Structured Light and White Light Scanning: Full-Field 3D Capture for CNC Manufacturing

What is actually being measured or controlled

Structured light scanning (SLS) and white light scanning (WLS) are non-contact, full-field 3D measurement technologies that capture the complete surface geometry of a physical object in a single scan. To understand how they work, first define two core terms:

• Full-field capture: Unlike coordinate measuring machines (CMMs) that measure discrete points one at a time, SLS/WLS captures millions of 3D points (called a point cloud) across an object’s entire visible surface in seconds.
• Structured light: A pattern of light (e.g., stripes, grids, or coded dots) projected onto an object’s surface. When the object’s shape distorts the pattern, the scanner’s camera system uses trigonometry to calculate the 3D coordinates of every point on the surface.

White light scanning is a subset of structured light scanning that uses broadband white light (instead of laser or monochromatic light) for projection. For the purposes of manufacturing, the two terms are often used interchangeably, with WLS referring to the most common industrial implementation of SLS.

At their core, SLS/WLS measure two critical sets of data:

1. The as-manufactured 3D shape of a part, including all surface features, curves, and deviations from design intent.
2. The spatial relationship between multiple parts in an assembly, enabling analysis of fit and alignment.

Why it matters

The value of SLS/WLS lies in its ability to bridge the gap between digital design (CAD) and physical production. For CNC manufacturing, this impacts three key areas: part function, assembly performance, and cost.

Part Function

Many parts require precise geometric compliance to meet functional requirements. For example:

• A turbine blade machined on a 5-axis center like Gree CNC’s GA-FA500 must have a specific airfoil profile to achieve optimal aerodynamic efficiency. Even a 0.1mm deviation in profile can reduce turbine output by 2–3%.
• A medical implant machined on the high-precision GA-UHD500 must match its CAD model within 0.05mm to ensure biocompatibility and proper integration with surrounding tissue.

SLS/WLS captures the full surface of these parts, allowing engineers to identify deviations that would be missed by discrete point measurements.

Assembly Performance

For multi-part assemblies, even minor deviations in individual parts can compound to cause fit issues. For example:

• A chassis machined on a large-format gantry center like the GA-GF4028 must have precise mounting hole positions and surface flatness to ensure that electronic components align correctly. A 0.2mm deviation in a mounting hole’s position can cause a circuit board to be misaligned, leading to electrical failures.
• An automotive powertrain assembly requires tight alignment between the engine block, transmission, and crankshaft. SLS/WLS can scan the entire assembly to identify misalignments that would be difficult to detect with traditional measurement methods.

Cost

SLS/WLS reduces manufacturing costs in two key ways:

1. Faster inspection: Traditional CMM inspection of a complex part can take hours, while SLS/WLS can scan the same part in minutes. This reduces inspection time and allows for more frequent quality checks, reducing the risk of producing large batches of defective parts.
2. Root cause analysis: By comparing the as-manufactured part to its CAD model, engineers can quickly identify where deviations are occurring. For example, if a part machined on the GA-DV750 consistently deviates in a specific area, it may indicate a tool wear issue or a problem with the machine’s positioning accuracy. This allows for faster corrective action, reducing scrap and rework costs.

How it is measured or achieved

Industrial SLS/WLS systems typically consist of three core components: a light projector, one or more cameras, and a computer for data processing. The measurement process follows four key steps:

1. System Calibration

Before scanning, the system is calibrated to establish the precise relationship between the projector, cameras, and a reference coordinate system. Calibration involves projecting a series of patterns onto a calibration target (a plate with known 3D features) and using the captured images to calculate the system’s intrinsic parameters (e.g., camera focal length, lens distortion) and extrinsic parameters (e.g., the relative position and orientation of the projector and cameras). Calibration is critical for ensuring measurement accuracy, and it must be repeated periodically to account for changes in the system (e.g., temperature fluctuations, mechanical drift).

2. Pattern Projection and Image Capture

The projector projects a sequence of structured light patterns onto the part’s surface. These patterns are designed to encode unique information into each point on the surface, allowing the system to identify and track individual points across multiple images. For example, a common pattern sequence involves projecting a series of vertical stripes of varying widths, followed by a series of horizontal stripes. The cameras capture images of the part as each pattern is projected.

3. 3D Point Calculation

For each point on the part’s surface, the system uses the distorted pattern information to calculate its 3D coordinates. This process, called triangulation, works by determining the angle at which light from the projector hits the point and the angle at which the reflected light is captured by the cameras. Using the known distance between the projector and cameras (the baseline), the system can calculate the 3D position of the point.

4. Point Cloud Processing

The raw 3D points are combined into a point cloud, which is then processed to remove noise and outliers. The point cloud can be aligned to a coordinate system (e.g., the part’s CAD model) and used to generate a 3D mesh (a surface representation of the part) for further analysis.

Accuracy Limits

The accuracy of SLS/WLS systems varies widely depending on the system’s design, the size of the part being scanned, and the measurement conditions. Typical industrial systems have an accuracy range of 0.01mm to 0.1mm for parts up to 1m in size. For example, a high-precision SLS system used for inspecting small parts machined on the GA-DV750 (positioning accuracy 0.003mm) may have an accuracy of 0.01mm, while a system used for scanning large parts machined on the GA-GF4028 (positioning accuracy 0.02mm) may have an accuracy of 0.05mm.

It is important to note that SLS/WLS accuracy is often specified as the system’s maximum error, not the error across the entire part. For large parts, the error can vary across the surface, with higher error at the edges of the scan volume.

Sources of error

SLS/WLS measurements are affected by a variety of error sources, which can be grouped into four categories:

1. System-Related Errors

These errors are inherent to the SLS/WLS system and are influenced by its design and calibration:

• Calibration error: Inaccurate calibration is one of the most common sources of error. Calibration can drift over time due to temperature changes, mechanical vibration, or physical damage to the system. For example, a 1°C change in temperature can cause the projector and cameras to shift relative to each other, leading to a calibration error of up to 0.02mm.
• Lens distortion: All lenses introduce some level of distortion, which can cause errors in the calculated 3D coordinates. High-quality systems use distortion correction algorithms to minimize this error, but residual distortion can still contribute to measurement error.
• Camera resolution: The resolution of the cameras determines the smallest feature that can be measured. For example, a camera with a resolution of 2 megapixels may not be able to accurately measure a feature smaller than 0.1mm, while a 12-megapixel camera can measure features as small as 0.02mm.

2. Part-Related Errors

These errors are caused by the properties of the part being scanned:

• Surface reflectivity: Highly reflective surfaces (e.g., polished metal) can cause glare, which can wash out the projected pattern and lead to errors in the calculated 3D coordinates. Dark or matte surfaces can also cause errors, as they reflect less light back to the cameras. To mitigate this, parts are often coated with a thin layer of matte spray before scanning.
• Surface texture: Rough surfaces can scatter the projected light, leading to noise in the point cloud. For example, a part with a surface roughness of Ra 6.3µm will have a noisier point cloud than a part with a surface roughness of Ra 0.8µm.
• Transparency: Transparent or translucent parts (e.g., glass, plastic) can cause errors, as the projected light can pass through the surface and reflect off internal features.

3. Environmental Errors

These errors are caused by the conditions in which the scan is performed:

• Temperature fluctuations: As mentioned earlier, temperature changes can cause the SLS/WLS system to drift. They can also cause the part itself to expand or contract, leading to errors in the measured dimensions. For example, a steel part will expand by approximately 0.012mm per meter for a 1°C increase in temperature.
• Vibration: Vibration from nearby machinery (e.g., a CNC machine operating in the same room) can cause the part or the SLS/WLS system to move during scanning, leading to blurry images and measurement errors.
• Ambient light: Bright ambient light can interfere with the projected pattern, leading to errors in the captured images. SLS/WLS systems are typically used in controlled environments with low ambient light to minimize this error.

4. Process-Related Errors

These errors are caused by how the scan is performed:

• Scan alignment error: When scanning large parts, multiple scans are often taken from different angles and then aligned to form a complete point cloud. Errors in the alignment process can lead to gaps or misalignments in the final point cloud.
• Occlusion: Occlusion occurs when part of the part’s surface is hidden from the projector or cameras. For example, a deep hole or a complex internal feature may not be visible to the scanner, leading to missing data in the point cloud.

Honest limitations

SLS/WLS is a powerful measurement technology, but it has several important limitations that users must be aware of to avoid overconfidence in the results:

1. Cannot measure internal features without access

SLS/WLS is a line-of-sight technology, meaning it can only measure surfaces that are visible to the projector and cameras. It cannot measure internal features that are not accessible, such as the internal diameter of a deep hole or the internal structure of a closed assembly. For these features, other measurement technologies (e.g., CMMs with touch probes, bore gauges) are required.

2. Accuracy is lower than high-precision CMMs for discrete points

While SLS/WLS can capture a large number of points quickly, its accuracy for individual points is typically lower than that of high-precision CMMs. For example, a high-precision CMM may have an accuracy of 0.001mm for discrete point measurements, while a high-precision SLS system may have an accuracy of 0.01mm. This means that for applications requiring extremely high accuracy for specific features (e.g., the position of a critical hole), SLS/WLS may not be sufficient.

3. Point cloud noise can affect small feature measurements

The point cloud generated by SLS/WLS contains noise, which can be caused by surface texture, vibration, or other error sources. This noise can make it difficult to accurately measure small features, such as sharp edges or small holes. For example, a 0.1mm diameter hole may appear as a 0.15mm diameter hole in the point cloud due to noise.

4. Cannot measure material properties or subsurface defects

SLS/WLS only measures the surface geometry of a part. It cannot measure material properties (e.g., hardness, tensile strength) or subsurface defects (e.g., cracks, porosity). For these measurements, other technologies (e.g., ultrasonic testing, X-ray inspection) are required.

5. Overconfidence in "full-field" data

A common misconception is that SLS/WLS captures the entire surface of a part. In reality, parts with complex geometries often require multiple scans, and even then, some areas may be occluded. Users must carefully inspect the point cloud to ensure that all critical features are captured before relying on the measurement results.

Summary Table

Practical Checklist for Applying SLS/WLS Correctly

To ensure accurate and reliable SLS/WLS measurements, follow this checklist:

1. Define measurement requirements upfront:

   • What is the required accuracy for the part?
   • Which features are critical and must be measured?
   • Is the part’s geometry complex enough to justify full-field scanning?

2. Select the appropriate SLS/WLS system:

   • Choose a system with an accuracy that meets or exceeds your measurement requirements.
   • Consider the part size: Ensure the system’s scan volume is large enough to accommodate the part.

3. Prepare the part and environment:

   • Clean the part’s surface to remove dirt, oil, and debris.
   • Coat highly reflective or dark surfaces with a thin layer of matte spray (ensure the coating is uniform and does not add significant thickness).
   • Control the environment: Ensure the temperature is stable (within ±1°C), minimize vibration, and reduce ambient light.

4. Calibrate the system:

   • Calibrate the system before each scan session, or if the temperature changes by more than 1°C.
   • Verify calibration by scanning a known reference part and comparing the results to the reference’s certified dimensions.

5. Plan the scan strategy:

   • Identify potential occlusions and plan multiple scan angles to ensure full coverage of critical features.
   • Use fixturing to secure the part and ensure it does not move during scanning.

6. Process and validate the data:

   • Remove noise and outliers from the point cloud.
   • Align the point cloud to the part’s CAD model or a coordinate system.
   • Inspect the point cloud for missing data or misalignments.
   • Validate critical feature measurements with a secondary method (e.g., CMM, caliper) if required.

FAQ

1. Q: Can SLS/WLS be used to inspect parts machined on high-precision CNC centers like the GA-UHD500? A: Yes, but with caveats. The GA-UHD500 has a positioning accuracy of 0.006mm and repeatability of 0.004mm. High-precision SLS/WLS systems can achieve accuracy of 0.01mm, which is sufficient for many applications. However, for critical features requiring accuracy better than 0.01mm, a high-precision CMM is still recommended.

2. Q: How does SLS/WLS compare to CMM for inspecting large parts machined on gantry centers like the GA-GF4028? A: SLS/WLS is often faster and more cost-effective for large parts. A CMM inspection of a large part can take hours, while SLS/WLS can scan the same part in minutes. SLS/WLS also provides full-field data, allowing engineers to identify deviations across the entire part, not just at discrete points. However, CMMs still offer higher accuracy for individual point measurements.

3. Q: Can SLS/WLS be used for reverse engineering parts? A: Yes, SLS/WLS is widely used for reverse engineering. The point cloud generated by SLS/WLS can be used to create a 3D CAD model of the part, which can then be used for manufacturing, repair, or analysis.

4. Q: What is the maximum part size that can be scanned with SLS/WLS? A: The maximum part size depends on the system’s scan volume. Small systems can scan parts up to 1m in size, while large systems can scan parts up to 10m or more. For very large parts, multiple scans are often taken and then aligned to form a complete point cloud.

5. Q: How often should an SLS/WLS system be recalibrated? A: Recalibration frequency depends on the system’s usage and environmental conditions. As a general rule, systems should be recalibrated before each scan session, or if the temperature changes by more than 1°C, or if the system is moved or damaged. It is also recommended to perform a full recalibration at least once a month for systems used in production environments.

← Back to home