Comparitive Study Between Different Fire Softwares

[Mohamed]

Dear All

Do anyone knows the difference or make a comparitive study
between FDS and other fire software (e.g. Fluent, Jasmine,SOFIE).

Regards,
Mohamed

[Fabian Braennstroem]

Hi,

Mohamed wrote:
> Dear All
>
> Do anyone knows the difference or make a comparitive study
> between FDS and other fire software (e.g. Fluent, Jasmine,SOFIE).

I would like to add OpenFOAM, which is a general purpose Open-Source CFD
program with some adjustments fire simulation:
http://www.openfoamworkshop.org/2009/4th_Workshop/1_Automotive_and_Combustion/OFW4_2009_Wang_slides.pdf

Best Regards!
Fabian

[robert.meroney]

1.The best initial comparison of FDS and other software would be the
validation exercise performed by the US NUREG-1824 and EPRI 101 1999
titled Verification and Validation of Selected Fire Models for Nuclear
Power Plant Applications Vol. 1-7, 2006. In this report is found an
extensive comparison of FDT-S & Five-REV1 (library approach), CFAST &
MAGIC (zone models) and FDS (field or finite volume model). A large
number of experimental data were used to do comparative evaluation of
these models. Table 3-1 in Vol 1 provides a succinct summary of their
results. These reports are available on the NIST FDS web site as well
as elsewhere.

Here are a few extra references that compare FDS and other cfd
software:

1.

[robert.meroney]

AS mentioned in the earlier post before I accidentally hit the send
key here are some additional references. All (and other) references
can be found and downloaded at http://www.engr.colostate.edu/~meroney/index.html

1. Meroney, R.N. VIRTUAL FIRES via COMPUTERS, 1st American
Association for Wind Engineering Workshop, Session 2: CWE and
Analytical, 21-22 August 2008, Vail, CO, 8 pp (CEP08-09-1)

This paper contains a discussion about the relative values of CFAST
and FDS and how they might be used together. It tries to answer the
question when is a fast approximate model appropriate, and when should
one take the time to predict fire and flow details?

2. Chang, Cheng-hsin, Meroney, R.N. and Banks, D. COMPUTATIONAL FLUID
DYNAMICS SIMULATION OF THE PROGRESS OF FIRE SMOKE IN LARGE SPACE,
BUILDING ATRIA, Tamkang Journal of Science and Engineering, Vol. 6,
NO. 3, (2003), pp. 151-157. (CEP03-04-7)

This paper discusses the results of calculations for an example
building atrium based on zone (ASMET) and field (FLUENT and FDS) based
models.

3. Chang, C.H. and Meroney, R.N. CONCENTRATION DISTRIBUTIONS FROM A
POINT SOURCE WITHIN URBAN STREET CANYONS: WIND TUNNEL AND
COMPUTATIONAL DATA, J. of Wind Engineering and Industrial
Aerodynamics, Vol. 91, No. 9, pp. 1141-1154, September 2003.
(CEP03-04-CHC-RNM-4.)

This paper uses FDS and FLUENT to predict flow and concentrations in a
generic urban city made up of arrangements of blocks and street
canyons. The results were compared with detailed measurements from a
wind tunnel study.

On Jun 26, 1:25 pm, Mohamed <mohamedal...@gmail.com> wrote:

[Mohamed]

Dear Sir

Could you please take a look to that report published by lund
university named " Risks in using CFD-codes for analytical fire-based
design in buildings with a focus on FDS:s handling of under-ventilated
fires" in the following link:

http://www.brand.lth.se/publikationer/examensarbeten/

it has some doubts in the fds operation in the under ventilated fires.

Regards,
Mohamed

On Jun 28, 6:01 pm, "robert.meroney" <robert.mero...@colostate.edu>
wrote:

> AS mentioned in the earlier post before I accidentally hit the send
> key here are some additional references.  All (and other) references

> can be found and downloaded athttp://www.engr.colostate.edu/~meroney/index.html

> > Mohamed- Hide quoted text -
>
> - Show quoted text -

[FPESC...@aol.com]

In a message dated 6/28/2009 2:45:20 P.M. Central Daylight Time, mohame...@gmail.com writes:

   Could you please take a look to that report published by lund
university named " Risks in using CFD-codes for analytical fire-based
design in buildings with a focus on FDS:s handling of under-ventilated
fires" in the following link:

http://www.brand.lth.se/publikationer/examensarbeten/

it has some doubts in the fds operation in the under ventilated fires.

Regards,
Mohamed

Mohamed, thank you for the link.  I'm reading the paper now.  Based upon the introduction to the paper, it looks like a little more support for the "Wild, Wild West" hypotheses regarding the FDS.

 

Richard Schulte

Schulte & Associates

Building Code Consultants

Chicago/New Orleans

fpesc...@aol.com

 

---

Make your summer sizzle with fast and easy recipes for the grill.

[FPESC...@aol.com]

"The computer program that is validated in this report, FDS (McGrattan 2008), is to some extent [emphasis added] verified and validated (see section 4.1 for definition) by its developer (National Institute of Standards and Technology – NIST). Despite this, there are many functions and models in the program that are linked with uncertainty and error.  [Emphasis added.]  In FDS one of these models, the mixture fraction combustion model, is linked with both uncertainties and errors when used for under-ventilated fires."

 

"When it comes to risks related to the use of CFD the previous work really consist of a wide spread theory that it is the user handling that stands for the biggest risk, see for example “An introduction to CFD” (Rubini 2008)."

 

"Risks in using CFD-codes for analytical fire-based design in buildings with a focus on FDS:s handling of under-ventilated fires"

Anders Björklund
Department of Fire Safety

Lund University

Sweden

 

Certainly, Dr. Craig Beyler's use of CFD (LES3D) as a plaintiff's expert in the McFrugal's Warehouse litigation is an excellent example of the second point above.  Below are a few more excerpts from Dr. Beyler's deposition in June, 1999 which demonstrate the point:

 

[Deposition] Page 1018

 

Q. Okay.  Is it possible to determine that?

A. The model could be run with different [top of storage to sprinkler] clearances to establish based on the model what the results, at what clearance problems would arise.

 

Q. As I understand your modeling results, we know that under the scenario you modeled for ten foot clearance, it would control the fire.  Correct?

 

A. Correct.

 

Q. We also know that based upon your model and your configuration, that it would not control with a 50 foot clearance.

 

A. That’s what the model indicated.

 

Q. But we don’t know where in between those two clearance numbers you draw the line that says it will control versus it won’t control?  Is that right?

 

A. Well, we have other hints, but nothing directly from the model.

 

Q. What are the hints?

 

A. Well, we have - - we have testing that was done, and I think we talked about this the last time, in support of [NFPA] 231C where, you know, there were in-racks and the first two tiers were controlled by the in-racks, and we simply had a ten foot high, you know, rack storage configuration with a 30 foot clearance and - -

 

Q. Wouldn’t the density make a difference whether a fire is controlled or not?

A. Yes.  My recollection - - yeah.  It certainly does.  And my recollection was in that test, and if we want to look at it more closely, obviously we should look it up in the appendix, that those were run at design densities that it would have been expected to work at, quote, ordinary clearances.  I have to check that.

 

Q. Weren’t the [sprinkler] densities in those tests run at a lower density that what we had here?

 

A. I would expect so.

[Deposition] Page 1022

 

Q. 0.13 [gpm/SF].  In connection with your investigation, did you determine what density would be required to control the fire with a 50 foot clearance involving two stacked pallets high?

 

A. No.

 

Q. Is that possible to calculate?

 

A. Presumably you could change the parameters of the calculation to seek out the density at which control was achieved.

 

Q. Based upon your model, does the heat release under your scenario with the 50 foot clearance and a 0.45 [gpm/SF] density increase over time?

A. The heat release increases over time.

 

Q. Does it ever decrease at any point?

 

A. Not during the simulation as I - - Well, let me check.  I remember it’s - - it stopped growing so fast, but I’m - - at the ten minute mark its rate of growth is slowing, but it is continuing to grow.

 

Q. Would you point out where you are referring to in the report for that information.

 

A. Figure 17 on page 68.

 

Q. If you carry those lines out, do you ever reach a point where the heat release level off?

 

A. Do I believe it would if the calculation had continued?  Is that the question?

 

Q. Yes.  That’s correct.  All right.  Go ahead.

 

A. At the ten minute mark on the simulation, the number of heads activated is right about at the design area, 2,500 square feet.  We’ve got - - if you look at Figure 18, we have I think it’s 33 heads operating at that time.  And that’s basically 2,500 square foot of area where sprinklers are operating, which is the design area.  And it would be my judgment since the fire is continuing to grow, though at a reduced rate, that I would expect the fire not to be - - if you don’t contain it by the time you reached the design area, you’re not - - you won’t.

 

Q. Right.  But the amount of heat release gradually levels off at some point, does it not, based upon the amount of fuel that you have available to burn?  I mean it’s not going - -

 

A. Yeah.  If we’re talking about will you ultimately get to a peak, I mean all fires, you know, die out at some point, yes.  I mean they grow to a size, they burn, they die out.

 

Q. Okay.

 

A. And I guess all I’m saying here is that since we have a fire that continues to grow at a reduced rate, certainly not continuing to escalate, it would be my - - my judgment that since the fire is still growing at the design area, you know, the number of heads consistent with the design area, that this fire won’t be controlled and the - - you know, we’ll get right over, you know, a large area such that control will not have occurred.

 

Q. Would you agree with me that under the scenario you modeled with two pallet high that the sprinkler system does have some effect on the increase in the heat release over time?

 

A. Yes, it does.

 

Q. And would you agree with me that the dynamics of the fire that occurs as you modeled it with the two pallet high is significantly different than the dynamics for the fire that you modeled with three racks high?

 

A. Yes, there’s “significantly” in two regards.  One is that since the three tier represents a more severe challenge, its heat release rates are inherently larger.

 

Q. Okay.

 

A. And secondly, you know, it is the case that the pallets stacked, one, you see a change - - you see a reduction in the rate of growth which is not at all apparent in the three tier Class II in Figure 10.

 

Not a word about the capability of CFD to actually make the predictions being discussed by Dr. Beyler in his sworn testimony.  "Houston, we have a problem."

 

Richard Schulte

Schulte & Associates

Building Code Consultants

Chicago/New Orleans

 

 

 

 

In a message dated 6/28/2009 2:45:20 P.M. Central Daylight Time, mohame...@gmail.com writes:

Could you please take a look to that report published by lund
university named " Risks in using CFD-codes for analytical fire-based
design in buildings with a focus on FDS:s handling of under-ventilated
fires" in the following link:

http://www.brand.lth.se/publikationer/examensarbeten/

it has some doubts in the fds operation in the under ventilated fires.

Regards,
Mohamed

[Laika]

Dear Fabian,

that's interesting news. I use Fluent and FDS for fire modelling, but
OpenFOAM for other applications. Do you have more information on that
fireFOAM solver for the OpenFOAM toolbox? Have you tried it?

thanks!

Laika,
still orbiting

On 27 jun, 13:48, Fabian Braennstroem <f.braennstr...@gmx.de> wrote:
> Hi,
>
> Mohamed wrote:
> > Dear All
>
> >        Do anyone knows the difference or make a comparitive study
> > between FDS and other fire software (e.g. Fluent, Jasmine,SOFIE).
>
> I would like to add OpenFOAM, which is a general purpose Open-Source CFD

> program with some adjustments fire simulation:http://www.openfoamworkshop.org/2009/4th_Workshop/1_Automotive_and_Co...
>
> Best Regards!
> Fabian

[Guillermo Rein]

Dear Mohamed,


Regarding your original request for inter-comparison of fire CFD
models: in the 9th IAFSS (Karlsruhe 2008) there was a very interesting
poster comparing a posteriori different codes to simulate a small fire
in an enclosure. The work was lead by Lund University, Sweden. I think
a follow up was presented at the Fire and Materials 2009 conference. I
am not sure at this point (I do not have at hand copy of the
proceedings), but the paper could had been:

- Determination of Uncertainty of Different CFD Codes by Means of
Comparison with Experimental Fire Scenarios
Patrick van Hees, G Holmstedt, Z Yan, Lund University, S Bengtsson, B
Hägglund, T Dittmer, Brandskyddslaget and P Blomqvist, H Tuovinen, A
Lönnermark, SP Technical Research Inst, Sweden
http://www.intercomm.dial.pipex.com/html/events/fm09d.htm

I would suggest to contact Patrick van Hees.

Cheers
G.

PD: Regarding the problems of FDS with underventilated/post-flashover
fires, I note this quote from Chp 11 in the book "The Dalmarnock Fire
Tests: Experiments and Modelling" (http://www.era.lib.ed.ac.uk/handle/
1842/2404).

"The different fire phases required different modelling approaches. In
the growth phase,
the fire is better simulated as a prescribed HRR. But during
ventilation-control phases, it
is more convenient to predict the HRR, at least partially, because an
imposed HRR tends
to err the flame locus. Imposing the HRR decouples the production of
pyrolyzates from
the flame heat feedback. The decoupling together with the use of a
mixture fraction
model tends to locate flames at the vents during ventilation-control
conditions. In turn,
this leads to temperature and heat flux underpredictions away from the
vents and
overpredictions near them. The flame-locus defect leads to erroneous
fire predictions
during post-flashover in large compartments, increases with the volume
of the enclosure
and decreases with the vent size."

[Mohamed]

Thanks all.....for your answers......thanks.............

On Jun 29, 1:41 am, Guillermo Rein <rei...@gmail.com> wrote:
> Dear Mohamed,
>
> Regarding your original request for inter-comparison of fire CFD
> models: in the 9th IAFSS (Karlsruhe 2008) there was a very interesting
> poster comparing a posteriori different codes to simulate a small fire
> in an enclosure. The work was lead by Lund University, Sweden. I think
> a follow up was presented at the Fire and Materials 2009 conference. I
> am not sure at this point (I do not have at hand copy of the
> proceedings), but the paper could had been:
>
> - Determination of Uncertainty of Different CFD Codes by Means of
> Comparison with Experimental Fire Scenarios
> Patrick van Hees, G Holmstedt, Z Yan, Lund University, S Bengtsson, B
> Hägglund, T Dittmer, Brandskyddslaget and P Blomqvist, H Tuovinen, A

> Lönnermark, SP Technical Research Inst, Swedenhttp://www.intercomm.dial.pipex.com/html/events/fm09d.htm