Search

This paper develops a novel interpolation algorithm to compensate the error caused by elastic deformation which can achieve higher tracking accuracy even without the signal from the linear scale on the table. First, a dynamic model which considers flexural modes of a feed drive system, servo control loops, and friction effects simultaneously is constructed. Based on the dynamic model, the equation of elastic deformation is then derived. The equation is further simplified to form a formulation in computing the deformations which is mainly caused by inertial and viscous forces during motion. Experiments and simulations are conducted to validate both dynamic model and the simplified elastic deformation equation. Based on the simplified equation, a novel elastic deformation compensation interpolation (MEDCI) algorithm is proposed to generate the modified position commands such that the tracking errors caused by elastic deformation can be reduced. Experimental results on the linear trajectory demonstrate that the proposed algorithm could reduce the maximum elastic deformation error from 23.4 μm to 6.4 μm and the root mean square (RMS) error from 19.5 μm to 3.1 μm which correspond to 84.1% and 72.6% reduction, respectively. Furthermore, the MEDCI can achieve more than 10% reduction as compared to the results in published literature. [ABSTRACT FROM AUTHOR]

In this paper, an integrated dynamic acceleration/deceleration (IDAD) interpolation algorithm which combines the acceleration/deceleration before interpolation (ADBI) and acceleration/deceleration after interpolation (ADAI) is proposed. The IDAD scheme could confine the maximum contour error at the junction of a linear line (G01) block under consideration of the command and servo dynamic errors. The contour error formulation for the IDAD is derived analytically and then utilized for interpolation design. Because the error formulation could predict the command and dynamic errors analytically, the IDAD could adjust the connecting velocity of two blocks such that the overall contour errors are bounded within the given tolerance. The simulation results validate the proposed algorithm that can reduce the contour errors by 50 % but only an increase of 10 % machining time as compared to the integrated acceleration/deceleration (IAD) interpolation technique. [ABSTRACT FROM AUTHOR]

Feedrate scheduling with kinematic constraints is a challenging problem for five-axis machining. Linear and rotary drives of five-axis machine tools have different capabilities and limits. The main issue is that the feedrate profile generated by the existing five-axis feedrate scheduling method might violate kinematic constraints since the tangential and axial jerk could not be directly correlated due to nonlinear kinematics transformation. In this paper, a novel master-based feedrate scheduling (MBFS) approach with real-time and drive constraints is developed to generate smooth trajectories along five-axis linear toolpaths. Based on inverse kinematic equations, tooltip feed parameters are evaluated by a five-axis feed regulation formula (FFRF) such that axial velocities, accelerations, and jerks could be constrained within given limits simultaneously. The tooltip or rotary axis having the largest motion duration is selected to be the master, and the others are regarded as the slaves which move with the master synchronously. Rather than planning the feedrate profile using the curve parameter algorithm, linear and angular trajectories are planned by matching the same motion duration to realize time synchronization of five-axis motion. The local corner smoothing method is adopted to blend the velocities of the motion axes and generate cornering trajectories while satisfying corner position tolerance and drive constraints. Finally, experiments are conducted on a five-axis engraving machine to verify that the proposed method can satisfy the kinematic constraints, and improve tracking and contour accuracy. [ABSTRACT FROM AUTHOR]

Modern motion control adopts acceleration/deceleration before interpolation (ADBI) motion planning to eliminate path command errors. However, the individual velocity profiles might not be continuous at the junction of the blocks. Acceleration/deceleration after interpolation (ADAI) method may provide an alternative for solving the discontinuous problems, but it causes path command errors. In this paper, an integrated acceleration/deceleration interpolation (IAD) scheme which integrates the ADBI and ADAI modules is proposed. The ADBI provides a look-ahead function which plans the feedrate profiles based on chord errors, command errors, curvatures, and acceleration limits. Within the look-ahead function, the command error equation is utilized to determine the feedrate at the junction of adjacent blocks. Then the ADBI performs non-uniform rational B-spline (NURBS) interpolation using the planned feedrate profile and outputs the position points to the ADAI module. The ADAI module processes the points by a digital convolution technique such that the continuity of the block junction velocity is ensured. Finally, the IAD is applied to the multi-block NURBS interpolation to validate its effectiveness. Simulations and experiments are conducted to demonstrate the IAD scheme. It is shown that the IAD scheme can reduce the acceleration significantly at the junctions of the blocks under the given tolerance of the command error. Furthermore, the proposed algorithm can improve tracking and contour accuracies as compared to the hybrid multi-blocks look-ahead approach. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, a dynamics-based interpolator with real-time look-ahead (DBLA) algorithm is proposed to generate a smooth and jerk-limited acceleration/deceleration (ACC/DEC) feedrate profile. The interpolator consists of three modules: geometric, dynamics-based, and jerk-limited modules. The geometric module can detect the local maximum/minimum (max/min) curvatures, and divide a NURBS curve into small segments according to the information of sharp corners. The feedrates at the sharp corners are determined based on confined chord errors and curvatures of the curve. The dynamics-based module utilizes a dynamics feedrate modification equation (DFME) to estimate contour errors at the sharp corners and adjusts the feedrates at the locations of the sharp corners. The jerk-limited module plans the feedrate profile of the curve according to the segments’ length and the given jerk limit. Simulations are performed to verify real-time performance of the look-ahead algorithm. Experiments using a PC-based motion controller and an X–Y table are conducted to demonstrate that high-accuracy can be achieved with the proposed dynamics-based interpolator as compared to the adaptive-feedrate and the curvature-based feedrate interpolation algorithms. [Copyright &y& Elsevier]

This paper presents an integrated servo-feed-drive model including the cutting force and structural effect to predict tracking errors in an end-milling process. Most conventional approaches consider the cutting force to be an equivalent torque to servo feed drive. However, in addition to acting as the equivalent torque to the servo feed drive, cutting forces also cause the machine table to vibrate. This paper considers the aforementioned cutting-force effects to predict tracking errors and then verifies the tracking errors using experimental results. Experiments are conducted on a 3-axis computer numerical control (CNC) machining center to validate the tracking errors predicted by the proposed servo-feed-drive model. For one case study, the peak-to-peak tracking errors from the experimental, proposed, and traditional models are 3 μm, 2.8 μm, and 0.5 μm, respectively, for the x-axis, and 2.1 μm, 1.7 μm, and 0.4 μm, respectively, for the y-axis. The experimental results illustrate that the tracking errors predicted using the proposed model are more accurate than those predicted using the traditional model without consideration of the transmission path. Therefore, it can be concluded that the proposed integrated model provides much accurate tracking-error prediction, and thus, the ball-screw and machine-table flexibilities should be considered. [ABSTRACT FROM AUTHOR]

vA generalized model that integrates the Navier- Stokes equation and coupling-of-modes (COM) model for biosensing SAW devices is developed in this paper. The SAW device is separated into three regions: interdigital transducer (IDT), substrate (delay line), and sensing regions. To evaluate the effects of metal thickness, mass loading caused by bioreaction, and different viscous fluid loading, the sensing region is further divided into three layers: piezoelectric substrate, metal layer, and fluid layer. In contrast to the conventional study, which is focused on the change of phase velocity, this model can evaluate the insertion loss and phase shifts under different sensing conditions. It can be shown that the integration of the COM model can provide guidelines for designing the bio-sensing device such as choosing the proper number of IDT, the width of the overlap, and the thickness of the metal layer. Furthermore, the generalized model can be utilized to evaluate the optimal thickness of the metal layer to achieve the maximum sensitivity. [ABSTRACT FROM PUBLISHER]

Flank wear is the most common wear that happens in the end milling process. However, the process of detecting the flank wear is cumbersome. To achieve comprehensively automatic detecting the flank wear area of the spiral end milling cutter, this study proposed a novel flank wear detection method of combining the template matching and deep learning techniques to expand the curved surface images into panorama images, which is more available to detect the flank wear areas without choosing a specific position of cutting tool image. You Only Look Once v4 model was employed to automatically detect the range of cutting tips. Then, popular segmentation models, namely, U-Net, Segnet and Autoencoder were used to extract the areas of the tool flank wear. To evaluate the segmenting performance among these models, U-Net model obtained the best maximum dice coefficient score with 0.93. Moreover, the predicting wear areas of the U-Net model is presented in the trend figure, which can determine the times of the tool change depend on the curve of the tool wear. Overall, the experiments have shown that the proposed methods can effectively extract the tool wear regions of the spiral cutting tool. With the developed system, users can obtain detailed information about the cutting tool before being worn severely to change the cutting tools in advance. [ABSTRACT FROM AUTHOR]

Flank wear is the most common wear that happens in the end milling process. However, the process of detecting the flank wear is cumbersome. To achieve comprehensively automatic detecting the flank wear area of the spiral end milling cutter, this study proposed a novel flank wear detection method of combining the template matching and deep learning techniques to expand the curved surface images into panorama images, which is more available to detect the flank wear areas without choosing a specific position of cutting tool image. You Only Look Once v4 model was employed to automatically detect the range of cutting tips. Then, popular segmentation models, namely, U-Net, Segnet and Autoencoder were used to extract the areas of the tool flank wear. To evaluate the segmenting performance among these models, U-Net model obtained the best maximum dice coefficient score with 0.93. Moreover, the predicting wear areas of the U-Net model is presented in the trend figure, which can determine the times of the tool change depend on the curve of the tool wear. Overall, the experiments have shown that the proposed methods can effectively extract the tool wear regions of the spiral cutting tool. With the developed system, users can obtain detailed information about the cutting tool before being worn severely to change the cutting tools in advance. [ABSTRACT FROM AUTHOR]

Flank wear is the most common wear that happens in the end milling process. However, the process of detecting the flank wear is cumbersome. To achieve comprehensively automatic detecting the flank wear area of the spiral end milling cutter, this study proposed a novel flank wear detection method of combining the template matching and deep learning techniques to expand the curved surface images into panorama images, which is more available to detect the flank wear areas without choosing a specific position of cutting tool image. You Only Look Once v4 model was employed to automatically detect the range of cutting tips. Then, popular segmentation models, namely, U-Net, Segnet and Autoencoder were used to extract the areas of the tool flank wear. To evaluate the segmenting performance among these models, U-Net model obtained the best maximum dice coefficient score with 0.93. Moreover, the predicting wear areas of the U-Net model is presented in the trend figure, which can determine the times of the tool change depend on the curve of the tool wear. Overall, the experiments have shown that the proposed methods can effectively extract the tool wear regions of the spiral cutting tool. With the developed system, users can obtain detailed information about the cutting tool before being worn severely to change the cutting tools in advance. [ABSTRACT FROM AUTHOR]

In the accepted paper, the authors' affiliation has been wrongly typeset as People's Republic of China, but this should read Taiwan. The correct affiliations are shown below. [ABSTRACT FROM AUTHOR]

Correction to: I Scientific Reports i https://doi.org/10.1038/s41598-021-97610-y, published online 7 October 2021 Hong-Tsu Young was omitted from the author list in the original version of this Article. The original article can be found online at https://doi.org/10.1038/s41598-021-97610-y. [Extracted from the article]