Re: [EnergyPlus_Support] Re: model non-standard building facade for thermal analysis

Computational time is heavily increased by the numbers or surfaces in the model. having each heaxagon modeled as 6 triangles is not advised. I previously said that you could make an hexagon with two poligons with 4 vertices. But i forgot one "small" detail: Eplus only accepts rectangular shapes for windows. So that's one less solution. Best you could do is 2 triangles and 1 rectangle. And that would leave you with 2 frames crossing the "ideal" glazed area of the hexagon, something that would certainly influence shadowing in your zone. If you do decide to try this, my advice is to model your window with as few hexagons as possible.

My other idea is for you to model your grid with rectangles instead of hexagons. you would have 3 times less surfaces and still might get some close behaviour comparing to the hexagons. The only thing you would have to be carefull about is to model your rectangular windows with the same area as the hexagons would have. I've attached a printscreen to show you what i'm talking about.

Pedro

On 18-05-2011 17:50, artitecture82 wrote:

Hi Pedro,

mhh not sure I got what you mean -- but if E+ accepts triangular faces for windows then an hexagon can be split in 6 sub triangles - my conernings with this is how much "computational" time will add to the calculation...
what do you think?

--- In EnergyPlus_Support@xxxxxxxxxxxxxxx, Pedro Peixeiro <ppeixeiro@...> wrote:
>
> You can try a simpler way to tackle the problem: instead of using single
> heaxagonal surfaces, you can try do it with single rectangles properly
> (dis)aligned to best aproximate your grid. From there, you can specify
> frame withs, inside and outside projections, material properties for
> both the frames and window, and so on. And for even a more simple model
> (which might prove usefull for preliminary studies) you might want to
> create a normal rectangular window with as many horizontal and vertical
> dividers as you need. This way you only model one window, and input the
> number of dividers you need. basically it would be the first approach
> but with the "individual" windows correctly side by side (rather hard to
> explain in words :)).
>
>
>
>
> On 18-05-2011 16:28, artitecture82 wrote:
> >
> > Hi,
> > thanks for your opinion. I suspected was hard...in that case I
> > probably need to find a simplified calculation model that takes into
> > account my geometry easily...especially for shadowing calculations; I
> > will implement it in Java. Do you maybe now a good calc method I could
> > look at for this problem (envelope thermal balance+shadowing)?
> >
> > --- In EnergyPlus_Support@xxxxxxxxxxxxxxx
> > <mailto:EnergyPlus_Support%40yahoogroups.com>, Pedro Peixeiro
> > <ppeixeiro@> wrote:
> > >
> > > Hi Giulio.
> > >
> > > In my opinion, will be hard to model such a frame. Windows in Eplus are
> > > limited to 3 or 4 vertices. Modelling one hexagonal surface considering
> > > it's a window is impossible. You can try to model each hexagonal as two
> > > windows, but that way you would end up having an _extra frame right
> > > across the hexagonal area_, and two window objects for each hexagonal
> > > window (and if your total window will have that many hexagonal surfaces
> > > as there is in your link, you're in for a lot of windows). If for your
> > > study you really need those frames in that geometry, honestly don't see
> > > any other way to do it.
> > >
> > > Pedro
> > >
> > >
> > > On 18-05-2011 12:49, artitecture82 wrote:
> > > >
> > > > Hi All,
> > > > Im new to E+ and I've usually run very basic analysis for thermal
> > > > analysis of traditional building ... often approximated with boxes
> > and
> > > > few window.
> > > >
> > > > Im doing a personal research as part of my master thesis and Im
> > > > interested to compare the performance of different facades tipologies
> > > > and shapes. The goal of the analysis is given a climate and
> > > > orientation for a typical vertical facade to understand the absorbed
> > > > solar radiation and Temperature and Humidity at the interior boundary
> > > > of the facade within closest zone. So single zone model.
> > > >
> > > > My question are these:
> > > > 1) Im interested to analyse the performance of the only exterior
> > > > facade and not the whole building. Is that possible to model in E+
> > > > ..probably modelling the building as a single box volume and changing
> > > > the parameter/geometry for the facade type I want to analyse?
> > > >
> > > > 2) The geometry input. The scope of the research is actually to
> > > > understand impact of different shapes/self shading combined with some
> > > > metrial properties. So standard flat plane doesnt wotrk in this case.
> > > > Take a look at this pdf
> > > > http://dl.dropbox.com/u/5212179/sampleEnvelope.pdf .
> > > > So typically I will have a frame structure (in this case hexagonal)
> > > > and some glazing. Let say I find an automated process to convert the
> > > > 3d model in E+ coordinates. The problem Im facing is how do you treat
> > > > the frame? as a wall?
> > > > and do I need to assign orientation for each of the frame panel ?
> > > > I need to model the frame because for my assumptions it is important
> > > > to understand the shading potential of facades I am going to study.
> > > >
> > > > Thank you very much in advance for the help and advice.
> > > >
> > > > Best,
> > > >
> > > > Giulio
> > > >
> > > >
> > >
> >
> >
>