Yongfeng Song

Novel path generation algorithm for high-speed pocket milling

A novel tool path generation algorithm is presented for high-speed pocket milling. A spiral fashion polyline is obtained by linear interpolation which is offset from the pocket boundary. The control points are then inserted into the polyline under exponential regularisation. Finally, the smooth tool path is generated via B-spline curve fitting. The tool path generated through this method can be directly used by the machine controller with a NURBS interpolation function, which avoids the time-consuming computation of dividing them into many short line segments. Compared with conventional methods of direction-parallel and contour-parallel path generation, the tool paths obtained by this algorithm have less cutting force fluctuation because they are Cncontinuous, smooth and have no direction jump. The engineering applications proved that the presented method can not only save machining time over the conventional algorithm, but also can avoid solving time-consuming scalar elliptic partial differential equations which are needed by existing spiral tool path generation method.

A self-reciprocity calibration method for broadband focused transducers

A procedure is developed for self-calibration of broadband, spherically focused ultrasonic transducers based on reciprocity. The input and received signals are measured in a pulse-echo configuration. These signals are used in conjunction with a multi-Gaussian beam model to obtain the electromechanical transfer function of the transducer. This calibration procedure is advantageous because it reduces the experimental configuration to a single transducer and a reflector. Experimental results indicate that the transfer function is insensitive to on-axis reflector placement. This result supports the feasibility of integrating the calibration procedure into actual testing in some situations.

Statistics associated with the scattering of ultrasound from microstructure

HIGHLIGHTSThe statistical behavior of ultrasonic scattering in polycrystals is investigated.Maximum scattering amplitudes can be theoretically predicted.Bounds and confidence levels on scattering amplitudes are derived.The extreme value statistics theory is applied to ultrasound scattering.The statistics are relevant for any finite number of waveforms N. ABSTRACT The spatial statistics of an ensemble of waveforms containing ultrasonic scattering from microstructure are investigated. The standard deviation of the waveforms is of primary interest, because it is related to the maximum scattering amplitudes in the extreme value statistics theory. Further statistical measures are employed to define theoretical confidence bounds, which bound the experimentally calculated maximum amplitude when a finite number of waveforms are included in the ensemble. These statistical measures are applied in conjunction with a previously developed ultrasonic backscatter model. It is validated through ultrasonic scattering measurements performed on a stainless‐steel pipe sample. These considerations are important for forward models related to the probability of detection (POD) of defects and inverse models used for characterization of polycrystalline microstructures.

Evaluating grain size in polycrystals with rough surfaces by corrected ultrasonic attenuation

HIGHLIGHTSWe study the effects of diffraction and surface roughness on ultrasonic attenuation.The MGB model is used to formulate the diffraction correction coefficient.We develop an approximate inverse function of Weavers model.Grain size evaluation results reveal the better accuracy of the presented method. ABSTRACT Surface roughness of a sample has a great effect on the calculated grain size when measurements are based on ultrasonic attenuation. Combining modified transmission and reflection coefficients at the rough interface with a Multi‐Gaussian beam model of the transducer, a comprehensive correction scheme for the attenuation coefficient is developed. An approximate inverse model of the calculated attenuation, based on Weavers diffuse scattering theory, is established to evaluate grain size in polycrystals. The experimental results showed that for samples with varying surface roughness and matching microstructures, the fluctuation of evaluated average grain size was ±1.17 &mgr;m. For polished samples with different microstructures, the relative errors to optical microscopy were no more than ±3.61%. The presented method provides an effective nondestructive tool for evaluating the grain size in metals with rough surfaces.

Diffuse ultrasonic backscatter using a multi-Gaussian beam model

Diffuse ultrasonic backscatter is widely used to evaluate microstructural parameters of heterogeneous materials. Recent singly scattered response (SSR) models utilize a single-Gaussian beam (SGB) assumption which is expected to have limitations. Following a similar formalism, a model is presented using a multi-Gaussian beam (MGB) assumption to characterize the transducer beam for longitudinal-to-longitudinal scattering at normal incidence through an interface with arbitrary curvature. First, the Wigner transform of the transducer field is defined using conjugate double-layer MGB expressions. The theoretical analysis shows that ten groups of Gaussian beams are sufficient for convergence. Compared with the SGB-SSR curve, the shape of MGB-SSR curve is positive skewed. Differences between the MGB-SSR model and the SGB-SSR model are quantified and shown to be complex functions of frequency, sample curvature, transducer parameters, and focal depth in the material. Finally, both models are used to fit experimental spatial variance data from a 304 stainless steel pipe with planar, convex, and concave surfaces. The results show that the MGB-SSR has some characteristics suggesting a better fit to the experiments. However, both models result in grain size estimates within the uncertainty of the optical microscopy suggesting that the SGB is sufficient for normal incidence pulse-echo measurements.

Flaw sizing method based on ultrasonic dynamic thresholds and neural network

A dynamic threshold method for ultrasonic C-Scan imaging is developed to improve the performance of flaw sizing: the reference test blocks with flat-bottom hole flaws of different depths and sizes are used for ultrasonic C-Scan imaging. After preprocessing, flaw regions are separated from the C-scan image. Then the flaws are sized roughly by 6-dB-drop method. Based on the real size of flat-bottom holes, enumeration method is used to get the optimal threshold for the flaw. The neural network is trained using the combination of amplitude and depth of flaw echo, the rough size of flaw and the optimal threshold. Finally, the C-Scan image can be reconstructed according to dynamic threshold generated by trained RBF NN. The experimental results show that the presented method has better performance and it is ideally suited for automatic analysis of ultrasonic C-scan images.

Combining physical shell mapping and reverse-compensation optimisation for spiral machining of free-form surfaces

Machining of free-form surfaces has an important role in industrial manufacturing, but conventional tool-path generation strategies for free-form surfaces machining have the drawbacks of serious flattening distortion and poor tool-path continuity. Therefore, a novel method is developed to generate a spiral tool path for the machining of free-form surfaces by improving surface-flattening distortion and tool-path continuity. First, physical shell mapping is presented to flatten a free-form surface into a plane, which takes stretching energy, bending energy, and global energy into account. Then, the spatial spiral polyline is rounded to generate a spiral path by proposing reverse-compensation optimisation. Therefore, the free-form surfaces can be quickly flattened with less distortion, remaining free of overlap, and can in addition be machined at high speed along a C2 continuous spiral tool path. Further, the flattening error, tool-path length, mean curvature, mean scallop-height error of the spiral path, machining time and surface roughness are obviously reduced. Finally, simulation results are given to show the effectiveness and feasibility of the presented strategy.

Evaluating the reinforcement content and elastic properties of Mg-based composites using dual-mode ultrasonic velocities.

&NA; Based on the wave‐mode‐converted principle, an immersion‐focused transducer is employed to determine the longitudinal wave and shear wave velocities. The experimental condition is then investigated to obtain the converted shear wave, which is used to analyze the relationship between the reinforcement content and the dual‐mode ultrasonic velocities. In addition, the elastic modulus is calculated. Magnesium‐based composite samples with different reinforcement contents are manufactured to conduct an ultrasonic experiment, wherein the dual‐mode velocities vary with the change in the reinforcement content; the correlation coefficient is 99.17%. An ultrasonic dual‐mode velocity model is developed to analyze the distribution of the reinforcement content. By employing the measured values obtained from the destructive method, the largest errors in the reinforcement content and elastic modulus evaluated using the proposed method are found to be −5.76% and 5.85%, respectively. The shear wave velocity determined using a normal‐incidence shear‐wave transducer reveals the accuracy with which the errors are measured. This method provides an effective tool to nondestructively evaluate the microstructure and elastic properties of Mg‐based composites. HighlightsThe water path is studied for a focused transducer to obtain the converted waves.We develop a dual‐mode ultrasonic velocity model to visualize the SiC distribution.The model is used to obtain the images of elastic modulus of the Mg‐based composites.The destructive evaluation results reveal the effectiveness of the presented method.

Effect of diffraction on evaluation of grain size in curved component using ultrasonic attenuation method

The ultrasonic diffraction attenuation in curved component was derived by the acoustic elasticity function and pressure field simulated based on Multi-Gaussian beam theory. The diffraction correction was introduced into the frequency-domain attenuation spectrum experimentally measured to control the systematic error. An experimental study on stainless steel blocks with different grain sizes and surface curvatures has been performed. The result shows that, the variation coefficient of the attenuation value for the curved blocks reduced by 31.6% after correcting the diffraction coefficient.