SHUANGXI, G., WANPING, L.: Inverse energy cascade and
self-organization in 2-D turbulent channel flow 255
JAVADI, H., RAJABI, I., BAKHSHANDEH, K.: Simultaneous consideration
of the influence of geometry and material properties on theoretical
stress concentration factors 267
YUAN, A. F.: Virtual design of movable tooth gear box for
SAFARUDIN, M., SETIAWAN, J., SIHOMBING, H.: Vehicle active roll
control using roll moment rejection algorithm 301
BHARTI, P. S., MAHESHWARI, S., SHARMA, C.: A comparative study
of Artificial Neural Network training algorithms for modelling
of Electric Discharge Machining process 323
G. SHUANGXI, L. WANPING
Inverse energy cascade and self-organization in two-dimensional (2-D) turbulent channel flow have mainly been investigated in the present paper. Direct numerical simulations (DNS) of forced 2-D turbulent channel flow are performed first, and the results verified the correctness and validity of the numerical methods and computation codes. Then the self-organization and inverse energy cascade are investigated with the decaying 2-D turbulent channel flow. The results indicate that an obvious self-organization of vortices exists in 2-D turbulent channel flow. The adjacent and same sign vortices interact with each other and merge into larger ones, and a pair of opposite sign largest vortices is formed and maintained ultimately. At the same time, the inverse energy transfer from small structures to larger ones is along with the self-organization process. In the end the differences of mean velocity profile of 2-D and 3-D turbulent channel flow and the characteristics of oceanic circulation are explained with the inverse energy cascade and self-organization mechanism.
H. JAVADI, I. RAJABI, K. BAKHSHANDEH
A. F. YUAN
In this paper, according to the requirements of aerogenerator, a kind of movable tooth gear box was designed and its performances in contacting force and movement stability of output shaft were simulated by ADAMS. Analysis results show that this kind of movable tooth gear box can meet the requirements of aerogenerator and it can operate normally in specific conditions.
M. SAFARUDIN, J. SETIAWAN, H. SIHOMBING
In this paper, a passenger vehicle active anti-roll control was developed and simulated using a proposed anti-roll algorithm based on 14 DOF full vehicle model. The Magic Formula was used for tire model. The full vehicle model was developed and the simulation was compared with vehicle dynamics simulation software and validated using an instrumented experimental vehicle. The active roll control algorithm was then introduced to the vehicle model. Combined with PID control, the results were then simulated and analysed. From the simulation, it was found that the algorithm can significantly reduce the roll angle and roll rate of the vehicle and eventually prevent the vehicle from rollover. The improvement of the roll motion also reduced the load transfer from the inner wheels to outer wheels and hence increased the road holding during cornering.
P. S. BHARTI, S. MAHESHWARI, C. SHARMA
Optimization of process parameters of Electric Discharge Machining (EDM) is done by mathematical techniques like fuzzy logic, grey relation, genetic algorithm, data mining, etc. Process modelling of EDM is done before applying these techniques to optimize the process parameters. Artificial Neural Networks (ANN) are very suitable for the same. This work is an attempt to highlight the importance of the selection of training algorithm, from a number of available Back Propagation (BP) algorithms, in training of ANN. This paper compares the results of training of ANN model of EDM process by taking thirteen different BP algorithms available in MATLAB Neural Network Toolbox. Sensitivity analysis of the algorithms, with respect to inputs, is also done to finally select the algorithm. Thus an attempt is made to objectively select the best BP algorithm for the training of ANN model of EDM process.