Enter your workpiece dimensions and the grain-size target. The pipeline runs the PINN processing-map surrogate (auto-detects the safe (T, ε̇) optimum), training-data correction, pass-schedule optimisation, and a mesh-graded 3D preform in sequence — one click, one result page.
Round-bar geometry before forging
What should the predictor aim for?
Drives the void-closure polynomial and flow-stress model
Original NSMLab calibration. Cr-Ni-Mo low-alloy steel, typical heavy-section forging.
Auto-finds the safe (T, ε̇) optimum
Choose which stages to run
Auto-detect safe optimum (T, ε̇)
~15–30 s on first run · mesh-quality graded