Search results

The purpose of this paper is to propose a quasi-three-dimensional (3D) thermohydrodynamic (THD) model for oil film bearings with non-Newtonian and temperature-viscosity effects. Its performance factors, including precision and time consumption, are investigated.

Two-dimensional (2D), 3D and quasi-3D numerical models are built. The thermal and mechanical behaviors of two types of oil film bearings are simulated. All the results are compared with solutions of commercial ANSYS CFX.

The 2D THD model fails to predict the temperature and pressure field. The results of the quasi-3D THD model coincide well with those of the 3D THD model and CFX at any condition. Compared with the 3D THD model, the quasi-3D THD model can greatly reduce the CPU time consumption, especially at a high rotational speed.

This quasi-3D THD model is proposed in this paper for the first time. Transient mechanical and thermal analyses of high-speed rotor-bearing system are widely conducted using the traditional 3D THD model; however, the process is very time-consuming. The quasi-3D THD model can be an excellent alternative with high precision and fast simulation speed.

The simulated temperature profile of the printed circuit board assembly (PCBA) during reflow soldering process deviates from the actual profile. To reduce this relative deviation, a new strategy based on the Kriging response surface and the Multi-Objective Genetic Algorithm (MOGA) optimizing method is proposed.

The simulated temperature profile of the PCBA during reflow soldering process deviates from the actual profile. To reduce this relative deviation, a new strategy based on the Kriging response surface and the MOGA optimizing method is proposed.

Several critical influencing parameters such as temperature and the convective heat transfer coefficient of the specific temperature zones are selected as the correction parameters. The hyper Latins sampling method is implemented to distribute the design points, and the Kriging response surface model of the soldering process is constructed. The updated model is achieved and validated by the test. The relative derivation is reduced from the initial value of 43.4%–11.8% in terms of the time above the liquidus line.

A new strategy based on the Kriging response surface and the MOGA optimizing method is proposed.