3D printing quality control using 3D scanning

The printed part may look flawless, but this does not mean that its geometry corresponds to the original digital model. A slight deformation of the landing surface, a violation of roundness, or a local excess of material can lead to problems during assembly, increase the amount of machining, and even cause failure of the finished assembly.

It is especially difficult to control products with curved surfaces, ribs, blades and internal cylindrical elements. Measuring such a part with a caliper at several points is not enough: local defects may go unnoticed.

3D scanning of printed parts allows you to solve this problem. It digitizes the actual surface of the product, after which the resulting model can be combined with the original CAD model and deviations can be seen throughout the available geometry.

The effectiveness of this approach is shown in the scientific work "Investigation of deviations of the actual geometry of parts manufactured using 3D printing technology from the nominal geometry of their digital counterparts." For control, the researchers used a RangeVision PRO 3D scanner and 3D Systems Control X software.

When preparing for 3D printing, a solid-state CAD model is usually converted to a polygonal format, such as STL, and then transferred to a slicer. Already at the export stage, there may be a slight decrease in the accuracy of the surface description.

After that, the result is affected by the printing parameters and the condition of the equipment:

• the height of the layer and the diameter of the nozzle;
• print head movement speed;
• extrusion temperature and stability;
• feed rate of the material;
• shrinkage and thermal deformation;
• condition of the drive belts and guides;
• calibration of stepper motors;
• nozzle wear or contamination;
• the position of the part on the printing platform.

Some of the defects are visible visually, but many geometric deviations cannot be reliably estimated without measurements. For example, a round surface may become slightly oval, a blade may deviate from the design angle, and overextrusion may occur in a separate area.

Such changes are especially critical for functional products. Even a small discrepancy with the digital twin can affect the assembly, load distribution, tightness and service life of the node.

The main task of 3D scanning in additive manufacturing control is to objectively compare the actual geometry of the printed part with its nominal CAD model.

Instead of several separate dimensions, the specialist gets a digital representation of the entire available surface. This allows you to determine

• where the part is larger or smaller than the nominal value;
• which surfaces went beyond the established tolerance;
• are there deformations and distortions;
• is the roundness of the cross sections broken;
• are there local surges or shortages of material;
• how are the deviations distributed along different axes?;
• is mechanical processing required;
• which printer settings need to be adjusted.

In other words, 3D scanning not only detects a marriage, but also provides data to find its cause.

The process of checking a printed product consists of several stages.

1. 3D scanning of the part
The product is scanned from different angles. Based on the data obtained, a point cloud is formed, which is then combined into a single polygonal 3D model. The quality of the result depends not only on the capabilities of the scanner, but also on the correct preparation of the object, calibration of the equipment and the chosen shooting strategy. Please note that an optical 3D scanner captures surfaces that are in the camera's field of view. Completely closed cavities and inaccessible areas require a separate approach to control.

2. Loading the reference CAD model
The initial model is imported into the metrological software, which acts as a digital reference. The study used 3D Systems Control X software for analysis.

3. Combining models
The scan and the CAD model are located in different coordinate systems, so they must be correctly combined before analysis. Depending on the task, alignment can be performed over the entire surface, basic elements, or functionally significant areas. The method of combining is of fundamental importance. For example, the global best fit is convenient for the overall shape assessment, and the alignment across the bases better reflects the installation conditions of the part in the real node.

4. Building a color deviation map
After the alignment, the program calculates the distances between the actual and nominal surfaces. The result is displayed as a colormap.

It allows you to quickly see:

• areas with excess material;
• areas where the actual surface area is less than the nominal one;
• areas within the tolerance range;
• local maxima and minima of deviations.

The color map turns a large array of measurement data into a visual picture that is understandable to the technologist, designer and quality specialist.

5. Анализ сечений и подготовка отчета
Общей карты не всегда достаточно для принятия решения о пригодности детали. Поэтому дополнительно исследуют рабочие, посадочные и сопрягаемые поверхности, строят поперечные и продольные сечения, проверяют круглость и другие геометрические параметры. Результаты можно оформить в виде отчёта с числовыми значениями, изображениями, гистограммами и отметками о соответствии допускам.

The researchers selected the turbine pump guide as a test product. This is a part with a complex spatial geometry: it includes a hub part, an outer body, guide vanes, cylindrical, flat and curved surfaces.

The model was made using the FDM method from PETG with the addition of 12% chopped fiberglass. A nozzle with a diameter of 0.5 mm, a layer with a height of 0.15 mm, a speed of 30 mm/s and a 100% filling were used for printing.

The printed part was scanned using RangeVision PRO, and the resulting polygonal model was compared with the original geometry in Control X.

A tolerance of ±0.1 mm was set for three-dimensional analysis. The results showed that

• the average deviation was 0.082 mm.;
• The maximum positive deviation reached 1.2108 mm.;
• the maximum negative is -1.2111 mm;
• 27.98% of the analyzed surface was in the tolerance field;
• 72.02% of the surface was outside the tolerance.

These indicators relate to a specific part, material, printer, and printing mode, so they cannot be considered a universal characteristic of FDM technology. However, the experiment clearly shows the main thing: the average value alone does not give a complete picture of the quality.

With a relatively small average deviation, significant local defects were present on the surface. Without a colormap and cross-section analysis, they could go unnoticed.

What defects were found

A three-dimensional comparison showed that the actual dimensions of the part have mostly increased relative to the nominal ones. This means that there is an allowance that can theoretically be removed by subsequent machining. But the distribution of deviations turned out to be uneven.

The maximum deflection zone was located at the junction of one of the guide vanes to the hub part. Since the defect was not repeated on the other blades, the researchers attributed it to local printing disorders. A possible cause was overextrusion, which is the accumulation of material on the nozzle, followed by the transfer of excess to the surface.

Another characteristic distribution was found on the stiffeners: negative deviations prevailed on the one hand, and positive ones on the other. This pattern indicates an angular displacement during printing.

Cross-sectional analysis revealed a violation of roundness. It may be related to the condition of the drive belts and guide shafts, or to incorrect calibration of printer movements along mutually perpendicular axes.

The longitudinal section study also confirmed a probable violation of the mechanical calibration and showed a local excess of material.

Thus, 3D scanning helped not only to fix the size discrepancy, but also to link the nature and location of the deviations with possible problems with the equipment and the printing process.

Control of the entire available surface
Contact measuring tools allow you to check only the selected dimensions and points. The 3D scanner collects data over the entire visible surface, including complex curved sections.

Visual identification of local defects
The color map immediately shows where the deviation is and how large it is. This is especially useful when analyzing details for which one average value is not enough.

Checking complex geometry
Blades, ribs, junctions, organic surfaces, and other elements are difficult to control by traditional means. In a digital model, you can analyze the shape, cross-sections, and relative positions of surfaces.

Search for systematic errors
If the same deviations are repeated on a series of products, the scan data helps to detect a pattern and check the printer settings, the state of mechanics or the stability of the technological process.

Reducing the number of trial iterations
Objective information about deviations allows you to purposefully change the print settings. The technologist does not have to adjust the process based solely on visual assessment or collect several unsuccessful samples.

Reducing the cost of marriage and processing
The control helps to avoid printing obviously unsuitable products, to assign allowances more precisely and to reduce the amount of mechanical work. This is especially important for large parts, expensive materials, and long printing cycles.

Digital traceability
The test results can be saved as measurement models and reports. The company receives a documented quality history for different batches, printers, materials, and production modes.

The most promising application scenario for 3D scanning is not a one-time inspection of the finished product, but the creation of a closed quality management cycle:

CAD model → 3D printing → 3D scanning → deviation analysis → parameter adjustment → reprinting.

In this cycle, the scanner becomes a source of feedback. Based on the measurement results, you can:

• adjust the displacement coefficients along the axes;
• control the wear of the nozzle and mechanical components;
• specify the temperature and feed of the material;
• choose the optimal printing speed;
• evaluate the effect of layer height and filling;
• calculate shrinkage compensation;
• change the orientation of the product on the platform;
• specify allowances for machining.

When controlling a series of identical parts, it becomes possible to statistically evaluate the stability of the process. If the geometry gradually changes from product to product, this may indicate equipment wear, temperature drift, or a change in material properties.

As we move from prototyping to mass production of functional products, the requirements for 3D printing control will increase. It is important for the industry not only to print the part, but also to confirm that it meets the digital double and the specified tolerances.

In the future, 3D scanning can be used for:

• automated input and output control;
• stability assessments of serial 3D printing;
• settings for new materials and modes;
• creating digital product passports;
• predicting the technical condition of the printer;
• automatic compensation of systematic deviations;
• control of parts after heat treatment and mechanical refinement;
• inspections of repaired and refurbished products;
• learning algorithms for defect analysis.

The integration of scanning with CAD/CAM systems, metrological software and production management systems is especially important. The accumulated inspection results can form a database linking geometric deviations to a specific printer, material, part orientation, and printing mode.

This opens the way to predictive control, in which the system not only detects a defect, but also helps determine its probable cause before the next batch is manufactured.

The study showed that 3D printing performance cannot be considered separately from geometric accuracy. An attempt to speed up the process without objective control can lead to an increase in the number of defects, the cost of refinement and a decrease in the reliability of products.

The combination of the RangeVision PRO 3D scanner and metrological software allowed the researchers to digitize a complex part, compare it with the original CAD model, quantify deviations and localize possible causes of defects.

For an enterprise, this means moving from a subjective assessment of quality to working with measurable data. 3D scanning helps not only to answer the question "does the part match the model", but also to understand why there was a discrepancy and what needs to be changed during the production process.

That is why technology is becoming an important element in the development of additive manufacturing.: It reduces the number of failed iterations, helps to use materials more efficiently, and allows for the production of 3D-printed products with controlled and documented geometry.