Area Calculation Error

Yet another try… Explanation is tedious…

I created an Arc 3Pt on the XZ plane. The ‘ArcOffset’ slider (blue group) sets the offset for the middle point. That will be relevant in the benchmarks below.

I oriented that arc on PFrames around the base curve. The ‘Count’ slider (white group) sets the number of PFrames, which I left at 62 for all of my tests, but you should try other values to see the effect.

Then I used three different methods for getting the curved surface: Swp2, NetSurf and Loft. All three are executed quickly and there is a Value List (‘SURFACE’ in orange group) to choose one of them before the Data Dam (red group). The surface and diagonal lines are visible before clicking the Data Dam, which prevents the slow SrfSplit and Area components from executing.

The purple group creates the extrusion and RuleSrf below the “arc surface”.

Though you didn’t ask for it, I added a green group to join and cap the three breps, to test that it results in a “Closed Brep” as I think it should. The loft option requires special treatment for this, replacing the bottom “rail” with the bottom edge of the lofted surface. This is common when using Loft as the lofted surface can be slightly different from the curves used to create it.

Finally, I did some benchmarks. Some are extremely slow, minutes (m) instead of seconds (s)
Loft is the fastest but gets 22 ‘Nulls’ when ArcOffset = 450.
NetSurf always gets zero ‘Nulls’ but can be extremely slow on Area.

ArcOffset (blue group) = 150

	Swp2   : SrfSplit = ~8.3s, Area = ~15.2s, Nulls = 0
	NetSurf: SrfSplit = ~16.6s, Area = ~6.2m, Nulls = 0
	Loft   : SrfSplit = ~7s, Area = ~2s, Nulls = 0

ArcOffset (blue group) = 250

	Swp2   : SrfSplit = ~12.1s, Area = ~1.2m, Nulls = 0
	NetSurf: SrfSplit = ~11.7s, Area = 25.5m, Nulls = 0
	Loft   : SrfSplit = ~7s, Area = ~2s, Nulls = 0

ArcOffset (blue group) = 450

	Swp2   : SrfSplit = ~17s, Area = ~1.6m, Nulls = 16
	NetSurf: SrfSplit = ~12s, Area = ~2.1m, Nulls = 0
	Loft   : SrfSplit = ~14s, Area = ~3.4s, Nulls = 22

Would be nice to repeat these benchmarks to double-check them but it can take a long time and overheat my laptop, which I don’t like to do.

Why do you need area Because all these methods work fairly fast except for Area.

P.S. I just restarted my laptop and re-ran the NetSurf benchmark:

ArcOffset = 250

	NetSurf: SrfSplit = ~10.4s, Area = 57s, Nulls = 0

25+ times faster

So clearly the benchmark results I posted earlier are suspect. Don’t know why.

Read full topic