Development of new algorithms for data processing for topological rationalization

Existing algorithms for processing the results of topological rationalization are imperfect. They can effectively treat the results only in the case of simple geometry, and for the evenly sized finite elements. They take a lot of time during the formation of the necessary matrices and inefficiently operate on large volumes of data. That is why the need to search for new and improved methods of processing the results of topological rationalization arises. Methodology. The proposed algorithm for processing the results of topological rationalization is based on the vectoring of coordinates of the centers elements and then on subsequent transfer of the basis of the coordinate system from one center of the element to another. To implement this idea, the program code in the MATLAB was developed. Results. This algorithm has been tested on the test problems of different size and of different geometrical complexity. The results confirm the effectiveness of the discussed method. A time saving on matrix formulation is observed to vary between factors of 2 and 6 depending on the size of the problem being solved. It is also found that along the increase of size of the problem the efficiency of the method also increases. Also the assumption about the possibility of a significant acceleration of the calculation of the topological rationalization has been confirmed. This was achieved by reducing the time spent on solution of finite element model through the implementation of iterative methods for solving systems of linear equations and usage of memory savings technology. Scientific novelty. Improvement in processing of the results of topological rationalization by applying data vectoring and memory preallocation. The practical significance. Usage of the proposed method can help to accelerate the introduction of topological methods for rationalization in everyday civil engineering design practice.