NURBS Representation Of3D Hull Surface Based On Reduced Data

Hull surface plays a crucial role in the performance of the ship, and it is the foundation of the calculation of floatation and stability, free surface correction, subdivision, general arrangement, structural design and the calculation of hydrodynamic. Currently, in the hull surface design field, NURBS is widely used for it can express the hull surface accurately. Its matching technology (such as node insertion and elimination and degree reduction, etc.) is very mature, and there will be a good performance to apply NURBS to hydrodynamic analysis, structure analysis and manufacturing simulation. However there are still some problems with the application of NURBS technology to build the hull surface, such as difficulties in fairing and modification caused by a large amount of data.For the reason that hull surface modeling based on NURBS needs a large amount of data, which lead to difficulties in fairing and modification, and inconvenience to combine with hydrodynamic performance, a method that modeling hull surface of3D based on a small amount of data is presented, in order to speed up the design process, reduce the design cycle, and lay the foundation for combine hydrodynamic performance with the design of hull surface later.First of all waterline approximation model by reducing the control points was build, it is based on the known data of hull surface. In order to realize modeling hull surface with less data, this paper will divide the waterlines into two categories according to whether there is a arc at the end of the curve and then get the waterlines segmented, build the control polygon for the free curve segment of waterline and determine the design variables and constraints, respectively. Through the repeated calculation, it can be finally found that using five control points for the waterline of free curve segment approximation expression can achieve good results, the middle three control points of the five have weight that not equal to1. Eventually the coordinates of control points and its weight are determined as the design variables, fore and aft endpoint and the tangent vector are the constraint conditions, and the objective function adopt two method:the maximum relative error of the new waterline and the original waterline or the sum of squares of all the transverse sections.Secondly, genetic algorithm program for the calculation of waterline approximation is written. Then, using the genetic algorithm class library GAlib, the author builds the basic genetic algorithm based on binary coding projects, and then writes ARX program depends on the genetic algorithm to find the solution to the design variables, outputs them and draws and displays the waterline. We can fulfill the constraints by constructing penalty function.At last, using the two groups of waterlines generated by the two kinds of objective function, the author generates lofting surface in AutoCAD operation respectively, and compares with the original ship lofting surface. It can be found that the line approximation has a better effect, but the surface lofted by waterlines generated by the second objective function is not fair, to analyze the reason, several waterlines have a bigger deviation in different positions may cause the result. But the surface lofted by waterlines generated by the first objective function has a better visual effect.In this paper, the results show that the free curve segment of the waterline can be represented using five weighted control points as constraint, in turn, the generated lofting surface can achieve a good effect, which has a certain engineering significance.