个人信息Personal Information
教授
博士生导师
硕士生导师
性别:男
毕业院校:北京航空学院
学位:硕士
所在单位:能源与动力学院
电子邮箱:xmz@dlut.edu.cn
Large eddy simulation and proper orthogonal decomposition analysis of turbulent flows in a direct injection spark ignition engine: Cyclic variation and effect of valve lift
点击次数:
论文类型:期刊论文
发表时间:2014-03-01
发表刊物:SCIENCE CHINA-TECHNOLOGICAL SCIENCES
收录刊物:SCIE、EI
卷号:57
期号:3
页面范围:489-504
ISSN号:1674-7321
关键字:in-cylinder flow; turbulence; proper orthogonal decomposition (POD); large eddy simulation; cycle-to-cycle variations (CCV)
摘要:Large eddy simulation (LES) is used to calculate the in-cylinder turbulent flow field in a direct injection spark ignition (DISI) engine. The computations are carried out for three different maximum valve lifts (MVL) and throughout 100 consecutive engine cycles. The simulated results as well as corresponding particle image velocimetry (PIV) measurement database are analyzed by the proper orthogonal decomposition (POD) method. Through a new developed POD quadruple decomposition the instantaneous in-cylinder flow fields are decomposed into four parts, named mean field, coherent field, transition field and turbulent field, respectively. Then the in-cylinder turbulent flow characteristics and cycle-to-cycle variations (CCV) are studied separately upon the four part flow fields. Results indicate that each part exhibits its specific characteristics and has close connection with others. The mean part contains more than 50% of the total kinetic energy and the energy cascade phenomenon occurs among the four part fields; the coherent field part possesses the highest CCV level which dominates CCV of the bulk flow. In addition, it is observed that a change in MVL affects significantly the in-cylinder flow behavior including CCV, especially for the coherent part. Furthermore, the POD analysis demonstrates that at least 25 sample cycles for the mean velocity and 50 sample cycles for the RMS velocity are necessary for obtaining converged and correct results in CCV.