How Rowan University used quantitative, layer-by-layer inspection to make a confident engineering decision on its very first independent metal AM build.
The first build on a new metal additive manufacturing system rarely goes exactly to plan.
Even with careful preparation, unexpected issues can arise that force operators into difficult decisions. Do you stop the build immediately? Or do you continue without knowing whether the process has already been compromised?
During the first independent build on Rowan University's newly installed DMG MORI LASERTEC 30 SLM US system, the team encountered exactly that situation. Rather than relying on visual inspection alone, they used quantitative process data from Phase3D's Fringe Inspection™ to understand what was actually happening inside the build.
Alex Kinoian, Undergraduate Research Student (left) and Andrew Holliday, Phase3D Applications Engineering Manager (right), after Fringe Inspection Installation on the DMG MORI LASERTEC 30 SLM US metal 3D printer
For manufacturers, these moments are often the hardest to manage. Cancelling a build too early can waste machine time, material and production capacity. Continuing without understanding the condition of the build introduces its own risks. The challenge isn't simply detecting that something unexpected has happened, it's knowing whether the process can safely continue.
An unexpected build preparation error
Early in the print, a build preparation output error caused approximately the first 60 layers to print only the part contours instead of the intended infill and support structures.
This type of error presents a genuine manufacturing risk.
Without the correct supporting geometry, transitioning into full bulk infill can result in:
- Part swelling
- Peeling
- Surface protrusions
- Recoater interference
- Potential build failure or machine damage
In many production environments, the safest response would simply be to cancel the build.
The challenge was that nobody actually knew whether the build had already become unsafe.
Turning uncertainty into measurable evidence
Rather than making a judgement based solely on machine images or operator experience, Rowan and Phase3D established a clear engineering decision point.
The build would continue only if Fringe Inspection confirmed there were:
- No abnormal part protrusions
- No swelling
- No evidence of peeling
- Healthy powder spreading after each recoater pass
If any of those conditions appeared, the build would be stopped immediately.
This changed the decision from "What do we think is happening?" to "What do the measurements tell us?"
Why quantitative measurements matter
Standard machine monitoring and layer images can often show that something looks different, but they cannot quantify whether the build surface itself has begun changing in a way that threatens part quality or machine safety. Fringe Inspection continuously measured the surface geometry of every layer, giving the Rowan team objective data throughout the build.
Throughout the transition from contour-only layers to full bulk infill, the Rowan and Phase3D teams monitored the height of the melted areas after exposure and checked subsequent powder layers for signs of recoater-related issues.
Because every layer was measured, they could see exactly how the build was behaving rather than relying on assumptions.
The result: evidence to keep building
The heightmaps captured the transition from contour-only layers into the bulk part geometry.
Critically, the measurements showed:
- Successful transition into bulk infill
- No measurable swelling
- No part protrusions
- No evidence of peeling
- Healthy powder layers throughout the transition
With quantitative evidence supporting the process, Rowan's team continued the build to successful completion.
The value was immediate. Instead of stopping the build solely because of uncertainty or accepting an unmeasured risk, the team had a quantitative basis for their decision.
As Phase3D CEO Niall O'Dowd explains:
"Day one is when inspection has to prove its value. This build did not follow the plan, but Rowan did not have to choose between canceling reflexively and hoping for the best. Fringe Inspection gave the team a measurement-driven basis to continue, without the possibility of a build crash or printer installation delay."
Looking beyond a single build
While this build demonstrated the immediate value of quantitative inspection, it also marked the beginning of a broader collaboration between Phase3D and Rowan University's Digital Engineering Hub.
Students and researchers will use calibrated, unit-based heightmaps generated during the metal additive manufacturing process to support research into process understanding, anomaly detection, qualification and engineering education.
For Phase3D, that's an important part of advancing additive manufacturing. Giving researchers and future engineers access to quantitative process data helps build a deeper understanding of how metal AM processes behave and how better decisions can be made during production.
“Our goal at DEHub is to connect advanced manufacturing with trusted data. This first build showed our students and researchers how real-time, quantitative inspection can support decisions at the moment they matter most.”
Antonios Kontsos, Ph.D., Director, Digital Engineering Hub at Rowan University
Better decisions through better data
Unexpected process events are an inevitable part of additive manufacturing. The value of in-situ inspection isn't that it prevents every anomaly from occurring it's that it gives engineers the information they need to respond with confidence.
In this case, quantitative measurements showed the build could continue safely. Rather than cancelling because of uncertainty, Rowan's team made an evidence-based engineering decision and completed the build successfully.
It's a simple example, but an important one. Real-time inspection isn't just about identifying defects. It's about giving manufacturers the confidence to distinguish between builds that need intervention and those that can safely continue, using measured evidence rather than assumption.
Rowan University
Rowan University's Digital Engineering Hub (DEHub) is advancing research into data-driven manufacturing by combining additive manufacturing, sensing and digital engineering. Working alongside Phase3D, the team is using quantitative, layer-by-layer inspection data to support process understanding, qualification research and hands-on engineering education.
→ More on Rowan University DEHub
Learn more about Fringe Inspection™
Discover how Phase3D helps manufacturers measure every layer, identify anomalies in real time and make more confident production decisions.
Frequently Asked Questions
What is in-situ inspection in metal additive manufacturing?
In-situ inspection measures the build as it is being printed, allowing manufacturers to identify process anomalies in real time rather than waiting until the part has been completed. This provides engineers with data that can support production decisions before defects become costly.
Why would a manufacturer cancel a metal additive manufacturing build?
Builds may be cancelled if there is evidence of part deformation, swelling, recoater interference or other conditions that could compromise part quality or damage the machine. The challenge is determining whether those risks are actually present or simply suspected.
What are heightmaps in additive manufacturing?
Heightmaps are quantitative measurements of the build surface captured layer by layer. Rather than relying solely on visual images, they provide calibrated surface height data that can reveal protrusions, swelling and other geometric changes during a build.
Can real-time inspection reduce unnecessary build cancellations?
Yes. By providing objective measurement data during production, real-time inspection can help engineers distinguish between anomalies that require intervention and those that can be safely monitored, reducing decisions based purely on uncertainty.
How does Fringe Inspection support process qualification?
Fringe Inspection generates calibrated, repeatable measurements throughout the build process. This data can support process understanding, anomaly detection, qualification activities and manufacturing research.
Share: