A Study On Radial Steam Turbine Characteristics Modeling And Perfomance Optimization

Under the background of carbon neutrality,radial turbines became a hot topic in the innovation of energy technology due to its extensive application prospect.And the research on the related theory will provide strong support to realize carbon peaking and carbon neutrality goals.An accurate radial turbines performance prediction model and optimization strategy is of great significance to turbine design,manufacture and operating,which will directly affect the role that radial turbines can play in the future energy system.Therefore,a systematic study of loss models,off-design characters under throttling adjusting and blade optimization in a steam radial has been conducted by numerical and theory ways in this paper.Firstly,a numerical model of radial turbine with partial admission and blade tip was established using the reference parameters.And two mean-line models based on flow and enthalpy checking were established with thermodynamic methodology.With those models,the loss correlation as well as characters of partial admission and blade tip were comparative studied.The result shows those established models can accurately predict the radial turbine performance under the design condition.For partial admission loss,most correlations lack precision with a small admission section.And for the blade tip loss,Dambach’s model has the most accurate result,whose relative error is less than 1.46%.Then,with above established models,the performance prediction model under throttling adjusting off-design conditions was studied.In order to boost the precision of prediction model,the calibration parameter was proposed,which can be set by the particle swarm optimization(PSO)algorithm.Moreover,the change of volume flow was taken into consideration,and the correction theory between volume flow and calculation area was built up.This theory can reflect the complex 3D flow character in a 1D calculation model,which reduce the prediction error further.The final model keeps the prediction error under 1%,and can completely meet the modeling demand of a running turbine.At last,a single passage numerical model of radial turbine blade profile optimization was constructed.With the thickness and angle curve,the profile can be parameterize by the simplified uniform cubic spline curve.The profile-efficiency response surface surrogate model was established through the design experiments of control points.And the blade profiles in nozzle and rotor were optimized by sequential quadratic programming.Due to the complex flow characteristics,the design of the protrusion at the rotor hub should be avoided for efficiency improvement.The optimized blade profiles were applied to partial admission turbine for verification.Result shows,with the rotor optimization,the efficiency of turbine raised 1% for turbine load span.