In the aerospace industry, the demand for high-performance components has led to the widespread use of precision castings, such as those made from titanium alloys, high-temperature alloys, aluminum alloys, and magnesium alloys. These aerospace casting parts are characterized by their large size, complex geometries, and internal cavities, making them susceptible to defects like shrinkage pores, gas pores, cracks, and inclusions during manufacturing. Traditional radiographic testing methods for these castings often require extensive trial exposures to avoid missing defects, which is time-consuming and costly. To address this, computer-based radiographic simulation technologies have emerged as efficient tools for optimizing testing processes without material consumption. In this study, I focus on the application of a radiographic simulation software, XRSIM, to simulate and verify the testing of aerospace casting parts, particularly a complex titanium alloy support plate. The research aims to validate the software’s accuracy in simulating exposure parameters, density representation, and defect detectability, ultimately reducing the need for physical trials and enhancing detection reliability.
The fundamental principle of radiographic testing involves the interaction of X-rays or γ-rays with matter, leading to attenuation through effects such as photoelectric absorption, Compton scattering, pair production, and Rayleigh scattering. The intensity of transmitted radiation follows the exponential attenuation law: $$I = I_0 e^{-\mu x}$$ where \(I_0\) is the incident intensity, \(I\) is the transmitted intensity, \(\mu\) is the linear attenuation coefficient, and \(x\) is the material thickness. In film-based radiography, the optical density \(D\) of the developed film relates to the exposure \(H\) as \(D = \lg(I_0 / I)\), and for the linear portion of the characteristic curve, it can be expressed as: $$D = G \lg H + k$$ where \(G\) is the film gradient and \(k\) is a constant. The XRSIM software leverages these principles by using ray-tracing algorithms on 3D STL models of castings aerospace components to compute thickness-dependent intensity distributions, simulating the radiographic image without physical exposure.
XRSIM offers several key functionalities essential for optimizing the testing of aerospace casting parts. First, the thickness analysis feature provides a visual distribution of material thickness, allowing for quick assessment of exposure parameters. For example, a complex support plate can be analyzed to identify regions with varying thickness, which informs voltage selection based on exposure curves. Second, the software enables optimization of testing conditions, such as incident angles and exposure parameters. By adjusting the orientation of the 3D model, users can simulate different perspectives to minimize obstructions and improve defect visibility. Third, the minimum detectable defect function allows users to insert virtual defects of various sizes and materials into the model, determining the smallest detectable flaw under given conditions. Lastly, the Probability of Detection (POD) simulation divides the component into a grid and evaluates detectability based on density differences, highlighting areas where defects might be missed. For instance, a density difference threshold of \(\Delta D = 0.08\) is used to flag undetectable regions in red, while \(\Delta D > 0.12\) indicates detectable areas in green.
To validate the accuracy of XRSIM in simulating real-world radiography, I conducted experiments focusing on density representation, spatial resolution, and contrast sensitivity. Initially, the software’s simulated density values deviated from actual film densities due to differences in environmental conditions and equipment aging. For example, in tests on a titanium alloy wedge block with thicknesses from 1mm to 9mm, the simulated densities at 78kV, 95kV, and 110kV showed significant discrepancies compared to physical film measurements, as summarized in Table 1. After applying a density correction factor of 1.5 to the exposure parameters in XRSIM, the simulated densities aligned closely with the actual values within the acceptable density range of 1.5 to 4, reducing deviations to within ±0.3. This correction ensured that the software could reliably guide parameter selection for aerospace casting parts testing.
| Voltage (kV) | Thickness (mm) | Simulated Density (Uncorrected) | Actual Film Density | Simulated Density (Corrected) |
|---|---|---|---|---|
| 78 | 3 | 1.90 | 2.40 | 2.52 |
| 5 | 0.72 | 1.43 | 1.02 | |
| 7 | 0.29 | 0.90 | 0.43 | |
| 95 | 4 | 2.46 | 3.12 | 3.43 |
| 6 | 1.29 | 2.08 | 1.80 | |
| 8 | 0.69 | 1.45 | 1.02 | |
| 110 | 5 | 2.80 | 3.70 | 3.95 |
| 7 | 1.69 | 2.65 | 2.53 | |
| 9 | 1.03 | 1.88 | 1.60 |
Spatial resolution in XRSIM is determined by the pixel size of the simulated detector, which can be adjusted based on the required image clarity. For most applications, a balance is struck between computation time and resolution. Contrast sensitivity was evaluated using simulated flat-bottom hole IQIs (Image Quality Indicators) integrated into 3D models of titanium blocks. The results demonstrated that the software could achieve a contrast sensitivity of 2-2T, meaning it could detect holes with diameters equal to 2% of the material thickness, which meets the standards for aerospace casting parts inspection. This is critical for ensuring that defects like small pores or inclusions in castings aerospace components are identifiable in simulated images.
In a practical application, I applied XRSIM to the radiographic testing of a titanium alloy support plate, a key aerospace casting part used in engine combustion chambers. This component features complex geometries with varying thicknesses and internal cavities, posing challenges for full coverage detection. The initial testing process involved designing a radiographic procedure card based on standards like ASTM E192, which specifies acceptance criteria for investment steel castings of aerospace applications. Using XRSIM’s thickness analysis, I identified optimal exposure angles and voltages for different sections of the support plate, such as the curved surfaces and internal ribs. The POD simulation was then employed to analyze detectability for minimum defect sizes derived from the standard reference radiographs. For instance, defects as small as 0.5mm diameter gas pores were set as thresholds for thin sections (0–6.35mm), while 1.4mm defects were used for thicker regions (>12.7mm). The simulation revealed undetectable areas, primarily due to thickness variations and obstructions, which were addressed by modifying the testing protocol.
To eliminate these undetectable regions, I implemented strategies such as dual-film techniques and parameter adjustments. For the curved surfaces of the support plate, which had thickness ranges of 6–12mm, the initial single-film approach at 100kV left areas with insufficient sensitivity. By switching to a dual-film setup using MX125 and AA400 films at 110kV, the thinner and thicker regions could be inspected simultaneously, as the different film sensitivities covered the density range effectively. Similarly, for the internal ribs, reducing the voltage from 124kV to 115kV improved detectability in high-thickness areas. The integrated POD analysis after these modifications showed that over 95% of the component was detectable, meeting the requirements for comprehensive inspection of aerospace casting parts. The final testing parameters are summarized in Table 2, which outlines the exposure conditions for each section of the support plate.
| Test Region | Thickness (mm) | Voltage (kV) | Angle (°) | Film Type | IQI Sensitivity |
|---|---|---|---|---|---|
| Curved Surface | 6–12 | 110 | 0 | MX125 & AA400 | 2-2T |
| Base Section 1 | 19 | 138 | 10 | MX125 | 2-2T |
| Base Section 2 | 28 | 173 | 10 | MX125 | 2-2T |
| Base Section 3 | 37–44 | 223 | 10 | MX125 | 2-2T |
| Internal Ribs | 8–12 | 115 | 35 | MX125 | 2-2T |
The experimental verification involved comparing simulated results with actual radiographs of the support plate. For the curved surfaces, the dual-film approach produced images with densities between 1.5 and 4 across all regions, confirming the elimination of undetectable areas. Similarly, the adjusted voltage for the ribs resulted in densities of 2.13–2.93, within the acceptable range. The overall testing sensitivity achieved 2-2T, complying with ASTM E192 standards for aerospace casting parts. This case study demonstrates that XRSIM can effectively guide the design of radiographic procedures for complex castings aerospace components, reducing trial exposures by over 50% and ensuring reliable defect detection.
In conclusion, the integration of radiographic simulation software like XRSIM into the testing workflow for aerospace casting parts offers significant advantages in terms of cost savings, efficiency, and accuracy. By validating density corrections, contrast sensitivity, and POD analyses, this research confirms that the software can replicate real-world testing conditions for castings aerospace applications. Future developments should focus on enhancing computation speed for large components, incorporating more physical factors like scatter radiation, and expanding defect libraries for diverse materials. As the aerospace industry continues to evolve, such digital tools will play a crucial role in achieving high-quality, sustainable manufacturing processes for critical components.