+4Hello,
I'm trying to convert an IFC file into a ArcGIS webscene, via I3S. This worked fine for other models in the past, but this time, the writer crashes after using up ~27 gb of RAM.
It's a tiny IFC, just some ~40 Mb. Even weirder, I can convert the IFC solids into multipatches via SHP, or GDB. Just no I3S. :(
I would really appreciate some hints.
+25Hi @davidrasner5 Try to add a GeometryValidator before the writer, and choose All tests - to see if there are any "difficult" geometries.
+47Souds like a geometry issue for sure - however, another possibility is if the data from the IFC are in very different area - I've seen cases where there are data at the origin of a coordinate system but also with data at the correct (projected) location. Meaning that the bounds of the data are Huuuuge. I'm not sure what king of effect this would have on the i3s writer, however, this could be playing a role is the increased memory usage.
Â
Â
It's easy to rule out, you just need to look at the IFC in the data inspector.
+19David, if you can share your data please open a support call with Esri and request the 3D team product engineer for i3S be attached to the issue, he is interested in seeing this behavior.
+4Hi @davidrasner5 Try to add a GeometryValidator before the writer, and choose All tests - to see if there are any "difficult" geometries.
Thanks for the reply. The IFC seems to be very messy. I tried the "basic integrity" preset and the geometry validator just logs a "error running translation".
+4David, if you can share your data please open a support call with Esri and request the 3D team product engineer for i3S be attached to the issue, he is interested in seeing this behavior.
I will try that, thanks.
+25Thanks for the reply. The IFC seems to be very messy. I tried the "basic integrity" preset and the geometry validator just logs a "error running translation".
You can try to add a Deaggregator before the GeometryValidator. And then try to "weed out" the bad objects.
