Fringe Inspection is a real-time inspection system for metal additive manufacturing. Through use of structured light scanning, Fringe Inspection measures the 3D height profile of every powder layer and melted layer during the print. The advantage of using a metrology system for inspection is the ability to ensure accuracy, repeatibility and precision.
Before an in-situ inspection system can be trusted to support production decisions, manufacturers need objective evidence that its measurements are accurate, repeatable and traceable.
This case study explains how Phase3D validates and verifies Fringe Inspection™ before every deployment, providing NIST-referenced evidence of measurement performance across the full build area. The resulting data gives manufacturers confidence that the system is ready for qualification, process development and production use.
Fringe Inspection™ optical module for retrofit in metal additive manufacturing machines.
Why Validation & Verification Matters
The V&V procedure delivers three things every customer needs before deployment:
- Quantifies Fringe Inspection™ measurement uncertainty before qualification begins
- Provides NIST-referenced evidence for quality and acceptance workflows
- Creates a traceable baseline for accuracy, repeatability and sensor drift across the build area
The V&V procedure is standardized across all installations, agnostic of printer make and model. Every customer receives the same testing methodology, the same pass/fail criteria referenced against NIST-certified gauge blocks, and the same structured output. The result is a documented performance baseline that supports customer acceptance processes and provides a foundation for ongoing system health monitoring.
Validation & Verification Results at a Glance
Finding |
Description |
| Standardized V&V Methodology | V&V follows a defined, repeatable procedure executed identically across every Fringe Inspection installation. |
| NIST-Traceable Documentation | All measurements are referenced against a NIST-certified gauge block, producing traceable records that support customer qualification workflows. |
| Baseline for Performance Monitoring | V&V output establishes a documented performance baseline, enabling ongoing system health monitoring and future comparisons. |
| Reduced Pre-Production Uncertainty | Customers receive objective evidence of system accuracy before production decisions depend on Fringe Inspection data. |
| Accelerated Acceptance | A structured, pre-defined pass/fail framework shortens the customer acceptance timeline compared to ad hoc testing approaches. |
The Challenge: Uncertainty Before Production Deployment
When a new metrology system is installed in a manufacturing environment, customers face a fundamental question: how do they know the system is performing as specified before they begin relying on its data for production decisions? Unlike AI-based decision-making systems or vision technologies that rely on grayscale imagery, Fringe Inspection produces quantitative height data: measurements with defined units, known uncertainty, and traceable calibration. That distinction matters when it comes to qualification. Objective, data-based evidence of system performance is something a structured V&V procedure can provide; a confidence score or visual flag cannot.
For metal additive manufacturing, where dimensional data informs process qualification, build acceptance and part conformity, measurement uncertainty introduces real production risk.
A system that has not been formally validated can introduce measurement doubt into the qualification process, slow customer acceptance, and undermine confidence in the data at exactly the moment it matters most.
Phase3D recognized this gap early. The V&V procedure was designed to address it directly: by giving every customer a standardized, documented, traceable test of system performance before production deployment begins.
What is the V&V Procedure?
The Fringe Inspection™ Validation & Verification (V&V) procedure is a standardized measurement accuracy and repeatability test performed during every installation. It combines two complementary test protocols:
1. System-Level Noise and Repeatability Assessment
Following calibration, the system performs 50 consecutive scans of the calibration plate and analyzes each of five spatial positions across the build area: center, top-left, top-right, bottom-left, and bottom-right. This calibration and 50-scan procedure is then run two more times, resulting in 3 runs total. This generates a high-density dataset that characterizes measurement noise, spatial uniformity, and height stability across the full field of view.
For each position, the procedure calculates:
- Standard deviation of height measurements (noise floor)
- Average height bias relative to the reference plane
- 98th percentile absolute error
- Height drift trend over the 50-capture sequence
Together, these metrics answer the core question: is the system producing stable, low-noise, spatially uniform measurements at every point across the build area?
2. Gage Block Measurement Accuracy Test
A NIST-certified gage block of known height is placed at each of the five measurement positions. The system captures and measures the gage block height at each location. Measured values are compared against the known reference height to assess absolute measurement accuracy across the full field of view.
Figure 1 : Phase3D Calibration Plate with Gage Block
This test verifies that the system is not only repeatable, but accurate: that the heights it reports correspond to physical reality within the specified tolerance of the instrument.
The complete V&V procedure follows a structured sequence: calibrate the system, perform the 50-scan repeatability runs (repeated three times in full for statistical confidence), then conduct the gage block accuracy test at each position. Each run produces a dedicated output folder, and all results are compiled into a final report.
The procedure evaluates both measurement repeatability and measurement accuracy, ensuring the system consistently reports the correct height across the full build area.
What the V&V Output Demonstrates
The V&V procedure produces a structured set of outputs – composite heightmaps, noise distribution histograms, height drift plots, and a summary results table that together characterize system performance across multiple dimensions.
1. Quadrant and Full Build Area
For each of the four quadrants of the build plate, plus the full build area, the V&V output reports the average error and standard deviation of height measurements. This allows performance to be evaluated both locally, confirming no quadrant is systematically worse than others and globally across the entire field of view.
Figure 2: Full build area heightmap (left) and height error histogram (right).
Each quadrant is represented by a heightmap showing the spatial distribution of measurement error alongside a histogram of that distribution. The heightmap makes any spatial gradients or localized anomalies immediately visible, while the histogram provides the quantitative noise statistics used to assess pass/fail against specification.
Figure 3: Top-left quadrant heightmap (left) and height error histogram (right).
For LPBF and Binder Jet users, this matters in a practical way. The plate scan data establishes a measurement baseline at the zero-height reference, characterizing the noise inherent in the system before any part geometry or powder is introduced. When Fringe Inspection later detects a significant height deviation or anomaly during a build, users can interpret that signal with confidence. They know what the system's background noise looks like at every location across the build area, so a flagged deviation isn't an artifact of where it appeared - it's real.
2. Line Scan Data
The 50-scan repeatability runs produce three complementary analyses for the six line scan positions (corresponding to two cross-section lines at each of the five sample point locations): a noise floor assessment, a height stability check, and a drift analysis. Together they characterize both the magnitude and the time behavior of measurement variability.
For LPBF and Binder Jet users, the line scan data provides a layer-by-layer analogue to the plate scan baseline. Each build layer is a new measurement event, and the line scans confirm that the system's noise characteristics remain stable across repeated captures, the same way they will across hundreds or thousands of build layers. If a height anomaly appears mid-build, users can trust it isn't the result of the system drifting or becoming noisier over time. The measurement environment at layer 500 is as reliable as it was at layer 1.
Figure 4: The six line scan measurement positions evaluated during V&V
Noise Floor / Measurement Distribution
For each line scan position, a histogram of all height measurements across the 50-capture sequence characterizes the noise floor. Tight, symmetric, near-zero-centered distributions indicate low noise and minimal systematic bias.
Figure 5: Height value distribution histograms for six line scan positions
Height Stability Over Time
The line plots track height measurements pixel-by-pixel across the full 50-capture sequence for each position. Stable, bounded noise without progressive drift confirms that the system maintains consistent performance over many repeated measurements – an important property for builds spanning hundreds or thousands of layers.
Figure 6: Height measurement line plots across 50 captures for the six line scan positions
Drift Analysis
The drift plots track the estimated Z-height at each of the five sample points across all 50 captures, and a box-and-whisker summary quantifies the spread and extremes at each region. The absence of systematic upward or downward trends, and the near-zero means across all regions, confirms that the system does not exhibit thermal drift or mechanical settling behavior that would degrade measurement accuracy over the course of a build.
Figure 7: Height drift over 50 captures by region (left) and box-and-whisker plot of region differences (right)
3. Gage Block Scan Data
The gage block test measures a NIST-certified 500 µm gage block placed at each of the five positions across the build plate. For every measurement, the output includes a heightmap showing the gage block location with its regions of interest (ROIs) annotated, and a cross-section profile plot showing the measured height across the block and adjacent plate surface.
Figure 8: Top-right position gage block: heightmap with ROI annotation (left) and cross-section profiles (right).
The cross-section profiles directly demonstrate the system's ability to resolve the known 500 µm step height, and the reported block mean and standard deviation at each position quantify absolute accuracy and repeatability across the full build area.
Results Summary
The results from the verification and validation for Phase3D's Fringe Inspection is directly translated to the commercial specifications of each product.
Figure 9: EOS M290 Fringe Inspection Kit Spec Sheet
Repeatability - At the Calibration Plane
The worst-case standard deviation of average height error across all four quadrants of the plate scan data. This reflects how consistently the system measures the same surface across repeated scans at the calibration height.
Repeatability - At Range Extremities
The worst-case standard deviation of gage block height measurements across all five positions. This reflects measurement consistency when the target surface is displaced 500 µm from the calibration plane.
Planar Measurement Accuracy - At the Calibration Plane
The worst-case average error across all four quadrants of the plate scan data. This reflects how close the system's reported height is to the true reference plane across the full build area.
Planar Measurement Accuracy - At Range Extremities
The worst-case deviation from the nominal 500 µm gage block height across all five measurement positions. This answers the question: how far from the true height are the system's gage block measurements?
Absolute Measurement Accuracy - At the Calibration Plane
The worst-case deviation from nominal across all point scan measurements, with the standard deviation reflecting the spread of those measurements. This characterizes accuracy at a specific point rather than averaged across a region.
Customer Impact
Reduced Risk During System Adoption
V&V de-risks system adoption at every stage. Rather than deploying a new metrology system and discovering performance questions after the fact, customers receive structured, traceable evidence upfront that Fringe Inspection is performing within specification before any production decisions depend on its data.
Because the test methodology, pass criteria, and output format are standardized across every installation, quality teams don't need to design their own acceptance tests from scratch. The documentation V&V produces NIST-referenced measurements, quantitative noise and accuracy statistics, and a structured report and is ready to slot directly into existing qualification packages, reducing the time and internal effort required to move from installation to production-ready status.
The V&V output also serves a purpose beyond initial acceptance. It establishes a documented performance baseline that can be revisited as the system ages. If performance metrics shift in a future V&V run, the original data provides an objective reference point for identifying what changed and by how much, giving quality teams the evidence they need to make informed decisions about recalibration, maintenance, or continued deployment.
Takeaway
V&V is the foundation of confident Fringe Inspection deployment. Customers don't have to take system performance on faith, they get objective, documented proof the system is ready before production ever depends on it. For quality-critical applications, that assurance isn't just useful; it's the difference between a smooth qualification process and one built on uncertainty.
Frequently Asked Questions
What is Validation & Verification (V&V) in additive manufacturing?
Validation & Verification (V&V) is the process of demonstrating that an inspection system measures accurately, repeatably and within its specified tolerance before being used to support production decisions. It provides objective evidence that the system is performing as intended.
Why is Validation & Verification important for in-situ inspection?
In-situ inspection systems are used to identify build anomalies and support quality decisions during production. Validation & Verification provides documented evidence that the measurements produced by the system are reliable, giving manufacturers confidence before qualification and production deployment.
How does Fringe Inspection™ verify measurement accuracy?
Fringe Inspection™ undergoes a standardized Validation & Verification procedure during installation. Using NIST-certified gage blocks, repeatability testing and drift analysis, the procedure confirms measurement accuracy, repeatability and stability across the full build area.
Why is measurement repeatability important in LPBF?
Laser Powder Bed Fusion (LPBF) processes involve hundreds or thousands of layer-by-layer measurements. Consistent measurement repeatability ensures that detected height deviations represent genuine process changes rather than measurement noise or sensor drift, improving confidence in build quality decisions.
Share: