Except in the case of martensitic transformation during quenching and age-hardening, the mechanical properties (tensile strength and hardness) of many metallic materials are often determined by its chemical composition. If mechanical properties can be predicted from the chemical composition of molten metal before casting, it can contribute to the stabilization of quality and the reduction of the testing process of tensile strength and hardness. In the case of gray cast iron, mechanical properties are often discussed with five main elements (C, Si, Mn, P and S). Multiple regression shows low performance in terms of correlation coefficient. Therefore, trace elements other than the five main elements should be considered since the influence of trace elements on mechanical properties is mostly nonlinear, making it difficult to analyze by multiple regression. Given that deep neural network (DNN) can take nonlinear cases into consideration, we investigated whether mechanical properties can be predicted from chemical compositions including trace elements, and obtained the following findings. For comparison, we also analyzed mechanical properties by multilayer perceptron (MLP) and multiple regression (MR). As a result, the prediction accuracy of DNN, MLP and MR improved by the consideration of not only the five main elements but also 18 other elements including trace elements. Prediction error of tensile strength analyzed by DNN was less than half of MR. Increasing the number of layers and the number of nodes in DNN improved the prediction accuracy of mechanical properties, demonstrating the effectiveness of DNN.
Refinement of primary Si grains of Al-21%Si alloy was investigated by applying mechanical vibration during solidification. The more the frequency or half amplitude was increased, the more the size of primary Si grains decreased and the number of primary Si grains increased. Because each frequency and half amplitude affected grain refinement, it is thought that the primary Si grain size can be summarized by the excitation force equation which takes both parameters into account, and evaluated based on its relation with excitation force, where the primary Si grain size decreases with increasing excitation force. In order to clarify the mechanism of grain refinement by applying vibration, the cooling rate during solidification and the starting time of vibration were changed. The cooling rate did not affect the grain refinement in the range of this study. The primary Si grain became refined when vibration was applied during casting. On the other hand, it did not become refined when vibration was applied to the mold after casting. These results suggest that grain refinement is promoted by many primary Si grains crystallizing at the wall of the mold and upper surface of molten metal, and moving continuously into the molten metal by the convection of the molten metal due to the vibration.
The purpose of this study is to investigate a nondestructive method for predicting the fatigue limit of spheroidal graphite cast iron using high resolution X-ray CT. Axial load fatigue test specimens were cut out from a large spheroidal graphite cast iron equivalent to FCD 350, and graphite and defects in the material were detected using high resolution X-ray CT for all specimens. Fatigue limit was estimated from the graphite and defect sizes using the fatigue limit estimation formula based on the four-parameter method.
Axial load fatigue test was performed in accordance with JIS (Japanese Industrial Standards). Repetition frequency was 17Hz, stress ratio was R ＝ －1, and number of cycles during the test was 1.0 × 107. The specimen used was JIS type 1 of 8.00mm in diameter. Fracture origins were observed in all fatigue fracture surfaces using a scanning electron microscope (SEM) in order to compare the results between the defects observed by X-ray CT and the fracture origins observed in the fatigue test.
The fatigue limit estimated by the defect with the largest volume detected by X-ray CT was 5% lower than the experimental fatigue limit of 125MPa, which is considered safe estimation. However, in the fatigue test, the fracture origin was not necessarily the defect with the largest volume. Therefore, the fatigue limit was estimated by the average defect size when the cumulative distribution function of ten defects with the largest volume of each test piece was F ＝ 50%. The result was 11% larger than the experimental fatigue limit, which is considered a dangerous estimation. These results indicate that estimation of fatigue limit using a nondestructive method is feasible.
The multi-component white cast iron has been widely used as a rolling-mill roll material for hot rolling owing to its excellent wear resistance. However, when applied to roughing trains for bar and wire rod rolling at low speed under high thermal and mechanical loads, damage was impaired due to crack propagation. To address this problem, the control of microstructures such as finer microstructures and dispersion of the granular type MC carbide were carried out by designing the alloy composition, adding a small amount of Ti and rapid solidification. As a result, the improvement of fracture toughness and the suppression of crack propagation were achieved, and both abrasion resistance and high-temperature oxidation resistance were provided. This material was adopted as the shell material for a composite roll manufactured by the continuous pouring process for cladding (CPC) method to actual rolling operation, and the durability was found to be improved to more than three times that of conventional rolls.