Study On Safety Of Key Components For Large Nuclear Steam Turbine

As a low-carbon and environmental protection way of power generation,nuclear power generation is getting more and more attention.In China,nuclear power capacity is increasing year by year.Compared with thermal power units,half-speed nuclear power steam turbine units have the characteristics of long final stage blades,large size,low stiffness and easy deformation of casing and rotor.Obviously,the safety of key components such as blade,casing and rotor of nuclear steam turbine in off-design working conditions directly affects the efficient and safe operation of the unit.Therefore,in this dissertation,the aeroelastic stability of long blade,the distribution and optimization of casing support load,and the variation regularity of rotor-gland seal clearance under off-design conditions are studied theoretically,and which are verified with the measured data in-site.Specific work includes:Aiming at the aeroelastic stability of the long last stage blade(LSB),taking a LSB with 1828 mm length of nuclear steam turbine as the research object,the steady-state flow field of the blade was numerically calculated at first,and the accuracy of the numerical simulation method was verified by the experimental measurement of the static pressure on blade surface.Based on this,the flutter critical condition of the blade was predicted by using harmonic balance method.The effect of inter-blade phase angles on the aerodynamic work distribution on blade surface was studied,and the aeroelastic instability area on blade surface under flutter condition was determined.Then,the influence of interference of upstream static blades on aerodynamic stability of the blade was analyzed,and the aerodynamic damping was calculated.Thirdly,the aeroelastic stability of the blade was studied based on the aerodynamic reduced order model of Volterra series,and the applied technology of this reduced order method on the research of the aeroelastic characteristics of ultra-long steam turbine blades was explored.The research result shows that,for the studied blade,the inter-blade phase angle has a great influence on the aerodynamic damping of the blade.When it is in the forward traveling wave mode,the aeroelastic stability is better.The critical condition for the blade flutter is between 22% and 32% volume flow.At 22% volume flow,the area above 80% of the blade height section is the aeroelastic instability zone.The influence of inter-blade phase angle on the aerodynamic work distribution is mainly located in the suction surface near the tip.Under certain conditions,increasing the spacing between the moving and stationary blades moderately can improve its aeroelastic stability.The constructed aerodynamic reduced-order model can effectively predict the aeroelastic stability of the blade.Concerning about casing support load distribution,the three-dimensional model of the low-pressure(LP)inner casing of a certain 1000 MW nuclear steam turbine was firstly established,the deformation of the LP inner casing and the load distribution at support points were studied under cold state and different working conditions based on the thermal-structural coupling analysis method.Then,on account of response surface analysis method,the support point load distribution was predicted,the influence of the support point height on the casing load distribution was analyzed,and the sensitivity of the support point load distribution to its height was studied.Thirdly,the multi-objective genetic algorithm was used to optimize the casing support point load distribution at full power to make it uniform.The simulation result shows that,when the casing support points are at the same height,the load distribution is uneven: the load on the outer support point is the largest,on the support point in the middle is the second,and on the supports between the above two is the smallest.Under cold state,the difference between maximum and minimum load is 6.4 times.Under different working conditions,the load distribution is different,and it is quite different with the cold state.With load increasing of unit,the load distribution tends to be average.At full load,the difference of load decreases to 3.7 times.The method named response surface analysis can accurately predict the support load distribution,and the prediction error is less than 4%.The multiobjective genetic optimization algorithm is adopted to adjust the level of support point,which can make the difference of load distribution at generator end support less than5%.Concerning about rotor-seal clearance prediction,a three-dimensional model of LP inner casing,diaphragm and rotor of a 1000 MW nuclear steam turbine was established.Firstly,the deformation of casing,diaphragm and rotor in cold installation process was analyzed by using finite element method.The stiffness characteristics of the upper and lower half casing under different conditions were studied,and the change regularity of the rotor-seal clearance was revealed.Secondly,the influence of different pin arrangements of center split face on casing deformation,tip clearance and diaphragm clearance was studied.Thirdly,the calculation method of blade tip clearance and diaphragm clearance under whole casing assembling was put forward,and the software of dynamic and static clearance adjustment was developed.The research result shows that the rotor-seal bottom clearance first increases and then decreases during installation.The rotor-seal bottom clearance increases under casing fully assembled state compared with the half casing state,and the increase value is around 0.45 mm.Among them,the third stage of tip clearance changes mostly,and the third stage of diaphragm clearance changes the least.The pin arrangement of center split face has great influence on the deformation of casing and diaphragm.The locating pin at exhaust guide ring of LP has the greatest influence on the deformation of casing and diaphragm.On the one hand,the pin arrangement of center split face changes the stiffness characteristics of the upper and lower casings,on the other hand,it can transfer load between the upper and lower casings.The developed dynamic and static clearance adjustment software can accurately calculate the rotor-seal clearance of nuclear steam turbines under casing fully assembled state,which meets the engineering requirement.Concerning about LP rotor-gland seal rubbing fault,this dissertation firstly studied the rotor-gland seal clearance during startup.The influencing factors of clearance change were analyzed,and the most dangerous condition of rubbing fault during startup condition was found out.Then,based on the sensitivity calculation of clearance change under this condition,the rubbing reason was analyzed.The rotor-gland seal clearance was predicted by neural network,and the formula for predicting clearance was put forward.Thirdly,the rotor-gland seal rubbing control method was proposed from multiple dimensions stand point.The study shows that the vacuuming phase is the most dangerous time for rotor-gland seal rubbing fault during the startup process.Under this condition,the gland seal is significantly lifted,resulting in a significant reduction in the clearance at the bottom of rotor.The increase of gland seal temperature is the main factor causing the reduction of its bottom clearance.The artificial neural network algorithm method can effectively predict the clearance between the rotor and gland seal during vacuuming,and the prediction error is less than 3.5%.When the initial bottom clearance is 0.54 mm and the gland seal temperature reaches 112℃,the rubbing may happen.The method proposed from multiple dimensions stand point for rotor-gland seal rubbing,can effectively reduce the risk of rubbing.The above study on safety of key components for nuclear steam turbine is of great significance in theory and engineering application.