[esp-r] Re: ribber outer wall

[Jon Hand]

A question about walls that are ribbed...

::how can a ribbled outer wall be modelled? We have a student project where the volumes to be modelled are regular ::containers used for bulk transport on boats. These have ribbled outer walls...

If the wall has a constant thickness (even though the shape is complex)  then the nature of
the wall is

a) some % of the surface area is parallel with the length
of the container and other portions are at an angle (so solar
radiation and convection might differ at different facets
of the wall.

b) the overall surface area is much greater than a flat surface.

One technique would be to take a literal approach - create the
container out of LOTS of surfaces.  This allows issues a & b
to be dealt with explicitly.  Essentially there are a few polygon
shapes that need to be created and then the use of copy and
transform can be used to gradually build up the container. Plan
carefully - each zone can only have 62 surfaces so there will
need to be several zones.

The second technique is to be slightly abstract.  Lump several
ribs together so that the orientations and surface area are
correct but there are fewer ribs.

A third technique is to use a 2D conduction application to model
the walls as an extrusion and then compose one or more
constructions to represent the heat transfer characteristics
of such a wall.

If the wall has a non-constant thickness - say if there is an
inner flat plate for a floor then this probably requires two
types of constructions - one to represent those parts where
the air gap between the floor and the rib is deep and one where
the air gap is shallow.  Such approaches are approximations
because the problem is really 3D rather than 1D.

-Jon Hand