Machine learning improves knockdown factor predictions for spherical shell caps
A Hybrid Gaussian Process-Extreme Gradient Boosting Machine Learning Framework for Robust Knockdown Factor Prediction and Reliability-Based Design of Thin Spherical Shell Caps
Machine Learning
Summary
Buckling is a problem that weakens thin spherical shells used in engineering, and traditional methods often underestimate how likely this is to happen, making designs too cautious. The researchers used machine learning to better predict a key measure called the knockdown factor, which adjusts for real-world imperfections. By combining two advanced machine learning techniques, they created a model that predicts this factor more accurately than older methods. This helps design shell caps that are safe but less wastefully strong.
What this means in practice
- •For structural engineers: Use the hybrid machine learning model to predict knockdown factors more accurately for lighter, safer spherical shell designs.
- •For manufacturing quality teams: Incorporate improved knockdown factor predictions to set realistic quality controls on geometric imperfections in shell production.
Authors
Budhaditya De, Rohan Majumder, Devasmit Dutta, Aman Deep Gupta, Sudib Kumar Mishra
Abstract
Elastic buckling remains a critical concern in the design of deep spherical shell caps, given that experiments reveal a significant reduction of critical loads from those predicted by linearized eigen-buckling analysis. This discrepancy has been attributed to initial geometric imperfections and is often reconciled through a dimensionless parameter, namely the knockdown factor (KDF). However, conventional KDFs (NASA SP-8007; Eurocode 3) are highly conservative, as they represent the lower bound of an extensive body of experimental data. This study aims to narrow the conservatism in the KDFs for spherical shell caps through machine learning (ML) regression models trained on experimental data on spherical shell buckling. A hybrid model is proposed by combining Gaussian Process Regression (GPR) with Extreme Gradient Boosting (XGBoost), and its improved performance is demonstrated over conventional ML algorithms, i.e., the Artificial Neural Network (ANN), Support Vector Regressor (SVR), Random Forest Regressor (RFR), and Histogram-based Gradient Boosting (HGBoost). Each of these models uses six input features and a single output, i.e., the KDF. The most critical feature affecting the KDF is ascertained through a sensitivity analysis. The study concludes with a reliability-based design framework that relies on the predictive ML model to obtain safety-consistent, yet less conservative, KDFs.