Thermal/structural analysis of radiators for heavy-duty trucks

A thermal/structural coupling approach is applied to analyze thermal performance and predict the thermal stress of a radiator for heavy-duty transportation cooling systems. Bench test and field test data show that non-uniform temperature gradient and dynamic pressure loads may induce large thermal stress on the radiator. A finite element analysis (FEA) tool is used to predict the strains and displacement of radiator based on the solid wall temperature, wall-based fluid film heat transfer coefficient and pressure drop. These are obtained from a computational fluid dynamics (CFD) simulation. A 3D simulation of turbulent flow and coupled heat transfer between the working fluids poses a major difficulty because the range of length scales involved in heavy-duty radiators varies from few millimeters of the fin pitch and/or tube cross-section to several meters for the overall size of the radiator. It is very computational expensive, if not impossible, to directly simulate the turbulent heat transfer between fins and the thermal boundary layer in each tube. In order to overcome the computational difficulties, a dual porous zone (DPZ) method is applied, in which fins in the air side and turbulators in the water side are treated as porous region. The parameters involved in the DPZ method are tuned based on experimental data in prior. A distinguished advantage of the porous medium method is its effectiveness of modeling wide-range characteristic scale problems. A parametric study of the impact of flow rate on the heat transfer coefficient is presented. The FEA results predict the maximum value of stress/strain and target locations for possible structural failure and the results obtained are consistent with experimental observations. The results demonstrate that the coupling thermal/structural analysis is a powerful tool applied to heavy-duty cooling product design to improve the radiator thermal performance, durability and reliability under rigid working environment.