It is necessary to use the integrated stainless steel pipe having two fitting bodies without welds while train travelling at high speed. In order to form this type of integrated stainless steel pipe, the method of preforming combined finish forming process is developed. The preforming process is characterized by flaring combined upsetting for left fitting body which is like a flange, and is characterized by tube axial compressive process under die constraint for right fitting body which is like a double-wall pipe. The finite element simulations of the processes are carried out by software package DEFORM, and the results indicate that: 1) left or right fitting body can be formed by a two-step forming process without folding and under-filling defects; 2) by using two-step forming, strain and stress in left fitting body are larger than those in right fitting body, and deformation in right fitting body is more homogenous than the deformation in left fitting body; 3) two or more preforming steps may be needed for left fitting body considering the distributions of strain and stress.
A remote open-path laser-induced breakdown spectroscopy(LIBS) system was designed and studied in the present work for the purpose of combining the LIBS technique with the steel production line. In this system, the relatively simple configuration and optics were employed to measure the steel samples at a remote distance and a hot sample temperature. The system has obtained a robustness for the deviation of the sample position because of the open-path and alloptical structure. The measurement was carried out at different sample temperatures by placing the samples in a muffle furnace with a window in the front door. The results show that the intensity of the spectral lines increased as the sample temperature increased. The influence of the sample temperature on the quantitative analysis of manganese in the steel samples was investigated by measuring ten standard steel samples at different temperatures. Three samples were selected as the test sample for the simulation measurement. The results show that, at the sample temperature of 500 ℃, the average relative error of prediction is 3.1% and the average relative standard deviation is 7.7%, respectively.
