SEOs & file speed

[Anonymous]

Hi folks,

I'm working on a house on a steep site with a LOT of ground works, piles, foundations, you name it.

To accurately get the cut & fill volumes, correct sections, etc all of the ground works are modeled and extracted from the mesh with SEOs. The issue is this slows the file down an insane amount; it can take up to a minute just to refresh a section.

Our IT department has upped my computer spec with only minimal improvements:

16GB RAM
Intel Xeon CPU E5-1620 @ 3.5GHz
NVIDIA Quadro K2200 4GB

My question is if anyone else has these issues? Is there a better workflow to get the cut & model correct, or is this something we're just going to have to deal with?

Cheers!

[Carstenem]

Yes a lot of SE's do slow down ARCHICAD a lot. Options:

1. leave it as is and cope with it !
2. convert the terrain after SEO's into morph object but this might be equally slow, worth a try ( make a copy of the file first ) Draw back : changes won't be reflected anymore.

Good luck.

[Richard Morrison]

Are you just selecting everything on the foundation level and using an SEO with upwards extrusion? If so, alternatively, you might try using a single (or a couple, if necessary) slab at the lowest level around the foundation perimeter, so you are using maybe less than 5 SEOs total. You can keep the original model for cut & fill calculations, but for day-to-day working, you'll have a faster model. You can touch up sections with 2D fills, if necessary.

[Eduardo Rolon]

Another item that slows terrain SEOs is PBCs, so first change the Intersection number of the layer were the mesh is to 50 and see if it helps.

[Laszlo Nagy]

You could also try to divide your Mesh element into 3 or 4 Mesh elements. Then perform the SEOs on them. The number of total SEOs will be the same, but the number of total SEOs per Mesh element will be much lower. This may improve performance.

My experience has been that beyond a certain complexity the performance hit increases exponentially. The above trick tries to not reach that complexity by dividing complex elements into less complex ones. Here by complexity I mean number of SEOs performed on elements.

[Anonymous]

All great hints, thanks folks. We'll do some testing & see how it goes!

Many thanks