Table 2 A list of features used to train BiLSTM model for predicting Φcrit 23

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

Feature

Description

λ1, λ2, λ3, λ4

Angles formed by the trace of the slip planes 1 to 4 with \(\overrightarrow{r}\) on the radial plane

Ωmax, Ωmin

Maximum and minimum tilt angles of the four slip planes relative to the crack orientation

Γ

A binary variable indicating whether a CFP is at the intersection of a grain boundary

λmax(Ω), λmin(Ω)

Slip plane angle having the maximum and minimum tilt angles

\(\left\vert \gamma \right\vert\)

Absolute angular position of CFP

ngrains

Number of grains intersecting the radial probe

λmax1(m), λmax2(m), λmax3(m), λmax4(m)

Slip plane angles sorted based on the maximum Schmid factor of the respective slip plane

Φk−1

Kink angle at crack growth increment k − 1

ω1, ω2, ω3, ω4

Disorientation angle between the grain containing the CFP and the grain containing a probe point that intersects with the trace of the slip planes 1 to 4

\({\Phi }_{{\rm{max}}({\epsilon }_{{\rm{eq}}}^{p})}\)

Angle along the radial probe having the maximum equivalent plastic strain extracted using the micromechanical fields at k = 0

\({\Phi }_{{\rm{max}}({M}_{{\rm{micro}}})}\)

Angle along the radial probe having the maximum micromechanical Taylor factor extracted using the micromechanical fields at k = 0