CLT - Cross Laminated Timber Fire protection.

Transcription

1 CLT - Cross Lamnated Tmber Fre protecton

2 C O N T E N T S Verson 01/2014 AG It should be noted that ths fact sheet on fre protecton s merely ntended to support the user. Stora Enso Wood Products does not assume any responsblty for the accuracy or completeness of ths document. 1 Introducton Fre protecton The buldng materal, wood when exposed to fre Reacton-to-fre performance of buldng products Fre resstance of buldng components Classfcaton Fre protectve claddng Fre resstance of CLT components Verfcaton of the fre resstance of CLT elements based on classfcaton reports accordng to EN CLT external wall structures CLT wall structures CLT celng structures Verfcaton of the fre resstance of CLT elements based on calculatons accordng to EN :2011 (Eurocode 5) Verfcaton method for actons n the fre stuaton accordng to EN : Verfcaton method for mechancal resstance n the fre stuaton accordng to EN : Charrng rates for Stora Enso CLT Desgn value of charrng rates 0 for CLT on surfaces whch are unprotected throughout the duraton of the fre Desgn value of charrng rates 0 for CLT on surfaces whch are ntally protected from exposure to fre by gypsum plasterboard Page 1 of 50

3 C O N T E N T S 5.4 Determnng the load-bearng capacty (R) of CLT elements accordng to EN : Determnng the ntegrty (E) and nsulaton (I) of CLT elements The nsulatng and protectve layers of a component Determnng a layer s basc tmes Calculatng the poston coeffcent k pos Determnng the jont coeffcent k j, Calculaton model/applcaton for CLT Techncal fre protecton of detalng Jonts and connectons between components Installatons and fxtures Fre-retardng sealng n CLT constructons Fasteners Double-layered components Bblography Annex 1 Desgn examples A 1.1 Determnng the charrng rate of Stora Enso CLT (non-faced) A 1.2 Determnng the charrng rate of Stora Enso CLT (faced) A 1.3 Determnng the notonal charrng depth of Stora Enso CLT (non-faced) A 1.4 Determnng the ntegrty (EI) of Stora Enso CLT (faced) Annex 2 Informaton on the resdual cross-secton and ntegrty of Stora Enso CLT components A 2.1 Calculated resdual cross-secton and ntegrty of CLT wall structures (non-faced) A 2.2 Calculated resdual cross-secton and ntegrty of CLT wall structures wth sngle layered fre protecton plasterboard claddng (12.5 mm) on the fre-exposed sde of the component A 2.3 Calculated resdual cross-secton and ntegrty of CLT wall structures wth servce cavty (50 mm, rock fbre) and sngle layered fre protecton plasterboard claddng (12.5 mm) on the fre-exposed sde of the component A 2.4 Calculated resdual cross-secton and ntegrty of CLT celng structures (non-faced) Page 2 of 50

4 I N T R O D U C T I O N 1 Introducton 1.1 Fre protecton Fre protecton n ts entrety s a complex system whch can be broken down nto several areas: Preventve fre protecton o Organsatonal fre protecton o Plant-specfc fre protecton o Constructonal fre protecton Defensve fre protecton Ths document only deals wth the subject of buldng materal- and component-specfc propertes of CLT elements whch, n reference to the fre protecton groups outlned above, are consdered to belong to the constructonal fre protecton group. 1.2 The buldng materal, wood when exposed to fre If the buldng materal, wood s exposed to fre and thus to an elevated supply of energy, ts temperature rses and the water molecules embedded wthn start to evaporate at approx. 100 C. At C, the long-chan molecules n the cell walls splt, producng gaseous and flammable compounds and the gas subsequently enters the surface of the wood where t reacts wth oxygen n the ar, and combusts. [1] These chemcal compounds decompose n a process known as pyrolyss (whereby gas emssons from combustble components n the wood burst nto flame), gradually spreadng along the wood, leavng a charrng area behnd t. Ths char layer s formed from the carbonaceous resdue of pyrolyss, whch burns, generatng embers. Ths layer s propertes n partcular, low densty and hgh permeablty act as heat nsulaton and protect the underlyng, undamaged wood. Page 3 of 50

5 I N T R O D U C T I O N Fg. 1: cross-secton of an 80 mm CLT element, orgnally clad wth fre protecton plasterboard, after a large-scale fre test Fgure 1 shows the cross-cut secton of a CLT element, clad wth fre protecton plasterboard after a large-scale fre test. It s possble to dentfy the dfferent layers on ths cross secton: the charred area (black area), followed by the pyrolyss area (brown area) caused by the spreadng fre or pyrolyss and the undamaged wood. Fg. 2: char layer of an 80 mm CLT element, orgnally clad wth fre protecton plasterboard, after a large-scale fre test Page 4 of 50

6 R E A C T I O N - T O - F I R E P E R F O R M A N C E O F B U I L D I N G P R O D U C T S 2 Reacton-to-fre performance of buldng products The reacton-to-fre performance of buldng products s classfed accordng to EN Euro classes: A1, A2, B, C, D, E, F (Crtera: gntablty, flame propagaton, heat release) Smoke classes: s1, s2, s3 (s1 > lowest smoke producton) Burnng droplets classes: d0, d1, d2 (d0 > no flamng droplets) The reacton-to-fre performance of Stora Enso CLT s classfed accordng to [26] as D-s2, d0. When usng CLT for raw floors (.e. wthout any floor structure), D fl -s1 apples. When usng flame retardants whch can delay the combuston of derved tmber products and reduce the subsequent release of energy, the fre behavour of CLT can, dependng on the retardant used, be classfed as class C or also B. When used outdoors wth the related lkely effects of humdty and drect exposure to weather, t must be ensured that the product has the necessary propertes and resstance. Page 5 of 50

7 F I R E R E S I S T A N C E O F B U I L D I N G C O M P O N E N T S 3 Fre resstance of buldng components 3.1 Classfcaton The performance characterstcs and fre resstance duraton are defned as follows accordng to the classfcaton standard EN : R (Load-bearng capacty) The performance characterstc R s assumed to be satsfed f the load-bearng functon of a component subjected to a mechancal load s mantaned durng the requred tme of fre exposure. E (Integrty) The performance characterstc E of a separatng element descrbes ts capacty to resst exposure to fre on one sde so that the spread of fre as a result of flames or hot gases to the sde not exposed to fre s prevented. I (Insulaton) The performance characterstc I s assumed to be satsfed f the average temperature rse over the whole of the non-exposed sde caused by one-sded exposure to fre does not exceed 140 C, and the maxmum temperature at any one pont does not exceed 180 C, above ambent temperature. Heat transmsson must be lmted so that the non-exposed component surface and any neghbourng materals do not catch fre, and so that any persons n the vcnty are protected. Addtonal characterstcs are W (thermal radaton), M (resstance to mechancal acton), C (self-closng capablty), S (smoke leakage), G (soot fre resstance) and K (fre protecton ablty), whch, n general, are not relevant for conventonal CLT components. The classfcaton tme s graduated n perods of 10, 15, 20, 30, 45, 60, 90, 120, 180, 240 and 360 mnutes. The drecton of the classfed fre resstance duraton s descrbed wth the followng abbrevatons accordng to EN : Classfcaton of façades (curtan-walls) and external walls o o o classfed fre resstance duraton from nsde to outsde classfed fre resstance duraton from outsde to nsde classfed fre resstance duraton from nsde to outsde and from outsde to nsde Classfcaton of celngs wth ndependent fre resstance a b a b a b classfed fre resstance duraton from top to bottom classfed fre resstance duraton from bottom to top classfed fre resstance duraton from top to bottom and from bottom to top Page 6 of 50

8 F I R E R E S I S T A N C E O F B U I L D I N G C O M P O N E N T S 3.2 Fre protectve claddng The desgnatons K 1 and K 2 for fre protectve claddng are descrbed n accordance wth EN as follows: The term fre protectve claddng refers to the outermost layer on vertcal components and to the lowest layer on horzontal or nclned components. The claddng defned by the desgnaton K 1 or K 2 must provde the protecton descrbed n accordance wth EN for the layer drectly backng the fre protectve claddng throughout the correspondng classfcaton perod (K 1 : 10 mn.; K 2 : 10, 30 or 60 mn.). For example, fre protectve claddng wthout an underlyng cavty wth the classfcaton K 2 60 must provde the followng protecton for a perod of 60 mnutes: Durng the classfcaton perod, the fre protectve claddng must not collapse n whole or n part. The average temperature recorded on the undersde of the supportng plate (on whch the fre protectve claddng to be classfed s tested) must not exceed the ambent temperature by more than 250 C. The maxmum temperature recorded on (at any one pont) on the undersde of the supportng plate (on whch the fre protectve claddng to be classfed s tested) must not exceed the ambent temperature by more than 270 C. After the test, no burned or charred materals should be apparent on any one pont of the supportng plate (on whch the fre protectve claddng to be classfed s tested). Informaton on the fre protectve claddng to be classfed can be obtaned from gypsum plasterboard manufacturers, among others. 3.3 Fre resstance of CLT components Components wth hgh fre resstance can be produced wth multple layer CLT elements. For example, wth a non-clad, three-layer CLT element, the fre resstance REI 60 s already obtaned, and wth a CLT element clad wth a sngle layer of plasterboard, the fre resstance REI 90 s obtaned. In prncple, ncreased requrements for fre resstance can be compensated by the followng measures: Increase the thckness of the CLT element Increase the number of layers of the CLT element Apply the correspondng claddng The verfcaton of fre resstance of tmber components can ether be based on classfcaton reports n accordance wth EN on the bass of large-scale fre tests, or on calculatons accordng to EN , performed n conjuncton wth the respectve natonal applcaton documents. Page 7 of 50

9 V E R I F I C A T I O N O F F I R E R E S I S T A N C E 4 Verfcaton of the fre resstance of CLT elements based on classfcaton reports accordng to EN Stora Enso commssoned dfferent accredted test nsttutes to test the fre resstance of dfferent CLT elements wth dfferent component desgns accordng to EN or EN The results of the classfcaton reports ([27], [28], [29], [30], [31], [32] and [36]) from the fre resstance tests, performed n accordance wth EN , are as follows. 4.1 CLT external wall structures Classfcatons of the fre resstance REI 90 of load-bearng cross-lamnated tmber elements as external wall elements: Internal claddng 12.5 mm fre protecton plasterboard 12.5 mm fre protecton plasterboard 12.5 mm fre protecton plasterboard 12.5 mm fre protecton plasterboard 2 15 mm fre protecton plasterboard 2 15 mm fre protecton plasterboard 12.5 mm fre protecton plasterboard 12.5 mm fre protecton plasterboard Servce cavty Cross-lamnated tmber element External claddng Desgnaton Lamella structure CLT 100 C3s mm wood wool slab, 15 mm plaster CLT 100 C3s mm rock fbre, 4 mm plaster CLT 100 C5s mm wood wool slab, 15 mm plaster CLT 100 C5s mm rock fbre, 4 mm plaster CLT 80 C3s mm wood wool slab, 15 mm plaster CLT 80 C3s mm rock fbre, 4 mm plaster rock fbre (40 mm) rock fbre (40 mm) CLT 100 C3s mm wood wool slab, 15 mm plaster CLT 100 C3s mm rock fbre, 4 mm plaster Table 1: classfcatons of the tested components n accordance wth [32] Test load [kn/m] Classfcaton o 35 REI REI REI REI 90 EI 90 EI REI REI 90 Page 8 of 50

10 V E R I F I C A T I O N O F F I R E R E S I S T A N C E 4.2 CLT wall structures Classfcatons of the fre resstance REI 60 of load-bearng cross-lamnated tmber elements as wall elements: Claddng Servce cavty Cross-lamnated tmber element Test load Classfcaton Desgnaton Lamella structure [kn/m] CLT 100 C3s REI 60 CLT 100 C5s REI mm fre CLT 80 C3s REI 60 protecton plasterboard 12.5 mm fre rock fbre (40 mm) CLT 80 C3s REI 60 protecton plasterboard 50 mm Heraklth BM, 5 mm plaster CLT 80 C3s REI 60 Table 2: classfcatons of the tested components n accordance wth [27] Classfcatons of the fre resstance REI 90 or EI 90 of load-bearng cross-lamnated tmber elements as wall elements: Claddng Cavty Cross-lamnated tmber element Test load Classfcaton Desgnaton Lamella structure [kn/m] 12.5 mm fre CLT 100 C3s REI 90 protecton plasterboard 12.5 mm fre CLT 100 C5s REI 90 protecton plasterboard 12.5 mm fre rock fbre (40 mm) CLT 100 C3s REI 90 protecton plasterboard 2 15 mm fre protecton plasterboard CLT 80 C3s EI 90 Table 3: classfcatons of the tested components n accordance wth [28] Claddng Cavty Cross-lamnated tmber element Test load Classfcaton Desgnaton Lamella structure [kn/m] 35 mm ProCrea clay board, 5 mm ProCrea clay plaster wth renforcement fabrc, 5 mm ProCrea clay plaster CLT 140 C5s REI 90 Table 4: classfcaton of the tested components n accordance wth [36] Page 9 of 50

11 V E R I F I C A T I O N O F F I R E R E S I S T A N C E Classfcatons of the fre resstance REI 120 of load-bearng cross-lamnated tmber elements as wall elements: Claddng Cavty Cross-lamnated tmber element Test load Classfcaton Desgnaton Lamella structure [kn/m] 12.5 mm fre protecton plasterboard rock fbre (40 mm) CLT 100 C3s REI 120 Table 5: classfcaton of the tested components n accordance wth [29] 4.3 CLT celng structures Classfcatons of the fre resstance REI 60 of load-bearng cross-lamnated tmber elements as celng/roof elements: Claddng Suspended celng Cross-lamnated tmber element Test load Classfcaton Desgnaton Lamella structure [kn/m 2 ] 12.5 mm fre CLT 100 L3s *1 REI 60 protecton plasterboard (on the unexposed sde) or floor structure CLT 140 L5s REI 60 Table 6: classfcatons of the tested components n accordance wth [30] Note *1: The 12.5 mm fre protecton plasterboard appled to the unexposed sde durng the fre resstance test was used to smulate a floor structure. Classfcatons of the fre resstance REI 90 of load-bearng cross-lamnated tmber elements as celng/roof elements: Claddng Suspended celng Cross-lamnated tmber element Test load Classfcaton Desgnaton Lamella structure [kn/m 2 ] CLT 160 L5s REI mm fre CLT 140 L5s REI 90 protecton plasterboard 35 mm Heraklth CLT 140 L5s REI 90 EPV 12.5 mm fre protecton plasterboard rock fbre (40 mm) CLT 140 L5s REI 90 Table 7: classfcatons of the tested components n accordance wth [31] The classfcaton reports can be downloaded from Page 10 of 50

12 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E 5 Verfcaton of the fre resstance of CLT elements based on calculatons accordng to EN :2011 (Eurocode 5) For the requred tme of fre exposure t, t must be demonstrated that: E d, f Rd, t, f where: E d,f R d,t,f s the desgn value of actons for the fre stuaton ( load effect) (Wth materals other than wood, thermal expanson must also be taken nto account.) s the correspondng desgn resstance n the fre stuaton ( resstance) 5.1 Verfcaton method for actons n the fre stuaton accordng to EN :2011 The desgn value for actons n the fre stuaton should be determned for tme t 0 usng combnaton factors 1,1 or 2,1 accordng to EN :2002, clause (See also EN , clause ) E da Gk, j" + " P" + " Ad " + " Qk,1 ( ψ 1,1 oderψ 2,1)" + " ψ 2, Qk, where: G k,j s the characterstc value of a permanent acton j P s the decsve representatve value of a pre-load A d s the desgn value of an exceptonal acton Q k,1 s the characterstc value of a decsve varable acton 1 Q k, s the characterstc value of a decsve varable acton 1 s the combnaton factor for frequent values of varable actons s the combnaton factor for quas-permanent values of varable actons 2 It s up to the user to choose/use 1,2 or 2,1. Page 11 of 50

13 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E For smplcty, the desgn value for actons n the fre stuaton E d,f from the calculaton of the desgn value for actons at normal temperature E d may be determned thus: E d, f η E f d where: E d,f f E d s the desgn value for actons for the fre stuaton s the reducton factor for the desgn value of actons n the fre stuaton s the desgn value for actons at normal temperature for the fundamental combnaton of actons For the load combnaton n accordance wth EN , the reducton factor f should be taken as follows, whereby the smallest value s gven by the followng two equatons: Gk + ψ f Qk,1 η f γ G + γ Q G k Q,1 k,1 Gk + ψ f Qk,1 η f ξ γ G + γ Q G k Q,1 k,1 where: Q k,1 G k G Q,1 f s the characterstc value of the leadng varable acton s the characterstc value of permanent actons s the partal safety factor for permanent actons s the partal safety factor for the leadng varable acton s the combnaton factor for frequent values of varable actons n the fre stuaton, gven ether by 1,1 or 2,1, see EN s a reducton factor for unfavourable permanent actons G (see EN , clause A.1.3.1) As a smplfcaton, for the reducton factor f, as an alternatve to the above equaton, the recommended value s f 0.6 accordng to EN :2011, clause Exceptons here are areas wth larger mposed loads accordng to category E gven n EN :2002, where the recommended value s f 0.7. When comparng the optons for determnng actons, t s clear that the smplfed assumpton wth the acton E d,f results n a greater load than the actons n the exceptonal desgn stuaton. Page 12 of 50

14 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E 5.2 Verfcaton method for mechancal resstance n the fre stuaton accordng to EN :2011 For verfcaton of mechancal resstance, the desgn values of strength and stffness propertes shall be determned from: f d k f20, f mod, f γ M, f where: f d,f k mod,f f 20 f k k f M,f s the desgn value of strength n fre s the modfcaton factor n the fre stuaton for the reduced cross-secton method: k mod,f 1.0 (as per EN ) s the 20% fractle value of a strength property at normal temperature; f 20 k f f k s the 5% fractle value of a strength property s the coeffcent for convertng 5% to 20% fractle values; k f for CLT 1.15 (as per EN ) s the partal safety factor for tmber n fre M,f 1.0 (as per EN ) For the calculaton n the fre stuaton, nstead of the 5% fractle values, the 20% fractles are used. The reason for ths assumpton les n the extremely low probablty of occurrence of a fully developed fre durng the lfetme of a supportng structure, and does not depend on the materal. [25] (Hence the coeffcent for convertng the fractle value k f wth 1.15.) S d k S20, f mod, f γ M, f where: S d,f k mod,f S 20 S 05 k f M,f s the desgn value of the stffness property (modulus of elastcty or shear modulus) n the fre stuaton s the modfcaton factor n the fre stuaton for the reduced cross-secton method: k mod,f 1.0 (as per EN ) s the 20% fractle of a stffness property (modulus of elastcty or shear modulus) at normal temperature S 20 k f S 05 s the 5% fractle of a stffness property (modulus of elastcty or shear modulus) at normal temperature s the coeffcent for convertng 5% to 20% fractle values; k f for CLT 1.15 (as per EN ) s the partal safety factor for tmber n fre M,f 1.0 (as per EN ) Page 13 of 50

15 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E 5.3 Charrng rates for Stora Enso CLT Durng exposure to fre and to the resultng effect of temperature on the CLT cross-secton, the use of polyurethane adhesves between ndvdual layers can lead to softenng. A possble consequence of ths may be that small sectons of the heat-nsulatng char layer fall off, and the protectve functon of ths layer may be lost at certan ponts. [2] Therefore, n the case of celng elements and other horzontal components, possble delamnatons must be taken nto account, and, for the subsequent fre-exposed layers, t s necessary to mathematcally estmate an ncreased charrng rate untl the formaton of a new 25 mm-thck char layer Desgn value of charrng rates 0 for CLT on surfaces whch are unprotected throughout the duraton of the fre The followng charrng rates for Stora Enso CLT were determned as part of [8] by the accredted nsttute Holzforschung Austra, and may be used for the calculaton of the fre resstance of constructons wth dfferent loads and/or layer thcknesses accordng to EN (wth reference to the respectve natonal annex). Celng and roof elements (horzontal components): o 0.65 mm/mn., f only one layer s affected by exposure to fre. [33] o 1.3 mm/mn. for any addtonal layers affected by exposure to fre untl charrng or the formaton of a 25 mm-thck char layer. Thereafter, a charrng rate of 0.65 mm/mn. can be appled up to the next bonded jont. [33] Fg. 3: dagram llustratng an example of charrng or the charrng rate of a horzontal CLT component (CLT 180 L5s), whch explans the mathematcally estmated charrng rate of 1.3 mm/mn. for each addtonal layer affected by fre untl the formaton of a new 25 mm-thck char layer. Page 14 of 50

16 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Wall element (vertcal components): o 0.63 mm/mn., f only one layer s affected by exposure to fre. [33] o 0.86 mm/mn. for each addtonal layer affected by exposure to fre. [33] Fg. 4: dagram showng an example of charrng or the charrng rate of a vertcal CLT component (CLT 100 L5s), whch explans the mathematcally estmated ncreased charrng rate of 0.86 mm/mn. from the second layer affected by fre. Page 15 of 50

17 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Desgn value of charrng rates 0 for CLT on surfaces whch are ntally protected from exposure to fre by gypsum plasterboard The fre resstance ratng of components s determned durng exposure to fre on the nsde of a room predomnantly by nteror claddng. To ncrease the fre resstance of structures such as wall, celng or roof elements, plaster buldng materals/gypsum plasterboards are generally used as, even f they are not very thck, they provde effectve protecton. Effectve protecton s based partcularly on the combned crystal water n the panels gypsum core whch has a concentraton of approx. 20%. Energy s consumed by the evaporaton of ths crystal water, and a protectve steam curtan s also formed on the fre-exposed sde of the component. In addton to delayng the spread of fre, the dehydrated gypsum layer also acts as nsulaton through the declnng thermal conductvty. Fre protecton plasterboard also contans glass fbre whch renforces the gypsum core and ensures structural coheson when exposed to fre. [3] Fg. 5: two-ply fre protecton plasterboard claddng exposed to fre durng a large-scale fre test Fgure 5 llustrates the behavour of fre protecton plasterboard when exposed to fre; n ths case there are two layers of claddng. As can be seen, after crazng and detachng of the char layer, as tme progresses, large gaps appear between the jonts, the jont plaster compound fals and the frst secton of the frst plasterboard layer falls off. If larger panel sectons fall away from the frst layer, crazng also occurs n the second layer. After gaps appear n ths layer s jonts, the flames spread through the ncreasng gaps n the jonts towards the underlyng CLT whch leads to the producton and emsson of wood gas. Charrng starts on the ntally protected CLT element. The followng correlaton or equvalence wth regard to gypsum plasterboard desgnatons should be noted: Desgnaton accordng to EN 520 Gypsum plasterboard, type A Gypsum plasterboard, type F or DF Gypsum plasterboard, type H2 Gypsum plasterboard, type DFH2 Desgnaton accordng to ÖNORM B 3410 and DIN Plasterboard claddng Fre protecton plasterboard Plasterboard claddng waterproofed Fre protecton plasterboard waterproofed Table 8: comparson of gypsum plasterboard desgnatons by EN 520 and ÖNORM B 3410 or DIN Page 16 of 50

18 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E In the case of ntally protected components, the tme of start of charrng behnd the protectve layer or claddng t ch and the falure tme of the protectve claddng t f s essental. Accordng to EN :2011, the followng must be taken nto account: The start of charrng s delayed untl tme t ch ; Charrng can occur before falure of the fre protectve claddng, however untl the falure tme t f, the charrng rate s lower than the value accordng to [22], table 3.1 or the value accordng to [33]; The charrng rate after the falure tme t f of the fre protectve claddng untl tme t a s greater than the value accordng to [22], table 3.1 or the value accordng to [33]; The charrng rate from tme t a, where the charrng depth corresponds to the lowest value ether the charrng depth of a smlar component wthout fre protectve claddng or 25 mm agan takes the values accordng to [22], table 3.1 or the values accordng to [33]. The followng llustratons from EN :2011 are provded to ad understandng of the above ponts: Key: 1 Relatonshp for components whch are unprotected throughout the tme of fre exposure wth the notonal charrng rate n (or 0) 2 Relatonshp for ntally protected components after falure of the fre protectve claddng 2a After the fre protectve claddng has fallen off, charrng starts at an ncreased rate 2b After the charrng depth exceeds 25 mm or the tme t a s exceeded, the charrng rate reduces to the normal rate Fg. 6: llustraton of the charrng depth dependng on the tme for t ch t f and a charrng depth of 25 mm at tme t a [22] Key: Fg. 7: llustraton of the charrng depth dependng on the tme for t ch < t f [22] 1 Relatonshp for components whch are unprotected throughout the tme of fre exposure wth the notonal charrng rate n (or 0) 2 Relatonshp for ntally protected components on whch charrng starts before falure of the fre protectve claddng 2a Charrng starts at t ch, at a reduced rate for as long as the fre protectve claddng remans ntact 2b After the fre protectve claddng falls off, charrng starts at an ncreased rate 2c After the charrng depth exceeds 25 mm or the tme t a s exceeded, the charrng rate reduces to the normal value Page 17 of 50

19 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Charrng rates for ntally protected components For tme t ch t t f, accordng to [22], the charrng rates gven n EN , table 3.1 or accordng to the statement of expert opnon of the Holzforschung Austra should be multpled by a factor k 2 ; for sngle layer gypsum plasterboard, type F, ths s calculated as: k 2 1 0, 018 h p where: h p s the thckness of the layer n mm For several layers of gypsum plasterboard, type F, h p should be taken as the thckness of the nner layer. If the tmber component s protected by rock fbre batts (thckness: 20 mm, bulk densty: 26 kg/m 3, meltng pont: 1000 C), the factor k 2 may be taken from table 9. For thcknesses between 20 and 45 mm, lnear nterpolaton may be appled. Table 9: values of k 2 for tmber components protected by rock fbre batts [22] For the stage after falure of the fre protectve claddng gven by t f t t a, accordng to [22], the charrng rates gven n EN , table 3.1 or accordng to the statement of expert opnon of Holzforschung Austra should be multpled by a factor k 3 2. For t t a, the charrng rates should be appled wthout multplcaton by the factor k 3. The tme lmt t a (see fgure 6) should for t ch t f, n accordance wth [22], be taken as: t a 2 t f mn 25 + t k3 βn Or for t ch < t f : 25 ( t f tch ) k2 β n t a + t k β where: 3 n f f n s the desgn value of the notonal charrng rate n mm/mn. (In the case of one-dmensonal charrng, n s replaced by 0.) Page 18 of 50

20 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Start of charrng on ntally protected components Sngle layer gypsum plasterboard, type A, F or H: For claddngs consstng of one layer of gypsum plasterboard, type A, F or H, accordng to EN 520, outsde of jonts or at locatons adjacent to flled jonts, or unflled gaps wth a wdth of 2 mm or less, n accordance wth [22], the start of charrng t ch should be taken as: t ch 2,8 h 14 p In locatons adjacent to jonts wth unflled gaps wth a wdth of more than 2 mm, the tme of start of charrng should be calculated as: t ch 2,8 h 23 p where: t ch h p s the tme of start of charrng of a protected component n mnutes s the thckness of the fre protectve claddng n mm Two-layer gypsum plasterboard, type A or H: For claddngs consstng of two layers of gypsum plasterboard, type A or H n accordance wth EN 520, accordng to [22], the tme of start of charrng t ch should be determned accordng to the formula n , where the thckness h p s taken as the thckness of the outer layer and 50% of the thckness of the nner layer. Ths s subject to the condton that the spacng of fasteners n the nner layer s not greater than the spacng of fasteners n the outer layer. Claddngs consstng of two layers of gypsum plasterboard, type F: For claddngs consstng of two layers of gypsum plasterboard, type F n accordance wth EN 520, accordng to [22], the tme of start of charrng t ch should be determned accordng to the formula n , where the thckness h p s taken as the thckness of the outer layer and 80% of the thckness of the nner layer. Ths s subject to the condton that the spacng of fasteners n the nner layer s not greater than the spacng of fasteners n the outer layer. Cavty nsulaton materal: If the tmber component s protected by rock fbre batts (thckness: 20 mm, bulk densty: 26 kg/m 3, meltng pont: 1000 C), for the tme of start of charrng t ch, the followng equaton must also be taken nto account: t 0,07 ( 20) ch h ns ρ ns where: t ch h ns s the tme untl the start of charrng of a protected component n mnutes s the nsulaton materal thckness n mm ns s the nsulaton materal bulk densty n kg/m 3 Page 19 of 50

21 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Falure tme of fre protectve claddng In prncple, the charrng or mechancal degradaton of the claddng materal, the spacng of, and dstances between, fasteners and/or a possbly nsuffcent penetraton length of fasteners nto the uncharred cross-secton could be responsble for the falure of the fre protectve claddng. Claddng consstng of gypsum plasterboard, type A or H: For gypsum plasterboard, type A or H n accordance wth EN 520, accordng to [22], the falure tme t f s equal to the tme at the start of charrng t ch. For gypsum plasterboard, type A or H, after the start of charrng and after the claddng smultaneously falls off, charrng occurs at double the rate untl tme t a. After formaton of a 25 mm-thck char layer, the charrng rate reduces to the normal rate. (See also fg. 6) t f t ch Claddng consstng of gypsum plasterboard, type F: However, n the case of gypsum plasterboard, type F or fre protecton plasterboard, accordng to [22], there s less charrng from the start of charrng t ch to the tme t f. Untl the subsequent formaton of a 25 mm-thck char layer, charrng occurs at double the rate, after whch, the charrng rate reduces to the normal rate. (See also Fg. 7.) EN :2011 does not provde any nformaton regardng the falure tme of gypsum plasterboard, type F or fre protecton plasterboard. Accordng to ÖNORM B :2011 (Austran natonal specfcatons), the falure tmes t f for claddng consstng of fre protecton plasterboard n accordance wth ÖNORM B 3410 or gypsum plasterboards, type DF accordng to EN 520 and gypsum fbreboard GF-C1-W2 accordng to EN can be determned as follows: Wall components: t 2,2 h + 4 f p Celng components: t 1,4 h + 6 where: f p t f h p s the falure tme of the fre protectve claddng n mnutes s the thckness of the fre protectve claddng n mm In determnng the falure tme of multple-layer claddng consstng of gypsum plasterboard, type F, the rules specfed n secton apply correspondngly, accordng to whch, the thckness h p corresponds to the thckness of the outer layer and to 80% of the thckness of the nner layer. Page 20 of 50

22 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Penetraton length of fasteners for gypsum plasterboard In addton to thermal degradaton of the claddng materal, the fre protectve claddng can also fall off due to the pull-out falure of fasteners. Accordng to [22], the requred mnmum length of the fasteners should also be determned n order to elmnate the fact that pull-out of the fasteners s a relevant factor for the falure of the fre protectve claddng. The mnmum penetraton length of the fastener l a nto the unburnt cross-secton should be taken as 10 mm. The requred penetraton length of the fastener l f, req s calculated as follows: l + l f, req hp + dchar, 0 a where: h p d char,0 l a s the panel thckness n mm s the charrng depth n the tmber component s the mnmum penetraton length of the fastener nto the unburnt wood For more nformaton on claddng fasteners/penetraton lengths, see [22], secton Falure tmes t f of fre protecton plasterboard on Stora Enso CLT confrmed by statements of expert opnon: In addton to the equatons demonstrated above and specfed n [22] and [25] to determne the start of charrng t ch and the falure tme t f of gypsum plasterboards, Stora Enso has a statement of expert opnon on falure tmes whch must be referred to durng dmensonng accordng to EN Accordng to [35], based on varous tests, the falure tmes t f lsted n table 10 were gven for fre protecton plasterboards n accordance wth ÖNORM B 3410 or gypsum plasterboards, type DF n accordance wth EN 520. (Compare the calculated values accordng to the equatons of [25], whch were orgnally worked out for tmber frame structures.) Fre protecton plasterboard CLT wall structure (vertcal component) HFA statement ON B of expert opnon t f [mn.] t f [mn.] Table 10: falure tmes for fre protecton plasterboards or gypsum plasterboards, type DF drectly appled to CLT elements n accordance wth [35] (cf. the falure tmes calculated accordng to EN ) The falure tmes gven n table 10 only apply to fre protecton plasterboards or gypsum plasterboards, type DF drectly appled to Stora Enso CLT elements. The fre protecton plasterboard must be appled and sealed accordng to the manufacturer s nstructons. [35] Page 21 of 50

23 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E 5.4 Determnng the load-bearng capacty (R) of CLT elements accordng to EN :2011 When determnng the load-bearng capacty (R) of tmber components exposed to fre, or when calculatng crosssectonal values, n addton to determnng the charrng area, the underlyng area affected by temperature must also be taken nto account because the wood s strength and stffness propertes decrease as the temperature rses. As an alternatve to the calculaton opton specfed n EN , annex B, the cross-sectonal values can also be calculated usng two smplfed methods. We recommend the frst method: Reduced cross-secton method For verfcaton n the fre stuaton, ths method uses a reduced cross-secton or resdual cross-secton, calculated on the bass of ncreased charrng (roundngs or corner charrng), and an addtonal area affected by temperature (reducton of mechancal propertes due to the effect of temperature). Reduced propertes method As an alternatve to the reduced cross-secton method calculated wth charrng speed and corner charrng ths method takes nto account the reducton of mechancal propertes dependng on the type of load and cross-secton. The verfcaton of load-bearng capacty n the fre stuaton s performed for CLT on the bass of the reduced cross-secton method. Reduced cross-secton method The cross-secton whch s reduced by the charrng depth s further reduced by removng a layer wth zero strength and stffness k 0 d 0. Thus, the notonal resdual cross-secton s calculated by deductng the notonal charrng depth d ef from the orgnal cross-secton. d e f d char, 0 + k0 d 0 where: d ef d char,0 n t k 0 d 0 s the notonal charrng depth s the desgn value of the charrng depth for one-dmensonal charrng d char,0 0 t s the desgn value of the one-dmensonal charrng rate under normal load s the tme under fre exposure s the coeffcent for takng the duraton of the fre nto account; t < 20 mn.: k 0 t / 20 t 20 mn.: k s the depth of a layer (close to the char lne) wth assumed zero strength and stffness. d 0 7 mm 2 Page 22 of 50

24 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Note *2: The value specfed for d 0 7 mm s based on [26] (relatng to [22]). The value d 0 of 7 mm (for the smplfed calculaton method of the reduced cross-secton method) s currently beng dscussed by scentsts around the world, however no unfed opnon has been establshed. Possble natonal regulatons on d 0 must be taken nto account. Wth regard to assumptons about charrng rates for Stora Enso CLT, the followng must be observed: When usng CLT for flat components (wall and celng structures), one-dmensonal charrng rates n accordance wth [33] (see secton 5.3.1) should be used. When usng CLT for supports (edgewse), for example, proceed accordng to [22], secton In dong so, for CLT cross-sectons wth orgnal wdths whch do not meet requrements, ncreased charrng rates should be expected. When verfyng the load-bearng capacty n the fre stuaton of Stora Enso CLT components, the followng must be observed: Charrng on both sdes must be taken nto account on load-bearng elements wth no separatng functon. [22] Possble addtonal, eccentrc load applcatons due to one-sded charrng must be taken nto account partcularly on thnner CLT elements. Resdual cross-sectons of layers 3 mm are not used n the remanng calculatons. (Ths assumpton takes nto account the generally non-lnear nature of the char lne.) The remanng calculaton steps and verfcatons are performed n the same way as the cold calculatons. Page 23 of 50

25 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E 5.5 Determnng the ntegrty (E) and nsulaton (I) of CLT elements The followng optons exst for verfcaton of ntegrty (E) and nsulaton (I): Calculaton method accordng to EN :2011, annex E ([22]) Model accordng to ÖNORM B :2011, 14.3 ([25]) or the European gudelne Fre safety n tmber buldngs ([9]) or the dssertaton by Ms Schlefer ([4]) Structures wthout addtonal verfcatons accordng to ÖNORM B :2011 ([25]) Verfcaton of the ntegrty and nsulaton of CLT can be performed usng the model specfed n ÖNORM B :2011 ([25]) or n the European techncal gudelne Fre safety n tmber buldngs ([9]) and elaborated by [4], whch have the same approach/support the same theory. If we compare ths model wth the calculaton method specfed n EN :2011, annex E, the possblty of an unlmted varaton of materals and number of layers can be consdered to be a sgnfcant advantage. Extended method for determnng the ntegrty (EI) of wall and celng structures n accordance wth ÖNORM B :2011 ([25]) or the European gudelne Fre safety n tmber buldngs ([9]) In prncple, the followng apples to the calculaton: The nfluence of temperature accordng to the unform temperature curve as per EN provdes the bass for the calculaton model. Accordng to [25], the calculaton model s lmted to a fre resstance duraton of 60 mnutes. Valdaton calculatons performed by the accredted nsttute Holzforschung Austra as part of large-scale fre tests show that ths model can also be used for a fre resstance duraton of 90 mnutes. [5] The requrement for ntegrty (E) s consdered satsfed f the requrement for the nsulaton (I) crteron s shown to be postve. The requrement for nsulaton (I) s satsfed f the average temperature rse on the unexposed sde of the component does not exceed 140 C, or 180 C at any one pont (above the ambent temperature). In order to ensure the calculated fre resstance or ntegrty, n the case of composte tmber components, the surplus nsulaton must be prevented from fallng out after falure of the claddng by mechancal means, where approprate. It must also be ensured, through correct nstallaton accordng to the manufacturer s nstructons (e.g. nformaton relatng to spacng between fasteners and penetraton lengths), that the claddng on the unexposed sde cannot fall off at an early stage. Page 24 of 50

26 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E The component may be composed of any of the followng panel and nsulaton materals and may be desgned wth a cavty: Panel materals (fasteners accordng to the manufacturer s nstructons): Sold wood panels of at least C 24 n accordance wth EN 338 OSB panels n accordance wth EN 300 Partcle board (chpboard) n accordance wth EN 309 Gypsum plasterboard, type A, H and F n accordance wth EN 520 Gypsum fbreboard n accordance wth EN Insulaton materal (surplus nstallaton accordng to the manufacturer s nstructons): Rock fbre n accordance wth EN Glass wool n accordance wth EN The requred ntegrty of a component s consdered satsfed f the followng equaton s satsfed: tns t req where: t ns t req s the tme untl falure of the separatng functon of the entre component, n mnutes. s the requred fre resstance duraton for the separatng functon of the entre component, n mnutes The nsulatng and protectve layers of a component Based on smulaton calculatons or FE modellng, t was demonstrated that the nsulaton (I) requrements were satsfed. If we consder a tmber structure wth multple layers, the ndvdual layers are arranged wth a protectve (protectve n terms of the underlyng layers) and nsulatng functon (last layer of the unexposed sde). [4] Fg. 8: desgn of a multple layer tmber constructon to defne the protectve and nsulatng layers [4] Page 25 of 50

27 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E The tme of the separatng functon (EI) of the component under consderaton s the tme untl the temperature crteron T MW / T Max 140 / 180 C s reached on the unexposed sde. Ths crteron wth the maxmum temperature to be mantaned of T 160 C (20 C room temperature K) s only relevant for the unexposed sde of the last layer. The precedng protectve layers must satsfy the claddng crteron n accordance wth EN , whereby the temperature crteron T MW / T Max 250 / 270 C must be satsfed. Thus, n applyng the average value, a maxmum temperature to be mantaned of T 270 C (20 C room temperature K) s obtaned. When the temperature crteron of 270 C s reached ( protecton tme t prot, s reached), t s assumed that the tested layer wll fall off the structure and the protecton tme of the drectly underlyng layer t prot,+1 begns. Fg. 9: method for determnng the contrbutons of ndvdual layers [4] As a result, the protectve layers lose ther protectve functon as soon as a temperature of T 270 C s reached on ther unexposed sde. The tme of the separatng functon of a complete component t ns s thus obtaned by addng together the contrbutons of the ndvdual layers the protecton tmes of the protectve layers and the nsulaton tme of the nsulatng layer whle observng the above-mentoned temperature crtera of 270 or 160 C. t prot, 1 + t t ns ns, where: t ns t prot, t ns, s the tme untl falure of the separatng functon of the entre component, n mnutes s the protecton tme of the layer, n mnutes s the nsulaton tme of the layer, n mnutes Page 26 of 50

28 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E When determnng the protecton and nsulaton tme, t s mportant to note that the precedng and backng layers nfluence the layer under nvestgaton dependng on ts poston n the component. Ths s taken nto account n the calculaton model wth the poston coeffcent k pos. In the process, k pos,exp s the poston coeffcent resultng from the nfluences of the layers precedng the layer under nvestgaton and k pos,unexp s the poston coeffcent resultng from the nfluences of the layers backng the layer under nvestgaton. The nsulaton and protecton tme of ndvdual layers s determned wth the followng equatons: t prot, ( t prot,0, k pos,exp, k pos, unexp, + t ) k j, where: t prot, t prot,0, k pos,exp, k pos,unexp, t, k j, s the protecton tme of the layer under nvestgaton, n mnutes s the basc nsulaton tme of the layer, n mnutes s the poston coeffcent for the layer under nvestgaton (nfluences from the precedng layer) s the poston coeffcent for the layer under nvestgaton (nfluences from the backng layer) s the tme dfference for the layer under nvestgaton, n mnutes. (To take nto account the nfluence of the precedng gypsum plasterboard, type F or gypsum fbreboard) s the jont coeffcent for the layer under nvestgaton t ns, ( tns,0, k pos,exp, + t ) k j, where: t ns, t ns,0, k pos,exp, t, k j, s the nsulaton tme of the layer under nvestgaton, n mnutes. s the basc nsulaton tme of the layer, n mnutes. s the poston coeffcent for the layer under nvestgaton (nfluences from the precedng layer) s the tme dfference ( delayed fall off tme) for the layer under nvestgaton, n mnutes; (To take nto account the nfluence of the precedng gypsum plasterboard, type F or gypsum fbreboard) s the jont coeffcent for the layer under nvestgaton Page 27 of 50

29 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Determnng a layer s basc tmes The basc tme t 0, whereby a dstncton s made between the basc protecton tme t prot,0, and the basc nsulaton tme t ns,0, descrbes the fre behavour of a layer wthout takng the nfluence of adjacent layers nto account. The basc protecton tme t prot,0, and basc nsulaton tme t ns,0, of dfferent materals can be defned usng table 11 (from [25]), takng the layer s thckness and bulk densty nto account. Materal Gypsum plasterboard n accordance wth ÖNORM EN 520 (apples to: type A, type H, type F, type DF) Gypsum fbreboard n accordance wth ÖNORM EN (apples to: GF-C1-W2) Sold wood board n accordance wth ÖNORM EN Partcle board (chpboard) n accordance wth ÖNORM EN 312 OSB board n accordance wth ÖNORM EN 300 Basc nsulaton tme t ns,0, mn *3 Basc protecton tme t ns,0, mn Rock fbre n accordance wth ÖNORM EN Glass wool n accordance wth ÖNORM EN Where: h p, h ns, ns, for h ns, < 40 mm: 0 for h ns, 40 mm: (0,001 ns, + 0,035) h ns, + 8,5 30 s the thckness of the fre protectve claddng, n mm s the thckness of the nsulaton, n mm s the bulk densty of the nsulaton, n kg/m³ *4 for h ns, < 40 mm: 0 for h ns, 40 mm: (0,0007 ns, + 0,046) h ns, Table 11: equatons to determne the basc tmes t 0, n mnutes [25] Note *3: Note *4: The equaton hghlghted n table 11 (taken from ÖNORM B :2011) for calculatng t ns,0, for OSB panels s not correct. Accordng to [4], the correct equaton should be: t B h 16 A 20 ns,0, The equaton hghlghted n table 11 (taken from ÖNORM B :2011) to calculate t ns,0, for rock fbre s not correct. 1,4 0,224 2 Accordng to [4], the correct equaton should be: t 0,01 ρ 0,02 h ns, 0, ( ) ns, ns, Page 28 of 50

30 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Calculatng the poston coeffcent k pos To enable any combnaton of layers n a tmber structure, the nfluences of adjacent layers on the layer under nvestgaton must be taken nto account. The nfluence of the precedng layer s descrbed or calculated wth poston coeffcent k pos,exp and the nfluence of the underlyng layer wth poston coeffcent k pos,unexp Poston coeffcent k pos,exp for takng nto account the nfluence of the precedng layers The poston coeffcent k pos,exp s assumed to be 1.0, f the layer under nvestgaton s exposed to fre from the start of the fre and f t s not protected by any precedng layers. If the layer under nvestgaton s protected from drect exposure to fre by precedng layers, the poston coeffcent k pos,exp must be defned accordng to the equatons n the followng table. When determnng the poston coeffcent k pos,exp, the sum of the protecton tmes of the precedng layers t prot,-1, the materal of the layer under nvestgaton, the thckness of the layer under nvestgaton and the bulk densty of the layer under nvestgaton must be taken nto account. However, the propertes of the layer under nvestgaton have already been calculated durng determnaton of the basc protecton tme t prot,0, or the basc nsulaton tme t ns,0, and therefore the poston coeffcent k pos,exp for claddng and nsulaton can be calculated dependng on the sum of the protecton tmes of the precedng layers t prot,-1 and the basc tme (basc protecton tme t prot,0, or basc nsulaton tme t ns,0, ). Materal k pos,exp for t ns, k pos,exp for t prot, Rock fbre see formulae for glass wool see formulae for claddng Glass wool for h 40 mm for for Claddng for Table 12: equatons to determne the poston coeffcent k pos,exp [25] for When usng the equatons n table 12, t must be consdered whether an nsulaton tme or a protecton tme should be calculated,.e. the basc protecton tme or nsulaton tme must be used accordngly for the basc tme. Page 29 of 50

31 A B CDE CDE V E R I F I C A T I O N O F F I R E R E S I S T A N C E Through the delayed fall off of protected layers from the gypsum plasterboard, type F or DF n accordance wth EN 520 or gypsum fbreboard GF-C1-W2 n accordance wth EN , the protecton or nsulaton tme of the underlyng layer s ncreased. If the layer under nvestgaton s protected by the precedng gypsum plasterboard or gypsum fbreboard of the above-mentoned types, the protecton tme or nsulaton tme of the layer under nvestgaton must be ncreased va tme dfferences t, accordng to table 13. If ths s not the case, t, 0 apples. Materal t for celng structures mn t for wall structures mn Claddng for for Insulaton for for *5 for Table 13: equaton to determne the tme dfferences t, for layers n celng and wall structures protected by precedng gypsum plasterboard or gypsum fbreboard [25] for Note *5: The equaton hghlghted n table 13 (taken from ÖNORM B :2011) for calculatng t, for nsulaton n celng structures s not correct. Accordng to [4], the correct equaton should be: t, 0,1 035 t prot, 1 0, t0, Poston coeffcent k pos,unexp for takng nto account the nfluence of the backng layers When determnng the poston coeffcent k pos,unexp, a dstncton s made between layers wth underlyng claddng and layers wth underlyng nsulaton. However, underlyng layers should only be taken nto account for layers wth a protectve functon, whch, n turn, means that when determnng the nsulaton tme, the basc tme of the nsulatng layer must only be used wth the poston coeffcent k pos,exp for takng the nfluence of the precedng layers nto account. The slght nfluence of the underlyng gypsum boards or tmber materal claddng on the layer under nvestgaton can be gnored, whereby, n ths case, the poston coeffcent k pos,unexp s assumed to be 1.0. The poston coeffcent k pos,unexp for underlyng nsulaton materals s assumed, dependng on the thckness of the layer h and bulk densty A, accordng to table 14. Page 30 of 50