Monitoring and Analytics to Enable AM Part Qualification

Darren Beckett, CTO of Sigma Additive Solutions (sigmaadditive.com) delivered this informative presentation at the ASTM Additive Manufacturing Center of Excellence Workshop on In-Situ Technology Readiness for Applications in AM Qualification and Certification. The event was supported by NASA’s Marshall Space Flight Center and held in collaboration with America Makes.

With the advent of Lagrangian and Eulerian optical sensing technologies for LBP that capture irradiance from the melt-pool plume system and that have derived energy density, temperature and cooling rate metrics, NDT to ML predicted metrics have been made to become synchronous with spatial-temporal galvo position and produce a unique maturation in visualization strategies and analytics that have enabled machine health, process monitoring and part quality determinations and diagnosis against a baseline to occur.

The in-process metric maturity and the nascent confidence in the technology now relates energy density to Archimedes density, bichromatic calibrated temperature ratio metrics to part thermal history, and layer-wise material cooling rate to microstructural exploration as a result of high frequency sample acquisition rates.

With a technology base that comprehends optics and irradiance transmission, sensor and spatial data acquisition, high performance edge computing architecture, novel metric generation that has now been contributed to global ASTM repositories, flaw visualization and diagnosis and multiple years of investigation into anomalous data, a process towards AM qualification and certification can now finally be realized. What is also treated is the realization that as the technology matured, autonomous in process monitoring ‘AIPQM’ data analytics for part qualification and production build quality sustaining, had to arrive out of a labor and skill necessity, to reduce the engineering analytical burden.

Qualification strategies using within IPQM process signature data rolls up, in machine, in process scan/vector, layer, part and part metrics that are used to make “digital golden build(s)” family baseline which can then be compared to qualification test build part families using a series of part comparison statistical tools. The “digital golden build(s)” references enable autonomous part to part, build to build, and machine to machine qualification and certification standards across a fleet of AM machines.

Mr. Beckett’s talk is titled,

Topics include: