Title: Development of a multiobjective design optimization procedure for marine propellers
Abstract: The design of a marine propeller is characterized by its complexity rather then its shortcoming of knowledge how to asses its performance. There are several constraints to satisfy and all are in a different field. The three major considerations are strength, performance (efficiency) and cavitation behavior. There is not a perfect design methodology for the engineer how to use the different analysis tools available to come to a final design. The process is in most cases an iterative one that ends with a satisfying design rather then an optimal design. Optimal in this case means optimal in the sense of the best compromise possible. The application of a multiobjective optimizer makes it possible to visualise the trade-off among different conflicting objectives to guide the engineer in making his compromise. Furthermore it gives more insight in the problem at hand. The goal of this thesis is therefore two-fold. First one is the implementation of a multiobjective optimiser to show what the gains are when an analysis tool can be turned into a design tool. The second one is to apply it and investigate what the trade-off is between cavitation performance and efficiency for a test case based on a container vessel. The implemented algorithm is the non-dominated sorting genetic algorithm (NSGA) which is currently used in many other practical design problems. Genetic Algorithms are in general robust but not very efficient when it comes to the number of design evaluations it has to do. The algorithm is able to give a good approximation of the trade-off with a good diversity among its solutions. The propeller is analysed by the lifting surface program ANPRO which has an extension to predict sheet cavitation and bubble cavitation at both sides of the blade. In the test case considered there were additional constraints for a minimum blade thickness, the avoidance of bubble cavitation and a maximum allowance of sheet cavitation at the pressure side. The algorithm is able to converge to a trade-off between the two conflicting objectives. The size of the population is chosen as 80 and the algorithm run for 300 generations. From statistical data of the average age of the individuals in the population, converge could be determined. The simulations were not able to converge completely but sufficiently enough to obtain an approximation of the Pareto front. The trade-off show a decrease from maximum efficiency of 4% while gaining a reduction of sheet cavitation at the suction side of 27%. The amount of reduction of efficiency was roughly the same for different test cases but the cavitation percentages showed more variation. The method showed to be sensitive to the interpretation of the cavitation data produced by the propeller analysis program ANPRO.
Publication Year: 2008
Publication Date: 2008-01-16
Language: en
Type: article
Access and Citation
Cited By Count: 2
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot