Problems encountered in the stress analysis of large-diameter high-temperature heat exchangers

Thread Content

For a project I am currently working on, there is a large-diameter high-temperature heat exchanger. The design parameters are as follows: Heat exchanger type: BEM; Diameter: Φ1600 mm. Design pressure and temperature for the shell side: 0.35 MPa/FV and 505°C; Design pressure and temperature for the shell side: 0.35 MPa and 580°C. Average metal wall temperature on the shell side: 328°C; average metal wall temperature on the tube side: 404°C. Specifications of the heat exchange tubes: Φ25×2 mm; length of each heat exchange tube: 6000 m; number of heat exchange tubes: 3514. The material used for the main structure is S30409. During routine design using SW6, the following issues were identified: 1. S30409 does not have a high-temperature yield strength at 580°C ； 2. The dimensionless width k of the area around the tube sheet where no tubes are present is 1.59332, which is greater than 1; moreover, the ratio Dt/Di = ρt = 0.766122, and this value is above 0.8. These values fall outside the range applicable to the calculation model specified in GB151, so it is not possible to perform calculations using the methods outlined in GB151. Finally, ANSYS was used to carry out calculations for the tube sheet. Due to the large temperature difference between the tube side and the shell side, a thermomechanical coupling analysis was necessary: mesh generation, mesh quality assessment, temperature field analysis, and thermomechanical coupling analysis. The thermomechanical coupling analysis was performed using a 1/4 scale model, and the resulting stress value was 2864 MPa; obviously, this stress value is inaccurate. However, I checked the settings of the material parameters, as well as the boundary conditions and applied loads, and found no issues. Additionally, the temperature field calculated using this method was exactly the same as that obtained through ANSYS’ conventional calculation methods, which means there is nothing wrong with the temperature field calculation. Since there are no problems with the boundary conditions and applied loads in the thermomechanical coupling analysis, it’s unclear why the results obtained are unreasonable. This is likely due to the fact that the finite element model is too large, with a high number of elements – around 1,900,000 – and further expansion of the model makes computation impossible. The highest stress values were observed at the fillets where the tube sheet connects to the shell, as well as on those rows of heat exchange tubes to which symmetric constraints were applied. The stress levels on the other heat exchange tubes, which did not have symmetric constraints applied, were much lower. Despite numerous attempts to fix the issue, the cause of the calculation errors could not be identified. My guesses regarding the cause are as follows: 1. The use of Solid185 elements led to shear self-locking, resulting in high stress levels. I considered using Solid186 elements for further calculations, but due to the extremely large size of the model, even switching to Solid186 elements made computation impossible on the computer ； 2. Stress singularities have appeared at the locations where constraints are applied; I really can’t figure out where exactly the problem lies. So I’m turning to the experts on Daohai Chuan for help. Thank you~