how to generate cylinder filled with insulating materia

[s_desai]

We want to generate the cylinder filled with insulating material.We
have generated the hollow cylinder but how to fill hollow portion. pls
help.

Regards..
Saumil

[Jonna]

I don`t think there is an easy solution to this.. Perhaps tiny blocks
of &obst with the same surface directly next to each other?
Cover a horisontal plane in the cylinder with blocks first, and then
copy with a shift in the Z-axis to fill rest of hole automatically
(you can use a script, a spreadsheet or Pyrosim for this function)

Regards

Jonna

[Kris]

What are you trying to do with it? You could try making it a single
solid cylinder and defining the different material layers for the
outside and the insulated inside (&MATL - see Chapter 8.4 of the user
guide).

Cheers,
Kris

[Tony Bova]

I've posted a Matlab m-file that allows you to create a text file
describing a solid, hollow or 'hole' cylinder. You can just copy and
paste the text into your FDS input file. It's named fds_cylinder_v2.m
and it's at

http://sites.google.com/site/bovaprojects/

You create the cylinder by defining it's center point in your domain.
If you choose to make a hollow cylinder, it will also ask for an inner
radius. Cylinders are created essentially by rotating obstructions (or
holes) around a central axis. The cell size you enter (the program
assumes you're using cubic cells) determines the number of rotation
steps and therefore the number of OBST in the final text file. You may
want to enter a smaller cell size than you're actually using just to
make sure the cylinder is defined through 360 deg.) All units are in
meters, and the script prompts you for radii, not diameters.

At the moment, the long axis of the cylinder must be parallel to one
of the cartesion axes (the script will prompt you for this
information). I may get around to making a slanted or curving cylinder
some day.

You'll probably want to change properties in the m-file (the SURF
property is currently INERT) as well as the output file name
(currently cylinder.obst). You could define a hollow cylinder of some
material first (or use the one you have), and then a second solid
cylinder made of insulating material that has the same dimensions and
location as the hole. Feel free to tweak and improve the code as you
need. If you don't have Matlab, the script may run in Octave. I
haven't tried that yet--it may need some tweaking.

Feel free to contact me if you have questions.

~tony

> > Saumil- Hide quoted text -
>
> - Show quoted text -

[dr_jfloyd]

FDS does not conduct between different OBST lines. i.e.

&SURF ID='foo1',MATL_ID='foo1'.../
&SURF ID='foo2',MATL_ID='foo2'.../
&OBST XB=0,1,0,1,0,1,SURF_ID6='foo1',.../
&OBST XB=1,2,0,1,0,1,SURF_ID6='foo2',.../

is not the same as

&SURF ID='foo3',MATL_ID='foo1','foo2'/
&OBST XB=0,2,0,1,0,1,SURF_ID6='foo3',.../

In the first case the OBST from x=0 to 1 does not transfer any heat to
the OBST from x=1 to 2

If you have a multiple layer surface, it needs to be defined as a
single OBST with a multiple layer SURF.

[Tony Bova]

Hello-

Good point. I wasn't thinking about conduction when I sent that
response. You could create a surf line with two layers and thicknesses
that would apply to the surface cells of a solid cylinder (pp 58-59 in
the user guide).

Depending on the size of the cylinder and the depth of the temperature
or flux profile you wish to model (assuming that is what you want to
do), your results may be very rough regardless of how the cylinder is
created. As I understand it, heat flux is one-dimensional in FDS (no
lateral fluxes) so you could model the response to heat flux only at,
or just below, the surface of a cylinder. As you move radially inward,
lateral heat transfer becomes much more important and the temp profile
will depart from that of a semi-infinite slab (Dr. Floyd - is that
basically correct?).

~tony

[dr_jfloyd]

Yes, there would be an error as one moved inward. The SURF line has
the keyword GEOMETRY and one can set that to CYLINDRICAL or SPHERICAL;
however, at the moment that option assumes that the OBST is filled to
radius=0 (i.e. you can't do a thin cylinder wall coupling to gas phase
cells). So if the insulation mentioned above completely fills the
cylinder this option would work. If it doesn't, and there is an air
gap in the inside that is being modeled, then there will be an error.
If the overall thickness of the cylinder wall and insulation is thin
compared to the overall cylinder radius, the error will be small.

[dr_jfloyd]

Another observation is that one can be clever and still get an OK
answer for a cylinder where the wall thickness is large compared to
the radius. The error in the 1D Cartesian solution is because as one
moves inward, the smaller radius means less mass is present along the
circle defined by the radius. For a thick walled cylinder, one could
define it as many smaller layers. For each layer, the material would
have the true value of the conductivity, but the specific heat (or the
density) would be reduced to account for the decrease in mass. For
example if I had a 1 m radius cylinder and it had a 0.5 m wall
thickness, I would reduce the specific heat for the innermost layer by
1/2 as a radius of 0.5 m has half the circumference (and hence half
the mass) of a radius at 1 m.

[saumil desai]

Thanks for the mat file and valuable inputs .I was unaware that conduction can not be possible between different 2 surf.

I think I have to put solid cylinder in fire and find out boundry heat flux which will be BC for simulation in CFX.

FYIP:We are studying structural safety when subjected to open fire.

Thanks to Mr Tony and Dr Floyd

Regards.

Saumil

--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To post to this group, send email to fds...@googlegroups.com.
To unsubscribe from this group, send email to fds-smv+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/fds-smv?hl=en.

--
Regards...
Saumil Desai