GB/T 41123.3-2021Non-destructive testing - Radiation methods for industrial computed tomography - Part 3: Qualification (English PDF)

4.1 Overview

Industrial CT is used for defect detection and size measurement. The size measurement value (such as aperture or wall thickness) cannot be directly given by CT scanning, which shall be calculated according to the X-ray attenuation data characterized by CT gray value. Feature detectability and detection accuracy depend on the detection task, detection equipment, and the employed analysis and evaluation methods. When the detection task, detection equipment and the employed analysis and evaluation methods are determined, the CT system shall be verified accordingly. See 4.2 and 4.3 for the verification methods. Validation should be performed by trained personnel. Trainees shall demonstrate that they are trained and qualified in digital radiography or computed radiography.

4.2 Verification of defect detection

4.2.1 Overview

In the detection verification, it should confirm whether the CT detection technology meets the detection accuracy requirements. The following verification is a typical process for determining applicability for industrial CT.

4.2.2 Quality characteristics

Typical detected features include the sizes of pores, holes, cracks, inclusions, contamination, as well as material distribution and assembly and mounting locations of components. Since the detection performance of a CT system depends on the type, location, and size of the test sample and features to be detected, the following information should be known.

a) Detected object:

1) Size;

2) Weight;

3) Material;

4) The thickness of X-ray transillumination in the material.

b) Detected characteristics:

1) Type;

2) Location;

Determine the CT system settings and image quality parameters as described in 4.2 and 5.1 of

GB/T 41123.2-2021.

4.2.4 Verification of applicability

4.2.4.1 Overview

A reasonable description of the defect detection sensitivity and defect detectability of the CT system shall be reflected in the detection accuracy (tolerance, fluctuation degree) required by the test. Below are a few options.

4.2.4.2 Reference samples for natural defects

If there are reference samples with known defects, verify the defect detectability by testing the samples.

If the reference sample has unquantified defects, test the sample first, and then re-verify the defect detectability, for instance, using a destructive test for verification.

4.2.4.3 Reference samples for artificial defects

If the detected feature can be simulated by artificial defects (such as holes), the defect detectability shall be verified by the method of 4.2.4.2.

4.2.4.4 Reference samples with unknown parameters

If the reference sample status is unknown and revalidation is impossible, use the system sensitivity for detection. Defect detectability can be estimated by sample structure (e.g., wall thickness, external dimensions, etc.). Other wire and spherical reference samples of known size can also be used.

4.2.5 Consistency checking

CT scanning steps are complex, so it cannot be guaranteed that errors shall be completely eliminated during the scanning process. After the CT scan is complete, possible sources of error can be traced through the following points:

--- Reconstruction: size, CT slice position, possible artifacts;

--- CT image scale;

--- Sinogram (CT gray value and curve changes) or CT projection sequence (comparison between projections, such as projected image quality, intensity changes);

--- System status (error message).

If there is an error, the error shall be corrected or the source of the error shall be eliminated, then re-tested.

4.2.6 Documentation

All parameter settings and verification results in the verification steps are described and presented in the verification report. The CT image archive period is determined according to the end user requirements. Save test parameters for repeated testing of parts and features using the same test flow.

4.3 Verification of dimensional measurement

4.3.1 Overview

CT detection provides the three-dimensional structural information of the sample; and the external contour and geometric structure data of the sample can be obtained through the three-dimensional structural information. Since these data are based on differences in the physical absorption of X-rays at contour transitions, small differences in measured values can occur compared to conventional contact methods or optical measurement processes. The following Clauses describe CT scan parameters that affect the results and process steps that affect the accuracy of the results.

4.3.2 Detection and measurement tasks

Dimensional measurement tasks include single dimension measurement, wall thickness measurement, surface extraction, volume extraction or nominal-actual comparison. The measurement accuracy is specified for each task and, if necessary, separately specified for different parts of the sample.

4.3.3 Dimensional measurement/detection system/system parameter setting

The achievable measurement accuracy depends on the object to be measured, the limitations of the physical principle of X-ray measurement, and the data processing method. The following parameters are used to preliminarily estimate the CT dimensional measurement accuracy.

a) Spatial resolution of the measured object:

1) Image size;

2) Geometric magnification, voxel size;

3) detector resolution;

4) Focus size.

b) X-ray transmission characteristics of the measured object:

1) Materials;

2) Maximum penetration thickness of X-ray;

accuracy. Using spherical and dumbbell-shaped reference test pieces is also a method of estimating measurement accuracy.

Typical parameters include the following:

a) Reference dimensions;

b) Compare the measurement procedure information with the information on the different test areas within the sample;

c) Record the standard deviation of the measurement error as the reference data.

4.3.5 Consistency Checking

See 4.2.5.

4.3.6 Documentation

See 4.2.6.

5.1 Overview

The ability of a CT system to guarantee high quality, stable, and reproducible result depends on the consistency of all system components and how they work together. To ensure this in daily operations, periodic system verification should be carried out in accordance with the relevant requirements.

A distinction should be made BETWEEN short-period (e.g., weekly) verification based on "overall performance" testing, AND long-term (e.g., annual) verification based on quality level descriptions and testing of possible changes to individual system components.

5.2 Verification of overall system

For routine system monitoring, the reference sample should be similar to a typical sample used by CT systems. During the overall validation process, system parameters similar to those used for testing typical samples should be used.

To evaluate system quality, the current test results are compared with reference measurement results. The measurement results of different object structures, such as material defects (pores, cracks), the thinnest and thickest parts of the reference test block, and wall thickness should be used as the quality evaluation criteria.

If combined systems (two ray tubes and/or detectors) are used, separate reference blocks shall be used for each system combination (e.g., application of microfocus and small focus).

Test results and system status shall be recorded and archived.

If discrepancies are found, further examination shall be performed to determine the cause (see 5.3). After system maintenance or major adjustment, system performance verification should be carried out before using the system.

5.3 System component checking

5.3.1 Overview

Periodically and when changes to the system are suspected (after maintenance or in the event of a crash), the following system components that may be affected shall be checked.

5.3.2 Mechanical system

The positioning accuracy of trajectories and axes shall be checked, and the measuring equipment can use coordinate measuring machines (CMMs).

5.3.3 Image magnification

According to Figure 1 in GB/T 41123.2-2021, it should use a high-accuracy sphere group (such as a club, dumbbell) with a known spatial structure to check the CT image magnification.

Differences in the CT grayscale thresholds used by these samples did not affect the obtained size results.

5.3.4 Perpendicularity of beam

Use a suitable test sample (e.g., tungsten wire or tip, sphere, etc.) to check the perpendicularity of the beam axis to the detector.

5.3.5 Focus position

The focal position of the radiation source shall be checked using a suitable method, for example, by checking the consistency of the dimensions (within a specified error range) obtained from CT scans at different magnifications to ensure the focal position.

5.3.6 Dose stability

Measuring the dose rate can check the stability of the X-ray tube output.

5.3.7 Detectors

The dynamic behavior of the detector can be checked by comparison with delivery conditions, e.g., by imaging a stepped reference block. Detectors should be regularly checked for pixel failures.

The stability of the detector can be checked using time-varying intensity measurements.

Reference test piece shall be suitable for specific equipment, such as microfocus systems and high-energy systems. In general, the reference test piece should be as close as possible to the measured object in attenuation and size. If necessary, customize more reference test pieces to meet the requirements.

The following Clauses describe the method of partial implementation of the comparative system on the reference test piece and several devices with different designs, fabrication methods, and ages.

The recommended method should be selected according to the detection situation. Guidelines for making reference test pieces are given in 6.3.

Theoretically, the measurement values given by the following methods are suitable for all cases.

6.2 Acquisition parameters

Since each CT system has its own method of image acquisition and reconstruction, it is important for each system to establish resolution standard measurement method based on optimal voltage, voxel size, and division increment (where possible).

6.3 Recommendations for making reference test pieces

The recommended method uses two types of reference test pieces, including the following types:

--- It is consisted by the components containing a row of calibrated holes for measuring spatial resolution described in Figure A.1 of GB/T 41123.2-2021;

--- It is consisted by the components containing additives for measuring density resolution, see Figure 1.

Since all measurements are related to the properties of the reference test piece, great care should be taken when defining and creating them.

In order to ensure optimal measurement conditions, the reference test piece should meet specific requirements. To avoid edge artifacts due to angular effects as described in GB/T 41123.2-2021, a cylindrical geometry is chosen.

In order to measure the density resolution, the linear attenuation coefficient of the additive shall be close to that of the base material to avoid the "edge effect" artifacts described in GB/T 41123.2-2021. Differences in attenuation coefficients between additives shall be kept low to ensure higher detection sensitivity. Appropriate beam hardening correction should be performed during CT image reconstruction. In addition, to ensure applicability, the material of the reference test piece shall be chemically and densely similar to the object to be tested. This is due to the fact that CT images measure X-ray attenuation coefficients that are related (indirectly proportional) to material density.

The materials that make up the matrix and additives should be uniform; density variations shall