Deflection of the tool and workpiece caused by cutting forces usually leads to machining errors of thin wall workpieces. Monitoring this kind of force-induced errors plays an extremely important role in controlling and compensating the deflection-related machining failures in real time. However, accompanied by time consuming and complicated computations, now available analytical prediction methods cannot satisfy the requirements of online machining errors prediction. Hence, data-driven regression methods are introduced to online predict the machining errors. The challenges lie in that: (a) the regression model must have high generalization performance to adapt varied cutting conditions; (b) the spatial continuous distribution of machining errors needs to be constructed via limited measured points; (c) the model complexity should be restrained to improve the real-time performance. To tackling these challenges, a knowledge embedded regression is presented to model the relationship between machining error and cutting condition, cutting location, as well as online measured cutting forces. The physical mechanism about machining errors caused by the cutting forces and location-varied rigidity is integrated into the model for improving the generalization accuracy. A Gaussian prior distribution over the model weights is introduced to reduce the model redundancy for the feasibility of learning the weight variables with limited samples and increasing the prediction efficiency. Results have indicated that the predicted machining errors by the proposed model accords with the measurement better than those predicted by a purely data-dependent regression.

In five-axis milling process, the tool path generated by a commercial software seldom takes the dynamics of the machining process into account. The neglect of process dynamics may lead to milling chatter, which causes overcut, quick tool wear, etc., and thus damages workpiece surface and shortens tool life. This motivates us to consider dynamic constraints in the tool path generation. Tool orientation variations in five-axis ball-end milling influence chatter stability and surface location error (SLE) due to the varying tool-workpiece immersion area and cutting force, which inversely provides us a feasible and flexible way to suppress chatter and SLE. However, tool orientations adjustment for suppression of chatter and SLE may cause drastic changes of the tool orientations and affects surface quality. The challenge is to strike a balance between the smooth tool orientations and suppression of chatter and SLE. To overcome the challenge, this paper presents a minimax optimization approach for planning tool orientations. The optimization objective is to obtain smooth tool orientations, by minimizing the maximum variation of the rotational angles between adjacent cutter locations, with constraints of chatter-free and SLE threshold. A dedicated designed ball-end milling experiment is conducted to validate the proposed approach. The work provides new insight into the tool path generation for ball-end milling of sculpture surface; also it would be helpful to decision-making for process parameters optimization in practical complex parts milling operations at shop floor.

Surface topography plays a key role in the service performance of parts, and is often used as a metric for detecting machining states as well. Hence, it is of great importance to online predict the surface topography to avoid surface deterioration and machining faults. However, alternative solutions for this purpose remain to be developed. This paper presents a generative adversarial network to predict the statistical distribution of surface topography conditioned on cutting parameters, measured cutting forces, as well as vibrations. The local dynamic characteristics and global quasi-static trend of measured process quantities are extracted by short-time spectrum technique, then mapped into the local surface features patch by patch. To improve the network performance, we embed several useful tricks, such as recursive residual block, skip connection and global residual architecture into the surface generator so as to enhance its capacity of learning the complicated non-linear mapping against limited training data. The model performance is evaluated and validated using the structure similarity index, the distance of extracted features by the visual geometry group (VGG) network, as well as the perceptual index. The results indicate that the recursive architecture performs better than nonrecursive one on learning the non-linearity between surface topography and measured process quantities. Besides, by applying adversarial learning, the texture characteristics can be well-captured to avoid the blur of high frequency details.

This paper presents a new methodology to study the orthogonal cutting mechanism based on the displacement and temperature fields obtained by the particle image velocimetry technique and thermal infrared imager, respectively. With the displacement and temperature fields as inputs, the thermo-elasto-viscoplastic constitutive model is adopted to calculate the orthogonal cutting strain and stress fields. The formation of built-up edge is detected using the velocity fields. The geometries of the primary shear zone for two different cutting speeds are determined by the strain rate field. The shear stress distributions along the shear plane are obtained according to the calculated results