Available online at www.sciencedirect.com Construction

www.elsevier.com/locate/conbuildmat

Construction and Building Materials 22 (2008) 2165–2169

and Building

MATERIALS

Combustion properties of alder (Alnus glutinosa L.)Gaertn. subsp. barbata (C.A. Mey) Yalt.) and southern pine

(Pinus sylvestris L.) wood treated with boron compounds

Ali Temiz *, Engin Derya Gezer, Umit Cafer Yildiz, Sibel Yildiz

Karadeniz Technical University, Faculty of Forestry, Department of Forest Industrial Engineering, 61080 Trabzon, Turkey

Received 10 November 2006; received in revised form 26 August 2007; accepted 28 August 2007Available online 24 October 2007

Abstract

Samples from alder (Alnus glutinosa L.) Gaertn. subsp. barbata (C.A. Mey) Yalt.) and southern pine (Pinus sylvestris L.) wood wereimpregnated with boric acid, borax and their mixed solutions according to ASTM D 1413-88 in order to determine their combustionproperties.

In this study, the fire resistance properties of wood treated with boron compounds were investigated. In addition, the models for eachwood species and fire retardant solution were also determined. The results demonstrated that the lowest mass losses for both alder andsouthern pine specimens treated with a mixture of 5% boric acid and borax aqueous solutions were found to be 68.72% and 72.37%,respectively. It was found that 5% of borax was the most effective treatment in terms of lengthening the time of glowing.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Boron; Fire resistance; Alder wood; Southern pine

1. Introduction

The borate chemicals offer substantial advantages forwood protection, providing fire resistance as well as effi-cacy against both fungi and insects, low cost, ease of han-dling and treatment. Boron wood preservatives have beenused for many years and there is growing interest in theirlow mammalian toxicity and environmental acceptability[1]. Boric acid and borax are most common boron com-pounds which have found many application areas in thewood preservation industry in order to get the benefit oftheir biological effectiveness and fire retardancy [2,3].

Boric acid and borax mixtures have some efficacy inretarding the spread of fire on wood surface. In additionto the usual char-forming catalytic effect, they have a rather

0950-0618/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.conbuildmat.2007.08.011

* Corresponding author. Tel.: +90 462 377 28 51; fax: +90 462 325 7499.

E-mail address: alitemiz@gmail.com (A. Temiz).

low melting point and form glassy films when exposed tohigh temperatures in fires. Borax tends to reduce fire spreadbut can promote smoldering or glowing. On the otherhand, boric acid suppresses smoldering but has little effecton fire spread. Therefore, these compounds are normallyused together [2,3].

Wood has many good properties from the point of viewof processing, physical and mechanical properties, aes-thetic, environmental and health aspects. In many coun-tries, wood is widely used as a building material and insome areas as the main construction material [4]. Wood(unlike steel and concrete) is a combustible material, andcertain types of constructions (defined by the Code) donot permit the use of combustible materials. There arearguments for and against this type of restriction, but theselimitations do exist. The fire-resistive requirements are veryimportant to the building designer [5]. For this reason, pre-dicting the models for thermal degradation of untreatedand boron-treated wood is very important.

2166 A. Temiz et al. / Construction and Building Materials 22 (2008) 2165–2169

In this study, the fire resistance properties of alder andpine wood treated with boron compounds and their modelswere investigated.

2. Material and method

2.1. Material

Sapwood of alder wood (Alnus glutinosa (L.) Gaertn.

subsp. barbata (C.A. Mey) Yalt.) and southern pine (Pinus

sylvestris L.) samples were used. The wood samples wereobtained from northeastern part of Turkey, along theBlack Sea coast. The densities of alder wood and southernpine samples were 0.503 and 0.52 g/cm3, respectively. Thesamples were cut into 13 · 13 · 76 mm (radial by tangentialby longitudinal). The wood samples were prepared accord-ing to ASTM E 160 [6].

The solutions of boron compounds, (boric acid, borax)alone or in mixed forms, were used. The concentrationsof chemicals used in this study are given in Table 1.

Table 1Retentions of treatment chemicals

Boron compounds Retention (kg/m3)a Retention (kg/m3)b

Mean Mean

Control – –Boric acid (1%) 6.53 (0.55)c 5.29 (1.32)Boric acid (3%) 19.09 (2.04) 17.67 (2.38)Boric acid (5%) 32.11 (3.11) 32.85 (3.27)Borax (1%) 6.30 (0.38) 6.24 (0.98)Borax (3%) 19.42 (1.96) 19.61 (2.41)Borax (5%) 32.24 (3.13) 31.99 (3.34)Boric acid + borax (1%) 6.12 (0.77) 5.93 (0.98)Boric acid + borax (3%) 18.13 (1.91) 17.43 (2.19)Boric acid + borax (5%) 29.88 (2.94) 31.02 (3.21)

a Retention for alder wood.b Retention for southern pine wood.c The values in the parenthesis indicates standard deviation.

Fig. 1. Fire test

2.2. Method

2.2.1. Impregnation procedure

Test samples were impregnated with boric acid (1%, 3%,5%), borax (1%, 3%, 5%) and boric acid + borax (1%, 3%,5%) with vacuum for 30 min and at ambient pressure for30 min according to ASTM D 1413 [7]. The wood sampleswere then removed from the treatment solution, wipedlightly to remove solution from the wood surface, andweighed (nearest 0.01 g) to determine gross retentions foreach treating solution and sample. The gross retentionsfor each treatment solution were calculated according tothe following formula:

R ¼ G� CV� 10 kg=m3; ð1Þ

where G: (T2–T1) is grams of treating solution absorbed bysamples (initial weight of block subtracted from the initialweight plus the treating solution absorbed); C is the gramsof preservative or preservative solution in 100 g of thetreating solution; and V is the volume of samples in cubiccentimeters.

2.2.2. Fire test method

Fire test for the wood specimens was performed accord-ing to ASTM E 160. Specimens were conditioned at27 ± 2 �C and 30–35% relative humidity to the targetedequilibrium moisture content of 7% prior to fire testing.The 24 specimens were arranged into 12 layers shaped likea square prism (Fig. 1).

The heating flame was connected to an LPG tank con-trolled by a sensitive bar-gauged valve. The flame was bal-anced the standard height before the fire test samples’frame was put in position. Then, fire testing was performedwith a flame source, without a flame source and at theglowing stage according to ASTM E 160. Temperaturewas recorded as a celsius degree (�C) at the combustion col-umn by a thermocouple at 15, 30 s and thereafter at 30 s

apparatus.

A. Temiz et al. / Construction and Building Materials 22 (2008) 2165–2169 2167

intervals for combustion with a flame source, without aflame source and at glowing stage, respectively. A lightmeter was also used to measure the amount of light (lightdensity). Mass losses of samples after fire test were calcu-lated according to the following equation:

Mass losses ð%Þ ¼ ððW bf � W afÞ=W bfÞ � 100; ð2Þwhere Wbf is the weight (g) of wood samples before fire testand Waf is the weight (g) of wood samples after fire test.

3. Results and discussion

The gross retentions of treatment chemicals are given inTable 1.

Weight losses, the temperature and light density valuesof samples during fire test are shown in Table 2. Tempera-ture degrees in celsius were recorded by thermocouple at15, 30 and 30 s intervals for combustion with and withoutheat source and at glowing stage, respectively. TWHS is thesymbol of combustion time without heat source and TAGafter glow time as listed in Table 2.

The results indicated that the weight loss of untreatedpine wood (control) was higher than that of untreated alderwood (control) because pine contains resin and extractivematerials which accelerate combustion. In addition, alderwood is very rich in terms of tannin contents whichdecrease the weight losses. It was also found that inorganicboron compounds with aqueous solutions were very effec-tive as fire retardant and weight losses reduced while boronconcentration increased in Fig. 2.

Table 2Weight losses and values of wood samples in combustion

Treatment chemicals Weight loss (%) Values of with heat source V

Temp (�C) Light density T

Means Means Means M

Alder wood

Control 89.13 478 900 7Boric acid (1%) 85.90 538 915 5Boric acid (3%) 75.46 378 955 4Boric acid (5%) 70.30 330 976 5Borax (1%) 88.61 482 910 6Borax (3%) 83.16 369 951 5Borax (5%) 77.72 284 980 4Boric acid + borax (1%) 84.08 473 922 6Boric acid + borax (3%) 78.89 345 981 5Boric acid + borax (5%) 68.72 275 981 4

Southern pine

Control 92.06 448 917 7Boric acid (1%) 86.90 538 883 5Boric acid (3%) 77.69 373 959 4Boric acid (5%) 75.12 319 965 4Borax (1%) 90.66 465 913 6Borax (3%) 88.01 340 969 5Borax (5%) 85.09 276 977 4Boric acid + borax (1%) 89.13 366 945 5Boric acid + borax (3%) 79.79 333 958 4Boric acid + borax (5%) 72.37 317 972 4

BA + BX mixture (especially 5%) (7:3, as weight) wasmore effective than individual boric acid or borax.

After combustion, the lowest mass losses found fromalder and pine samples treated with the mixture of boricacid and borax (5%) were 68.72% and 72.37%, respectively.

Researchers reported that the wood samples treatedwith boron compounds reduced the weight losses and wereeffective as a fire retardant [8–13].

In order to determine the relationship between the con-centrations of fire retardant solutions and weight losses anddetermine the model to predict the weight losses, regressionanalysis for each wood species was performed by using sta-tistical software package (SPSS). Relationships between theconcentration of fire retardant solutions and weight lossesfor alder and pine are depicted in Figs. 3 and 4,respectively.

As seen in Figs. 3 and 4, for both wood species, while therewas a linear correlation between BX concentrations andweight losses, the correlation between BA and BA + BXconcentrations and weight losses was more complicatedthan that of BX. Once BA concentration reached 4%, theweight losses tended to stay constant. In other words,increasing the concentrations of BA by more than 4% didnot help much in terms of fire resistance and did not lowerthe weight losses. For BA + BX mixture, the weight lossesshowed a slight decrease until their concentration reached3%. Once the concentration reached that level, it wasfound that there was an improvement in terms of fire resis-tance. Therefore, at least 3% or higher concentrations ofBA + BX mixture should be used to provide fire resistance.

alues of without heat source Values of after glow twhs Tag

emp (�C) Light density Temp (�C) Light density sec. sec.

eans Means Means Means

16 829 451 1000 120 45082 888 172 1000 165 64590 910 136 999 210 66005 911 140 999 195 58554 842 225 992 150 37557 846 252 985 165 25578 872 210 987 240 51031 865 206 996 150 43513 899 135 998 180 87060 910 144 998 210 555

46 778 246 986 225 42074 880 152 1000 165 82596 909 137 1000 240 76536 903 135 1000 255 67550 811 233 991 165 42030 892 189 999 270 58528 888 174 998 330 47539 892 168 999 210 69094 890 134 999 195 84043 906 135 999 210 690

Fig. 2. Weight losses of treated wood with boron compounds.

Fig. 3. Relationship between the concentration of fire retardant solutionsand weight losses for alder wood.

Fig. 4. Relationship between the concentration of fire retardant solutionsand weight losses for pine.

2168 A. Temiz et al. / Construction and Building Materials 22 (2008) 2165–2169

The models for each wood species and fire retardantsolution were also determined (Figs. 3 and 4). Thanks tothe determined models in this study, the weight losses canbe predicted as a given concentration of each fire retardantsolution.

4. Conclusion

In this study, the fire resistance properties of wood trea-ted with boron compounds were investigated. In addition,a simple empirical correlation for each wood species andfire retardant solution were also determined.

The results indicated that inorganic boron compoundswith aqueous solutions were very effective as fire retar-dant. A mixture of boric acid and borax showed the high-est fire resistance in this study. As expected, the highestTWHS results were determined from the samples treatedwith borax (5%) since borax tends to promote smolderingtime.

The strength of a wood member in service is affectedby the temperature. Strength is increased as the tempera-ture goes down below the normal temperature rangefound in most buildings. On the other hand, the strengthdecreases as temperatures are increased. Tabulated designvalues are applied to wood used in the ordinary tempera-ture range and are perhaps occasionally heated up to74 �C. Prolonged exposure to temperatures above 74 �Cmay result in a permanent loss of strength. Reductionsin strength caused by heating below 74 �C are generallyreversible. Temperature factor (Ct) is one of the mostimportant adjustment factors for building design.Ct = 1.0 is normally used in the design of ordinary woodframe buildings; however, in an industrial plant theremaybe operations that cause the temperatures to be con-sistently elevated [5]. Therefore, temperature factor is verycrucial in terms of designing wooden building materials.Since dielectric constant and heat flow of treated woodwith boron compounds is lower than untreated wood,boron compound not only provides an excellent fire resis-tance as well as efficiency against both fungi and insectsbut also does not cause any strength losses in woodmaterials.

References

[1] Laks PE, Manning MJ. Inorganic borates as preservative systems forwood composites. In: Proceedings of second pacific rim bio-basedsymposium, Vancouver, Canada; 1994. p. 236–44.

[2] Le Van SL, Tran HC. The role of boron in flame retardant treatment.In: Proceedings of first international conference on wood protectionwith diffusible preservatives, 28–30 Nov. Tennessee, USA; 1990. p.39–41.

A. Temiz et al. / Construction and Building Materials 22 (2008) 2165–2169 2169

[3] Baysal E. Determination of oxygen index levels and thermal analysisof scots pine (Pinus sylvestris L.) impregnated with melamineformaldehyde-boron combinations. J Fire Sci 2002;20(5):373–89.

[4] Grexa O. Flame retardant treated wood products. In: The proceed-ings of wood & fire safety 2000 (part one) Zvolen, Slovak Republic;2000. p. 101–10. ISBN 80-228-0774-5.

[5] Breyer DE. Design of wood structures. USA; 1993. p. 902, ISBN 0-07-007678-2.

[6] ASTM E 160. Standard test method for combustible properties oftreated wood by the crib test. 1975. p. 809–13.

[7] ASTM D 1413. Standard method of testing wood preservatives bylaboratory soil-block cultures. D 1413-76. In: Annual book of ASTMstandards, vol. 4.09 Wood, Philadelphia, PA; 1988. p. 239–45.

[8] Temiz A, Yıldız UC. Fire resistance of alder wood treated with somechemicals. Part II. Effects of other chemicals on the combustionproperties, International Research Group on Wood Preservation 33.Annual Conference, IRG/WP 02-40235, 12–17 May, Cardiff, Wales,UK; 2002.

[9] Temiz A. Effects of the wood preservatives on the physical andmechanical properties of alder wood (Alnus glutinosa (l.) gaertn.subsp. barbata (C.A. Mey) Yalt.), M.Sc. Thesis. Trabzon, Turkey:

Graduate School of Natural and Applied Science Karadeniz Tech-nical University; 2000.

[10] Yıldız UC, Temiz A, Gezer ED, Yıldız S. Combustion properties ofalder wood treated with new environment friendly natural extractives.Part I. Effect of natural tannins on the combustion properties,International Research Group on Wood Preservation 33. AnnualConference, IRG/WP 02-40234, 12–17 May. Cardiff, Wales, UK;2002.

[11] Baysal E. The effect of boron compounds and water repellents onthe some physical properties of brutia pine MS Thesis. Trabzon,Turkey: K.T.U, Graduate School of Natural and Applied Science;1994.

[12] Yalinkilic MK, Demirci Z, Baysal E. Fire resistance of douglas fir[Pseudotsuga menziesii (Mirb.) Franco] wood treated with somechemicals. J Eng Sci. Engineering College: Pamukkale University;1998;4(1–2). p. 613–25.

[13] Colak S, Temiz A, Yıldız UC, Colakoglu G. Fire retardant treatedwood and plywood: A comparative study. Part III. Combustionproperties of treated wood and plywood, International ResearchGroup on Wood Preservation 33. Annual Conference, IRG/WP 02-40236, 12–17 May. Cardiff, Wales, UK; 2002.