Based on the ensemble mean outputs of the ensemble forecasts from the ECMWF （European Centre for Medium-Range Weather Forecasts）, JMA （Japan Meteorological Agency）, NCEP （National Centers for Environmental Predic...Based on the ensemble mean outputs of the ensemble forecasts from the ECMWF （European Centre for Medium-Range Weather Forecasts）, JMA （Japan Meteorological Agency）, NCEP （National Centers for Environmental Prediction）, and UKMO （United Kingdom Met Office） in THORPEX （The Observing System Research and Predictability Experiment） Interactive Grand Global Ensemble （TIGGE） datasets, for the Northern Hemisphere （10~ 87.5~N, 0~ 360~） from i June 2007 to 31 August 2007, this study carried out multimodel ensemble forecasts of surface temperature and 500-hPa geopotential height, temperature and winds up to 168 h by using the bias-removed ensemble mean （BREM）, the multiple linear regression based superensemble （LRSUP）, and the neural network based superensemble （NNSUP） techniques for the forecast period from 8 to 31 August 2007. A running training period is used for BREM and LRSUP ensemble forecast techniques. It is found that BREM and LRSUP, at each grid point, have different optimal lengths of the training period. In general, the optimal training period for BREM is less than 30 days in most areas, while for LRSUP it is about 45 days.展开更多
Because of the randomness of many impact factors influencing the dynamic assembly relationship of complex machinery, the reliability analysis of dynamic assembly relationship needs to be accomplished considering the r...Because of the randomness of many impact factors influencing the dynamic assembly relationship of complex machinery, the reliability analysis of dynamic assembly relationship needs to be accomplished considering the randomness from a probabilistic perspective. To improve the accuracy and efficiency of dynamic assembly relationship reliability analysis, the mechanical dynamic assembly reliability（MDAR） theory and a distributed collaborative response surface method（DCRSM） are proposed. The mathematic model of DCRSM is established based on the quadratic response surface function, and verified by the assembly relationship reliability analysis of aeroengine high pressure turbine（HPT） blade-tip radial running clearance（BTRRC）. Through the comparison of the DCRSM, traditional response surface method（RSM） and Monte Carlo Method（MCM）, the results show that the DCRSM is not able to accomplish the computational task which is impossible for the other methods when the number of simulation is more than 100 000 times, but also the computational precision for the DCRSM is basically consistent with the MCM and improved by 0.40-4.63% to the RSM, furthermore, the computational efficiency of DCRSM is up to about 188 times of the MCM and 55 times of the RSM under 10000 times simulations. The DCRSM is demonstrated to be a feasible and effective approach for markedly improving the computational efficiency and accuracy of MDAR analysis. Thus, the proposed research provides the promising theory and method for the MDAR design and optimization, and opens a novel research direction of probabilistic analysis for developing the high-performance and high-reliability of aeroengine.展开更多
Studying and understanding of the surface topography variation are the basis for analyzing tribological problems,and characterization of worn surface is necessary.Fractal geometry offers a more accurate description fo...Studying and understanding of the surface topography variation are the basis for analyzing tribological problems,and characterization of worn surface is necessary.Fractal geometry offers a more accurate description for surface roughness that topographic surfaces are statistically self-similar and can be quantitatively evaluated by fractal parameters.The change regularity of worn surface topography is one of the most important aspects of running-in study.However,the existing research normally adopts only one friction matching pair to explore the surface topography change,which interrupts the running-in wear process and makes the experimental result lack authenticity and objectivity.In this paper,to investigate the change regularity of surface topography during the real running-in process,a series of running-in tests by changing friction pairs under the same operating conditions are conducted on UMT-II Universal Multifunction Tester.The surface profile data are acquired by MiaoXAM2.5X-50X Ultrahigh Precision Surface 3D Profiler and analyzed using fractal dimension D,scale coefficient C and characteristic roughness Ra ＊based on root mean square（RMS） method.The characterization effects of the three parameters are discussed and compared.The results obtained show that there exists remarkable fractal feature of surface topography during running-in process,both D and Ra ＊increase gradually,while C decreases slowly as the wear-in process goes on,and all parameters tend to be stable when the wear process steps into the normal wear process.Ra ＊illustrates higher sensitivity for rough surface characterization compared with the other two parameters.In addition,the running-in test carried with a set of identical surface properties is more scientific and reasonable than the traditional one.The proposed research further indicates that the fractal method can quantitatively measure the rough surface,which also provides an evidence for running-in process identification and tribology design.展开更多
基金Supported by the China Meteorological Administration Special Public Welfare Research Fund (GYHY(QX)2007-6-1)National Key Basic Research and Development (973) Program of China (2012CB955204)
文摘Based on the ensemble mean outputs of the ensemble forecasts from the ECMWF （European Centre for Medium-Range Weather Forecasts）, JMA （Japan Meteorological Agency）, NCEP （National Centers for Environmental Prediction）, and UKMO （United Kingdom Met Office） in THORPEX （The Observing System Research and Predictability Experiment） Interactive Grand Global Ensemble （TIGGE） datasets, for the Northern Hemisphere （10~ 87.5~N, 0~ 360~） from i June 2007 to 31 August 2007, this study carried out multimodel ensemble forecasts of surface temperature and 500-hPa geopotential height, temperature and winds up to 168 h by using the bias-removed ensemble mean （BREM）, the multiple linear regression based superensemble （LRSUP）, and the neural network based superensemble （NNSUP） techniques for the forecast period from 8 to 31 August 2007. A running training period is used for BREM and LRSUP ensemble forecast techniques. It is found that BREM and LRSUP, at each grid point, have different optimal lengths of the training period. In general, the optimal training period for BREM is less than 30 days in most areas, while for LRSUP it is about 45 days.
基金supported by National Natural Science Foundation of China(Grant Nos.51175017,51245027)Innovation Foundation of Beihang University for PhD Graduates,China(Grant No.YWF-12-RBYJ008)Research Fund for the Doctoral Program of Higher Education of China(Grant No.20111102110011)
文摘Because of the randomness of many impact factors influencing the dynamic assembly relationship of complex machinery, the reliability analysis of dynamic assembly relationship needs to be accomplished considering the randomness from a probabilistic perspective. To improve the accuracy and efficiency of dynamic assembly relationship reliability analysis, the mechanical dynamic assembly reliability（MDAR） theory and a distributed collaborative response surface method（DCRSM） are proposed. The mathematic model of DCRSM is established based on the quadratic response surface function, and verified by the assembly relationship reliability analysis of aeroengine high pressure turbine（HPT） blade-tip radial running clearance（BTRRC）. Through the comparison of the DCRSM, traditional response surface method（RSM） and Monte Carlo Method（MCM）, the results show that the DCRSM is not able to accomplish the computational task which is impossible for the other methods when the number of simulation is more than 100 000 times, but also the computational precision for the DCRSM is basically consistent with the MCM and improved by 0.40-4.63% to the RSM, furthermore, the computational efficiency of DCRSM is up to about 188 times of the MCM and 55 times of the RSM under 10000 times simulations. The DCRSM is demonstrated to be a feasible and effective approach for markedly improving the computational efficiency and accuracy of MDAR analysis. Thus, the proposed research provides the promising theory and method for the MDAR design and optimization, and opens a novel research direction of probabilistic analysis for developing the high-performance and high-reliability of aeroengine.
基金This work was supported by the Medical Development Foundation of Soochow University (No. EE134033) and the China-JapanInteruniversity Cooperative Research Foundation (No. EE134005).
基金supported by National Natural Science Foundation of China (Grant No.50975276,Grant No.50475164)National Basic Research Program of China (973 Program,Grant No.2007CB607605)Doctoral Programs Foundation of Ministry of Education of China (Grant No.200802900513)
文摘Studying and understanding of the surface topography variation are the basis for analyzing tribological problems,and characterization of worn surface is necessary.Fractal geometry offers a more accurate description for surface roughness that topographic surfaces are statistically self-similar and can be quantitatively evaluated by fractal parameters.The change regularity of worn surface topography is one of the most important aspects of running-in study.However,the existing research normally adopts only one friction matching pair to explore the surface topography change,which interrupts the running-in wear process and makes the experimental result lack authenticity and objectivity.In this paper,to investigate the change regularity of surface topography during the real running-in process,a series of running-in tests by changing friction pairs under the same operating conditions are conducted on UMT-II Universal Multifunction Tester.The surface profile data are acquired by MiaoXAM2.5X-50X Ultrahigh Precision Surface 3D Profiler and analyzed using fractal dimension D,scale coefficient C and characteristic roughness Ra ＊based on root mean square（RMS） method.The characterization effects of the three parameters are discussed and compared.The results obtained show that there exists remarkable fractal feature of surface topography during running-in process,both D and Ra ＊increase gradually,while C decreases slowly as the wear-in process goes on,and all parameters tend to be stable when the wear process steps into the normal wear process.Ra ＊illustrates higher sensitivity for rough surface characterization compared with the other two parameters.In addition,the running-in test carried with a set of identical surface properties is more scientific and reasonable than the traditional one.The proposed research further indicates that the fractal method can quantitatively measure the rough surface,which also provides an evidence for running-in process identification and tribology design.
