Publisher’s version / Version de l'éditeur:
Construction Technology Update, 1999-12-01
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la
première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez pas à les repérer, communiquez avec nous à PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca.
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
NRC Publications Archive
Archives des publications du CNRC
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Fire performance of high-strength concrete structural members
Kodur, V. K. R.
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
NRC Publications Record / Notice d'Archives des publications de CNRC:
https://nrc-publications.canada.ca/eng/view/object/?id=23ed8799-b7c9-4d70-aca1-c1c65a6d6f00 https://publications-cnrc.canada.ca/fra/voir/objet/?id=23ed8799-b7c9-4d70-aca1-c1c65a6d6f00C o n s t r u c t i o n T e c h n o l o g y U p d a t e N o . 3 1
High -stren gth con crete (HSC) p rovid es a h igh level of stru ctu ral p erform an ce, esp ecially in stren gth an d d u rability, com p ared to trad ition al, n orm al-stren gth con crete (NSC). Previou sly em p loyed in brid ges, offsh ore stru ctu res an d in frastru ctu re p rojects, HSC h as seen in creased u se in h igh -rise bu ild in gs, esp ecially for colu m n s. Th e h igh er com p ressive stren gth of HSC allow s for th e u se of smaller-diameter columns, which increases th e am ou n t of u sable sp ace in a bu ild in g.
In bu ild in g d esign , th e p rovision of ap p rop riate fire resistan ce for stru ctu ral m em bers is a m ajor requ irem en t. Th e fire resistan ce of a stru ctu ral m em ber is d ep en d en t on its geom etry, th e m aterials u sed in its con stru ction , th e load in ten sity an d th e ch aracteristics of the fire itself. The recent version of CSA Stan d ard CSA-A23.3-M94 (“Design of Con crete Stru ctu res”) p rovid es d etailed p rovision s for th e d esign of HSC stru ctu ral m em bers. How ever, th ere are n o sp ecific gu id elin es for evalu atin g th e fire p erform an ce of HSC in eith er th is stan d ard or th e Nation al Bu ild in g Cod e of Can ad a 1995 (NBC).
T he Spa lling Proble m
Gen erally, con crete stru ctu ral m em bers p erform well under fire situations.
Studies show, h ow ever, th at th e p erform an ce of HSC d iffers gen erally from th at of NSC an d m ay n ot exhibit good fire performance. Spalling under fire con d ition s is on e of th e m ajor con cern s w ith HSC, a p roblem th at is d u e to its low w ater/ cem en t ratio.1
Sp allin g h as been observed u n d er both laboratory an d actu al fire con d ition s (see Figu re 1). A fire in th e En glish Ch an n el tu n n el in 1996, for exam p le, cau sed severe d am age to tu n n el rin gs ow in g to th e sp allin g of con crete an d resu lted in in ju ries to eigh t p eop le an d a p rop erty loss of £50 m illion .2
Th e sp allin g w as attribu ted to th e h igh stren gth of th e con crete.
Sp allin g resu lts in th e rap id loss of th e su rface layers of th e con crete d u rin g a fire. It exp oses th e core con crete to fire tem p era-tu res, th ereby in creasin g th e rate of tran s-m ission of h eat to th e core con crete an d th e rein forcem en t. Sin ce th e sp allin g occu rs in th e in itial stages of a fire, it m ay p ose a risk to evacu atin g occu p an ts an d firefigh ters.
Sp allin g is attribu ted to th e bu ild -u p of p ore p ressu re d u rin g h eatin g. HSC is believed to be m ore su scep tible to th is p ressu re bu ild -u p becau se of its low p erm e-ability, com p ared to th at of NSC.3,4 Th e
extrem ely h igh w ater vap ou r p ressu re, gen erated d u rin g exp osu re to fire, can n ot escap e becau se of th e h igh d en sity (an d low p erm eability) of HSC. Th is p ressu re often
Fire Performance of
High-Strength Concrete
Structural Members
b y V.K.R. Kod ur
The increased use of high-strength concrete in buildings and structures has led
to concerns about its fire performance, especially the problem of spalling.
This Update presents research results comparing the fire resistance of
high-strength and normal-high-strength concrete columns and offers guidelines for
improving the fire resistance of high-strength concrete structural members.
Figure 1.View of (a) normal-strength concrete column and (b) high-strength concrete column after fire-resistance tests
a
2
reaches the saturation vapour pressure, which at 300°C is abou t 8 MPa. Su ch in tern al p ressu res are often too h igh to be resisted by th e HSC, w h ich h as a ten sile stren gth of abou t 5 MPa.
Fire Pe rform a nc e of H SC a nd N SC Research is in p rogress at NRC’s In stitu te for Research in Con stru ction (IRC) to evalu ate th e fire p erform an ce of HSC an d to d evelop solu tion s th at w ill m in im ize sp allin g an d en h an ce fire resistan ce.
Typ ical resu lts from an IRC fire-resistan ce test in volvin g NSC an d HSC colu m n s are sh ow n in Figu res 2 an d 3 (asid e from th e con crete stren gth , th e tw o typ es of colu m n s w ere sim ilar an d w ere su bjected to com p a-rable load s).3,4
Figure 2 shows that temperatures, recorded at variou s d ep th s from th e su rface, at m id -h eig-h t of t-h e colu m n , are gen erally low er for th e HSC colu m n th an for th e NSC colu m n th rou gh ou t th e fire exp osu re. Th ese d ifferen ces can be attribu ted p artly to th e d ifferen ces in th e th erm al p rop erties of th e tw o con cretes an d to th e h igh er com p actn ess (low er p orosity) of HSC. Th e low er p orosity of HSC retard s th e rate of in crease in tem p eratu re u n til th e cracks w id en an d sp allin g occu rs. Large cracks occu rred in th e HSC colu m n after ap p roxim ately th ree h ou rs of fire exp osu re.
Figu re 3 sh ow s th at th e d eform ation of th e HSC colu m n is sign ifican tly low er th an th at of th e NSC colu m n . Th is can be exp lain ed in p art by th e low er th erm al exp an sion of HSC an d th e slow er rise in tem p eratu re in th e HSC colu m n d u rin g th e in itial stages of fire exp osu re d u e to th e com p actn ess of HSC.
As th e tem p eratu re con tin u es to rise, th e steel rein forcem en t in both th e NSC colu m n an d th e HSC colu m n grad u ally yield s an d th e colu m n con tracts. Wh en th is h ap p en s, th e con crete carries a p rogressively in creasin g p ortion of th e load . Th e stren gth
of th e con crete also d ecreases w ith tim e an d , u ltim ately, w h en th e colu m n can n o lon ger su p p ort th e load , failu re occu rs.
At th is stage, th e colu m n beh aviou r is d ep en d en t on th e stren gth of th e con crete (Figu re 3). Th ere is sign ifican t con traction in th e NSC colu m n lead in g to grad u al (d u c-tile) failu re. Th e low er con traction in th e HSC colu m n can be attribu ted to th e fact th at at elevated tem p eratu res HSC becom es brittle an d loses stren gth faster th an NSC.
An im p ortan t fin d in g in th e research w as th at w h ile th ere w as n o sp allin g in th e NSC colu m n , th ere w as sign ifican t sp allin g at th e corners of the HSC column just before failure. Th ere w as also som e sp allin g in th e HSC colu m n in th e in itial stages of fire exp osu re. Figure 1 shows the condition of the two types of colu m n s after th e fire resistan ce tests. In th e HSC colu m n , th e rein forcem en t (both lon gitu d in al an d tran sverse) is com p letely exp osed an d th ere is sign ifican t sp allin g.
For th e NSC colu m n s, th e fire resistan ce w as ap p roxim ately 366 m in u tes, w h ile for HSC colu m n s it w as 225 m in u tes. Th e low er fire resistan ce for th e HSC colu m n can be exp lain ed by th e th erm al an d m ech an ical p rop erties of HSC. Fu rth erm ore, th e sp allin g p h en om en on , w h ich resu lted in a d ecrease in th e cross-section of th e colu m n , also con tribu ted to th e low er fire resistan ce of th e HSC colu m n s.
At IRC, both exp erim en tal an d n u m erical stu d ies are bein g carried ou t to d evelop firresistan ce d esign gu id e-lin es for in corp oration in to cod es an d stan d ard s. In th e exp erim en tal stu d ies, 20 fu ll-scale load ed HSC colu m n s an d a n u m ber of HSC blocks are bein g exp osed to fire in a sp ecially bu ilt fu rn ace, in ord er to exam in e th e in flu en ce of variou s p aram eters on th e fire p erform an ce of HSC. Th e research is bein g con d u cted in p artn ersh ip w ith th e Portlan d Cem en t Association , th e Can ad ian Portlan d Cem en t Association , Con crete Can ad a, CANMET, MOBIL an d Nation al Ch iao Tu n g Un iversity, Taiw an .
Figure 2.Temperature distribution at various depths during fire exposure in normal-strength concrete (NSC) and high-strength concrete (HSC) columns
Figure 3.Axial deformation of normal-strength concrete (NSC) and high-strength concrete (HSC) columns during fire exposure
3 Factors Affecting Fire Performance
Research at IRC an d elsew h ere sh ow s th at th e fire p erform an ce of HSC, in gen eral, an d sp allin g, in p articu lar, is in flu en ced by th e follow in g factors:
• origin al com p ressive stren gth of th e con crete
• m oistu re con ten t of th e con crete • d en sity of th e con crete
• fire in ten sity
• d im en sion s an d sh ap e of sp ecim en s • lateral rein forcem en t
• load in g con d ition s • typ e of aggregate
Th e stu d ies u n d erw ay at IRC are seek-in g to qu an tify th e seek-in flu en ce of th e above factors an d to d evelop solu tion s to en h an ce th e fire p erform an ce of HSC. So far it h as been p ossible to d raw som e con clu sion s w ith regard to th ese factors, as d iscu ssed below.
Concrete Strength
Wh ile it is h ard to sp ecify th e exact stren gth ran ge, based on th e available in form ation , con crete stren gth s h igh er th an 55 MPa are m ore su scep tible to sp allin g an d m ay resu lt in low er fire resistan ce.
Moisture Content
Th e m oistu re con ten t, exp ressed in term s of relative h u m id ity, in flu en ces th e exten t of sp allin g. High er RH levels lead to greater sp allin g. Fire-resistance tests on full-scale HSC columns h ave sh ow n th at sign ifi-can t sp allin g occu rs w h en th e RH is h igh er th an 80% . Th e tim e requ ired to attain an accep table RH level (below 75% ) in HSC stru ctu ral m em bers is lon ger th an th at requ ired for NSC stru ctu ral m em bers becau se of th e low p erm eability of HSC. In som e cases, su ch as in offsh ore structures, RH levels can remain high throughou t th e life of th e stru ctu re an d sh ou ld th erefore be accou n ted for in th e d esign .
Concrete Density
Th e effect of con crete d en sity w as stu d ied by m ean s of fire tests on n orm al-d en sity (m ad e w ith n orm al-w eigh t aggregate) an d ligh tw eigh t (m ad e w ith ligh tw eigh t aggre-gate) HSC blocks.5 Th e exten t of sp allin g
w as fou n d to be m u ch greater w h en ligh t-w eigh t aggregate is u sed . Th is is m ain ly becau se th e ligh tw eigh t aggregate con tain s m ore free m oistu re, w h ich creates h igh er vap ou r p ressu re u n d er fire exp osu res. Fire Intensity
Th e sp allin g of HSC is m u ch m ore severe in fires ch aracterized by fast h eatin g rates
or h igh fire in ten sities. Hyd rocarbon fires p ose a severe th reat in th is regard . Wh en HSC is to be u sed in facilities w h ere h yd ro-carbon fu els are p resen t, su ch as offsh ore d rillin g stru ctu res an d h igh w ay tu n n els, th e p robable occu rren ce of sp allin g sh ou ld be con sid ered in th e d esign .
Specimen Dimensions
A review of th e literatu re sh ow s th at th e risk of exp losive th erm al sp allin g in creases w ith sp ecim en size. Th is is d u e to th e fact th at sp ecim en size is d irectly related to h eat an d m oistu re tran sp ort th rou gh th e stru ctu re, as w ell as th e cap acity of larger stru ctu res to store m ore en ergy. Th erefore, carefu l con sid eration m u st be given to th e size of sp ecim en s w h en evalu atin g th e sp allin g p roblem ; fire tests are often con -d u cte-d on sm all-scale sp ecim en s, w h ich can give m islead in g resu lts.
Lateral Reinforcement
Th e sp acin g an d con figu ration of ties both h ave a sign ifican t effect on th e p erform an ce of HSC colu m n s. Both closer tie sp acin g (at 0.75 times that required for NSC columns) and th e ben d in g of ties at 135° back in to th e core of th e colu m n , as illu strated in Figu re 4, en h an ce fire p erform an ce. Th e p rovision of cross ties also im p roves fire resistan ce. Fire tests on HSC colu m n s, w ith ad d ition al con fin em en t th rou gh cross ties an d ben d in g of ties back in to th e core of th e colu m n , h ave sh ow n th at sp allin g is sign ifican tly red u ced an d fire resistan ce as h igh as 266 m in u tes can be ach ieved even u n d er fu ll service load s.3
Load Intensity
A load ed HSC stru ctu ral m em ber w ill sp all to a greater d egree th an an u n load ed m em -ber. Th e load ad d s to th e stresses created by th e p ore p ressu re gen erated by steam . Also, a h igh er load in ten sity lead s to low er fire resistan ce, sin ce th e loss of stren gth w ith a rise in tem p eratu re is greater for HSC th an for NSC.
Type of Aggregate
Of th e tw o com m on ly u sed aggregates, car-bon ate aggregate (p red om in an tly lim eston e) p rovid es h igh er fire resistan ce an d better sp allin g resistan ce in con crete th an d oes
Figure 4.Tie configuration for reinforced concrete column: (a) conventional tie configuration; (b) modified tie configuration
siliceou s aggregate (p red om in an tly qu artz).3 Th is is m ain ly becau se
carbon ate aggregate h as a su bstan -tially h igh er h eat cap acity (sp e-cific h eat), w h ich is ben eficial in p reven tin g sp allin g. Th is in crease in sp ecific h eat is likely cau sed by th e d issociation of th e d olom ite in th e carbon ate con crete.
Fibre Reinforcement
Th e ad d ition of p olyp rop ylen e fibres m in im izes sp allin g in HSC m em bers u n d er fire con d ition s.1
On e of th e m ost accep ted th eories on th is is th at by m eltin g at a relatively low tem p eratu re of 170°C, th e p olyp rop ylen e fibres create “ch an n els” for th e steam p ressu re in con crete to escap e, th u s p reven tin g th e sm all “exp losion s” th at cau se sp allin g. Th e stu d y sh ow ed th at th e am ou n t of p oly-propylene fibres needed to minimize sp allin g is abou t 0.1 to 0.25% (by volu m e). Fu rth er research is bein g carried ou t to d eterm in e th e op tim u m fibre con ten t for d ifferen t typ es of con crete. Th e effect of p olyp rop ylen e fibres on sp allin g is illu strated in Figu re 5, w h ich sh ow s HSC con crete blocks after tw o h ou rs of fire exp osu re.5
Steel fibres also red u ce sp allin g in HSC an d im p rove fire resistan ce. Th e fibres en h an ce th e ten sile stren gth of con crete, even at h igh temperatures, and help to with-stand th e p ore p ressu re gen erated d u e to vap ou rization of w ater u n d er fire exp osu re. With th ese fibres th e tensile strength
increases to between 5 an d 7 MPa, w h ich in m an y cases m ay be su fficien t to ach ieve tw o to three hours of fire resistance without sign ifican t sp allin g. Fu ll-scale tests h ave sh ow n th at th e u se of steel fibres in HSC significantly improves th e fire-resistan ce p erform an ce of HSC-filled tu bu lar steel colu m n s.6 How ever, w h en th e p ore p
ressu re exceed s th e ten sile stren gth of con -crete, sp allin g m ay still occu r.
Guide line s for Enha nc ing Fire Pe rform a nc e
High -stren gth con crete is a h igh -p erform in g m aterial th at offers a n u m ber of ad van tages. En gin eers can en h an ce its fire p erform an ce
by ad op tin g th e follow in g gu id elin es: • Use n orm al-w eigh t aggregate (in stead of
ligh tw eigh t aggregate) to m in im ize sp allin g.
• Ad d p olyp rop ylen e fibres to th e m ix to red u ce sp allin g.
• Ad d steel fibres to en h an ce ten sile stren gth an d red u ce sp allin g. • Use carbon ate aggregate (in stead of
siliceou s aggregate) to red u ce sp allin g. • Em p loy both closer tie sp acin g an d cross
ties to im p rove fire resistan ce.
• In stall ben t ties (at 135° back in to th e con crete core) in stead of straigh t ties. • Take ap p rop riate p recau tion s to p reven t
sp allin g w h en con crete stren gth exceed s 55 MPa.
Sum m a ry
Th e attractive p rop erties of h igh -stren gth con crete m u st be w eigh ed again st con cern s abou t its fire resistan ce an d its su scep tibility to sp allin g at elevated tem p eratu res.
Research at IRC is con tin u in g an d w ill resu lt in m ore p recise d esign gu id elin es. Re fe re nc e s
1. Kod u r, V.K.R. Sp allin g in h igh stren gth con crete exp osed to fire — con cern s, cau ses, critical p aram eters an d cu res, Proceed in gs, ASCE Stru ctu res Con gress, Ph ilad elp h ia, PA, 2000.
2. Ulm , F.J., Acker, P. an d Levy, M. Ch u n n el fire. II: an alysis of con crete d am age, Jou rn al of En gin eerin g Mech an ics, 125(3), 1999, p p . 283-289.
3. Kod u r, V.K.R. an d Su ltan , M.A. Stru ctu ral beh aviou r of h igh stren gth con crete colu m n s exp osed to fire, In tern ation al Sym p osiu m on High Perform an ce an d Reactive Pow d er Con cretes, Sh erbrooke, QC, 1998, p p . 217-232.
4. Lie, T.T. an d Woollerton , J.L. Fire resistan ce of rein forced con crete colu m n s: test resu lts, In stitu te for Research in Con stru ction , Nation al Research Cou n cil of Can ad a, In tern al Rep ort No. 569, Ottaw a, 1988, 302 p p . 5. Bilod eau , A., Malh otra, V.M. an d Hoff, G.C. Hyd rocarbon
fire resistan ce of h igh stren gth n orm al w eigh t an d ligh t w eigh t con crete in corp oratin g p olyp rop ylen e fibres, In tern ation al Sym p osiu m on High Perform an ce an d Reactive Pow d er Con cretes, Sh erbrooke, QC, 1998, p p . 271-296.
6. Kod u r, V.K.R. Perform an ce of h igh stren gth con crete-filled steel colu m n s exp osed to fire, Can ad ian Jou rn al of Civil En gin eerin g, 25(6), 1998, p p . 975-981.
Dr. V.K.R. Kod uris a research officer in th e Fire Risk Man agem en t Program of th e N ation al Research Cou n cil’s In stitu te for Research in Con stru ction .
“Construction Te chnology Up d a te s” is a se rie s of te chnica l a rticle s conta ining p ra ctica l inform a tion d istille d from re ce nt construction re se a rch.
For more information, contact Institute for Research in Construction, National Research Council of Canada, Ottaw a K1A 0R6
Telephone: (613) 993-2607; Facsimile: (613) 952-7673; Internet: http://irc.nrc-cnrc.gc.ca
© 1999
Nation al Research Cou n cil of Can ad a Decem ber 1999
ISSN 1206-1220
Figure 5.View of high-strength concrete blocks after two-hour hydrocar-bon fire test: (a) block without polypropylene fibres; (b) block with polypropylene fibres
a
