Aligning REVOLVE and BPRISM geometry

[shaun coomber]

Hi

I’ve been having some problems aligning a REVOLVE and a BPRISM_. The BPRISM_ seems to scale strangely when using RESOL. With fewer segments, it scales up slightly and loses alignment with other elements. When using RESOL 360 it aligns perfectly, with RESOL 36 it’s noticibly out of aligmnet, and at RESOL 8 it massively wrong.

The attached image shows the different between the BPRISM_ (the top section of the profile) and the REVOLVE (bottom section).

I’ve uploaded the test object to:
https://cloud.applecoredesigns.co.uk/s/YqpSC4obQ8ptoLW

Any ideas?

Cheers
Shaun

Attachments:

1. 

   geometry-problems.png
   

[shaun coomber]

It’s OK, I’ve figured it out

[Barry Kelly]

shaun coomber wrote:

It’s OK, I’ve figured it out

Any hints for those following along?

Barry.

Versions 6.5 to 22
Dell XPS- i7-6700 @ 3.4Ghz, 16GB ram, GeForce GTX 960 (2GB), Windows 10
Dell Precision M6800 - i7 4700MQ @ 2.40GHz, 16GB RAM, AMD FirePro M6100 (2GB), Windows 7 64bit

[shaun coomber]

Hi Barry

There were a few issues where BPRISM deals with things differently to other commands. The main problem was that ARCHICAD facets BPRISMs completely differently. The attached image shows the original geometry (circle) with how ARCHICAD usually facets (inner hexagon) and an example facet for the bprism (triangle extending to the outside of the circle).

To get around the problem I had to:
– multiple the radius of the BPRISM by * cos(180/resolution)
– multiple the thickness of the BPRISM by * cos(180/resolution)
– rotate the BPRISM by 180+180/resolution along its axis

GS HQ also suggested using REVOLVE{2} alongside BPRISM, instead of the basic version of REVOLVE. It seems to be undocumented but REVOLVE{2} behaves in the same way as BPRISM when facetted, instead of like the usual commands.

However, I’m having a couple of other issues with REVOLVE{2} so you might hear more from me shortly!

Attachments:

1. 

   curvesegments.png
   

[shaun coomber]

Right, my next problem is that REVOLVE{2} is missing the top and base surfaces, which were present with REVOLVE. Image attached, and example GDL below. Any ideas on this would be appreciated…

! Attributes

BUILDING_MATERIAL shellMaterial
n = REQUEST{2} (“Building_Material_info”, shellMaterial, “gs_bmat_surface”, shellSurface)

! Dimensions

diskWidth = 0.500
diskHeight = 0.200
lipWidth = 0.100

! REVOLVE

PUT 0, 0.5*diskWidth, 2,
0.5*diskHeight, 0.5*diskWidth + lipWidth, 2,
diskHeight, 0.5*diskWidth, 2

ROTY -90
REVOLVE NSP/3, 360, 15, GET(NSP)
DEL 1

! REVOLVE{2}

PUT 0, 0.5*diskWidth, 2, shellSurface,
0.5*diskHeight, 0.5*diskWidth + lipWidth, 2, shellSurface,
diskHeight, 0.5*diskWidth, 2, shellSurface

ADDZ 1
ROTY -90
REVOLVE{2} NSP/4, 0, 360, 15, shellSurface, GET(NSP)
DEL 2

Attachments:

1. 

   revolve2-test.png
   

[Barry Kelly]

The REVOLVE{2} will always close the polygon (first to last point).
So of you want the surface back to the rotation axis you need to add a point with a 0 (zero) y-position.

i.e.

PUT 0, 0.000, 2, shellSurface,
 0, 0.5*diskWidth, 2, shellSurface,
 0.5*diskHeight, 0.5*diskWidth + lipWidth, 2, shellSurface,
 diskHeight, 0.5*diskWidth, 2, shellSurface,
 diskHeight, 0.000, 2, shellSurface

ADDZ 1
 ROTY -90
 REVOLVE{2} NSP/4, 0, 360, 4*1 + 8*1 + 16*1 + 32*1 + 64*0 + 128*0 + 256*0, shellSurface, GET(NSP)
 DEL 2 

Barry.

Versions 6.5 to 22
Dell XPS- i7-6700 @ 3.4Ghz, 16GB ram, GeForce GTX 960 (2GB), Windows 10
Dell Precision M6800 - i7 4700MQ @ 2.40GHz, 16GB RAM, AMD FirePro M6100 (2GB), Windows 7 64bit

[shaun coomber]

Barry Kelly wrote:

The REVOLVE{2} will always close the polygon (first to last point).
So of you want the surface back to the rotation axis you need to add a point with a 0 (zero) y-position.

i.e.

<code class="hljs">PUT 0, 0.000, 2, shellSurface,
 0, 0.5*diskWidth, 2, shellSurface,
 0.5*diskHeight, 0.5*diskWidth + lipWidth, 2, shellSurface,
 diskHeight, 0.5*diskWidth, 2, shellSurface,
 diskHeight, 0.000, 2, shellSurface

ADDZ 1
 ROTY -90
 REVOLVE{2} NSP/4, 0, 360, 4*1 + 8*1 + 16*1 + 32*1 + 64*0 + 128*0 + 256*0, shellSurface, GET(NSP)
 DEL 2 

Barry.

That worked perfectly, thank you Barry

[Author]

Posts