Long-term life stress mapping algorithm of the deep-sea pressurized spherical shell based on digital-twin technology

Abstract

Aiming at the difficulty of the global stress field monitoring of the deep-sea pressurized spherical shell in real time, this paper proposes a long-term life stress field reverse mapping algorithm based on digital-twin technology. Specifically, a high-fidelity FEM model is established based on the titanium alloy data, structure construction data and model test data. The Elastic modulus E and Poisson's ratio υ are selected as the structure characteristic parameters based on the preliminary small sample test results. The sensitivity analysis is carried out. And then, stress monitoring location selection results is proposed considering the factors such as mutual independence, installation conditions, and economic costs. The Long-Short Term Memory network (LSTM) together with the FEM results is tailored and employed to build the reverse mapping model. The mapping error is below 1% and compared to the FEM model, the efficiency of the proposed model is improved by more than 60 times.