• Aucun résultat trouvé

Development of an FDS-SAFIR interface

N/A
N/A
Protected

Academic year: 2021

Partager "Development of an FDS-SAFIR interface"

Copied!
15
0
0

Texte intégral

(1)

SECURE WITH STEEL

Development of an FDS – SAFIR Interface

Nicola Tondini Jean-Marc Franssen

(2)

3 problems

have to be solved. Each of them is governed by

different equations.

1.

Fire development

=> Temperatures and flows in the

compartment. It requires a 3D model.

2.

Thermal response

=> Temperatures in the structural

elements. A 2D model is generally sufficient.

Elements across the compartment.

Elements on the boundaries of the compartment.

(3)

WEAK COUPLING STRATEGY

NOTE: Interaction 2 to 1

• It is complete if step 2 is performed by the CFD

• It is limited to the boundary of the compartment if step 2 is performed by the FE.

1. Compartment

2. Temperatures in elements

(4)

WEAK COUPLING STRATEGY ESSENTIAL FEATURES

1) The structural elements are not present in the CFD analysis, except for the boundaries of the compartment, possibly in an approximated manner.

2) The temperatures in the boundaries of the compartment are calculated by the CFD software.

3) The temperatures in load bearing elements are calculated by the FE software.

(5)

WEAK COUPLING STRATEGY TRANSFER FILE INFORMATION: •Temperatures •Convection factors •Radiant intensities FDS • Simulation of the fire development in the compartment SAFIR •Thermal response •Mechanical response

(6)

WEAK COUPLING STRATEGY ADVANTAGES

• The CFD calculation can be performed before and separately of the FE

analysis.

• Can be used with different combinations of CFD and FE software.

• Less demanding in terms of CPU and hardware.

If p structures must be evaluated under q fire scenarios, only q CFD

analyses must be performed, compared to p∙q coupled analyses in a two-way coupling approach.

ISSUES

• How to judge whether the interactions between 1 and 3 are really

(7)

INTERPOLATION

A three level interpolation must be performed.

1. A Cartesian interpolation in space to have the information at the locations that are relevant for the structure.

y

x

Structural point 2 Structural point 1 A B Points of the CFD domain Beam element

(8)

INTERPOLATION

2. An Interpolation in time to have the information at the time step that are relevant for the structure.

3. An Interpolation in spherical coordinates to have the radiant intensities in the appropriate directions on the surface of the structure.

INTEGRATION OF RADIANT INTENSITIES TO GET THE IMPINGING FLUX

View angle

CONCAVE ELEMENT

Integration on “clear sky”

CONVEX ELEMENT

(9)

Standard fire ISO 834: spherical integration vs. view factors EXAMPLES: Uniform radiation

X

Y

Z

Diamond 2011.a.1 for SAFIR

FILE: Massivi ty_CFD NODES: 75 ELEMENT S: 62 SOLIDS PLOT TEMPERATURE PLOT TI ME: 1200 sec 318.60 291.96 265.33 238.69 212.05 185.41 158.78 132.14 105.50 X Y Z

Diamond 2011.a.1 for SAFIR

FILE: Massivi ty_FISO NODES: 75 ELEMENT S: 62 SOLIDS PLOT TEMPERATURE PLOT TI ME: 1200 sec 319.80 293.00 266.20 239.40 212.60 185.80 159.00 132.20 105.40 50 100 150 200 250 300 Te m p e ra tu re C ] LARGE GROOVE Node 57 CFD Node 60 CFD Node 65 CFD Node 67 CFD Node 57 FISO em*vw Node 60 FISO em*vw Node 65 FISO em*vw Node 67 FISO em*vw

FDS-SAFIR coupling Modified view factors

50 100 150 200 250 300 Te m p e ra tu re [ °C ] DEEP GROOVE Node 4 CFD Node 6 CFD Node 7 CFD Node 10 CFD Node 4 FISO em*vw Node 6 FISO em*vw Node 7 FISO em*vw Node 10 FISO em*vw

Large groove

(10)

Compartment with one hot wall and a HE400M section EXAMPLES: Non-uniform radiation

H o t w a ll T = 1 2 0 0 ° C Cold wall T= 20 °C Cold wall T= 20 °C C o ld w a ll T = 2 0 ° C 2 1 1

(11)

Temperature development in the section EXAMPLES: Non-uniform radiation

(12)

EXAMPLES: Compartment pool fire (FDS vs. HASEMI)

1

POOL FIRE (Impacting the ceiling) D = 1.5 m q = 1727 kW/m² Q = 3051 kW 3 3 0,5 Beam 3,5 1,5 z y

(13)

EXAMPLES: Compartment pool fire (FDS vs. HASEMI) Simulation with FDS

(14)

EXAMPLES: Compartment pool fire (FDS vs. HASEMI) 10000 20000 30000 40000 50000 60000 70000 F lu x [ W /m 2 ]

Impinging flux received by the ceiling with normal -z @ the position of the beam (2.5,1.5,3.0)

FDS Hasemi Avg. FDS

(15)

X

Y

Z

Diamond 2011.a.1 for SAFIR

FILE: Squ are30x30_CFD NODES: 289 ELEMENT S: 256 SOLIDS PLOT TEMPERATURE PLOT TIME: 18 00 sec >T max 600.00 542.00 484.00 426.00 368.00 310.00 252.00 194.00 136.00 78.00 20.00 <T min X Y Z

Diamond 2011.a.1 for SAFIR

FILE: Square 30x30_HASM NODES: 289 ELEMENT S: 256 SOLIDS PLOT TEMPERATURE PLOT TIME: 18 00 sec >T max 600.00 542.00 484.00 426.00 368.00 310.00 252.00 194.00 136.00 78.00 20.00 <T min

EXAMPLES: Compartment pool fire (FDS vs. HASEMI)

FDS @ t = 1800 s HASEMI @ t = 1800 s Node 1 Node 1 ADIABATIC CFD ADIABATIC HASEMI 200 300 400 500 600 700 Te m p e ra tu re [ °C ] Node 1 FDS HASEMI

Références

Documents relatifs

In this way, the definition of the direct kinematic model of the SCARA robot requires the identification of 4 parameters, for each rotational joint, which describe the

In this paper, we presented the principles of a multi-level isolation solution allowing Android end users to define fine grained security policy for installed applications..

Since an intensity is estimated as the difference between counts on peak aral counts on background, and each count is subject to statistical (Poisson) fluctuations, a reflexion

Based on the presented groups of indicators, the authors developed a corporate growth strategy model, evaluated the investment attractiveness of the growth

In order to assist in the discussion of the Strategy for the African Region in ~he International Development Strategy for the Third Development Decade and other relevant economic

The hypothesis is as follows: By putting factors and their dynamics into a multinomial logit model using the crisis definition of Bussière and Fratzscher, one may assume that

Numerical investigation has been realised for CFSHS columns of existing range (reinforcement from 3% to 6% of reinforcement and circular containing addition I-section) and

De plus, dans chaque site, 5 placeaux ont été installés en dehors des clôtures, de façon non systématique et uniquement en présence de régénération