Table 4 Performance metrics of the ensembled BiLSTM models predicting crack extension (Δa) and kink angle (Φcrit)

From: An interleaved physics-based deep-learning framework as a new cycle jumping approach for microstructurally small fatigue crack growth simulations

Metric

Δa BiLSTM model

Φcrit BiLSTM model

R2

0.72

0.75

RMSE

0.31

0.51

MDAE

0.14

0.22

  1. Root mean square error (RMSE) and median absolute error (MDAE) are reported in micrometers for the Δa model and in radians for the Φcrit model.
Back to article page