Furnace Simulation

[gforney]

FDS is primarily to simulate natural fires. Can a feature be added so that 
we can conduct furnace simulations?

Original issue reported on code.google.com by khalid.moinuddin@vu.edu.au on 2007-05-29 01:14:07

[gforney]

What do you feel is missing to be able to simulate a furnace?

Original issue reported on code.google.com by drjfloyd on 2007-05-29 12:42:51

• Labels added: Type-Enhancement
• Labels removed: Type-Defect

[gforney]

Hi Jason,
I am not sure what is missing to be able to simulate a furnace. Can we have the 
feature to input time-gas temperature curve (similar to time-HRR curve in FDS) which

will be the same for furnace walls as well (so that no heat loss to the furnace wall

takes place). At the same time this time-gas temperature curve should not be 
applicable to the voids within a test specimen (such as plasterboard encapsulated 
steel I beam).

More complicated way will be adding the ability to simulate pre-mixed combustion.

Last year, I proposed to Kevin to add the ability to simulate furnace tests. He 
replied to me that it was possible by adding some features ( sorry, I cann't 
remember what it was and I lost his email reply).

with regards

Khalid 

Original issue reported on code.google.com by khalid.moinuddin@vu.edu.au on 2007-05-30 00:56:31

[gforney]

You can specify time varying inlet boundary conditions of Z_2 and temperature (Z_2 =

products of combustion).  Premixed could also be done with the finite rate 
combustion model.  Surfaces can already be declared adiabatic. 

Given that we are only doing 1D heat conduction, is FDS the appropriate tool for a

detailed analysis of a plasterboard encapsulated steel I beam?  We may not do great

in terms of temperatures at the web-flange connection.

Original issue reported on code.google.com by drjfloyd on 2007-05-30 12:25:32

[gforney]

I will try your suggestion though I am bit confused. About I beam modelling, can I 
draw your attention to the following paper:

Liang H, Welsh S. Development of an engineering methodology for thermal analysis of

protected structural members in fire, Proceedings of Third International Conference

on Steel and Composite Structures, Manchester, UK, 2007 (to be appeared). 
(https://www.era.lib.ed.ac.uk/bitstream/1842/1531/1/Liang_ICSCS07.pdf)

Regards

Khalid

Original issue reported on code.google.com by khalid.moinuddin@vu.edu.au on 2007-05-31 06:34:00

[gforney]

Z_2 = burned fuel.  Say for example that 1 kg of the fuel being used requires 9 kg 
of air per kg of fuel.  Then at stoichiometry you would have 1 kg of fuel in a total

mass of 10 kg and if that was fully combusted Z_2 = 1/(9+1) = 0.1.  Lets also say 
that c_p of the gas mixture is 1 kJ/kg-K and delta_h is 10000 kJ/kg.  We can add 
that amount of energy by saying that the 10 kg/s of incomming mass (of which 10 % is

fuel) has a temperature of TMPA+1000 (10000 kJ/kg fuel* 1 kg fuel / (1 kJ/kg-K) / 
(10 kg)=1000 K).

Based on your comment about voids I was envisioning the following:

CCCCCCCCC
WSSSSSSSW
W   S   W
W   S   W
W   S   W
WSSSSSSSW
WWWWWWWWW

Where S is steel, W is wall board, and c is whatever the ceiling is made of.  FDS 
does 1D heat trasnfer so:

WS  
WW  

Here we cannot get the combined contributions to the beam from both sides of 
corner.  The "S" is treated as two seperate 1D calcs.

 S
SSS
WWW  

Same here.  The S at the junction of the flange will be treated as three seperate 
calcs: one for the W, one for the S to the left, and one to the right.

Add to that the uncertainties added by the additional convection calc within the 
void.  You may be better off (depeding on what question you are ultimately trying to

answer) of taking the effective adiabatic temperature to the wallboard and using it

as an input to a 22 or 3D thermal-structural code.

Original issue reported on code.google.com by drjfloyd on 2007-05-31 15:10:21

[gforney]

Since there has been no futher traffic in a while on this issue, I am going to 
assume that the current capabilities in FDS are sufficient for your problem and will

close this issue.  If that is not the case it can be reopened it at a later date.

Original issue reported on code.google.com by drjfloyd on 2007-06-06 19:58:30