With the rapid development of material science, more and more workpieces, especially workpieces with complex curved surfaces, are being made of composite materials. Robotic non-destructive testing (NDT) systems for complex curved surface composite material parts are being used more and more. Despite the emergence of such flexible NDT systems, the detection of semi-enclosed parts is also a challenge for robotic NDT systems. In order to overcome the problem, this paper establishes an NDT solution for semi-enclosed workpieces based on a dual-robot system of synchronous motion, in which an extension arm is installed on one of the robots and presents a trajectory planning method that always ensures the extension arm is parallel to the rotary axis of a semi-enclosed workpiece and that the ultrasonic probes are perpendicular to the workpiece surface. Trajectory analysis experiments and ultrasonic NDT experiments utilizing the optimal water path distance determined by simulation result of multi-Gaussian beam model for two types of semi-enclosed workpieces are performed with the dual-robot NDT system. Experimental results prove that the dual-robot NDT scheme functions well and the planned trajectories are correct. All the hole-shaped artificial defects with diameters ≥3 mm are detected by using 2.25 MHz ultrasonic probes through the transmission testing method. Vivid 3D C-scan image of a small diameter cylindrical workpiece based on the testing result is provided for convenience of observation.

Purpose: With the development of materials science and technology, composite workpieces are increasingly used. This paper aims to discuss a non-destructive testing (NDT) solution for semi-enclosed composite workpieces. A dual-robot system with one robot that grips an irregular-shaped ultrasonic probe (tool) is established. Design/methodology/approach: According to robotics, this paper defines the orientations of the discrete points coordinate frames in trajectory and proposes an orientation constraint rule between the tool coordinate frame and the scanning trajectory. A four-posture calibration method for calibrating the transformation relationship of the irregular-shaped tool frame relative to the robot flange frame is presented in detail. Findings: Calibration and verification experiments were performed, and good-quality C-scan images were obtained by applying the constraint rule and the calibration method. Experimental results show that the calibration method used to determine the tool centre point (TCP) position is correct, effective and efficient; the TCP orientation constraint rule can ensure the extension pole of the irregular-shaped ultrasonic probe is parallel to the axis of the semi-enclosed cylindrical workpieces; and the ultrasonic transducer axis is perpendicular to the surface of the workpiece. Originality/value: This paper proposes a constraint method for the posture of an irregular-shaped tool in this scheme. Theoretical foundations for the four-posture calibration method of the irregular-shaped tool for dual-robot-assisted ultrasonic NDT are presented in detail. This strategy has been successfully applied in the NDT experiment of semi-enclosed composite workpieces.

Composite workpieces, especially the complex-curved surfaces composite workpieces, have been increasingly used in different industries. Non-destructive testing of these parts has become an urgent problem to be addressed. To solve the problem, this article presents a dual-robot air-coupled ultrasonic non-destructive testing scheme and introduces the structure of the system and a general calibration method for the workpiece frame of a dual-robot system in detail. Importantly, this article proposes a tangential constraint method, which makes the probes completely aligned during the inspection process. Verification experiments and ultrasonic testing experiments for a glued multilayered composite workpiece were performed using the dual-robot air-coupled ultrasonic non-destructive testing system. A comparative experiment was also performed using a dual-robot water jet-coupling ultrasonic testing system. Experimental results show that the dual-robot non-destructive testing scheme and the tangential constraint method function well, and all the artificial defects on the sample can be detected by both kinds of testing methods. Vivid 3-D C-scan image based on the test result is provided for convenience of observation. In other words, a kind of flexible versatile testing platform with multiple degrees of freedom is established.

The injection of ultrasonic wave into a melt during casting can refine grain size, improve grain distribution, and thereby enhance casting performance. The available studies on ultrasonic-assisted casting are mostly about inserting a transducer directly into the melt. Such a method is not suitable for sand casting. Therefore, the study of ultrasonic sand casting by indirectly injecting ultrasonic waves into an aluminum alloy melt through a sand mold was proposed and carried out in this study. The effects of ultrasonic waves of different powers at different solidification stages on the microstructure and mechanical properties of the melt were studied. Compared to conventional sand casting, the samples prepared by ultrasonic sand casting have finer grains and better grain distribution, as well as higher mechanical properties. Moreover, the sample’s performance improves to different levels when the same ultrasonic wave is injected at different periods, when compared to the injection of ultrasonic waves with different powers in the same period.

Industrial ultrasonic Total Focusing Method (TFM) imaging has become a mainstream non-destructive testing technology for workpiece internal defects in petrochemical, nuclear power and aerospace fields due to its high resolution and full-range focusing performance. Nevertheless, TFM images are plagued by intense speckle noise and low signal-to-noise ratio(SNR) of defect edges, which hinders the distinction between defect features and background. These issues cause existing deep learning-based detection models to suffer from missed and false detections, with limited detection accuracy and generalization capability. To address these issues, this paper presents FDE-YOLO, an improved model for defect detection of industrial ultrasonic TFM images. A dual-branch Feature Edge Enhancement and Calibration (FEEC) module is embedded into the YOLO11n backbone to enhance weak defect edges and suppress speckle noise. Meanwhile, the Feature Cleanliness Coefficient (FCC) is proposed to quantitatively evaluate the noise suppression effect of heat maps. Dynamic upsampling is introduced into the neck network to mitigate the loss of defect details with low SNR. Furthermore, the Efficient Multi-scale Attention (EMA) module is integrated to strengthen the perception of small targets, amplify defect-related feature responses and suppress background noise interference. Experimental results on a self-built dataset show that FDE-YOLO achieves a Precision of 0.9227, Recall of 0.8256 and mAP@0.5 of 0.9399. Its GFLOPs rise slightly from 6.3 to 7.7, and the inference speed reaches 344.31 FPS, fully satisfying industrial real-time detection requirements. This work provides a reliable solution for TFM-based defect detection and promotes the intelligent upgrading of industrial non-destructive testing.