Comparison of the results of studies designed by Taguchi and full factorial methods for parametric investigation of flow pattern around spur dike

Full factorial investigation is necessary in the study of the hydraulic phenomena which are function of different variables with different levels. It is logical to use the full factorial method when the number of variables and their levels are low. However, sometimes due to lack of time and shortage of financial restrictions, using the full factorial method is not possible. The Taguchi method, which is used for design of experiments, uses the fractional factorial instead of full factorial. This method not only decrease the number of studies but also guaranties the correlated comparison of all variables. In this paper the Taguchi method is used for finding the optimized hydraulic parameters like length and location of spur dike in different Froude Numbers in 90 degree bend. Furthermore, the comparison between the Taguchi method and full factorial is done for the number of investigations, finding the optimized level for each parameter and the time needed for study. In order to get the results, the parameters of length, location of spur dike in 90 degree bend, and the Froude Number are considered with three different levels. The SSIIM numerical model is applied to simulate the studies designed by Taguchi and full factorial methods. The results show that the Taguchi method, could predict the optimum parameters only with 9 studies whereas with full factorial method 27 studies was necessary. Also, using Taguchi method leads to more than 66% decrease in the total running time. For studies designed by Taguchi method, the optimum value of length of spur dike is the as one designed by Full factorial method. Also, the length of spur dike is the most effective parameter on flow pattern around spur dike and the position of spur dike and Froude Number are next in rank, respectively. These results are the same for two methods used to design of studies. Using 9 studies designed by Taguchi method, investigation of the effect of other parameter such as the angle of the spur dike is possible without changing the number of studies, whereas 81 studies should be done by full factorial method.
Full factorial investigation is necessary in the study of the hydraulic phenomena which are function of different variables with different levels. It is logical to use the full factorial method when the number of variables and their levels are low. However, sometimes due to lack of time and shortage of financial restrictions, using the full factorial method is not possible. The Taguchi method, which is used for design of experiments, uses the fractional factorial instead of full factorial. This method not only decrease the number of studies but also guaranties the correlated comparison of all variables. In this paper the Taguchi method is used for finding the optimized hydraulic parameters like length and location of spur dike in different Froude Numbers in 90 degree bend. Furthermore, the comparison between the Taguchi method and full factorial is done for the number of investigations, finding the optimized level for each parameter and the time needed for study. In order to get the results, the parameters of length, location of spur dike in 90 degree bend, and the Froude Number are considered with three different levels. The SSIIM numerical model is applied to simulate the studies designed by Taguchi and full factorial methods. The results show that the Taguchi method, could predict the optimum parameters only with 9 studies whereas with full factorial method 27 studies was necessary. Also, using Taguchi method leads to more than 66% decrease in the total running time. For studies designed by Taguchi method, the optimum value of length of spur dike is the as one designed by Full factorial method. Also, the length of spur dike is the most effective parameter on flow pattern around spur dike and the position of spur dike and Froude Number are next in rank, respectively. These results are the same for two methods used to design of studies. Using 9 studies designed by Taguchi method, investigation of the effect of other parameter such as the angle of the spur dike is possible without changing the number of studies, whereas 81 studies should be done by full factorial method.