Establishment and control method of digital twin model for refractory alloy smelting and sintering process

Currently, the smelting and sintering process of refractory alloys has become an important method for controlling alloy quality. By controlling the sintering process, the smelting and sintering quality of alloys can be controlled, and the performance of alloys can be improved. However, traditional methods often use a single approach, which results in low control effectiveness of the smelting process. Therefore, this study proposes a digital twin model based on an adaptive fuzzy algorithm to improve the smelting and sintering quality of refractory alloys. The new model uses an adaptive fuzzy algorithm to control and analyze sintering parameters, to improve the control effect and accuracy of refractory alloy smelting and sintering. The results showed that the sintering accuracy of the adaptive fuzzy algorithm could reach up to 95.2%, which was 12.0% higher than that of the Bayesian neural network model. In the test with a tungsten content of 30%, the damage degree of the adaptive fuzzy algorithm was the lowest at only 7.1%, which was 4.5% lower than that of the Bayesian neural network model. The deviation between the hardness change curve using the digital twin model and the actual change curve was relatively small, with a decrease of 4.9 Hv/MPa compared to the deviation without using the model. At a pressure of 2500 MPa, the highest degree of damage using the digital twin model was 5.7%, while the degree of damage without the model was 7.2%. Therefore, the use of new methods can significantly improve the smelting and sintering quality of alloys, achieving quality control of alloys. This has good guiding significance for the quality control of smelting and sintering processes for refractory alloys.