THE DESIGN FOR FIRE PARTICULARLY FOR MULTI-STOREY STRUCTURES Patrick JE Sullivan City University London 1. Abstract Multi-storey and other indeterminate structures need special consideration when designing for fire. Prescriptive design is not suitable and will be described together with its shortcomings. Performance design is more appropriate for complex structures. Performance design can be carried out by using conventional ultimate limit state design in bending under ambient conditions and superimposing effects of fire loading. Alternately computer programs using iterative finite difference and/or finite element techniques can be adopted taking into account the losses of strength and other properties due to elevated temperatures including the transient effect of the fire load. Among the computer programs available for fire design, there are some which take moisture into account, which may be more appropriate for some types of concrete. Another method discretises the domain by random lattice modelling using irregular elements and energy considerations where cracking is determined by non-linear fracture mechanics. This latter simulates heterogeneous materials (like concrete) more closely than the use of triangular or rectangular elements. Also energy methods are more likely to predict explosive spalling than elastic methods. In addition to the above analysis, which deals with the effects of fire on structural members, it is recommended that a complete analysis of the structure is also performed. The results of a grid analysis of a 7 storey by 5 bay RC structure with a fire affecting a compartment in different parts of the structure carried out in the early 70’s will be described. The fire has been assumed to be subjected to the effects of a standard time-temperature curve representing a period from 0 to 2 hours and acts uniformly on both columns and beams in the compartment. This aggressive fire has been specially selected to accelerate the formation of plastic hinges in the structure. It will be shown that the behaviour of the structure depends on the location of the fire, the relative stiffness of beam and column, the applied live loading particularly alternate bay loading and is very much dependent on the detail and design of the building. The paper will stress the importance of overall design for fire, since fire requires reinforcing bars in areas where applied loading may not need them To emphasise the importance of detailing in design examples will be shown of a plastic covered lifeboat or survival craft, which fails a fuel fire test due to a minor detail. After incorporating the detail, the survival craft passes the test. A boat with an alternate design passes the fuel fire test the first time round 2. Suggested keywords Multi-storey structures, fire, survival craft, prescriptive, performance, design and detail 3. Biographical sketch of the author Patrick Sullivan is currently a visiting Professor and Senior Research Fellow at City University. He is also an assessor and monitor for Research and development work awarded to UK industry in conjunction with Academia by the Department of Trade’s Technology programme on Innovation. Prior to this he worked with consultants and contractors in design for 10 years and 25 years at Imperial College London, where his research interest in fire started. During his time at Imperial he also had his own commercial laboratory investigating deteriorating structures, developing new materials and testing equipment. He carried out many site test including load tests on bridges and floor slabs and realistic fire tests. He has been involved as an expert witness in a number of high profile litigation cases.

4. REFERENCES 1. Ingberg SH. Tests of the severity of building fires. National Fire Protection Quarterly.

Vol. 22, No. 1, 1928. p. 43. 2. ENV 1992-1-2 Eurocode 2 Design of concrete structures. Part 1- 2 General Rules –

Structural fire design. 3. Connor MA, Kirby BR & Martin DM. Behaviour of a multi-storey composite steel

framed building in fire. The Structural Engineer V 81 no. 3 January 2003, p 27-36 4. Introduction to the fire safety engineering of structures (2003). The Institution of

Structural Engineers London England, Ref. No. 413 60 pages. 5. Sullivan PJE Behaviour of building structures. Chapter XI "Fire and buildings" Editor.

Marchant, EW, Medical and Tech. Publication, 1972. p. 150-170. 6. Sullivan PJE Computer model of a fire in a building. Fire Protection Science and

Technology, No. 7 November 1973. p. 12-19. 7. Bandyopadhyay B. Computer analysis of skeletal structures with special reference to

deterioration due to fire. PhD, University of London, 1975. 8. Gustaferro AH. Temperature criterion at failure. Fire Test Performance, ASTM Special

Technical Publication STP 464, 1970, p 68-84. 9. Gustaferro AH. Design of pre-stressed concrete for fire. PCI Journal Nov./Dec.1973. 10. Design and Detailing of Concrete Structures for Fire. The Institution of Structural

Engineers London England April 1978, 59 pages. 11. Sullivan PJE and Dougill JW. Developments in design of structural concrete under fire

conditions. International Seminar on "Three Decades of Structural Fire Safety" - Borehamwood, Feb 1983. p. 261-277.

12. Dougill JW. Materials dominated aspects of fire design for structural fire resistance of concrete structures. ACI Special Publication SP 80-4. p. 151-174.

13. Hertz K Comments on simplified calculation method for fire exposed concrete columns. Department of Buildings and Energy, Technical University of Denmark, June 1998, 20pages

14. Bizri H. Structural capacity of reinforced concrete columns subjected to fire induced thermal gradients. Dept. of Civil Eng. Univ. of California, Berkeley (UCB), Report No. UC SESM 73-1, January 1973.

15. Becker J and Bresler B. "FIRES-RC". A computer program for the fire response of structures - reinforced concrete frames. Fire Research group Report UCB FRG 74-3 July 1974.

16. Bresler B. Response of reinforced concrete frames to fire. Fire research group report. FRG WP 76-4. UCB 1976. (Presented at IABSE Tokyo meeting, September 1976.

17. Sullivan PJE, Armaghani A and Ooi HB. Pre- and post-cracking behaviour of R.C. members subject to imposed and fire loads. Proceedings international conference on Applied Numerical Modelling, Southampton, UK, 1977.

18. Towler K, Khoury GA and Sullivan PJE. Computer modelling of the effect of fire on structures. Fire Safety Design Consultants Interim Report to Dept. of Environment England, July 1989.

19. Terro MJ. Numerical modelling of thermal and structural response of RC structures in fire. PhD , University of London, 1991.

20. Terro MJ and Sullivan PJE. Model of reinforced concrete under fire. Proc. of international conference on Materials and Design against Fire. IMechE London 1992.

21. Jeyarupilingam K. Steel, steel/concrete composite and RC beams and columns exposed to fire. PhD City University, London p. 181, 1996.

22. Sullivan PJE, Terro MJ and Morris AW. Critical review of fire dedicated thermal and structural computer programs. J. Applied Fire Science V 3 (2) 113-135, 1993-94.

23. Harada K. Heat and Mass transfer in an Intensely Heated Wall. Faculty of Engineering Kyoto University 1995, 10 pages.

24. Consolazio GR, McVay MC & Rish JW. Measurement and prediction of pore pressures in saturated cement mortar subjected to radiant heating. American Concrete Institute Materials Journal 95(5) 1998 p 525-536.

25. Consolazio GR & Chung JH. The simulation of near-surface moisture migration and stress development in concrete exposed to fire - accepted for publication in "Computers and Concrete", 37 pages.

26. Tenchev RT, Li LY & Purkiss JA. Finite Element Analysis of Coupled Heat and Moisture Transfer in Concrete Subjected to Fire. Numerical Heat Transfer Part A, 2001 p 685-710.

27. Khoury GA, Majorana CE, Pesavento F & Schrefler BA. Modelling of Heated Concrete Magazine of Concrete Research, 2002, April, p 77-101.

28. Okabe A, Boots B & Sugihara K. Spatial Tesselations - Concepts and Applications of Voronoi Diagrams. John Wiley and Sons Ltd., Chichester UK, 532 pages.

29. Bolander JE & Saito S. Fracture Analysis using Spring Networks with Random Geometry. Engineering Fracture Mechanics 1998 Volume 61 p 569-591.

30. Bolander JE & Berton S. Shrinkage Induced Cracking in Cement Composite Overlays. Fifth World Congress in Computational Mechanics, Vienna Austria July 2002, 12 pages.

31. Sullivan PJE. Review on a Structural Fire Engineering report on a fire in a structural steel building under construction entitled "Investigation of Broadgate Phase 8 fire" The Steel Construction Institute publication, August 1991.

32. Bailey CG Lennon T. & Moore DB. The behaviour of full-scale steel-framed buildings subjected to compartment fires. Journal of the I.Struct.E. V 77 No. 88 April 99 p. 15-21.

33. Nuttal H. Steel in Hong Kong. New Steel Construction, Feb/March 1999.p. 30-32. 34. Sullivan PJE Prescriptive or performance design for fire? Invited paper at ACI Nawy

symposium “Serviceability for Concrete” American Concrete Institute SP 226-2 April 2005 p 1- 26

35. Sullivan PJE Advantage of performance design for fire. Invited lecture Proc. International Workshop Proc. “Fire tests of concrete/steel frames & performance-based codes” June 2006 p 1- 18.

5. List of notations As Area of steel reinforcing bars fy Yield stress of steel g Load factor la Lever arm L Span Mc Moment of resistance at centre Mct Moment of resistance at centre at high temperature Ms Moment of resistance at supports Mst Moment of resistance at supports at high temperature r = Ms/ Mc or restraint at the supports W Uniformly distributed load over entire span 6. A list power point presentation.

1. Design for fire in particular multi-storey structures 2. Deformation and stresses of SS beam due to point load & temperature cross-fall 3. Deflections of structural members and frames due to fire 4. Objective if fire design 5. Prescriptive Design 6. Performance Design 7. Conventional ULS Fire Design 8. SS beam Example 9. Continuous beam Example

10. Improvement of fire resistance 11. Shortcomings in ULS assumptions 12. Principles of Stress analysis 13. Thermal & structural programs for fire design 14. Analysis of Frameworks 15. Assumptions for Frameworks 16. Plastic hinges formed at central top compartment under fire with beams carrying dead

loads 17. Topology of structural framework 18. Plastic hinges formed at central/middle compartment under fire with beams carrying

dead loads and hinges allowed to form in the centre of beam 19. Computed deformation diagram of compartment at 625oC in previous example 20. Stress diagram at the mid-section of upper beam at 625oC in previous example 21. Live load arrangement on alternate bay for example in 22 and 23. 22. Plastic hinges formed at central/middle compartment under fire with beams carrying

dead loads and alternate bay loading Hinges allowed to form at beam end only 23. Plastic hinges formed at central/middle compartment under fire with beams carrying

dead loads and alternate bay loading. Hinges allowed to form in the centre of beam 24. Plastic hinges formed at central/middle compartment under fire with beams carrying

dead loads and live loads on all beams. Hinges allowed to form in the centre of beam 25. Resume of results 26. Remarks 27. Conclusion for Fire Design 28. Additional Conclusions 29. Emphasis on importance of detail design 30. GRP Survival Craft with engine running at start of a fuel fire test. 31. GRP Survival Craft with engine running during the fuel fire test 32. Failure of GRP Survival Craft with engine running at end of a fuel fire test 33. Modified GRP Survival Craft with engine running at start of a fuel fire test. 34. Modified GRP Survival Craft with engine running during the fuel fire test 35. Modified GRP Survival Craft with engine running at end passes the fuel fire test 36. Alternative design of GRP Survival Craft with engine running at start of a fire test 37. Same GRP Survival Craft with engine running during the fire test 38. Same GRP Survival Craft with engine running still serviceable and therefore passes

test

Design for fire in particular multi-storey structures

Patrick JE SullivanCity University London

Objective of fire design

• Limit heat transmission through walls & floors.• Maintain structural stability for a sufficient periodCurrent codes allow use of Prescriptive - OK for SS &

common & garden members or: use of Performance - more suitable for indeterminate members.

Performance design can be carried out by:• ULS analysis with fire as additional load (rational)• Analytical methods using computer programs

Prescriptive design involves• Selection of tabulated values from C of P for cover

etc. - derived from tests on SS members at a given mc, heated in a furnace at a standard ISO rate.

• Generally conservative - design restrictive• A 2-hour FR does not imply a member lasts 2 hrsa) To maintain ISO curve, fuel input to furnace has to

be increased for higher thermal inertia material -Not a fire! Also curves dissimilar

b) ss members rarely found in structures & unlikely to have concrete members cured in standard manner

Performance Design Process involves

• Analysis of structural members for a given fire

• Estimating material Temp - material properties are modified. Hence thermal s/e

• These are added to the load-imposed s/e• Estimating additional stress resultants• Hence F.R. of structural member

Conventional ULS Fire Design

• Originally based for beams in bending. Hence steel in tension governs the design

• Temp curves for given fire periods - in COP & properties for concrete/steel at high temp

• Hence reduction in moment capacity - FR• The principle and advantage of the method

can best be illustrated by simple examples that follow.

Simply supported beam (subjected to a fire)

At normal temperature:gMc = gWL/8 = Mrs

= Asfyla ….1At high temperature Yfy

Hence beam survives ifWL/8 < AsYfyla ….2substituting in 1 gY > 1i.e. FR depends on f, l, g

gMcgMc

W

Trade off - more important practical members -Continuous beam

To survive after heating:Mct + Mst > WL/8

but Mst = Msand as before(Mrt) = Mct = YMcSubstituting above in rhs

1-r(g-1) < gY for r = 0 1 < gY as before for r =1 g =2 LHS =0 &g & Y = +n or (FR inf)

McMc

MsMs MsMs

Mc + Ms = Mc + Ms = gWL/8gWL/8

Ms/Mc = r (restraint)Ms/Mc = r (restraint)

W

Improvement of fire resistance

• increase cover to bottom steel (with care)• Increase lever arm (or depth/reduce cover)• Increase rebars at supports • change type and/or area of steel• reduce/increase fire intensity – (function

member)• change load factor (importance of structure)

Shortcomings in assumptions

• Flexure only considered• Full moment redistribution• Upper bound for assumed collapse • Shear loads not considered as we are not dealing

with short deep beams• Many advances since the 70’s and methods now

available to assess the FR of various members including axially & eccentrically loaded columns

Principles of Stress Analysis

• Temp at different time/depths assessed/FF -element properties revised hence FTS+LITS

• FTS & LITS deformations due to temp • Temp gradient induces bending/direct

stresses (sb +sd). Stresses (s) from applied loads & model, hence Total S = s + sb +sd

• At each iteration check total stresses etc. until end of fire or failure at sultimate / dultimate

Many thermal/structural programs available for FR

• Iterative methods. FD or FE for temp diffusion & FE for struct. part - (some with moisture)

• Critical assessment of 14 such programs developed in USA & EU – refs. in Journal Applied Fire & Science V3(2) ‘94

• There is a technique - discretises the domain by random lattice modelling (V tesselations) elements irregular shape, cracking determined by use of energy balance relationships within framework of non linear fracture mechanics

• Explosive spalling (brittle, sudden release of energy) -Energy methods rather than elastic methods, which so far can only predict gradual surface spalling

Main part - Analyses - framework

• All programs in 70’s on members or portal• Initiated grid analyses programs for 2 &

3-D skeletal frames • 7 storey x 5 bay RC structure designed for

normal loading subjected to a standard fire in a central compartment and determined:

• History of variation of BM, Axial & SF & deformations with temperature increase

Assumptions - framework

• Members heated uniformly along L.• Concrete/steel strength change with temp &

N/A assessed. Ec also reduces with temp• Creep modifies high axial loads & BMs due

to restraints as temperature increases• Mmt/Rotns multi-linear but perfectly plastic

at hinges. But M/Rs & Mp reduce with temp• Excess moments developed - redistributed

Although the members were subjected to unrealistically severe fires & member

design simplified - results• Hinge formation dependent on design, fire location

and manner of loading• Hinges may form remote from fire• Deformations and stresses dependent on loading and

relative stiffness of members• Serious implications in FD if all factors are not taken

into account in indeterminate structures and tall buildings

(Other reasons why cover is not sufficient protection for fire)

REMARKS

• Well detailed structures can have a reserve of strength particularly for accidental loads

• A good design can improve performance of structure independent of material

• By proper detailing, material can serve purpose of designer instead of being subjected to its weaknesses

For fire design we can conclude

• ULS more realistic than prescriptive design & allows greater economies & flexibility to designer - preferred for complex structures

Computer analysis of:• members more accurate• frameworks - a better understanding of O/A

behaviour

In addition we may conclude

• Use of relevant material properties appropriate models & particularly good

designreleases engineer from being a slave to

material in such a way that:• FLAMMABLE MATERIAL like plastic

can be made to behave as a FIREPROOF ONE

And to emphasise importance of detail design)

1. Slides of a GRP survival craft failing a fuel fire test due to a poor detail of spray nozzles in the hull

(boat in boomed off area + 1000 gals kerosene boat to last 10 + mins. fuel fire)

2. With minor modifications craft passes test.3. A better designed boat with alternative

details, which passes test first time.