ForestProductsLaboratoryFire Resistance ofStructural CompositeLumber ProductsResearchPaperFPL–RP–633Robert H. WhiteUnited StatesDepartment ofAgricultureForest Service

AbstractAcknowledgmentsUse of structural composite lumber products is increasing.In applications requiring a fire resistance rating, calculation procedures are used to obtain the fire resistance ratingof exposed structural wood products. A critical factor inthe calculation procedures is char rate for ASTM E 119 fireexposure. In this study, we tested 14 structural compositelumber products to determine char rate when subjected tothe fire exposure of the standard fire resistance test. Charrate tests on 10 of the composite lumber products were alsoconducted in an intermediate-scale horizontal furnace. TheNational Design Specification/Technical Report 10 designprocedure for calculating fire resistance ratings of exposedwood members can be used to predict failure times formembers loaded in tension. Thirteen tests were conducted inwhich composite lumber products were loaded in tension asthey were subjected to the standard fire exposure of ASTME 119. Charring rates, observed failure times in tensiontests, and deviations from predicted failure times of thestructural composite lumber products were within expectedrange of results for sawn lumber and glued laminatedtimbers.This study was suggested by Rodney McPhee of the Canadian Wood Council (CWC). Funds for this study were provided by the CWC. The following manufacturers providedthe structural composite lumber (SCL) test materials: BoiseCascade, Georgia–Pacific, Louisiana–Pacific, Trus JoistMacMillan (now Trus Joist, a Weyerhaeuser Business), andWillamette Industries (now part of Weyerhaeuser Company). Advice and suggestions were provided by the membersof the CWC/FPL SCL Charring Rate Study task group ofthe Wood I-Joist Manufacturers Association. This task groupwas chaired by Bruce Craig. Comazel Caldwell of the FPLfire research work unit made the specimens and conductedthe tests. The author greatly appreciates the support, assistance, and patience from the parties involved.Keywords: structural composite lumber, LVL, fire resistance, char rateApril 2006White, Robert H. 2006. Fire resistance of structural composite lumberproducts. Research Paper FPL-RP-633. Madison, WI: U.S. Department ofAgriculture, Forest Service, Forest Products Laboratory. 28 p.A limited number of free copies of this publication are available to thepublic from the Forest Products Laboratory, One Gifford Pinchot Drive,Madison, WI 53726–2398. This publication is also available online Laboratory publications are sent to hundreds of librariesin the United States and elsewhere.The Forest Products Laboratory is maintained in cooperation with theUniversity of Wisconsin.The use of trade or firm names in this publication is for reader informationand does not imply endorsement by the United States Department ofAgriculture (USDA) of any product or service.The USDA prohibits discrimination in all its programs and activities on thebasis of race, color, national origin, age, disability, and where applicable,sex, marital status, familial status, parental status, religion, sexual orientation, genetic information, political beliefs, reprisal, or because all or a partof an individual’s income is derived from any public assistance program.(Not all prohibited bases apply to all programs.) Persons with disabilitieswho require alternative means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGETCenter at (202) 720–2600 (voice and TDD). To file a complaint of discrimination, write to USDA, Director, Office of Civil Rights, 1400 IndependenceAvenue, S.W., Washington, D.C. 20250–9410, or call (800) 795–3272(voice) or (202) 720–6382 (TDD). USDA is an equal opportunity providerand e Resistance Calculations.1Char Rate of Wood.2National Design Specification TechnicalReport 10 Procedure.3Char Rate Experiments.3Materials.3Small Vertical-Furnace Tests.4Test method.4Test results.5Horizontal-Furnace Tests.10Test methods.10Test results.12Discussion.13Tension Tests.15Methods and Materials.18Failure Times.20National Design Specification Model Predictions.20Conclusions.26Literature Cited .26Appendix—Individual Test Results for Each SmallVertical-Furnace Test.28

Fire Resistance of Structural CompositeLumber ProductsRobert H. White, Research Wood ScientistForest Products Laboratory, Madison, WisconsinIntroductionProduction and use of structural composite lumber (SCL)products are increasing (McKeever 1997, Schuler and others 2001). Such products include laminated veneer lumber(LVL), parallel strand lumber (PSL), laminated strandlumber (LSL), and oriented strand lumber (OSL) (Greenand Hernandez 1998, Moody and others 1999, ASTM International 2005) (Fig. 1). Introduction of composite lumberproducts started in the 1970s with LVL and continued withintroduction of PSL in the 1980s and LSL in the 1990s(Yeh 2003). Oriented strand lumber is the latest of theseproducts. For each of these products, adhesives are utilized to manufacture the composite product from veneersor strands of wood. Questions are sometimes raised aboutthe performance of the adhesive when exposed to elevatedtemperature. The code and market acceptance of productsdepends on documentation of acceptable performance.Code acceptance of calculation procedures for determiningfire resistance ratings of exposed wood beams and columnspermitted their use in applications requiring structural members to have specified fire resistance ratings. As a result, thepotential market for such wood products increased. This report provides test data on the charring rate of three types ofcomposite lumber products (LVL, LSL, and PSL) and theirperformance when loaded in tension and subjected to thefire exposure specified in the standard fire resistance test.Several species of composite lumber products were includedin testing. In the first two phases of this project, we documented charring rates of composite lumber products. Firetests conducted in the first two phases of this project did notinclude any load being applied to the member. Results ofPhase One were initially published in a conference proceedings article, “Charring Rate of Composite Timber Products”(White 2000). These one-dimensional charring experimentswere conducted in the small vertical-furnace at the USDAForest Products Laboratory (FPL), Madison, Wisconsin. Inthe second phase, we measured temperatures in the interiorof a non-loaded beam subjected to the ASTM E 119 (ASTMInternational 2000) fire exposure in an intermediate-scalehorizontal furnace. Charring rates determined in the twofurnaces were compared. In the third and final phase ofthe study, a series of SCL test specimens were subjected toASTM E 119 fire exposure while loaded in tension.Figure 1. Examples of types of composite lumber products tested in this study: (left to right) laminated veneerlumber, parallel strand lumber, and laminated strandlumber.BackgroundFire Resistance CalculationsFire resistance ratings of structural members are normallydetermined by conducting the full-scale test described inspecifications of ASTM E 119 or similar standards. Calculation procedures for determining fire resistance rating ofwood members have code acceptance (White 2002). Lie(1977) developed the first procedure to gain code acceptance in the United States and Canada for wood beams andcolumns (American Institute of Timber Construction 1984,Canadian Wood Council 1997, American Forest & PaperAssociation 2000, 2003). A more recent calculation procedure, the National Design Specification (NDS) method, wasdeveloped by the American Wood Council (AWC) of theAmerican Forest & Paper Association. It was included in theNDS (American Forest & Paper Association 2001) startingwith the 2001 edition. This new, more explicit procedure isapplicable to other structural members besides beams andcolumns. Per NDS, the fire design procedure is applicable toall wood structural members and connections covered underNDS. These include wood members of solid-sawn lumber,structural glued-laminated timber, and structural composite

Research Paper FPL–RP–633lumber. Methodology and supporting data for solid-sawnlumber and structural glued-laminated timber are fully discussed in “Calculating the fire resistance of exposed woodmembers,” Technical Report 10 (TR 10) of the AmericanForest & Paper Association (2003).Tests of three glued-laminated specimens were conductedto verify calculation procedures for determining the fireresistance rating of axially loaded wood tension memberscontained in TR 10 (White 2004). Tests were conductedusing the horizontal furnace and tension apparatus at theFPL. Glued-laminated specimens were exposed to anASTM E 119 fire exposure while loaded in tension forthe full duration of the test.Other options for analytical methods for determining fireresistance of timber members are reviewed by White (2002).Analytical methods in the updated Eurocode 5 document ofEurope are discussed by König (2005).A critical parameter in calculation procedures is charringrate of the member while directly subjected to fire exposureof the time–temperature curve specified in ASTM E 119.Char Rate of WoodParameters affecting charring rate for solid wood havebeen extensively studied at the FPL (Schaffer 1967, Whiteand Nordheim 1992, White and Tran 1996) and elsewhere(White 1988, White 2002). For solid wood and gluedlaminated members, the value for charring rate generallyused in the United States and Canada is 0.635 mm/min. The0.60 mm/min value was used in developing calculation procedures for large wood members (Lie 1977).The following time–location models for the time needed toobtain a char depth (xc) were considered by White and Nordheim (1992) and White (1988):(1)(2)(3)or its linear form(4)and(5)Equation (1) is the simple linear single parameter (m1)model (zero-intercept) that is generally reported for thecharring rate of wood products. Equation (1) is also the oneparameter model assumed when charring rate is calculatedfrom residual section and duration of fire exposure. The val-2ue of m1 for the conventional 0.635-mm/min charring rateis 1.575 min/mm. Equation (2) is a two-parameter model(m2 and b) that allows a fast initial char rate followed by aslower linear char rate. Equation (3) is a nonlinear charringmodel with two parameters (m3 and a). Equation (4) is thelinear form of Equation (3). Results for Equation (3) can beobtained by linear regression of Equation (4). The model ofEquation (5) was developed by White and Nordheim (1992).This nonlinear model was used in the methodology of TR10. The value of m5 for the conventional 38-mm char depthat 1 h is 0.682 min/mm1.23.Charring data on composite lumber products in the publicdomain have been limited. In a series of cone calorimetertests of sawn lumber and wood composites, Mikkola (1990,1991) tested a 37-mm-thick laminated veneer lumber product with a density of 520 kg/m3. For an external exposureof 50 kW/m2, char rates were 1.05, 0.82, and 0.68 mm/minfor moisture contents of 0%, 10%, and 20%, respectively.For the 38-mm-thick spruce sawn lumber with density of490 kg/m3, char rates were 1.06, 0.80, and 0.60 mm/min formoisture contents of 0%, 10%, and 20%, respectively. Ignition properties and char rates from cone calorimeter tests ofradiata pine were reported by Lane and others (2004). Gettoand Ishihara (1998) found that fire resistance of untreatedwood or fire-retardant-treated wood was improved by compression of the board. Fire resistance of lathed-veneer laminated boards was greater than that of the solid-sawn board.In a study of LVL with different types of joints, Uesugi andothers (1999) obtained charring rate of 0.6 to 0.7 mm/minin tests of Douglas-fir and larch LVL and concluded thatthese materials showed the same performance in fire asheavy timber. Subyakto and others (2001) investigated fireresistance of a LVL–metal plate connection and its improvement with graphite phenolic sphere sheeting. Experimentalresults for parameters of both Equations (1) and (5) for sixwood composite rim board products were reported by White(2003). The six products included three oriented strandboards (OSB), a plywood, an LVL, and a com-ply product.In addition to data for the exposed rim boards, White (2003)also provided parameter estimates for rim boards protectedwith gypsum board. More recently, fire resistance of LVLwas investigated at the University of Canterbury in NewZealand (Lane and others 2004). In a report on variabilityof wood charring rates, Hietaniemi (2005) included unpublished data for Kertopuu LVL specimens obtained fromprivate communication with T. Oksansen. For LVL data obtained from T. Oksansen, Hietaniemi (2005) reported meancharring rates of 0.60 and 0.76 mm/min for charring perpendicular and parallel to the veneers, respectively. For eachdirection, coefficients of variation were 8% to 9%.Charring tests in the small vertical-furnace included in thisreport were previously discussed by White (2000). Resultsfor the composite lumber products were comparable to thosefor solid-sawn lumber. As with solid-sawn lumber, charring

Fire Resistance of Structural Composite Lumber Productsrate of a specific composite lumber product would dependon density, moisture content, and species.National Design Specification TechnicalReport 10 ProcedureThe NDS method described in TR 10 (American Forest &Paper Association 2003) is a reduced-section method forcalculating fire resistance of an individual structural woodmember (White 2002). For each surface of the member subjected to fire exposure, charring reduces the cross-sectionalarea of the member. In TR 10, a form of Equation (5) is usedto calculate char depth at time t. A nominal char rate, βn(mm/min or in/h), is initially assumed. This is a linear charrate based on char depth for 1-h exposure, and in most casesa value of 38 mm at 60 min is used. As noted in the NDS(American Forest & Paper Association 2001), this nominalchar rate of 38 mm/h is commonly assumed for solid-sawnand structural glued-laminated softwood members. The chardepth, xc, at time t is calculated from(6)To account for loss of strength from elevated temperatureswithin the residual cross-sectional area of the member androunding the corners of a rectangular member, the crosssectional area is further reduced to an effective cross-sectional area. In the TR 10 procedure, additional reduction indimensions is 20% of char depth. Thus, the effective charrate used to calculate section properties of the charred member at time t is(7)For the nominal char rate of 38 mm at 1 h, the effectivechar rates are 46 mm/h, 42 mm/h, and 40 mm/h for 60, 90,and 120 min, respectively. For a rectangular member withdimensions of B and D and all four sides exposed to the fire,the area of the cross section at time t, A(t), is(8)With this effective reduced cross-sectional area, the ultimateload bearing capacity of the member at time t is calculatedusing normal room temperature assumptions. The failurecriteria for tension members are(9)whereDLKRASDFtisdesign dead load,design live load,factor to adjust from nominal design capacity toaverage ultimate capacity,nominal allowable design capacity or FtAf,tabulated tension parallel-to-grain designvalue, andAfarea of cross section using cross-sectiondimensions reduced from fire exposurethat will result in failure of the tensionmember.For a member in tension, the allowable design stress to average ultimate strength adjustment factor, K, is 2.85 (American Forest & Paper Association 2003). For the fire designprocedure in the NDS, this value for design stress to member strength factor is considered valid for solid-sawn, gluedlaminated timber, and structural-composite lumber woodmembers (American Forest & Paper Association 2003).Calculated failure time or fire resistance rating is the timefor which the effective cross-sectional area has been reducedto Af per the failure criteria of Equation (9).Char Rate ExperimentsThis project included the determination of charring rate ofcomposite lumber products in two fire resistance furnacesat the FPL. The small vertical-furnace was used to obtainthe one-dimensional charring rates for the widest range ofproducts and included replicates for statistical comparison.Tests in the intermediate-scale horizontal furnace providedverification that data obtained in the smaller furnace wouldlikely be valid for the even larger test furnaces specified inASTM E 119.MaterialsFourteen different materials were tested in the small vertical-furnace experiments of Phase 1 (Table 1). Ten of thefourteen materials were tested in horizontal-furnace tests ofPhase 2 (Table 2). Three general types of products were tested: LVL, PSL, and LSL. No OSL product was tested. Species included two aspen products, four Douglas-fir products,three Southern Pine products, four yellow-poplar products,and one eucalyptus product (Table 1). Except for MaterialNumber 7 (Table 1), the products tested were thosecommercially available. The eucalyptus LVL (MaterialNumber 7) was a prototype sample of the product.Laminated strand lumber is made of strands of wood thatare glued together such that strands are parallel to the longitudinal or axial direction of the lumber product. Laminatedveneer lumber is a composite made by laminating sheets ofveneer with an adhesive so the longitudinal grain of the majority of veneers is parallel to the longitudinal or axial direction of the lumber product. The veneers are end-jointed witha lap, butt, or scarf joint. In typical applications, the gluelines are vertical. Parallel strand lumber is made of strandsobtained by clipping sheets of veneer with strands alignedparallel to the longitudinal axis. The wide face of the strandsis parallel to the tangential direction and perpendicular tothe radial direction of the wood itself. Aligned strands arepressed together with adhesive.Information on adhesives used in the products (Tables 1and 2) was provided by the manufacturers. Exterior-type3

Research Paper FPL–RP–633Table 1. Composite lumber products lyptusSouthern PineSouthern PineYellow-poplarYellow-poplarDouglas-firSouthern PineYellow-poplarMaterial numberin White (2000)19–243–6711105812aPF, phenol–formaldehyde.MDI, methylene diphenyl diisocyanate.bDensity calculated