MERRILL
Micromagnetic Earth Related Robust Interpreted Language Laboratory

FORCINN – Experimental FORC Inversion

Browser-based FORC preprocessing and CNN inversion — upload experimental data and retrieve particle size and aspect ratio distributions without any local installation

FORCINN accepts standard FORC output from AGM and VSM instruments (AGM .dat, CSV, XLSX). The underlying CNN was trained on ~50,000 synthetic FORC diagrams spanning grain sizes of 20–200 nm and aspect ratios of 1.0–2.0 for magnetite. Results include predicted particle size and shape distributions; processed data and the inversion figure are available for download after each run.

Run FORC inversion from experimental data

FORCINN applies a convolutional neural network trained on synthetic FORC diagrams to invert experimental measurements into grain size and aspect ratio distributions. Upload a FORC file from an AGM or VSM instrument, choose the appropriate mode, and the pipeline handles noise estimation, slope correction (optional), and inversion automatically.

For context on how the CNN model was trained and validated against experimental measurements, see the Publications page. For the underlying micromagnetic framework, refer to the About page or the user manual.

Configure your FORC inversion