Hi @jamal If you set the tolerance to 3.597246683841558e-7 then the polygons will dissolve together as you expect. The Automatic tolerance is 3.597246683841558e-10
What does the tolerance represent when it comes to my polygons? Does this indicate that the adjacent polygons have gaps?
Yes. You could use an AreaGapAndOverlapCleaner to remove the gaps before the Dissolver and keep its tolerance the default Automatic. The attached workspace also measures the gaps between the vertices, ranging from 4.243546074853105e-8 to 4.233246933918182e-8
In order to find out where the gabs are:
1) I added a boundary polygon around the whole 10 polygons I have (R1.dwg)
2) I subtracted the 10 polygons from the boundary polygon with the clip command
3) I disaggregated the result
4) There is no clue that there are slivers that should show the gaps
I tried the same steps but with other CAD file (k20.dwg) at which the gaps are exaggerated, the gabs are caught
What could be the issue here?
In order to find out where the gabs are:
1) I added a boundary polygon around the whole 10 polygons I have (R1.dwg)
2) I subtracted the 10 polygons from the boundary polygon with the clip command
3) I disaggregated the result
4) There is no clue that there are slivers that should show the gaps
I tried the same steps but with other CAD file (k20.dwg) at which the gaps are exaggerated, the gabs are caught
What could be the issue here?
The Clipper creates a donut polygon with the R1 dataset - not an aggregate. Use a DonutHoleExtractor to break it up. With the K20 dataset, the Clipper creates an outside feature which is an aggregate of 4 parts - the outer donut and 3 polygons inside the hole of the donut.
My aim her is to identify the gaps between the polygons in question that cause the dissolver tool to fail to get them dissolved as expected.
I’m not able to identify these gaps even with the “DonutHoleExtractor” as per the screenshot below. For example, I’m expecting to have gabs between polygon#1 and polygon#2 indicated in the screenshot. These gaps are not identified anyway