[Radiance-general] How to generate surface with gensurf program?

[Marija Velickovic]

Thanks for answers,

After detailed analysis of lamellas' SKetchUp geometry, I decided to
simplify it a little bit - so I replaced curved parts with really small
radius with 2 polygons.
When exported number of surfaces per lamella is much smaller, and I still
get correct reflectances and overall blind transmittance, so I think this
simplified lamella geometry will work.

Since I need in building lots of blinds, so I'll surely use instances and
xform to put it all in building.

CU,
Marija.

On Wed, Nov 9, 2011 at 4:11 PM, Lars O. Grobe <grobe at gmx.net> wrote:

> Hi Marija!
>
> > I'm trying to model one complex blinds' lamelas in Radiance.
> > I have lamela model in SketchUp but when exported to Radiance it gives
> > 70 surfaces per one lamela - which is too much for simulations.
> >
> > So I have an idea to try to model surface with gensurf, but no idea how
> > exactly.
> >
> > Since I have SketchUP export I can extract data about sruface points so
> > it is some array of (x,y,z) coordinates:
> > (x1, y1,z1), (x2,y2,z2)...(xn,yn,zn).
> > Points in x and z aren't equidistant, while y={0,1}.
> >
> > Problem is how to use these values in gensurf.
> > I want to have surface composed of multiple surfaces in x direction and
> > 1 surface in y direction (length of a lamela)
> >
> > I suppose appropriate command format would be:
> > /gensurf mat name dfile dfile dfile m n /
> >
> > What is the format of dat files I need here? I tried to understand
> > something from manpage but no sucess.
> >
> > Is this the right way to model complex blind lamela (it has few
> > curvatures and w shape etc.)?
>
> My impression is that this will get quite complicated. gensurf would
> still need the coordinates to come in the right ordering, and I do not
> know how you could achieve that when exporting from Sketchup. Without
> that, gensurf does not know which vertex locations to combine into one
> surface. Also, you would again get lots of surfaces, as gensurf would
> create flat surfaces between the vertex locations.
>
> Two alternatives:
>
> a) If you exactly know the profile of your lamellas, you could use a
> functional description instead. This would still lead to quite a lot of
> surfaces for each, but probably less. And (most important) you would
> have great control on the resolution of the facetting. This would e.g.
> make it possible to find out at what resolution you see no effect on
> your results any more.
>
> b) If you need to rely on your exported Sketchup model, I would advice
> looking into instances or the mesh object. Create a "short" lamella,
> something like one unit length, and export it. Make it a frozen instance
> (using oconv -f) or a mesh object (using obj2mesh). Create a small scene
> file containing only the mesh or instance primitive for this compiled
> geometry. Than build your blinds from this element using an xform array
> in your scene file. This will lead to very little memory consumption for
> these objects.
>
> If your shading systems consists of lots of blinds, I would even
> consider using instances when creating geometry using gensurf (case a).
> However in typical scenes, you do not reach critical polygon counts yet.
> Radiance is very good in handling lots of surfaces, unless memory gets
> used up.
>
> Cheers, Lars.
>
>
> _______________________________________________
> Radiance-general mailing list
> Radiance-general at radiance-online.org
> http://www.radiance-online.org/mailman/listinfo/radiance-general
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.radiance-online.org/pipermail/radiance-general/attachments/20111111/3c832e7a/attachment.html>