基于CFD的液力变矩器等效参数性能预测模型

CFD Analysis based Torque Converter Performance Prediction Model with Equivalent Parameters

摘要:

从保证冲击损失计算精度的角度出发，推导了一维束流理论性能预测模型中各参数基于计算流体力学（CFD）分析的等效表达式.从CFD分析结果中提取各叶轮的损失功率，运用最小二乘法确定各损失项的损失系数，进而得到液力损失的构成情况，为进一步改进设计提供依据.运用得到的等效参数与各项损失系数修正一维束流理论性能预测模型，修正后的等效参数性能预测模型的泵轮转矩系数与CFD分析最大偏差减小至3.2%以内.改变液力变矩器的叶形参数,使失速时泵轮转矩系数提高6.9%，最高效率降低2.5%.分别使用等效参数性能预测模型与CFD分析重新计算，等效参数性能预测模型的泵轮转矩系数与CFD计算结果偏差在7.3%以内，仍然保持较好的一致性.

Abstract:

The correlation between parameters of one dimensional flow model and computational fluid dynamics (CFD) results is derived from guaranteeing precision of estimation of incidence loss. Energy loss within each impeller passages are extracted from CFD results, and the exact energy loss coefficients are achieved by the least mean square method which minimizes the deviation of the calculated energy loss from CFD results. Then, the component of energy loss is analyzed, which provides reference for further modification of design. The one dimensional flow model is modified by taking account of the equivalent parameters and energy loss coefficients, and the deviation of predictive pump torque capacity from CFD results is reduced within 3.2%. Furthermore, this deviation still keeps less than 7.3% after modifying the shape parameters of blades which leads to 6.9% increment of torque capacity at rotating stall and 2.5% decrement of peak efficiency.