Parameter Modification Of SST K-ω Turbulence Model Based On Bluff Body Wind Pressure Data

Compared with wind tunnel test and field measurement,CFD(Computational Fluid Dynamics)numerical simulation has the advantages of time-saving and economical in the calculation of wind load of building structure,and it is widely used in practical engineering.RANS(Reynolds-Averaged Navier-Stokes)simulation with low computational cost and high efficiency is currently the most commonly used simulation method.However,compared with LES(Large Eddy Simulation),its calculation accuracy is poor.To solve this problem,a parameter optimization method based on wind tunnel data-driven is proposed,which is applied to the parameter optimization problem of SST(Shear-Stress Transport)k-ω model,and the effect of parameter optimization is evaluated.The basic performance of SST turbulence model in bluff body wind pressure simulation is analyzed.Numerical simulation results show that the SST model has good simulation accuracy in the positive wind pressure area,but there is a large error in the negative pressure area,that is,the model is insufficient in simulating fluid separation and backflow phenomena.Sensitivity analysis of model parameters was conducted through Morris screening,and the results showed that there are four parameters that have a greater impact on the calculation results:β2,β*,a1,β1,and are called key parameters.A data-driven model parameter optimization method is established.The method is based on the Bayesian framework and combines surrogate models and principal component analysis techniques.Using principal component analysis techniques,the data features of 400 points can be compressed to 3,and the amount of information lost in the compression process is less than 5%,which significantly improves the calculation efficiency.The results of the surrogate model error analysis show that the surrogate model established by the Gaussian process regression method has good prediction accuracy and robustness,with an average error of about 0.01 and an average relative error of less than 3%.The key parameters in the SST turbulence model were optimized,the goal of reducing the simulation error in the bluff body wind pressure simulation was achieved,and the applicability of the optimized parameters was verified.For the common high-rise and low-rise building models,the model parameters are optimized under different wind angles.The results show that there is a certain difference between the optimized values of the parameters and the standard values.In the high-rise building model wind pressure simulation,the simulation results after parameter optimization show that the simulation error of the wind pressure coefficient of the building model under different wind direction angles is reduced to varying degrees.The optimization effect is the most significant at 0° and 40°and the optimization effect is poor at 20°.In the wind pressure simulation of lowrise buildings,the simulation error of the SST model after parameter optimization is reduced.Finally,the calculation results of the SST model after parameter optimization are compared with the LES results,and it is found that the optimized SST turbulence model can reduce the simulation error of the bluff body wind pressure,but the error is still greater than the large eddy simulation.