GTS continuous Boolean operation failed ~

We have been paying attention to and solving the problem of Boolean operations on objects. This is a major issue in geometric modeling and is also the core of cutting, processing, and simulation. Although there are some technical methods that can simulate "Boolean operations", the results are not good, and the most direct is the true Boolean operation. However, this is not very easy.

Some time ago, I kept searching for open-source 3D Boolean libraries on the Internet. At that time, I felt satisfied with GTS (GNU Triangulated Surface library ), it took a long time to solve the compilation problem under vs2005, and then took some time to understand the "object-oriented" of "weird" C ".

Yesterday, GTS and openscenegraph were combined to convert each other in two formats. GTS is responsible for Boolean operations and openscenegraph is displayed. The initial experiment results were satisfactory yesterday, and various Boolean operations were normal.

It also shows the result of a Boolean operation on a cylindrical tool and a cube.

The next step is to allow these Boolean operations to be performed consecutively, so as to simulate the process of the item cutting on the model. However, the GTS library finally crashes. Boolean operations generate a large number of small triangles. However, when the tool moves a small distance, GTS cannot solve the Boolean operation problem between many small triangles, leading to the failure of assertions in GTS! Later, I checked it in the GTS mail list. However, N people still asked this continuous Boolean operation problem, and there was no good solution. To solve this problem, at least simplify the grid after the Boolean operation. Of course, it is not that easy to simplify a Web page. Recently, GTS cannot be applied any more.

In other libraries, cgal also supports 3D Boolean operations. However, the format of cgal is odd and you have not understood what is going on yet.CodeThe developer is required to publish the code. Of course, as a student, this can be ignored for the moment. It is not a problem to use it on your own. Opencascade brings together all functions, that is, it's too complicated ~~~ In addition, the recently found ACIS, I applied for a University Promotion Program license, but I still don't know when it will be approved. I think this estimate can be met. In addition, there is also a very useful sgcore. The free version does not support Boolean operations on any grid. It is annoying to use this library if it does not work. You have to test whether multiple Boolean operations can be performed ~

Come on, you have a long way to go ......