STRUCTURAL INTEGRITY FIRE TEST Wald, F.a, Simões da Silva, L.b, Moore, D.B.c, Lennon T.c

a Czech Technical University in Prague, Czech Republic b University of Coimbra, Portugal c Building Research Establishment, United Kingdom SUMMARY The paper describes the preparation, execution and major results from a large scale fire test on eight storey building in Cardington January 16, 2003 designed to study tensile membrane action and robustness of structural steel joints under a natural fire on real building. The major objectives of the test were to determine: the temperature distribution through the connections; the internal forces in the connections and the behaviour of the composite slab. The concept of using the fire protected columns and unprotected floor beams was confirmed by the fire test. Key words: Structural Engineering, Steel and Composite Structures, Full-Scale Tests, Composite Slab, Fire Design, Structural Integrity. 1 INTRODUCTION

The Cardington Laboratory serves as a facility for the advancement of the understanding of whole-building performance. Most aspects of a building’s lifecycle, from fabrication to fire resistance and explosions through to demolition, can be investigated on real buildings. This facility is located at Cardington, Bedfordshire, UK and consists of a former airship hangar 48 m x 65 m x 250 m, see Fig. 1. It is used by industrial organizations, Universities and research institutes, government departments and agencies. The Building Research Establishment (BRE) Cardington Laboratory contains three large experimental buildings [1]: a six storey timber structure, a seven storey concrete structure and an eight storey steel building.

a) b) Fig. 1 a) Cardington hangars with airships in 1921, b) hangar with open door, in front is

marked six storey timber building

The test steel structure was built in 1993. It is a steel framed construction using concrete slabs supported by a steel decking in composite action with the steel beams. It has eight storeys (33 m) and is five bays (5 x 9 m = 45 m) by three bays (6 + 9 + 6 = 21 m) in plan, see Fig. 2. The structure was planned as non-sway with a central lift shaft and two end staircases

providing the necessary resistance to lateral wind loads. The main steel frame was designed for gravity loads and the connections, which consist of flexible end plates for beam-column connections and fin plates for beam-beam connections, were designed to transmit vertical shear loads. The building simulates a real commercial office in the Bedford area and all the elements were verified according to British Standards and checked for compliance with the provisions of structural Eurocodes.

9000 9000 9000 9000 9000

9000

6000

6000

E

1

2

3

4

F D C B A

Fig. 1 The Cardington fire tests on steel structure

The building was designed for a dead load of 3,65 kN/m2 and an imposed load of 3,5 kN/m2, see [2]. The floor construction consist of steel deck and light-weight in-situ concrete composite floor, incorporating an anti-crack mesh of 142 mm2/m in both directions [3]. The floor slab has an overall depth of 130 mm and the steel decking has a trough depth of 60 mm. Seven large-scale fire tests at various positions within the experimental building were conducted; see Fig. 2 and Tab. 1 [4]. The main aim of the compartment fire tests was to assess the behaviour of structural elements with real restraint under a natural fire, se Fig. 3. Tab. 1 Fire test on steel structure in Cardington laboratory [15]

No. Test Fire compartment Load size, m area m Fire Mechanical

1 One beam heated by gas 8 x 3 24 Gas 30% 2 One frame heated by gas 21 x 2,5 53 Gas 30% 3 Corner compartment 9x 6 54 40 kg/m2 30% 4 Corner compartment 10 x 7 70 45 kg/m2 30% 5 Large compartment 21 x 18 342 40 kg/m2 30% 6 Office – Demonstration 18 x 9 136 45 kg/m2 30% 7 Structural integrity 11 x7 77 40 kg/m2 56%

Tab. 2 Summary of results from major fire tests in Cardington laboratory [16]

No. Org. Level Time, min Reached temperature °C Measured deformations to max. temp. gas steel maximal residual

1 BS* 7 170 913 875 232 113 2 BS* 4 125 820 800 445 265 3 BRE 3 114 1000 903 269 160 4 BS* 2 75 1020 950 325 425 5 BRE 3 70 - 691 557 481 6 BS* 2 40 1150 1060 610 - 7 ČVUT** 4 55 1108 1088 > 1000 925

*BS- British Steel (now Corus); **ČVUT – collaborative research proposed by Czech Technical University

2. FIRE COMPARTMENT The structural integrity fire test (large test No.7) was carried out in a centrally located compartment of the building, enclosing a plan area of 11 m by 7 m on the 4th floor [5]. The preparatory works took four months. The identification of the compartment is illustrated in Fig. 2. The fire compartment, see Fig. 3, was bounded with walls made of three layers of plasterboard (15 mm + 12,5 mm + 15 mm) with a thermal conductivity 0,19 - 0,24 W/mK. In the external wall (gridline 1) the plasterboard is fixed to a 0,9 m high brick wall. The opening of 1,27 m high and 9 m length simulated an open window to ventilate the compartment and allow for observation of the element behaviour. The ventilation condition was chosen to result in a fire of the required severity in terms of maximum temperature and overall duration. The columns, external joints and connected beam (about 1,0 m from the joints) were fire protected to prevent global structural instability. The material protection used was 18 - 22 mm of Cafco300 vermiculite-cement spray, with a thermal conductivity of 0,078 W/mK.

a) b) c) Fig. 3 Compartment - a) opening, b) plasterboard walls, and c) fire protection of columns

150

55

1560

D E

1

2

N356x171x51UB

305x165x40UB

356x171x51UB305x165x40UB

Fin plate connection

Secondary beamEnd plate connectionP8-260x140

End plate connectionP8-260x150

P10-260x100 Fin plate connectionP10-260x100

Secondary beam

Secondary beam

Primary beam

254060

6

10(9)

20

60

4060

6040

27406060

4060

40 50

4M208M20

6

14020

406060

4060

5040

8M20

6

Fig. 4 Arrangement of members in selected fire compartment including geometry of

connections and slab

The steel structure exposed to fire consists of two secondary beams (section 305x165x40UB, steel S275 measured fy = 303 MPa; fu = 469 MPa), edge beam (section 356x171x51UB), primary beams (section 336x171x51UB, steel S350 measured fy = 396 MPa; fu = 544 MPa) and columns (internal section 305x305x198UC and external 305x305x137UC, steel S350) as shown in Fig. 4 [6]. The joints were a cruciform arrangement of a single column with three or four beams connected to the column flange and web by the header plate connections, steel S275. The beam to beam connections were created by fin plates, steel S275. The composite behaviour was achieved by a concrete slab (light weight concrete LW 35/38; experimentally by Schmidt hammer 39,4 MPa) over the beams cast on shear studs (∅19-95; fu = 350 MP). The measured sections geometry and material property are summarized in [6].

The mechanical load was simulated using sandbags, each one of 1100 kg applied over an area of 18 m by 10,5 m on the 5th floor, see Fig. 5. Sand bags represent the mechanical loadings; 100% of permanent actions, 100% of variable permanent actions and 56% of live actions. The mechanical load was designed to reach the collapse of the floor, based on analytical and FE simulations. Wooden cribs with moisture content 14 % provided the fire load of 40 kg/m2 of the floor area.

1

2

D E

Sand bags 5th floorFire compartment level 4

Fig. 5 Mechanical load by sandbags

Fig. 6 Fire load of 40 kg/m2 by wooden cribs

The instrumentation used included thermocouples, strain gauges and displacement transducers. A total of 133 thermocouples monitored the temperature of the connections and beams within the compartment, the temperature distribution through the slab and the atmosphere temperature within the compartment. An additional 14 thermocouples measured the temperature of the protected columns. Two different types of gauge were used, high temperature and ambient temperature to measure the strain in the elements. In the exposed and un-protected elements (fin plate and end plate - minor axis) nine high temperature strain gauges were used. In the protected columns and on the slab a total of 47 ambient strain gauges were installed. 25 vertical displacement transducers were attached along the 5th floor to measure the deformation of the concrete slab. An additional 12 transducers were used to measure the horizontal movement of the columns and the slab. Ten video cameras and two thermo imaging cameras recorded the fire and smoke development, the deformations and temperature distribution. 3 COMPARTMENT TEMPERATURES

The quantity of thermal load and the dimensions of the opening on the facade wall were designed to achieve a representative fire in the office building. The maximum recorded compartment temperature near the wall (2 250 mm from D2-E2) was 1107,79 ºC after 54

minutes. The prediction by the simple model of Eurocode 1, Part 1-2 (eqs. A1 and A11 [7]) was 1078 °C after 53 min, as shown in Fig. 7.

0

200

400

600

800

1000

1200

0 15 30 45 60 75 90 105 120 135 150

Time, min

Temperature, °C

Back in fire compartment

Prediction prEN 1991-1-2, Annex A

In front of fire compartment

Average temperature

300

500 500

Fire compartment (DE, 1-2)

1108

1078

53 54

b)

c)

Fig. 7 a) Development of gas temperature; b) fire in progress, c) cooling phase 4 COMPOSITE SLAB

Temperature variation within slab over the rib is shown in Fig. 8 for cavity C4. The cameras above the fire compartment recorded the loss of the integrity limit state of the concrete slab after 54 min. The first opening of the composite slab occurred around column E2 by a punching mechanism due to the tension in the concrete slab in the edge compartment. Furthermore, many tiny cracks were observed in different areas of the concrete slab.

Temperature, °C0

20

40

60

80

100

120130

0 50 100 150 200 250 300

Depth of the slab, cavity C4, mm

Reinforcement

0 10 20 30 min.

40 50 60 70 min.

303070 E2D2

E1D1

N

4500 4500

1500 Cavity C4

Fig. 8 Temperature variation within slab over the rib, cavity C4 Fracture in the concrete slab was observed, with a large crack propagating from the face of the column flange parallel to the beam (D2-E2), see Fig. 9b. This crack developed due to the tension in the concrete slab, along the weak zone in the composite beam - flange extremity. After the concrete cracked the joint stiffness gradually decreased. Secondary cracks occurred perpendicular to, and continuous across, the connections on both sides of the slab, see Fig. 9c. Coincident with the maximum vertical displacement, the mesh was overlapped but not connected. Reinforcement mesh slippage was observed and the slab in this area was effectively unreinforced.

a)

D E

1

N

b)

c)

2

b)

c) largest crack in mid-span

Fig. 9 Cracks in concrete slab - a) geometry; b) slipping near to column D2

-900

-800-700-600-500-400

-300-200-100

00,00 2,25

Line 1 3/4

Deformation, mm

D1

N

1500 Line 1 3/4

D2

E1

E2

a) b) c)

c)d)

70

9,00 Position, m10 20

30

40

50

4,50 6,75

60

80*90*min.

100

´

Fig. 10 Vertical deformations - a) reached values; b) at drawn line 1 3/4, * deflection more then 900 mm may be affected by the limited travel of the transducers; b) compartment after

fire residual deformation 925 mm; c) no local collapse of structure 5 ELEMENTS

The maximum deflections were not recorded by the displacement transducers as their range was limited to 1000 mm minus the initial offset. From the video cameras on the 5th floor it is possible to recalculate a maximum deformation of approximately 1 200 mm. Fig. 10 shows the vertical displacement recorded in the beams D1-E1, D1/2-E1/2, D2-E2. Comparing the different secondary beams, it is observed that during the heating phase, the beam with a lower displacement is the beam near the window, due to the lower temperatures, while the beam near the internal wall shows the biggest displacement, see Fig. 8. In the cooling phase both of these beams partially recovered, see Fig. 10.

Time, min

Beam temperature, °C

Prediction, prEN 1991-1-2, 1993-1-2

0 15 30 45 60 75 90 105 120

Beam D1-E1 upper flangeBeam D1-E1 webBeam D1-E1 lower flange

Beam D2-E2 upper flangeBeam D2-E2 webBeam D2-E2 lower flange

3

1088 °C1067 °C

54 min.57 min.

0

200

400

600

800

1000

1200

Beam D21-E21 lower flange N

E2D2

E1D1

3

4

3

4

700 800 900 1000

3

Temperature in 54 min., °C

Fig. 11 Temperature variation within the beams D1-E1; D2-E2

The temperatures in the mid-span beams were measured on the bottom flanges, the web, and the upper flange. A summary of the temperatures recorded in the beams is presented in Fig. 11. The maximum reported steel temperature of 1087,5ºC occurred after 57 minutes of fire on the bottom flange of the beam D2-E2 in the middle of the section. By calculation a temperature of 1067 °C after 54 min. was predicted using an iterative procedure of the transfer of heat into the unprotected steel structure, see eq. 4.24 [9].

N

E2D2

E1D1

Fig. 12 Local buckling of beam lower flange

N

E2D2

E1D1

Fig. 13 Beam web in shear

Local buckling of the beam lower flange was one of the main observed mechanisms. It is seen on the beam lower flange and web adjacent to the joint, see Fig. 12, the concrete slab restrained the upper flange. This local buckling occurs first during the heating phase after 23 min. of fire (observed by thermo imaging camera) due to the restraint to thermal elongation provided by the adjacent cooler structure and the structural continuity of the frame. The heated lower flange of the beam is unable to transmit the high normal forces generated in the beams lower flanges to the adjacent beams/columns after closing the gap in the lower part of the connections. As the temperature and the associated deformations increase the capacity of the beam web in shear was exceeded, see Fig. 13.

The formation of a plastic hinge in the beam cross-section next to the protected zone was one of the visual mechanisms in the secondary beam D1/2, see Fig. 14. This hinge starts with a lateral-torsional buckling during the first stage of the heating phase due to the restraint to thermal elongation provided by the adjacent protected section. This behaviour is associated with the local loss of stability in the bottom flange. Essentially during the second stage of the heating phase the beam rotates around this point due to the large mid-span deflection.

N

E2D2

E1D1

Fig. 14 Plastic hinge in the unprotected beam cross-section close to the end of fire protection

N

E2D2

E1D1

Fig. 15 Local buckling of column flange in compression, column E2

The local buckling of the column flange in compression was observed in the major axis connection of the beam-to-column joints, see Fig. 15. This behaviour is characterized by the small column flange thickness (t = 21,4 mm) and the small distance between the bolts, the bolted end-plate behaved as a welded one. This behaviour was observed in both column flanges in the beam-to-column joints (D2; E2).

6 CONNECTIONS Fracture of the end-plate along the welds due to horizontal tensile forces generated during cooling of the connected beam under the large rotations associated with flexible end-plate joints (Fig. 16) was the major observed behaviour. The fracture occurred along one side of the connection only, while the other side remained intact. After one side has fractured, the flexibility increased allowing larger deformations without further fracture. This behaviour was observed in the major axis of the beam-to-column joints (D2-D1; E2-E1) and in the minor axis of the beam-to-column joints (D2-C2). At the end plate, in the level of fourth bolt row, reached the temperature 772,2°C after 75 min. of test. The maximum temperature of the lower flange of the secondary beam was 973,1 °C after 58 min. The prediction by the simple model of eq. D5 in prEN 1991-1-2 [7] (based on the mid span temperature) yields the value of 800,3°C, as shown in Fig. 17.

a) b)

Fig. 16 Fracture of the end-plate along the welds - a) connections D2-D1; b) deformation of end-plate in view from bottom

Time, min

Temperature, °C

E2D2

E1D1 N

of end-plate D2 - D1Crack on one side

0

200

400

600

800

1000

0 15 30 45 60 75 90 105 120 135 150

Measured plate 4th bolt row

Predicted plate 4th bolt row

Measured plate 1st bolt row

600 650 700 750 800Temperature in 75 min, end-plate D2-D1, °C

Measured midspan bott. flange

PlateBolt

Plate

Bolt

Fig. 17 Temperature variation within the beam-to-column major axes

end plate connection, D2-D1

The elongation of the holes in the web of the beam in the tension/compression part of the fin plate connection is due the associated large rotations, see Fig. 18. Elongation of the holes occurred on the web of the connected beam, while the fin-plate remained intact: the beam web thickness (6 mm) is smaller than the fin-plate (10 mm). After the elongation of the

holes of the beam web has taken place, the joint flexibility is increased, allowing larger deformations without further fracture.

N

E2D2

E1D1

Fig. 18 Elongation of holes in the beam web in fin plate connection

N

E2D2

E1D1

Time, min.

Temperature, °C

0

200

400

600

800

1000

0 15 30 45 60 75 90 105 120 135 150

120

Temperature in 63 min., °C800 850 900 950

Bolt

Bolt

Plate 1st rowUpp. flange

Bott. flangePlate 4th row

Midspan bott. flange

Fig. 19 Temperature variation within the beam-to-beam fin plate connection D1/2-E1/2

a) 220,0°C

1050,0°C

400

600

800

1000

b) 850,0°C

1050,0°C

900

950

1000

1050

LI01

LI02LI03

LI04

LI05LI06

Fig. 20 Visualisation of record of thermo imaging camera after 57 min. - a) record in full range, b) reading with limited temperatures, lines are marked

The termo imaging cameras (FLIR 695 PM) with lens 6° recorded the temperatures on the structure in blocks of approximately 24 x 24 mm, see Fig. 23b. The matrix obtained was calibrated against the visible thermocouples to increase the accuracy limited due to changes of emisivity. Fig. 20a illustrates the matrix of temperatures after 57 min., corresponding to the

maximum temperature measured on the lower flange of the secondary beam (1088°C). The matrix is visualised in range from 220°C to 1050°C. In Fig. 20b the range is shortened from 850°C to 1050°C. Detailed temperature readings along lines 01 to 06 are illustrated in Fig. 21. The maximum temperature on the fin plate connection was measured after 63 min. of the experiment. Fig. 22a illustrates the observed vertical lines. Fig. 23 shows the adequate measured temperatures. On cooling the joints are hotter than the surrounding structure. The highest difference was reported after 92 min., see Fig. 21b.

900

1000li01li02li03

Line Min Max ...li01 901,0°C 967,4°Cli02 926,0°C 991,0°Cli03 906,6°C 985,0°C

°C IR01

900

1000li04li05li06

Line Min Max ...li04 974,6°C 1022,2°Cli05 1007,4°C 1047,7°Cli06 1005,8°C 1041,2°C

°C IR01

Fig. 21 Reading by termo imaging camera after 57 min. temperature of lower flange in midspan 1088°C

a) 800,0°C

930,0°C

820

840

860

880

900

920

LI01LI02LI03LI04LI05

b) 390,0°C

597,1°C

400

450

500

550

Fig. 22 Pictures of thermo imaging camera - a) highest connection temperature after 63 min., b) picture after 92 min. in cooling phase of test

a)

850

900 li01li02li03li04li05

Line Min Max ...li01 860,3°C 884,9°Cli02 858,0°C 885,3°Cli03 844,3°C 860,3°Cli04 853,5°C 884,0°Cli05 853,5°C 887,6°C

°C IR01

b)

E2D2

E1D1

N

Fig. 23 a) Fin plate temperatures of connection D1-2 recorded by thermo imaging camera after 63 min.; b) position of termo imaging camera in front of the compartment

7 CONCLUSIONS The collapse of the structure was not reached and the test demonstrated that the structure remained intact subject to a fire load of 40 kg/m2, which represents a design fire load in a typical office building, together with a mechanical load greater than the design value for

the fire limit state. The test results fully support the concept of unprotected beams and connections with protected columns as a viable system for composite floors. The local buckling of the lower flanges of beams was observed after 23 minutes of the fire. The fracture of end plates occurred under cooling in the heat affected zones of welds without losing the shear capacity of the connections. The fin plate connections behaved in a ductile fashion due to elongation of holes in bearing. The experiment checked the high conservatism of the Eurocode fire design, see [8], [9]. The calculated values show good and conservative predictions of the temperature in the fire compartment, the transfer of heat into the structure and connection and the prediction of structural behaviour. The detailed behaviour of the composite slab, connections and columns is being investigated to refine the predictive analytical and numerical models. The boundary conditions of the elements measured on the Cardington frame test are applied to new programmes of laboratory tests on connections, columns and beams at participating laboratories. Acknowledgement

This paper reports the results from a collaborative research project, “Tensile membrane action and robustness of structural steel joints under natural fire”, involving the following institutions: Czech Technical University (Czech Republic), University of Coimbra (Portugal), Slovak Technical University (Slovak Republic) and Building Research Establishment (United Kingdom). The project has been supported by the grant of European Community FP5 HPRI - CV 5535 and COST C12. This paper was prepared as a part of project 103/04/2100 of the Czech Grant Agency. REFERENCES [1] Moore D.B.: Steel fire tests on a building frames, Building Research Establishment, Paper

No. PD220/95, Watford 1995, p. 13. [2] Lennon T.: Cardington fire tests: Survey of damage to the eight storey building, Building

Research Establishment, Paper No127/97, Watford 1997, p. 56. [3] Moore D.B. and Lennon T.: Fire engineering design of steel structures”, Progress in

Structural Engineering and Materials, No.1(1), 1997, pp. 4 - 9. [4] Bailey C.G., Lennon T., Moore D.B.: The behaviour of full-scale steel-framed building

subject to compartment fires, The Structural Engineer, Vol.77/No.8, 1999, p. 15-21. [5] Wald F., Santiago A., Chladná M., Lennon T., Burges I., Beneš M.: Tensile membrane

action and robustness of structural steel joints under natural fire, Internal report, Part 1 - Project of Measurements; Part 2 - Prediction; Part 3 – Measured data; Part 4 – Behaviour, BRE, Watford, 2002-2003.

[6] Bravery P.N.R.: Cardington Large Building Test Facility, Construction details for the first building, Building Research Establishment, Internal paper, Watford 1993, p. 158.

[7] ECSC 2002: Design recommendations for composite steel framed buildings in fire, Project 7210 PA, PB, PC, PD112, December 2002, p. 108.

[8] CEN, Eurocode 1, Draft prEN - 1991-1-2: 200x, Part 1.2: General actions – Actions on structures exposed to fire, Eurocode 1: Actions on structures, Final Draft, CEN, European Committee for Standardization, Brussels 2003.

[9] CEN, Eurocode 3, Draft prEN - 1993-1-2: 200x, Part 1.2: Structural Fire Design, Eurocode 3: Design of Steel Structures, CEN, European Committee for Standardization, Brussels 2003.