Model reduction techniques are applied to enable DT co-simulation running two times faster than the physical crane. The inputs from the standard crane instrumentation are processed for noise reduction and singularity removal and converted to hydraulic actuator inputs. A simple inverse method for estimation of the crane payload is implemented based on hydraulic pressures. Structural loads due to wave induced ship motions are predicted based on sensor signals from the ship IMU. Based on the standard ship and crane instrumentation, the digital twin allows for real-time determination of stresses, strains and loads at an unlimited number of hot spots. Hence, a digital twin can be an effective tool for predictive maintenance of real offshore knuckle boom cranes with minor additional costs. The presented approach is described in a general manner and is applicable for offshore cranes used in the industry.
