Abstract: This study deals with the problem of an algorithm development and optimization that allows to visualize the surface of the liquid jet in dynamics by approximating its triangles. This is so-called problem of triangulation. There are many algorithms to solve such problems but most of them can not reach the speed necessary for real-time applications. The algorithm should be used in extremely loaded simulation environment and the specificity is that the program should be called in every frame. In addition to all this, the environment has a lot of other logic and calculations. This imposes the significant restrictions during operation time, the number of generated triangles and the overall complexity of the algorithm. A set of points is supplied to the program input, the global coordinates which are the centers of the liquid particles. We should get 3D object data which contains the information about the vertices forming the desired surface, the method of their decomposition by triangles, the textural coordinates of the vertices and the normal values at the point. The result is that the original problem is divided into two independent sub-tasks: the calculation of the function value forming the liquid jets at each point and the triangulation of three-dimensional scalar field. These problems are solved with the help of different approaches that in combination with parallel technologies and optimization methods allowed to achieve the high efficiency of the algorithm and use the visualization algorithm of the physics engine successfully and the simulation environment associated with medicine.