Phase3D Scales In-Situ Monitoring to NASA's Quad-Laser Metal 3D Printer

NASA Wants Parts Certified While They Print

The current process for qualifying a metal 3D printed part for flight is slow. A part comes off the machine, goes through CT scanning and destructive testing, and only then does anyone know whether it is good. NASA calls the alternative "born qualified" manufacturing: a build record so complete that the part is already certified when the print finishes.

Phase3D is running a new program with NASA Marshall to test whether that goal is realistic at production scale. The centerpiece is an EOS M300-4, a quad-laser powder bed fusion system that aerospace contractors are already buying for production work. If the test works, manufacturers could set quantitative go/no-go thresholds during the build instead of waiting weeks for post-build analysis.

How Fringe Inspection Works

Phase3D's Fringe Inspection system captures a continuous measurement record across every layer of a build. It tracks powder spreading irregularities, recoater blade interactions, layer shifts, melt pool abnormalities, spatter accumulation, delamination, and surface height variation. The system does not sample a few layers and extrapolate. It measures all of them.

Previous research showed strong correlation between these in-situ anomalies and defects found in post-build CT scans. Tests with the US Air Force on Ti64 printed on an EOS M 290 flagged 81 percent of anomalies for depressions larger than 47 micrometers. NASA tests on GRCop-42 printed on a Colibrium Additive M2 hit 83 percent correlation, with 100 percent accuracy for depressions larger than 42 micrometers.

The Production-Scale Challenge

Those earlier results came from single-laser research machines with small build chambers. The EOS M300-4 changes the math. Four independent lasers, a larger build volume, and faster scan speeds mean more process disturbances and more data to interpret. The current program is designed to answer whether the earlier correlations hold when the machine is running at the speed production shops actually use.

The test plan calls for a 50,000-layer dataset on flight-representative brackets. That volume is intended to establish statistical confidence in the go/no-go criteria. If the layer-wise anomalies detected during production match the defects found in post-build CT, the next step is turning those patterns into formal qualification standards.

Standards and Scope

The program maps directly to NASA Civil Space Shortfalls 1490 through 1494, which cover in-situ monitoring, process qualification, and qualification of complex additive manufactured geometries. It also aligns with NASA-STD-6030, NASA-STD-6033, and SAE AMS7032, the existing standards for metal AM flight hardware.

Phase3D introduced its Fringe Qualification platform at Formnext 2024, moving the technology from single-machine inspection to fleet-wide quality management. The NASA program is the first test of that platform at the scale and speed required for actual production qualification.

Why This Matters

Metal additive manufacturing is increasingly common in aerospace, but qualification remains a bottleneck. Parts that pass every in-process check can still fail CT inspection, and parts that look fine on the surface can hide defects that only show up under load. A monitoring system that catches those problems during the build, rather than after, would cut weeks out of the qualification timeline and reduce scrap rates on expensive titanium and nickel alloy parts.

The aerospace industry has been chasing in-situ qualification for years. If Phase3D and NASA can prove it works on a machine like the M300-4, the next question is not whether to adopt it, but how fast the standards committees can write the rules around it.