

The AM Sasquatch: Challenging AM Qualification with Math, Physics, & AI
Fabian Alefeld welcomes back Harshil Goel, founder and CEO of Dyndrite, to discuss how current additive manufacturing (AM) qualification practices limit advanced laser powder bed fusion toolpaths by locking variables such as software, machine, strategy, layer height, and material. Harshil reviews MQ, IQ, OQ, and PQ, noting cost and rigidity increase toward part qualification. They discuss Dyndrite’s roles in multiple America Makes efforts: AI4AM to reduce the specimen counts needed for material allowables (starting with 17-4PH) using improved statistical methods; DeltaQual 2.0 to enable “bracketed qualification” so parameter changes within a defined process window don’t retrigger requalification (including varying layer heights within a part); and Gotham to transfer scalable, qualified parameter approaches across different machines and sizes using materials like Scalmalloy. Harshil describes “Sasquatch,” a physics-based KPV framework using ratios (e.g., energy density, enthalpy, melt pool dimensions) to bound process windows and increase confidence without relying solely on sensors or simulation, and highlights software and AI accelerating DOE and parameter iteration with a target program timeline of about two years. 01:34 What Dyndrite Builds 02:11 Why Qualification Blocks Innovation 03:58 IQ OQ PQ MQ Explained 08:22 PQ Locks Everything Down 09:00 America Makes Program Overview 09:57 AI4AM Materials Allowables 12:39 Process Windows Not Fixed Params 14:57 Delta Qual 2 Bracketed Qual 17:38 Partners Sensors And Confidence 19:25 Sasquatch New KPVs 25:33 Gotham Scaling Across Machines 27:07 Future Less PQ More Reuse 28:39 Industry Impact Of Faster Qual 29:52 Customer Owned Parameters 30:54 Additive As Moat 32:13 Microstructure Driven IP 33:12 Selling Machine Control 34:31 Breaking Vendor Silos 35:36 Two Year Timeline 37:57 Risks And Adoption 41:23 Software As Glue 42:11 AI Accelerated DOE 45:34 Math Meets Empiricism 49:48 Future Outlook Wrap

















