What it does
Also known as polygon reduction or mesh decimation, the Triangle Reduction tool basically does what it says on the tin: it reduces the amount of triangles in the STL file. Every triangle your design is made up of uses memory in your computer. Therefore, STL files with more triangles are heavier to process and slice. In many cases, the 3D printer won’t even be able to print a file over a certain amount of polygons, making the Triangle Reduction tool an essential component.
When the Triangle Reduction tool gets to work, it makes your STL file lighter and therefore the 3D Printing software will take less time to slice it before printing.
Why is triangle reduction necessary in the first place?
After reading all that, you would be forgiven for thinking “Why doesn’t the designer just make sure they design their object with less triangles in the first place?” There are two good reasons why even the most thrifty of designers inadvertently ends up creating an STL with too many triangles.
Your software isn’t doing its job
Sometimes, the case might be that the software you use to transform your CAD file into an STL file is simply not advanced enough. Instead of generating an accurate STL file, it might generate an STL file that uses way too many triangles than it should be using.
For example, the two shapes below both form a square. But the square on the right uses two triangles for this, and the one on the left uses a grand total of 13. That’s 11 triangles too much! The Triangle Reduction tool will get rid of those superfluous triangles, and leave you with your original – and now much more manageable – design.
You might be the problem
We hate to point fingers, but this option tends occur much more often as most software is quite advanced these days and well-adapted to making a solid STL file. It almost always boils down to the same thing though: when transforming their CAD file to the STL format, the user will have inadvertently used bad parameters.
There is also a perfectly good explanation for this. Usually, the designer will input parameters that are much too precise for the 3D printer to print, because they want their design to (understandably) be printed as precisely as possible.
The quest for precision
So why are high amounts of triangles associated with a more precise design? Essentially, a triangle in an STL file is always flat. So if you want to form a curved surface with triangles, it is possible to do that by using many tiny triangles together, and the more triangles, the more curved the surface will appear.
For example, take a look at these circles down below. The blue line represents the original design, in its CAD format. The red lines represent the triangles in the STL format. Since you are obliged to use straight lines in order to describe a curved shape, there will always be a certain disparity between the two file formats. This disparity is called the tolerance.
In its roughest form, only three straight lines are used to represent the original circle – which is clearly undesirable to any designer, as the shape resembles a triangle more than anything else. The more red lines you use, the more the shape will start to look like a genuine circle, and the smaller the disparity between the CAD file and the STL file will be. But at some point, you will have too many red lines. Your file will be needlessly precise, whereas your object would look just as good with a smaller amount of triangles.
Why it all goes wrong
A designer will often input an extremely precise tolerance, thinking that their design will only benefit from it. But often, a printer won’t even be capable of printing at that sort of resolution in the first place. When working on a design, the user gets used to working with extremely small values (this happens a lot while creating small, complex pieces such as jewelry). When it comes to making the STL file, they continue to use these small values. For example, a designer will generate a tolerance of 0.001mm for maximum precision. But it is perfectly possible to raise the tolerance to 0.005mm, which will result in a printed part that has no visible differences to an STL that has been generated with a value of 0.001mm. Moreover, the STL file will have a great deal less triangles than the STL file of 0.001mm, and will take up much less memory and processing time.
How it works
Materialise Cloud’s Triangle Reduction tool uses the following parameters:
smallest detail (0-10 mm; default 0.005): allowable deviation of the output surface from the input surface
maximum angle (0-90; default 5): angle between adjacent surface normals that determines which triangles can be combined into one
the number of iterations (1-1000); default 5): how many times the algorithm will subsequently run
The bigger the value of “smallest detail” and “angle”, the more degradation in the surface quality.