Neural Network-based Prediction Method for Low-speed Impact Damage Area in Composite Materials
Currently,composite laminates are widely utilized in the maritime sector.However,traditional experimental and finite element methods fail to accurately depict the actual damage situation when subjected to low-speed impacts.To address this issue,an equivalent impact damage prediction model based on BP neural network and CNN neural network is established by integrating low-speed impact test results and ultrasonic immersion scanning results.This model enables rapid prediction of the damage status of composite laminate structures during impact events.After training the model using experimental sample data,it efficiently and accurately predicts the damage area of composite laminates.This method is simple,practical,and can be widely applied.
composite materialslow-speed impactneural networkdamage area
NETL
NSTL
万方数据
