Material thickness and burn away

[Wally]

Hi all,

I want to model a fire burning through a timber floor which is
composed of three materials (20mm hardwood, 10mm masonite and 2mm
carpet).

I have a few questions regarding how to go about modelling this:
1) Layers:
In FDS materials can be specified as layers composed of different
components. Is there any difference between specifying my floor as
three layers or one layer with three components?

2) Thickness:
As I understand it, the thickness of the OBST is not relevant to the 1-
D heat transfer calculation and the thickness in the SURF line is
used, however for burn away the OBST thickness matters.
Does the thickness of the OBST have to be the actual thickness (in my
case 20+10+2=32mm), and my mesh has to been fine enough to model such
a thickness?

3) Applying SURF to the OBST:
User Guide Section 8.4.5: “The SURF should be applied to the entire
object, not just a face of the object because it is unclear how to
handle edges of solid obstructions that have different SURF_IDs on
different faces”
Does that simply mean I should use SURF_ID and not SURF_IDs or
SURF_ID6 when assigning the SURF to the OBST?

Thanks

[Kevin]

On Mar 2, 7:37 am, Wally <welso...@gmail.com> wrote:
> Hi all,
>
> I want to model a fire burning through a timber floor which is
> composed of three materials (20mm hardwood, 10mm masonite and 2mm
> carpet).
>
> I have a few questions regarding how to go about modelling this:
> 1) Layers:
> In FDS materials can be specified as layers composed of different
> components.  Is there any difference between specifying my floor as
> three layers or one layer with three components?

Yes. These are completely different ways of specifying the floor. You
want three layers.

>
> 2) Thickness:
> As I understand it, the thickness of the OBST is not relevant to the 1-
> D heat transfer calculation and the thickness in the SURF line is
> used, however for burn away the OBST thickness matters.
> Does the thickness of the OBST have to be the actual thickness (in my
> case 20+10+2=32mm), and my mesh has to been fine enough to model such
> a thickness?

The mesh doesn't matter. The obstruction should disappear when it has
run out of fuel. I suggest that you set up a very simple test case
where you have you put a hot object just off the floor and let it burn
out. You may even want to reduce the amount of fuel to make it go
faster. This is a good way to check that you have specified everything
properly.

>
> 3) Applying SURF to the OBST:
> User Guide Section 8.4.5: “The SURF should be applied to the entire
> object, not just a face of the object because it is unclear how to
> handle edges of solid obstructions that have different SURF_IDs on
> different faces”
> Does that simply mean I should use SURF_ID and not SURF_IDs or
> SURF_ID6 when assigning the SURF to the OBST?

Your case is tricky because of the layering. You need two SURF's --
one for the top of the floor and one for the bottom, with the layers
in reverse order and BACKING='EXPOSED'.

Word of warning -- this is a fairly difficult thing to do in FDS. Try
a very simple test case first so that you do not waste days waiting
for your floor to disappear. If you have trouble, submit the simple
case to the Issue Tracker.

[Wally]

Thanks for the response Kevin.

I'm a bit confused Re question (2). You said that the mesh doesn't
matter. From the guide: "a solid object with this SURF_ID disappears
from a
calculation as the mass of each of its mesh cells are consumed. The
mass of each mesh cell is the volume of the mesh cell multiplied by
the DENSITY of the materials making up the obstruction." This would
suggest that the mesh does matter? what am i missing here?

If an obstruction thickness is smaller than the mesh cell size , would
burn_away work?

If the thickness of an obstruction is smaller than the mesh cell size,
it would appear as a zero thickness obstruction in SMV. Does that mean
FDS treats it as a zero thickness (hence it has no volume?) or is it
simply a visual thing in SMV and the volume of the obstruction is
determined by the OBST XB coordinates?

[Hostikka Simo]

You are reading some outdated version of the user's guide. Please
update, and see section 8.4.5.

In the current version of FDS, the obst mass is determined by surface
density of the SURF type. And this can be overriden by BULK_DENSITY
parameter of the OBST.

Simo

[ESRI]

Hi Waled, I am also facing the exact problem and doubts. Can you please explain what you learnt after this ?

Thanks