Session: 02-06-01: Machine Learning in Structural Dynamics and Aeroelasticity

Paper Number: 152306

152306 - Assessment of Multi-Fidelity Structural Theories for Dynamic Analyses Using Machine Learning 

This paper presents a machine-learning approach to assessing structural theories' computational cost and accuracy and provides guidelines on the proper finite element modeling for various numerical cases. The proposed approach exploits the Carrera Unified Formulation (CUF) to obtain multi-fidelity refined structural theories and governing equations. Machine learning techniques are used to build surrogate models. CUF provides training data for neural networks, considering higher-order polynomial expansions of the displacement field and problem features, e.g., thickness, geometry, material properties, and boundary conditions. The surrogate model's training aims to estimate a structural theory's accuracy, i.e., the fidelity, when providing structural dynamics outputs for a given set of inputs. For instance, the trained network can establish the accuracy of a third-order shear deformation theory in detecting the natural frequencies and time response of thin-walled beams. Furthermore, indications of the most influential input parameters and generalized primary variables are obtained. Multi-fidelity structural theories are assessed by varying the number of generalized displacement variables, i.e., the nodal degrees of freedom, of finite element models. Low-, first-order theories are incrementally enriched with higher-order terms. The trained network indicates which terms must be retrieved to satisfy a given fidelity requirement, e.g., having less than 5% errors on the first ten natural frequencies.  Furthermore, perspectives on the best spatial distributions of structural theories are drawn via the Node-Dependent Kinematics (NDK) approach, i.e., the best distribution of structural theories over a set of finite element nodes, to minimize the computational overhead without affecting the accuracy.

Presenting Author: Marco Petrolo Politecnico di Torino

Presenting Author Biography: Marco Petrolo is an Associate Professor and a member of the MUL2 Lab in the Department of Mechanical and Aerospace Engineering of Politecnico di Torino (www.mul2.com). His current research activities involve the multiscale analysis of composites and micromechanics and the development of higher-order structural theories. He cooperates with various institutions, including the University of Washington and NASA. He is a member of the Governing Board of the Italian Association of Aeronautics and Astronautics (AIDAA).