Inline X-Ray CT for Metal AM Defects — PatSnap Eureka
Inline X-Ray CT for Internal Defect Detection in Metal AM Parts
Industrial computed tomography is the internationally recognised gold standard for volumetric defect inspection in metal additive manufactured parts — detecting porosity, cracks, lack-of-fusion, and oxide films simultaneously, without part destruction. Explore the patent landscape across 40+ filings from Boeing, Xi'an Jiaotong University, GE, and more.
The Gold Standard for Volumetric Defect Inspection in Metal AM
Industrial computed tomography — also known as micro-CT (μCT) or X-ray CT (XCT) — is broadly recognised across the patent literature as the most effective non-destructive technique for resolving internal defects in metal additive manufactured parts. Unlike 2D radiographic methods, which suffer from image overlap and cannot unambiguously locate embedded flaws in three dimensions, CT reconstructs volumetric data without superposition artefacts.
As documented by Shanghai Intelligent Manufacturing Functional Platform Co., Ltd. (2022), 3D ray detection technology overcomes the fundamental limitation of 2D radiographic methods — shadow overlap — and is internationally recognised as the most effective non-destructive inspection technique. Traditional contact-based ultrasonic inspection is unsuitable for complex-surface parts, while 2D X-ray is insufficient for detecting defects in geometrically complex structures.
Laser powder bed fusion (L-PBF) and directed energy deposition (DED) processes routinely introduce porosity (gas pores, keyhole pores), lack-of-fusion defects, cracks, inclusions, and residual powder in internal channels. CT can detect all such volumetric anomalies simultaneously in a single scan — a capability no alternative NDE method currently matches. The World Intellectual Property Organization has catalogued this technology across multiple international patent families, underscoring its global strategic importance.
A uniquely powerful CT capability is the characterisation of oxide films on defect inner walls — invisible to surface-only or lower-resolution techniques. Beijing University of Science and Technology (2024) describes a CT-based workflow using linear absorption coefficient (LAC) analysis and discrete least squares segmentation (DLSS) to reconstruct the three-dimensional spatial distribution of metal matrix, oxide film, and pore phases. This level of sub-defect characterisation is not achievable by ultrasound, eddy current, or optical methods.
Detecting Defects During the Build — Not Just After
The most significant frontier in the field is the integration of X-ray CT directly into the AM build process, enabling layer-by-layer defect detection and immediate process interruption.
Linear Aperture Real-Time CT: Pause, Rotate, Scan
Boeing's dominant inline CT patent family describes a system that pauses additive manufacturing, rotates the partially built object within the build chamber by at least half a full rotation, and directs an X-ray pulse through a linear aperture toward the object. A linear X-ray detector array captures transmission data to reconstruct an X-ray image of the entire object volume. Pattern recognition compares the image against a reference model. If a defect is found, manufacturing is halted immediately — saving time, material, and cost. Scintillating materials including CsI:Na, Gd₂O₂S, and CaWO₄ convert X-rays to visible light for the detector.
Detects defects of ~1 mm and larger during fabricationFully Integrated CT in EBSM: Simultaneous Print-and-Inspect
GUET's system integrates an industrial CT subsystem directly into an electron beam selective melting (EBSM) machine, achieving simultaneous print-and-inspect operation. A rotating platform, Y-axis, and Z-axis movement system give the X-ray source and detector six degrees of freedom, keeping them aligned with the part centre to guarantee imaging accuracy. The system maintains a defect database; when a defect is identified during printing, the database retrieves previously identified optimal parameter corrections, which are fed back to the electron beam system to prevent recurrence. This closed-loop inline CT architecture represents a significant step beyond purely passive inspection.
Closed-loop parameter correction from defect databaseTomographic Scanning at Intermediate Build Stages for Geometric Correction
GE's approach uses tomographic scanning (CT) at intermediate build stages to identify thermal distortion and geometric defects. A partial build is scanned; the resulting model is converted from scanner format to AM system format and compared against the intended 3D model to identify deviations. The corrected model is then used to adjust subsequent build layers. This approach — using CT to close the loop between as-built geometry and design intent — is particularly relevant for high-value aerospace components where dimensional accuracy is as critical as internal soundness.
As-built geometry vs. design intent comparisonCT-OT Fusion: Training Predictive Models Without Scanning Every Part
Boeing's process-monitoring patent explicitly integrates CT data with optical tomography (OT) data from the build process. A predictive model is trained by correlating OT-detected build anomalies with post-build CT-confirmed defects. After training, the model monitors production builds in real time using OT signatures to predict the likely CT-equivalent defect — enabling process parameter updates (laser power, scan speed, gas flow, powder temperature) without requiring every part to be CT-scanned. CT serves as the ground-truth reference for training and validating all alternative monitoring approaches.
CT as ground truth for OT predictive model trainingCT Inspection for Metal AM — Key Metrics from 40+ Filings
Quantitative analysis of the patent record from PatSnap Eureka, covering jurisdiction distribution and technology focus areas across 2017–2026.
Patent Filing Distribution by Jurisdiction (CT Inspection for Metal AM, 2017–2026)
Chinese institutions dominate the filing landscape, reflecting intense national investment in quality assurance for metal AM. Boeing accounts for the majority of US and EP filings.
Technology Focus Areas in CT Inspection Patent Filings (2017–2026)
Inline/real-time CT integration and post-build defect characterisation are the two largest technology clusters, with calibration standards and FEM integration growing rapidly post-2022.
From CT Scan to Structural Certification: FEM Integration and In-Situ Loading
A growing body of patents translates CT-measured defect data into quantitative structural performance predictions, bridging inspection and engineering acceptance decisions.
Boeing: CT Defects Embedded Directly into Finite Element Models
Boeing's method uses CT image data to detect and classify observable defects in metal parts — including 3D-printed and forged parts — and embed them directly into a finite element model (FEM). The FEM is then analysed using isotropic and/or anisotropic structural models to generate test results that inform quality control decisions. This approach explicitly addresses the limitation that traditional NDE applied to finished 3D-printed parts is sub-optimal for understanding how complex microstructural differences affect overall structural behaviour. Learn more about IP analytics for aerospace on the PatSnap platform.
Shenyang Aerospace University: Performance Coefficients from NDE Measurements
Shenyang Aerospace University addresses the relationship between CT-detected internal defects and mechanical performance. Coefficients for defect size, shape, and positional influence on mechanical performance are developed from non-destructive measurements, allowing engineers to estimate part mechanical properties without destructive testing once defect geometry is known from CT. This reflects the broader movement toward quantitative NDE as a substitute for extensive mechanical test programmes in regulated industries. The National Institute of Standards and Technology has similarly highlighted quantitative NDE as a priority for AM qualification.
Reference Specimens and Detection Sensitivity: The Foundation of CT Qualification
Reliable CT-based AM part qualification depends critically on system-specific detection sensitivity established through reference specimens and calibration protocols — an active area of patent development.
| Organisation | Year | Calibration Approach | Key Capability |
|---|---|---|---|
| Shanghai Aircraft Manufacturing Co. | 2022 | Modular CT reference block with embedded artificial defects at micrometer level; configurable as cylindrical or frustum geometries | Traceable reference standard for CT detection sensitivity at specific defect sizes and burial depths |
| Shanghai Intelligent Manufacturing Platform | 2022 | Comparison specimens with artificial defects of known dimensions; full sensitivity matrix across specimen diameter, height, defect position, and burial depth | High-precision three-dimensional evaluation and traceability for small defects |
| Hebei University of Technology | 2026 | Micro-CT (μCT) with prefabricated void cavities via Boolean-subtraction CAD modelling; controlled sizes, shapes, and positions | High-throughput batch qualification framework for AM parts with defined void-size ranges for accurate μCT characterisation |
| Beijing Xinghang Electro-Mechanical Equipment Co. | 2021–2022 | Reference blocks with composite layers incorporating defects of varied sizes and depths; CT transmission imaging compared to metallographic cross-sections until size agreement confirmed | Addresses detection sensitivity shift when system parameters and material types change in multi-layer AM structures |
| AECC Shenyang Liming Aero-Engine Co. | 2022 | X-ray inspection comparison specimens designed for internal channel defects in laser selective melting AM parts | Covers porosity, unfused powder, cracks, lack-of-fusion, foreign object debris, and residual powder in flow channels |
Find All Calibration and Reference Specimen Patents for Metal AM CT
PatSnap Eureka surfaces the full filing history, assignee relationships, and claim scope across all CT calibration patent families.
Key Assignees and Their Focus Areas in CT Inspection for Metal AM
Based on frequency and technical depth of filings across the reviewed dataset, these organisations lead the global CT inspection patent landscape for metal additive manufacturing.
Boeing Company
The most prolific international filer, with at least five distinct patent documents across US, EP, JP, and CN jurisdictions. Boeing's filings cover inline real-time CT inspection (the linear aperture/detector array architecture), CT-OT data fusion for predictive process monitoring, FEM integration of CT-detected defect data, and load-based verification of AM parts with internal flaws. Boeing's filings reflect a vertically integrated approach linking detection, process control, and structural certification. The European Patent Office has granted Boeing's EP family as recently as 2025.
US · EP · JP · CN — 5+ distinct documentsXi'an Jiaotong University
Holds multiple active Chinese patents focused on in-situ XCT under mechanical loading, defect tracking via point-cloud registration, gravitational hazard modelling, and forward-reverse defect traceability. Their work addresses failure prediction at the individual-defect level — a capability unique in the patent record. This work is highly relevant to life sciences and regulated-industry applications where individual component certification is required.
In-situ loading · Point-cloud tracking · Hazard modellingBeijing University of Science and Technology
Has filed two related active patents on CT-based oxide film characterisation in L-PBF alloy defects — a highly specialised application of industrial CT unique to powder bed AM. The use of LAC analysis and DLSS segmentation to reconstruct three-dimensional spatial distribution of oxide film, metal matrix, and pore phases represents a sub-defect characterisation capability not achievable by any other NDE method.
LAC analysis · DLSS segmentation · L-PBF alloysShanghai Aircraft Manufacturing Co. / COMAC Group
Has contributed calibration standards and reference block methodologies for CT inspection of AM aerospace parts. The modular reference block assembly with micrometer-level artificial defects provides traceable standards for validating CT detection sensitivity — a critical enabler for aerospace qualification programmes. Explore how aerospace customers use PatSnap for IP and R&D intelligence.
Micrometer-level artificial defects · Traceable standardsFrom Build Chamber to Structural Certification: The CT Inspection Workflow
The patent literature describes a clear escalation from post-build CT inspection toward inline CT integrated with the build process, and from qualitative defect detection toward quantitative coupling with finite element simulation and machine-learning classification. PatSnap's IP analytics platform enables R&D and quality teams to track this evolution in real time across all major assignees and jurisdictions.
In Boeing's inline system, the process pauses, the part rotates at least half a revolution, an X-ray pulse is directed through a linear aperture, and a detector array reconstructs the full volumetric image. Pattern recognition then compares this against the reference model. If a defect of approximately 1 mm or larger is identified, manufacturing halts immediately. This saves time, material, and cost — and prevents the build of additional layers on a defective foundation.
GUET's EBSM-integrated system adds a closed-loop dimension: when a defect is identified, the system retrieves optimal parameter corrections from a defect database and feeds them back to the electron beam system, preventing recurrence. GE's approach closes the loop on geometry rather than defects — comparing CT-reconstructed as-built geometry against the intended 3D model and adjusting subsequent layers accordingly.
At the post-build stage, Boeing's FEM integration method embeds CT-detected defects directly into finite element models for isotropic and anisotropic structural analysis. Xi'an Jiaotong University's in-situ XCT method goes further, tracking individual defects under mechanical loading and reverse-tracing from fracture surfaces to identify the causally responsible original defects. The ASTM International standards body has published guidance on CT for AM that aligns with these patent-documented approaches.
Across all approaches, CT serves as the ground-truth reference for training and validating all alternative monitoring methods — optical, acoustic, eddy current, and laser ultrasonics — reflecting its status as the most reliable volumetric NDE method available for metal AM. For teams building data-driven AM quality systems, PatSnap's open API provides programmatic access to this patent intelligence.
Inline X-Ray CT for Metal AM Defects — Key Questions Answered
Industrial CT can detect porosity (gas pores, keyhole pores), lack-of-fusion defects, cracks, inclusions, residual powder in internal channels, foreign object debris, and oxide films on defect inner walls. This diversity of defect types motivates the use of CT, which can detect all such volumetric anomalies simultaneously in a single scan.
Inline CT integrates X-ray scanning directly into the AM build process, enabling detection of defects layer-by-layer or at intervals during fabrication rather than only after build completion. Boeing's inline system pauses additive manufacturing, rotates the partially built object, and directs an X-ray pulse to reconstruct a volumetric image — detecting defects of approximately 1 mm and larger. If a defect is found, manufacturing is halted immediately, saving time, material, and cost.
Yes. Boeing's method uses CT image data to detect and classify observable defects in metal parts and embed them directly into a finite element model (FEM). The FEM is then analyzed using isotropic and/or anisotropic structural models to generate test results that inform quality control decisions. Shenyang Aerospace University has also developed coefficients for defect size, shape, and positional influence on mechanical performance from non-destructive measurements, allowing engineers to estimate part mechanical properties without destructive testing.
Reliable CT-based qualification requires traceable reference specimens with artificial defects of known geometry. Shanghai Aircraft Manufacturing Co. disclosed a modular CT reference block assembly with embedded artificial defects at the micrometer level. Shanghai Intelligent Manufacturing Platform uses comparison specimens with artificial defects of known dimensions to characterize how specimen diameter, height, defect position, and burial depth each affect the CT-measured defect size, providing documented detection limits.
Industrial CT reconstructs volumetric data without superposition artifacts that limit 2D radiographic methods. Traditional contact-based ultrasonic inspection is unsuitable for complex-surface parts, while 2D X-ray is insufficient for detecting defects in geometrically complex structures. Other inspection methods — ultrasound, eddy current, laser ultrasonics, and optical monitoring — appear throughout the patent literature but are frequently positioned as complements to, or alternatives limited in scope compared to, CT.
Boeing Company is the most prolific international filer, with at least five distinct patent documents across US, EP, JP, and CN jurisdictions. Xi'an Jiaotong University holds multiple active Chinese patents on in-situ XCT under mechanical loading and defect traceability. Beijing University of Science and Technology has filed patents on CT-based oxide film characterisation. Shanghai Aircraft Manufacturing Co. and General Electric have also made important contributions to calibration standards and tomographic-scanning-corrected build processes respectively.
Still have questions about CT inspection for metal AM? Let PatSnap Eureka answer them instantly from the patent record.
Ask Eureka Your CT Inspection QuestionsAccelerate Your Metal AM Quality R&D with CT Patent Intelligence
Join 18,000+ innovators already using PatSnap Eureka to accelerate their R&D. Search 40+ CT inspection patents, map assignee landscapes, and link defect data to structural performance — all in one AI-powered platform.
References
- Real time additive manufacturing process inspection — The Boeing Company, 2020 (US)
- Real time additive manufacturing process inspection — The Boeing Company, 2020 (EP)
- Real time additive manufacturing process inspection — The Boeing Company, 2025 (EP grant)
- Real time additive manufacturing process inspection — Boeing Company, 2020 (CN)
- Real time additive manufacturing process inspection — The Boeing Company, 2020 (JP)
- Progressive damage and failure analysis of metal parts using computed tomography — Boeing Company, 2023 (CN)
- Methods and systems for in-process monitoring of additive manufacturing — Boeing Company, 2025 (CN)
- A non-destructive testing method for spatial distribution of oxide films on alloy defect inner walls — Beijing University of Science and Technology, 2024 (CN)
- An industrial CT defect detection capability evaluation method for small internal defects in metal parts — Shanghai Intelligent Manufacturing Functional Platform Co., Ltd., 2022 (CN)
- A method and system for locating hazardous defect sources in metal interiors based on industrial CT — Xi'an Jiaotong University, 2024 (CN)
- A method and system for forward tracking of metal internal defects and reverse positioning of failures — Xi'an Jiaotong University, 2024 (CN)
- A method for evaluating the effect of internal defects on the performance of additively manufactured parts — Shenyang Aerospace University, 2024 (CN)
- A device and method for electron beam selective melting integrating an online CT inspection system — Guilin University of Electronic Science and Technology, 2024 (CN)
- Defect correction for additive manufacturing using a tomographic scanner — General Electric Company, 2017 (CN)
- An industrial CT inspection comparison block, manufacturing method, and inspection method — Shanghai Aircraft Manufacturing Co., Ltd., 2022 (CN)
- A high-throughput non-destructive testing method for prefabricated metal AM defect specimens — Hebei University of Technology, 2026 (CN)
- A method for determining X-ray detection sensitivity for additive manufacturing multi-layer structures — Beijing Xinghang Electro-Mechanical Equipment Co., Ltd., 2021 (CN)
- A method for designing and manufacturing X-ray inspection comparison specimens for internal channel defects in laser selective melting AM parts — AECC Shenyang Liming Aero-Engine Co., Ltd., 2022 (CN)
- World Intellectual Property Organization (WIPO) — International patent filings database
- European Patent Office (EPO) — EP patent grants for AM inspection technology
- ASTM International — Standards for computed tomography in additive manufacturing
- National Institute of Standards and Technology (NIST) — Quantitative NDE for AM qualification
All data and statistics on this page are sourced from the references above and from PatSnap's proprietary innovation intelligence platform. Patent analysis covers 40+ filings across China, the United States, Europe, Japan, and WIPO (2017–2026).
PatSnap Eureka searches patents and research to answer instantly.