Toolset profiling for rotary draw bending of waveguides
The paper discusses the problem of bending quality of thin-walled structures of waveguides of rectangular cross section. Modern methods of solving the bending problems have been studied, from empirical relations to the use of artificial intelligence methods to achieve a minimum bending radius withou...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2025-01-01
|
| Series: | EPJ Web of Conferences |
| Online Access: | https://www.epj-conferences.org/articles/epjconf/pdf/2025/06/epjconf_apitech-vii2025_01003.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The paper discusses the problem of bending quality of thin-walled structures of waveguides of rectangular cross section. Modern methods of solving the bending problems have been studied, from empirical relations to the use of artificial intelligence methods to achieve a minimum bending radius without product defects. It has been shown that the quality of the bend depends on a combination of many input parameters that are not always controllable. Even with the current level of computer performance, taking into account all these external factors, leads to the need to solve the probabilistic problem of a large number of variables, the unambiguous answer of which can currently only be given by approaches based on artificial intelligence. Despite the development and popularity of artificial intelligence methods, this way is still very expensive and long, so we propose a solution based on creating bending conditions that limit unwanted deformation and actually lead to die forming of the product. As a result, the resulting product and its quality parameters should be practically independent of various external factors and only the correctness of calculations and the accuracy of manufacturing the toolset of the pipe bending machine should be determined. |
|---|---|
| ISSN: | 2100-014X |