Temperature of Steel Columns under Natural Fire. Prof. Ing. František Wald, CSc., Ing. Petra Studecká, Ing. Lukáš Kroupa

Transcription

1 Temerature of Steel Columns under Natural Fire Prof. Ing. František Wald, CSc., Ing. Petra Studecká, Ing. Lukáš Kroua Czech Technical University in Prague, Praha 6, Thákurova , , wald@fsv.cvut.cz KEYWORDS Steel structures, Fire design, Fire test, Comartment temerature, Protected steel, Natural fire RESUME Current fire design models for the time-temerature develoment within the structural elements as well as for the structural behaviour are based on the isolated member tests subjected to the standard fire regimes, which serve as a reference heating, but do not model the natural fire. Only the tests on a real structure under the natural fire may evaluate the forthcoming models of the temerature develoments in fire comartment, of the transfer of heat into the structure and of the overall structural behaviour under fire. To study a global structural behaviour, a research roject was conducted on the eight storey steel frame building at the Cardington s Building Research Establishment laboratory January 16, 3. The fire comartment 11 x 7 m was reared at fourth floor. The fire load of 4 kg/m 2 was alies with 1% ermanent mechanical load and 65% of imosed one. The aer summarises the exerimental rogramme and show the temerature develoment of the gas in the fire comartment and of the fire rotected columns carrying the unrotected floors. a) b) c) d) Fig. 1a) Fire load in comartment; b) fire load around column ; c) rotection of internal column (after test), d) rotection of external column (after test), 1

2 INTRODUCTION The BRE s Cardington Laboratory is a unique worldwide facility for the advancement of the understanding of whole-building erformance, see [1]. This facility is located at Cardington, Bedfordshire, UK and consists of a former airshi hangar with dimensions 48 m x 65 m x 25 m. The Cardington Laboratory comrises three exerimental buildings: a six storey timber structure, a seven storey concrete structure, and an eight storey steel structure. The steel test structure was built in It is a steel framed construction using comosite concrete slabs suorted by steel decking in comosite action with the steel beams. It has eight storeys (33 m) and is five bays wide (5 x 9 m = 45 m) by three bays dee ( = 21 m) in lan. The structure was built as non-sway with a central lift shaft and two end staircases roviding the necessary resistance against lateral wind loads. The main steel frame was designed for gravity loads, the connections consisting of flexible end lates for beam-tocolumn connections and fin lates for beam-to-beam connections, 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 comliance with the rovisions of the Structural Eurocodes. The building was designed for a dead load of 3,65 kn/m 2 and an imosed load of 3,5 kn/m 2. The floor construction consists of steel deck and light-weight in-situ concrete comosite floor, incororating an anti-crack mesh of 142 mm 2 /m in both directions, see [2]. The floor slab has an overall deth of 13 mm and the steel decking has a trough deth of 6 mm. Seven large-scale fire tests at various ositions within the exerimental building were conducted, see [3], and there is still a lace for two more tests. The main aim of this comartment fire tests was to assess the behaviour of structural elements with real restraint under a natural fire. The structural integrity fire test (large test No.7) was carried out in a centrally located comartment of the building, enclosing a lan area of 11 m by 7 m on the 4 th floor [4]. The rearatory works took four months. The fire comartment was bounded with walls made of three layers of lasterboard (15 mm + 12,5 mm + 15 mm) with a thermal conductivity,19 -,24 W/mK. In the external wall the lasterboard is fixed to a,9 m high brick wall. The oening of 1,27 m high and 8,7 m length simulated an oen window to ventilate the comartment 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 temerature and overall duration. The steel structure exosed to fire consists of two secondary beams (section 35x165x4UB, steel S275 measured f y = 33 MPa; f u = 469 MPa), edge beam (section 356x171x51UB), rimary beams (section 336x171x51UB, steel S35 measured f y = 396 MPa; f u = 544 MPa) and columns, internal section 35x35x198UC and external 35x35x137UC, steel S35. The joints are a cruciform arrangement of a single column with three or four beams connected to the column flange and web by the header late connections, steel S275. The beam to beam connections were created by fin lates, steel S275. The comosite behaviour was achieved by a concrete slab (light weight concrete LW 35/38; exerimentally by Schmidt hammer 39,4 MPa) over the beams cast on shear studs ( 19-95; f u = 35 MPa). The measured sections geometry and material roerty are summarized in tab. A1, see [2] and [5]. The mechanical load was simulated using sandbags, 11 kg of each, alied over an area of 18 m by 1,5 m on the 5th floor. Sand bags reresent the mechanical loadings; 1% of ermanent actions, 1% of variable ermanent actions and 56% of live actions. The mechanical load was designed to reach the collase of the floor, based on analytical and FE simulations. Wooden cribs with moisture content 14 % rovided the fire load of 4 kg/m 2 of the floor area, see Fig. 1a,b. 2

3 The columns, external joints and connected beam (about 1, m from the joints) were fire rotected to revent global structural instability. The material rotection used was 2 mm of Cafco vermiculite-cement sray, based on vermiculite and gysum, see Tab. A2 and Fig. 1c,d. It was alied as a single ackage factory controlled remix, with a thermal conductivity of,78 W/m K. The aer describes the temerature develoment in the fire rotected columns. The temeratures redicted analytically and by 2D FEM simulation are comared to the measured ones. 1 INSTRUMENTATION The instrumentation used included thermocoules, strain gauges and dislacement transducers. A total of 133 thermocoules monitored the temerature of the connections and beams within the comartment, the temerature distribution through the slab and the atmoshere temerature within the comartment, see Fig. 2a. An additional 14 thermocoules measured the temerature of the rotected columns, see Fig. 5. Two different tyes of gauge were used, high temerature and ambient temerature to measure the strain in the elements. In the exosed and un-rotected elements (fin late and end late - minor axis) nine high temerature strain gauges were used. In the rotected columns and on the slab a total of 47 ambient strain gauges were installed. 25 vertical dislacement transducers were attached along the 5 th 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 develoment, the deformations and temerature distribution, see [5]. D Internal wall of the fire comartment E A - A major axis,8 2,25 2,25 2,25 2,25 2 G521 A - A B - B G522 G523 G525 G526 G527,5625 1,625 1,625 1,625 G529 G53 G531 N,8 G533 G534 G535 C415, C416, C417 C418, C419, C Fire rotected art B - B minor axis North view G524 G528 G532 G536, Window of the fire comartment 1,27 x 8,7 m C415 PLAN 4th floor C416 C417 C418, C419, C42 a) b) Fire rotected art Fig. 2a) Location of thermocoules in comartment mm below ceiling; b) thermocoules on the beam and column end round the connection 3

4 1 1 6 Temerature, C Prediction, EN , Annex B [1] 118 C 178 C G526 redicted 53 min. Back in fire comartment G525 recorded 54 min. G525 G526 3 * ,5 Everage temerature In front of fire comartment G528 Time, min G528 Fire comartment (DE, 1-2) 562,5 Fig. 3 Comarison of the rediction of the gas temerature to the measured temeratures, for values see tab. B1 Time: 15 min. Time: 3 min Time: 54 min. Time: 75 min Fig. 4 Isotherms in comartment by thermocoules mm under ceiling, the inut data are summarised in tab. B1 2 FIRE DEVELOPMENT The quantity of thermal load and the dimensions of the oening on the facade wall were designed to achieve a reresentative fire in the office building. The oenings allow the fire without a flashover managed by combustible timber sticks, see [4]. The temerature grow to 4

5 reach the lato of the time temerature curve in about 18 minutes with a eak at 54 min., when the cooling starts, see Fig. 3. The maximum recorded comartment temerature near the wall (2 25 mm from ) was 117,8 ºC after 54 minutes. Predicted was 178 C in 53 min, see [5]. During the heating was the temerature distributed regularly, see Fig. 4. The measured differences of gas temerature have decreased by cooling from C till 2 C in 12 min. The measured maximal gas temeratures are summarised in tab. B1 in the time intervals. The average gas temerature is calculated from all sixteen thermocoules. C49, C412, C414 C429, C432, C434 C411 C431 C48, C41, C413 C428, C43, C433 C417,C416,C415 C48 C49 C42,C419,C418 C429 C428 C41, C411 C412 C432, C431 C43 C42, C45, C47 C422, C425, C427 C C424 C437,C436,C435 C41,C43,C46 C421,C423,C426 C41,C42 C421,C422 C44,C439,C438 C44 C43,C45 C423, C425 C424 C413 C414 C434 C433 C46,C47 C426, C427 Colum Column E2 Column Column UC 35 x 35 x 137 UC 35 x 35 x 137 UC 35 x 35 x 198 UC 35 x 35 x 137 Fig. 5 Location of thermocoules on columns/beams, on flanges 2 mm from edge, on web in centreline 1 6 Temerature, C Gas temerature TC525 External column at mid height TC421 Secondary beam -E2; midsan, lower flange TC488 Internal column E2, mm under slab, TC428 Internal column E2, at mid height, TC43 Internal column E2, mm above floor, TC433 N E2 C488 C421 C428 C43 C433 C428 C43 C433 E2 C421 C Time, min. Fig. 6 Comarison of the measured temerature along the column length to the gas and connected beam temerature (column ) and external column temerature () 4 COLUMN TEMPERATURES The temeratures in the columns in the fire comartment were measured at middle of its high, mm from the floor, and mm below the ceiling at both flanges and at its web, and in the connections, see Figs 2b and 5, tab. B2. The columns were fire rotected excet the joint area, where the rimary and secondary beams were connected. A selection of the temeratures recorded at column and are resented in Figs 6 and 7, where are comared to the gas 5

6 temerature, the beam mid-san temerature, the beam end temerature as well as the column end temerature. The fire created a homogenous gas temerature the both columns were heated almost equally. The maximum reorted temerature in the insulated art of the middle column of 426, ºC, which occurred after 16 minutes of fire. The values reached at the middle of column height and in the uer art of the column are similar. The gradient of the temeratures along the column is changing during time. The differences of the measured temeratures cross sections were insignificant, see tab. B Temerature, C Gas temerature G525 Secondary beam -E2; midsan, lower flange, C488 External column, thermocoule C Time, min. C41 Internal column, mm under slab, C48 Internal column, at mid height, C41 Internal column, mm above floor, C413 C488 E2 N C48 C41 C413 C48 C41 C413 Fig. 7 Comarison of the measured temerature along the column length to the gas and connected beam temerature (column E2) and external column temerature () C41 C Steel temerature, C C415, measured C417, measured C418, measured C49, measured C414, measured Time, min. N E2 5 C418 C49 C C415 C417 C49 C418 C414 Fig. 8 Measured temeratures along a column length, column, and at a beam end C415 C417 The accurate and simle ste-by-ste calculation rocedure is based on the rincile that the heat entering the steel over the exosed surface area in a small time ste t (taken as 3 seconds maximum) is equal to the heat required to rise the temerature of the steel by θ a,t (at time t) assuming that the steel section is a lumed mass at uniform temerature, so that q& F t = ρa ca,t V θa,t, (1) where ρ a is the unit mass of steel, c a,t is the temerature deendant secific heat of steel, V is the volume of the member er unit length, and q& is the heat transfer at the surface, given by 4 4 q& = h ( θ θ ) + σ ε ( θ θ ), (2) c g,t a,t g,t a, t 6

7 where h c is the convective heat transfer coefficient, σ is the Stefan-Bolzman constant (56, kw/m 2 K 4 ), ε is the resultant emissivity and θ g,t is the increase of the ambient gas temerature during the time interval t. Eqs (1) and (3) may be rearrange to give Am / V 4 θ a,t = [ hc ( θ g,t θa,t ) + σ ε ( θ g,t θa,t )] 4 t, (3) ca ρa where A m / V is the section factor for unrotected steel members, A m is the surface area of the member er unit length. The convective heat transfer coefficient is recommended to have a value of 25 W/m 2 K. The iterative rocedure for rotected steelwork is similar to that for unrotected steel. The equation does not require heat transfer coefficients because it is assumed that the external surface of the insulation is at the same temerature as the fire gases. It is also assumed that the internal surface of the insulation is at the same temerature as the steel. The equation is A k c,t ρ θ a,t = ( θ g,t -θ a,t ) t, (4) V d c A d c a,t ρ a,t ρ ca,t ρ a + V 2 where A /V is the section factor for steel members insulated by fire rotection material, k is the thermal conductivity of insulation, d is the thickness of the fire rotection material, c,t is the temerature indeendent secific heat of the fire rotection material, λ is the thermal conductivity of the fire rotection system, ρ is the unit mass of the fire rotection material. ECCS, see [6], suggested to ignoring the heat caacity of the insulation if it is less than half of that of the steel section, such that ca, t ρ a A / 2 > c,t ρ Ai, where A i is the cross-section area of the insulating material and A is the cross-section area of the steel. The rediction is used in EN : 3 ar [7], taking constant 3 instead of constant 2 in the heavy insulations term to allow the calculations for the temerature gradient across the insulation material, in form λ A /V ( θ g,t -θ a, t ) θ a,t = t - (e φ / 1-1) θ g,t (but θ a,t if θ g,t > ) (5) d ca ρ a (1+ φ /3) c,t ρ with φ = d A /V, where A is the aroriate area of fire rotection material er unit ca,t ρ a length of the member, which should generally be taken as the area of its inner surface, but for hollow encasement with a clearance around the steel member the same value as for hollow encasement without a clearance may be adoted, For rediction was used the fire rotection material thickness d =,18 m; unit mass ρ = 31 kg m -3 ; secific heat c = 1 J kg -1 K -1 ; and thermal conductivity λ =,78 W m -1 K -1. Fig. 9 shows the comarison of the redicted and measured temeratures. The internal column was exosed from four sides, the external one from two sides only. The rediction is based on measured gas temerature in thermocoule G525, on calculated arametric temerature, see [8] and [9], as well as on nominal temerature, see [1]. 7

8 Prediction by EN , C41 Steel temerature, C from nominal curve EN Prediction by EN , C41 from arametric gas tem. EN E2 C41 G525 C45 N Measured, C45 external column Measured, C41 internal column Prediction by EN , C41 from ex. gas tem. G Prediction by EN , C45 from ex. gas tem. G525 Time, min. Fig. 9 Comarison of column calculated temerature based on rediction by EN to measured one, thermocoule C C41 C41 C45 C45 Steel temerature, C Exerimental curve column - C E2 C41 G525 N Prediction by 2D FEM from ex. gas tem. G Prediction by 2D FEM from nominal curve, EN C Prediction by 2D FEM from arametric curve, EN Time, min. Fig. 1 Comarison of column calculated temerature based on rediction by Temcalc code to measured one, thermocoule C41 The results of modelling of the heat transfer into the columns by the 2-dimensional FE code are shown on Fig. 1. The Suer-Temcalc code [11], taking into account the above listed arameters, was used for rediction. In the code is derived the differential equation for 2 - dimensional heat flow from conservation of energy, relating on the fact that total inflow er unit time equals the total outflow er unit time. The constitutive relation invoked is Fourier s law of heat conduction, which describes heat flow within the material. The satial and time domains are discretized by the weighted residual aroach. Boundary conditions imlemented include convective and radiative heat flow and heat exchange within enclosures. 3-node 1848 C41 8

9 triangular finite elements were used. Thermal roerties of materials are described as temerature deendent. The temerature distribution within the rotected column is resented during heating, after 3 min. of fire, and during cooling, after 12 min. of fire at Fig. 11. In the section was reached the temerature difference of 4 C only. The comarison of the analytical and numerical results confirms a good quality of the resented analytical model. d h 3 minutes t w 12 minutes b t f 1 C 7 C UC 35x35x137 Cafco, 18 mm C d a) Fig. 11a) Temerature distribution within the rotected column during heating (after 3 min.) during cooling (after 12 min.) by Temcalc code; b) local buckling of the column flange E2 CONCLUSIONS The collase of structure or its arts was not reached during the exeriment for the fire load of 4 kg/m 2, which reresents the fire load in a tyical office building, together with a mechanical load greater than standard aroved cases. The structure showed good structural integrity. The test results suorted the concet of unrotected beams and rotected columns as a viable system for comosite floors. The connections do not need to be fire rotected from its resistance oint of view, see Fig. 11b, where a moderate local buckling is visible only. The test in Cardington January 16, 3 documents, that the incremental analytical models in ren : 3 [7] allows resuming the column temerature from the gas temerature during the heating hase with a good accuracy, see Figs 3, 9 and 1. From the nature of the heat transferred from the connected unrotected beams it is clear, that the only 3D solution may describe the transfer of the heat into the rotected columns under the unrotected floors. An aroximation based on 2D calculations is accetable for design till 6 minutes of fire only. The accurate analytical rediction of the temerature of the structure during its cooling will allow otimising the alication of the fire rotective material on the comressed member of the structure only. ACKNOWLEDGEMENTS The authors would like to thank all nineteen members of the roject team working on this large scale exeriment in Cardington BRE laboratory from October 2 till January 3. Secial thanks go to Mr. Tom Lennon, Mr. Nick Petty, and Mr. Martin Beneš for careful measurement of data resented above. The roject has been suorted by the grant of Euroean Community FP5 HPRI - CV Paer was reared as a art of roject 13/4/21 of the Czech Grant Agency. b) 9

10 REFERENCES [1] Wang Y.C.: Steel and comosite structures, Behaviour and design for fire safety, Son Press, London 2, ISBN [2] Bravery P.N.R.: Cardington Large Building Test Facility, Construction details for the first building, Building Research Establishment, Internal aer, Watford 1993, [3] Moore D.B.: Steel fire tests on a building framed, Building Research Establishment, Paer No. P2/95, Watford 1995,. 13. [4] Lennon T.: Cardington fire tests: Survey of damage to the eight storey building, Building Research Establishment, Paer No. 127/97, Watford 1997,. 56. [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 reort, Part 1 - Project of Measurements; Part 2 - Prediction; Part 3 Measured data; Part 4 Behaviour, BRE, Watford, 2-3. [6] Buchanan A.H.: Structural design for fire safety, John Wiley & Sons, Chichester 3. ISBN X. [7] Eurocode 3: Design of Steel Structures - ren : 4, Part 1.2: Structural Fire Design, Final draft, 3, CEN, Brussels 3. [8] Wald F., Silva S., Moore D.B, Lennon T., Chladná M., Santiago A., Beneš M.: Exeriment with structure under natural fire, The Structural Engineer, in ress. [9] Wald F., Chladná M., Santiago A., Lennon T.: Temeratures of structure during Cardington structural integrity fire test, in Proceedings of ICSCS'4, Soul 4, in ress. [1] Eurocode 1: Actions on structures, ren : 4, Part 1.2: General actions Actions on structures exosed to fire, CEN, Brussels 2. [11] Anderberg, Y.: SUPER-TEMPCALC, A commercial and user-friendly comuter rogram with Automatic FE-generation for temerature analysis of structures exosed to heat, Fire Safety Design AB, Lund Annex A Measured Geometrical Proerties of Columns Tab. A1 Geometry of column, [5] Fig. 2 h l h r b u b low t w t f.u.l t f.u.r t f.low.l t f.low.r Profile Column mm mm mm mm nominal 32,5 39,2 13,8 21,7 318,1 316,6 38,2 39,6-21,2 21,9 21,2 21,6 UC35x35x ,1 317,6 37,2 39,8-21,4 21,6 21,2 21,7 E2 32,2 318,6 39,2 39,6-21,4 21,9 21,1 21,5 nominal 339,9 314,5 19,1 31,4 UC35x35x , 336,8 312,3 312,5-32, 3,8 31,4 31,1 Symbols, see Fig. 11: h l is the height of column section on left side, h r is the height of column section on right side, b is the width of column section, u is uer measured value, low is lower measured value, t w is the thickness of column web, t f is the thickness of column flange. 1

11 Tab. A2 Thickness of column thermal rotection d, mm, [5], Fig. 11 Column West (outer) flange East (inner) flange Thickness of Above Average of rot. thickness rot. thickness rot. on web floor cross section max. min. max. min. max. min E Average on column Annex B - Measured Temeratures Tab. B1 Maximal gas temeratures in time intervals C, thermocoules mm under ceiling, thermocoules numbers see Fig. 2 [5] Thermocoule C 521 C 522 C 523 C 524 C 525 C 526 C 527 C 528 Time interval, min. aver ,4 321, 349,5 37,4 399, 422,8 386, 358, ,6 66,1 698,3 762,6 86, ,6 782, ,5 777,3 834,8 851,1 935, 971,6 964,5 885, ,3 1 16,1 1 7,3 99,5 1 17,8 1 96,3 1 63,1 979, ,6 796,2 73,5 697,2 762,6 754,5 735, 662, ,1 579,7 576,9 528,7 56,3 535, 555,1 475,1 555 Tab. B2 Steel beam temeratures C, thermocoules numbers see Figs 2 and 5 [5] Column - E2 E2 E2 Time, min. C 41 C 48 C 49 C 41 C 411 C 415 C 418 C 428 C 43 C ,6 17,6 16,9 16,1 16,3 22,6 18,4 15,9 15,9 15, , 23,2 22,2 23,4 23,2 16,2 26,6 19,8 2,8 21,2 3 31, 85, 82,9 82,5 89,4 521,2 84,8 73,3 74,6 76, ,3 16,9 16,8 19,9 17,4 726, 141,7 17,3 17,6 19, 6 88,7 191,5 25,5 215,1 28,8 976, 266,6 174,3 197,8 22,7 9 95,2 48,8 41,3 385,8 421,7 74,6 489,4 377,8 38,3 385, ,5 426, 421,8 415,9 434,9 522,2 511,2, 42,3 416, ,6 413,1 41,8 42,5 48,7 365,6 495,5 392,2 39,9 48, ,2 367,4 367,3 355, 354,1 215,2 411,5 352,1 345,7 358,7 Maximal 95,3 426, 421,8 415,9 436,3 985,8 511,2 41,4 42,3 416,6 Position 3/4 SW 3/4 SW 3/4 SE 1/2 NW 1/2 N **BF *NW 3/4 SW 1/2 SW 1/4 SW Thermocoules were located 2 mm from the column/beam edge; * mm belowe secondary beam; ** on the beam lower flange, mm from the column face. 11