Graduate Research Assistant Oklahoma State University
Additive manufacturing (AM) delivers exceptional design freedom and material efficiency, but layer-wise deposition creates nonuniform thermal histories, microstructural heterogeneity, and location-dependent properties that bulk tests average. As a result, critical weak regions can be missed and property inputs for qualification and life prediction can be non-representative. We have developed a mini-specimen methodology to resolve location-specific properties inside AM builds and coupled the measurements to a build-aware prediction framework. Rectangular mini coupons (≈ 3.0 × 1.2 × 0.25 mm) were extracted from predefined sites including thin walls, fillets, and heat-accumulation zones that standard dog-bone samples cannot sample. For each location, we measured elastic modulus, yield strength, ultimate tensile strength, and strain to failure using consistent fixturing to ensure repeatability. The small gage enables dense sampling with minimal material removal and without compromising the parent part. Tests reveal pronounced spatial variation across a single component. Regions with higher cumulative heat input or slower cooling show changes in grain morphology, and porosity/defects, correlating with reduced stiffness/strength and altered ductility; favorable thermal histories show improved properties and tighter scatter. This work supports miniature coupons as valid proxies for bulk behavior and for deriving tensile properties in alloys such as IN718, while noting that post-processing (e.g., STA, HIP+STA) reduces but does not eliminate variation. Building on these insights, we assemble a location-resolved database and model linking thermal history to microstructure to local mechanical (currently) and fatigue properties (in the future). This could enable a model for Lifetime & Design certification: assigning each voxel/region database-backed estimates (e.g., E, σ_y, UTS, fatigue life) for first-pass life prediction at build completion, reducing broad destructive testing to targeted validation. The combined approach is material-efficient, high-resolution, and process-linked, yielding property maps that could feed QA, digital twins, and design allowables, and guiding qualification-by-analysis for reliable fielding of AM structures.
Learning Objectives:
Upon completion, participants will be able to select mini-tensile extraction sites in AM parts and measure E, σᵧ, UTS, εf with repeatable fixturing to build location-specific datasets.
Upon completion, participants will be able to correlate measured properties with thermal-history indicators (heat input, cooling rate) to generate location-resolved maps and identify weak regions missed by bulk tests.
Upon completion, participants will be able to apply BUILD-CERT: estimate first-pass life from maps, propose process windows (scan, power, speed, hatch, heat treatment), and plan targeted validation for qualification-by-analysis.
