FusionCert is the qualification package for fusion first-wall materials: a specimen registry, a diffraction and tomography analysis suite built on GSAS-II and scikit-image, and a certificate engine that issues only on recorded evidence. One system from raw frame to signed certificate.
Upload the pattern, the instrument file and the phases, and the refinement runs in the registry with the result bound to the specimen. The engines are the ones the community already trusts.
Staged refinement of background, cell, isotropic size and microstrain and sample displacement, with GSAS-II's own estimated standard deviations, Rwp and goodness of fit. The same powder mathematics as TOPAS and FullProf, scriptable and auditable.
Williamson-Hall and modified Williamson-Hall with contrast factors, instrument correction from the standard, domain size, microstrain and dislocation density by Williamson-Smallman and Wilkens.
Biaxial residual stress with hkl-specific elastic constants and d₀ solved rather than assumed; lattice-strain depth profiles in ppm from edge-on and taper-section scans.
Otsu or fixed threshold, distance-transform watershed to separate touching pores, porosity, pore-size distribution, largest pore and crack-like count, the Avizo chain on scikit-image.
The input file is hashed, the method and its parameters are stored with the numbers, and the fit quality sits beside each value in the record and on the certificate.
The same code runs on the beamline workstation for full tomograms and batch refinements, submitting through the API with identical provenance.
A result that cannot be traced to a specimen, a protocol and a raw file is an anecdote. The registry refuses to issue a certificate without all three.
Baseline and post-irradiation measurements run at P02.2 on PETRA III: transmission and grazing-incidence diffraction, propagation radiography and absorption tomography on the same coupon, with CeO₂ calibration refined once per energy.
IFUEL-W-L01-007 and its cut children, each with a QR label that resolves to its own record, paired as pristine and implanted twins from adjacent positions.
Technique, instrument, energy, calibrant, analysis software and the SHA-256 of the raw frame are stored with every number and its uncertainty.
Versioned criteria sets evaluate lattice strain, microstrain, porosity and pore size against agreed limits, and refuse to pass on absent data.
Every event is appended to a chain that the dashboard re-verifies on each load; certificates freeze their evidence and can be checked publicly by serial.
One value per specimen builds the envelope for each grade and quantity, so a batch from a new supplier is judged against the whole campaign.
Results flow from the beamline workstation through a token-authenticated API; CSV export and JSON records for every specimen.
Within the BMFTR "Fusion 2040" programme, the IFuEL project assesses candidate first-wall materials under fusion-relevant conditions, from tungsten grades carrying each European mitigation strategy to low-Z ceramics for inertial confinement.
Batch-to-batch variation in porosity, texture and inclusions is otherwise mistaken for radiation damage. A pristine baseline measured under one protocol is what makes a post-irradiation shift meaningful.
Register a lot, refine the pattern, segment the volume and issue the certificate without leaving the record.
Register specimens, record measurements, define criteria and issue certificates.
Enter workspace →Follow your own lots and download the records and certificates that belong to you.
Open partner view →Enter a serial to confirm it exists and that its hash is intact, with no data exposed.
Check a serial →