Pneumatic polishing tools are commonly used in traditional robot mold polishing systems, but they have problems with the stable control of mold surface roughness due to low precision and poor adaptability in polishing force adjustment. The integration of an adaptive hydraulic polishing (AHP) tool and robot system effectively solves the above problems, providing a robust solution for the high-precision polishing of various molds. This study systematically investigates the robotic polishing of NAK80 mold steel using an AHP-equipped robotic platform with 3M abrasive discs of progressively refined grit sizes (P180, P400, P800). Through single-factor experiments and response surface methodology, the effects of polishing force, rotational speed, and feeding speed on surface roughness were quantitatively analyzed. The relationship between surface roughness and the polishing parameters was derived to elucidate the roughness evolution before and after over-polishing. Orthogonal experiments combined with range analysis identified optimal parameter combinations for P180 (20 N polishing force, 5000 RPM rotational speed, and 5 mm·s⁻¹ feeding speed) and P400 abrasives (10 N polishing force, 4000 RPM rotational speed, and 5 mm·s⁻¹ feeding speed), achieving minimum surface roughness values of 0.08 µm and 0.044 µm, respectively. For P800 abrasives, a central composite design was used to develop a roughness prediction model with a ≤7.14% relative error, and the optimal parameters are a 20 N polishing force, a 5000 RPM rotational speed, and a 5 mm·s⁻¹ feeding speed. The sequential application of the optimized parameters across all the grit sizes can reduce the surface roughness from an initial 0.4 µm to a final 0.017 µm, representing a 95.75% improvement in the surface finish.

Shrinkage greatly influences the mechanical and fatigue properties of compacted graphite iron and it is necessary in order to study the causes of shrinkage in compacted graphite iron and to predict it effectively. In this paper, a kind of cylindrical necking test sample was designed to evaluate the shrinkage in compacted graphite iron, and a method to calculate the size of shrinkage was proposed. By observing the microstructure around the shrinkage zone, it is concluded that concentrated shrinkage mainly appears in the solidification region where the dendritic gap is closed, and the isolated shrinkage mainly occurs in the final solidification region, and the supersaturated carbon elements are gathered on the surface of the shrinkage. The cause of shrinkage in compacted graphite iron is caused by its solidification method, where the austenite dendrites and the eutectic clusters are generated close to the melt zone during the solidification process, leading to the inability to feed the shrinkage. Based on the thermodynamic analysis, the equations between the volume change of each phase, solid phase rate, and time during solidification of compacted graphite iron were established to theoretically explain the formation mechanism of the shrinkage. Taking nine parameters such as the chemical elements and characteristic values of thermal analysis as the input nods, a four-layer BP neural network model for predicting the size of shrinkage in compacted graphite iron was constructed, and the R-squared of the model reached 97%, which indicates it could be used to predict the shrinkage tendency.

Three-dimensional printing (3DP) is considered to be one of the important technologies for a new manufacturing mode. When ceramsite sand is used as a 3DP material to produce a mold (core), the printed layer is prone to deviation from the original location. In this study, the continuous stacking of the printed part deviation was termed as pushing dislocation, and a physical model was designed to investigate the pushing dislocation mechanism. When the gravity of the printing layer and the pressure of the sand scraper decreased, or when the supporting force increased, the angle of the sand scraper and the maximum friction of the prelaying layer on the printed part will reduce the pushing dislocation. To optimize the quality of the ceramsite sand mold, experiments on the pushing dislocation were conducted by altering the recoater speed, layer thickness, and bottom support condition (with or without bottom supporting plate). The sample dimensions were obtained by a 3D imaging scanner, and the gas evolution and ignition loss were measured. The results revealed that the dimensional difference of samples continuously decreased and the pushing dislocation was gradually reduced as the recoater speed and layer thickness increased. The pushing dislocation of the X-direction sample was more severe compared with that of the Y-direction sample. Increasing the layer thickness is an effective way of reducing the pushing dislocation. The bottom supporting plate can reduce the pushing dislocation, but the effect was insignificant.

Robotic grinding serves as a pivotal embodiment and key technological support of Industry 4.0. Elucidating the influence of robotic grinding parameters on the material removal depth (MRD) of 42CrMo steel and optimizing these parameters are critical to enhancing grinding efficiency and quality. In this study, the influences of revolution speed, feed speed, grinding force, and grit designation on MRD and surface Vickers hardness of 42CrMo steel were investigated by using an adaptive electro-hydraulic-actuated triangular abrasive belt in robot grinding. A predictive model for MRD of 42CrMo steel has been established using the orthogonal central composite design method. The results indicated that as the revolution speed or grinding increases, both MRD and surface hardness increase. However, as the revolution speed surpasses 4000 RPM or the grinding force exceeds 60 N, the increase of MRD becomes slower due to the increase in surface hardness. Both the MRD and surface hardness decrease continuously as the feed speed increases, and once it exceeds 15 mm·s⁻¹, the decrease of the MRD becomes slow. The rise in grit designation of the abrasive belt makes the MRD reduce gradually while the surface hardness rises slightly. The correlation coefficient of the predictive model is 0.9387, and the relative error between the predicted and experimental MRD is within 10%, indicating a relatively high accuracy. At the optimal grinding parameters (grinding force of 81 N, revolution speed of 4739 RPM, and feed speed of 7.6 mm·s⁻¹), the maximum MRD of 42CrMo steel achieved by an abrasive belt of 60 grit designation is 0.934 mm. This work provides a basis for high-precision robot abrasive belt grinding of 42CrMo steel.