Tensile membrane action in fire of composite slabs with ...fire-research.group.shef.ac.uk/steelinfire/downloads/OV_2012.pdfArcelorMittal Long Carbon Research & Development Tensile

ArcelorMittal Long Carbon

Research & Development

Tensile membrane action in fire of composite

slabs with cellular steel beams

Prof. Olivier Vassart


Research & Development 2

FICEB+ - Partnership



Additional fire resistance through 3D

membrane effect - Bailey's methods extended

to Long span Cellular Beams



Behaviour of slab and beam during a fire

Composite slab is one-way

spanning onto unprotected beam




Plastic hinge forms in unprotected beam

fan yield line forms in slab




Strength of composite beam

continues to reduce

resulting in the yield pattern shown




With increasing loss of strength for the beam

the slab behaviour tends towards a yield line pattern

given for the slab acting without the beam



Load

capacity

(bending &

membrane

action)

Temperature

Strength of slab based on lower

bound assuming no beam strength



X Y

Z

Diamond 2008 for SAFIR

FILE: RealEP

NODES: 658

BEAMS: 174

TRUSSES: 0

SHELLS: 432

SOILS: 0

N1-N2 MEMBRANE FORCE PLOT

TIME: 2540.639 sec

- Membrane Force

+ Membrane Force

Tension

Compression



Horizontal movement

Fracture

Compression failure Compression failure



15 m 9 m

3D fire Test



Bearing structure



Steel sheeting and reinforcement



Design Loads

Description Characteristics

kN/m2 Load Factor

Design Load

kN/m2

Partition 1.0 1.0 1.0

Services &

Finishes 0.5 1.0 0.5

Live Load 3.5 0.5 1.75

Total 3.25

The loads used within the structure are the same as those

which are commonly used in the design of office buildings.



Fire Loads Toome Test

Crib Heat Release Rate

of 1M Wide by 1M Long by 0.5M High Wooden Crib consisting of 44mm2 square section sticks

-200

0

200

400

600

800

1000

1200

1400

1 75 149 223 297 371 445 519 593 667 741 815 889 963 1037 1111 1185 1259 1333 1407 1481 1555 1629 1703 1777 1851 1925 1999 2073 2147 2221 2295

Time

Kilo

watts (

kW

)

Assuming the design for an office, the fire load density would

be 511 MJ/m2 according to Table E.2 of EN 1991-1-2.

For the test, a fire load of 40 kg of wood/m2 was used, which

corresponds finally to a fire load of about 700 MJ/m².



Fire Loads



Fire protection





Shape of the beam after the fire



Shape of the beam and connection




membrane effect – Final Method




membrane effect – Final Method

Material properties

0

0,2

0,4

0,6

0,8

1

0 200 400 600 800 1 000 1 200

Re

du

cti

on

fa

cto

rs

Temperature ( C)

kEa,θ

kap,θ

kay,θ

0,0

0,2

0,4

0,6

0,8

1,0

0 200 400 600 800 1 000 1 200

Re

du

cti

on

fa

cto

rs (x

1E

-3)

Temperature ( C)

kEa,θ

kap,θ

kay,θ

a) θ < 600 °C b) θ ≥ 600 °C and cooling phase



Ansys Numerical Model

B5

B4

a) 3D view b) 3D zoom view

B5

B4

B1

B3

B2

c) Bottom view

Test - beam

Model 1 – slab

Model1 – beamModel 2 – slab

Model 2 - beam



SAFIR Numerical Model



Numerical simulation

Vertical displacements

Realistic modelling?

Secondary beam: BEAM finite elements

Secondary beam: SHELL finite element

Ansys model comparison




Model 4: unprotected beams with shell elements

Temperatures at 90 min (ISO fire)

°C






Temperatures at 90 min (ISO fire)

°C






Vertical displacements at 90 min

-587 -447 -308 -169 -30 39 (mm) -100 -239 -378 -517






Vertical displacements at 90 min

-562 -430 -298 -165 -33 33 (mm) -99 -231 -364 -496






Inward lateral torsional buckling

End

109 °C ≤ θa ≤ 392 °C

Mid-span

400 °C ≤ θa ≤ 636 °C

t = 15 min







Mid-span

667 °C ≤ θa ≤ 800 °C

End

157 °C ≤ θa ≤ 540 °C

t = 30 min







t = 60 min

End

308 °C ≤ θa ≤ 883 °C

Mid-span

963 °C ≤ θa ≤ 1 001 °C







End

227 °C ≤ θa ≤ 710 °C

Mid-span

876 °C ≤ θa ≤ 937 °C

t = 90 min





Comparison of the 4 numerical models

Vertical deflections

«Simplified» kept model : model 4

Unprotected secondary beam

Slab

-500

-450

-400

-350

-300

-250

-200

-150

-100

-50

0

0 10 20 30 40 50 60 70 80 90Temps (min)

Flè

ch

e d

e l

a s

oli

ve

in

téri

eu

re (

mm

)

Modèle 1

Modèle 2

Modèle 3

Modèle 4b

Defl

ecti

on

of

the u

np

rote

cte

d b

eam

(m

m)

Time (min)

-700

-600

-500

-400

-300

-200

-100

0

0 10 20 30 40 50 60 70 80 90Temps (min)

Flè

ch

e d

e l

a d

all

e (

mm

)

Modèle 1

Modèle 2

Modèle 3

Modèle 4b

Defl

ecti

on

of

the s

lab

(m

m)

Time (min)




RFCS MACS+ Software


Research & Development 4/17/2012 36

THE NATIONAL SEMINARS OBJECTIVES

• to distribute the produced data to the practitioners in order that

they become aware of what are the advantages of the

membrane effect as of its applicability

SEMINARS

MACS+

RFCS MACS+ Seminars



Thank you for

your attention

Documents

Tensile membrane action in fire of composite slabs with ...fire-research.group.shef.ac.uk/steelinfire/downloads/OV_2012.pdfArcelorMittal Long Carbon Research & Development Tensile