Production of mahogany sawdust reinforced LDPE wood-plastic composites using statistical response surface methodology

Production of mahogany sawdust reinforced LDPE wood-plastic composites using statistical response surface methodology

Sofina-E-Arab?Md.Azharul Islam

We produced wood–plastic composite board by using sawmill wastage of mahogany(Swietenia macrophylla)wood and low density polyethylene.We used multi-response optimization to optimize the process parameters of composite board production including mixing ratio,fire retardant(%)and pressing time(min).We investigated the effects of these three process parameters in the mechanical and physical properties of the composite board.Afterwards,Box–Behnken design was performed as response surface methodology with desirability functions to attain the optimallevelof mixing ratio,fire retardantand pressing time(min).The maximum modulus of elasticity (MOE)and modulus of rupture(MOR)were achieved at the optimal conditions of wood plastic mixing ratio of 60:40,pressing time of 9 min and zero fire retardant percentage.The optimized MOR and MOE were 13.12 and 1,781.0 N mm-2,respectively.

Wood plastic composite·LDPE·Mechanical properties·Physical properties·Response surface methodology

Introduction

Advances in science and technology have led to more sophisticated and diverse uses of wood(Panshin and DeZeewe 1980).Increasing availability of composite products including plywood,oriented standard board(OSB),hardboard,particle board,fiberboard,and veneer counters the growing scarcity of timber and degradation of biodiversity due to timber harvest(Islam et al.2012).Enhanced utilization of wood requires more intelligent processing technologies(Panshin and De Zeewe 1980)such as improved composite products which are useful economic alternatives to solid wood.They are used both for interior and exterior construction and furniture manufacturing throughout the world(Leu et al.2011).

Most wood based composites were previously bonded with urea formaldehyde and phenolformaldehyde resin but emission of formaldehyde from these types of composites has become an environmental concern(Lee et al.2002). For this reason,wood–plastic composite(WPC)was introduced in 1970 in Italy to solve this problem(Pritchard 2004).Environmentally friendly alternatives to these wood–plastic materials include products that use polyethylene(Kuo et al.1998;Rahim 2009;Yuan et al.2013). Various types of polyethylene are used in preparing WPC including low-density polyethylene(LDPE),high density polyethylene(HDPE),and linear low-density polyethylene (LLDPE)(Chen et al.2007).Studies of wood composites based on low-density polyethylene(LDPE)are few(Yamashita et al.1999).LDPE has been used as a binding material in a combination with sawdust to manufacture wood composite product.Low density polyethylene (LDPE)is a thermoplastic made from petroleum.LDPE materials are strong,light-weight and durable(Pramila and Ramesh 2011).Low density polyethylene(LDPE)is a polyethylene with a density of less than 940 kg m-3and melting point of 120°C.It is produced by a high pressure process,and is often referred to as high pressure polyethylene.Recent studies have evaluated particle boardproduction using low density polyethylene(LDPE)(A-tuanya et al.2011;Zhang et al.2013).

Because polymersare now used to manufacture a wide range of wood–plastic composites(WPCs),the fire risk of polymers hasattracted considerable attention.Acyclic pattern isengaged in the process of combustion of polymers-preheating,decomposition,ignition,and combustion.In the presence of a flame source,the polymer becomes preheated.Boron based fire retardants,for example,a combination of Borax(Na2B4O7-10H2O)and Boric acid(H3BO3)have yielded improved performance(Stark etal.2010).The use offire retardantin WPCs has an impacton the mechanicalproperties of wood butcomposites with borax/boric acid have shown better dimensional stability and bending strength(Youngquistetal.1997).In our study,we used a mixture of Borax and Boric acid to study the effectoffire retardants on mechanicalproperties of wood.

When using low density polyethylene(LDPE)in composite board production there are many factors that influence board quality,including mixing ratio,temperature, time,pressure,and type of coupling agent(Atuanya et al. 2011;Benthien and Thoemen 2012;Zhang et al.2013; Yang et al.2013).In our study,the mixing ratio,pressing time and fire retardant were evaluated in production of sawdust LDPE composite.

Statisticalexperimentaldesign has been used in manufacturing ofwood compositesin severalstudies(Zhao etal.2008; Tabarsa etal.2011;Azizietal.2011;Nirdosha etal.2011). Butthe use of modern industrialoptimization tools has been limited in identifying the factors responsible for the quality production ofwood composites.This has resulted in wastage of raw materials and reagents,and increases in other production-oriented costs.This has adversely affected forest resources and biodiversity(Islam etal.2012).Response surface methodology(RSM)is a toolthatincludes collection of statisticaland mathematical techniques used for developing, improving and optimizing processes.It evaluates the relationships between a group of controlled experimentalfactors and the observed results of one or more selected criteria (Myers and Montgomery 2001).We aimed to optimize the process parameters of wood composite board production by using multi-response optimization process.Box–Behnken design was performed as the response surface methodology (RSM)to discoverthe optimum combination ofmixing ratio, pressing time and fire retardants that affect mechanical and physicalproperties of WPC production.

Materials and methods

Materials

Sawdust of mahogany(Swietenia macrophylla)was collected from a local sawmill,boric acid and borax from the local market of Khulna,and LDPE(low density polyethylene)was sourced from Dhaka,Bangladesh.The sawdust and plastic powder were screened through a 35 mesh(500 micron or 0.5 mm).A multi-daylight hot press that was digital and electrically heated(model-XLB,Manufacture no.120049,Qingdaoyadong,China)was used to manufacture the boards.A Universal Testing Machine(Model UTN100,Fuel Instruments and Engineers Pvt.Ltd.India) of Khulna University of Engineering and Technology (KUET)was used to determine the mechanical properties (MOR and MOE)of the boards according to the ASTM D790-10(2010)standard.The physical properties such as moisture content,density,linear expansion,thickness swelling and water absorption of manufactured WPC boards were determined according to the ASTM D1037-12 (2012).

Experimental design

Three levels of sawdust:LDPE mixing ratios(70:30,60:40, 50:50)were used to manufacture WPC boards.Fire retardants(Boric acid+Borax ata ratio of 50:50)were used as 0,3 and 6%of the total weight of raw materials.Three durations of pressing time were used(7,9 and 11 min). Press temperature and pressure were fixed at 150°C and 4 Mpa,respectively,according to the literature(Leu et al. 2011;Lee et al.2002;Pritchard 2004;Kuo et al.1998; Zhang et al.2013).

Table 1 shows three levels(low,medium and high)ofthree factors(mixing ratio,pressing time and fire retardant).The corresponding Box–Behnken design matrix is shown with results in Table 2.The Box–Behnken design is a widely used RSMapproach developed by Box and Behnken(1960)forthree levelfactorsthatfitssecond-ordermodelsto the response.Box–Behnken design avoidsallhigh values ofthe variables(extreme treatmentconditions).The applied Box–Behnken consisted of 15 experiments(N=2 k(k-1)+Co=2*3(3-1)+ 3=15 runs(i.e.,15 composite boards),where N is the total number of experiments required,k is the number of factors or variables and Cois center point’s runs.

STATISTICA(statistical software)was used for the design and analysis of the Box–Behnken design.To fit the response to the independent variables a second order polynomial model was used:

Table 1 Factors with range and level of manufacturing of WPC boards

Table 2 Design of studied variables used in the manufacture of the WPC boards

where,Y is the response(MOE and MOR),β0is the intercept parameter andβi,βiiandβijare parameters for linear,squared and interaction factor effects,respectively.

Results and discussion

Box–Behnken design analysis and development of process model

The processing variables of the experiment(mixing ratio, pressing time and fire retardantpercentages)had significant effects on the mechanical properties of boards.The interaction effects of different factors on bending properties of boards,particularly on MOR and MOE,were calculated by using the Box–Behnken design.

The sufficiency of the model was evaluated through analysis of variance(ANOVA).The lack of fit(LOF)was not significant relative to the pure error,indicating good response to the model.The modelregression co-efficientof determination of(R2)of 0.9282 for MOR and 0.9236 for MOE was in reasonable agreement with the experimental results,indicating 92.82 and 92.36%of the variability can be revealed by the model,leaving 7.18 and 7.64%residual variability for MOR and MOE,respectively.

We conclude from the above results that the Box–Behnken design was adequate to predict the board strength (MOR and MOE)within the range of variables studied. The final predicted process models in terms of actual significant factors for MOR and MOE are given in Eqs.2 and 3,respectively.

where,a is the mixing ratio,b is the pressing time,and c is the fire retardant.

Factors affecting MOR

The factors affecting the MOR of boards are shown in ANOVA(Table 3),and also represented by a standardized Pareto chart developed by the software and shown in Fig.1.The quadratic effect of pressing time and quadratic effect of mixing ratio had positive significant effects on MOR.The linear effect of the mixing ratio had a negative significant effect on MOR.The interaction between linear effects of pressing time and fire retardant on MOR had a positive significanteffecton MOR.The positive interaction effect of pressing time and fire retardant on MOR is presented in Fig.2.MOR attained the highest value of 12 N mm-2for pressing time of 8–9 min and fire retardant proportion of 2–3%(Fig.2).The increases of pressing time increased the MOR whereas increases of fire retardant decreased the MOR sharply.Similar observation was also reported by Ayrilmis et al.(2012)that the MOR of uncoupled(without polypropylene)specimens declines with increasing fire retardant content and uncoupled control specimens yield the highest MOR.In one study by Atuanya et al.(2011)investigated the production of WPCat140°C,pressure of 4 MPa,pressing time of 10 min and at 60/40 wood particles/LDPE ratios that yielded the highest MOR of 20.31 N mm-2.Zhang et al.(2013)also reported MOR value of 21 N mm-2when the hot pressing time was 10 min,hot pressing temperature was 180°C, and hotpressing pressure was 3.0 MPa in WPC production with low density polyethylene.

Table 3 ANOVA for MOR

Fig.1 Standardized Pareto Chart for MOR

Factors affecting MOE

The quadratic effects of pressing time and mixing ratio had positive significant effects on MOE and the linear effect of fire retardant and mixing ratio had a significant negative effect on MOE(see ANOVA and Standardized Pareto Chart,Table 4;Fig.3).The interaction between linear effect of mixing ratio and fire retardant on MOE had significant and negative effect on MOE.

Fig.2 The effect of pressing time and fire retardanton MOR

The negative interaction effect of mixing ratio and fire retardanton MOE is illustrated in Fig.4.MOE attained the highest value,1,800 N mm-2for mixing ratio ranges from 1.5 to 2.5%and fire retardantproportion from 2 to 4%.As the mixing ratio declined from 1.5 to 1%or increased from 2.5 to 3%and as the fire retardant decreased from 2%to 0 or increased from 4 to 7%,MOE declined sharply.This finding suggested that the addition of LDPE to the wood particles increased the stiffness of the composite boards.Similar observations for MOE (1,617 N mm-2)was found by the Atuanya et al.(2011)in the production of WPC where wood particles and LDPE mixing ratio was maintained of 70:30.In another report by Obidiegwu and Nwosu(2012)stated that the tensile properties of board increased with increasing wood fiber content from 0 to 20%while Young’s Modulus ofelasticity showed progressive increase after initial reduction of wood fiber content percentage.

Table 4 ANOVA for MOE

Fig.3 Standardized Pareto Chart for MOE

Fig.4 The effect of pressing time and fire retardant on MOE

Optimization

The optimization process was followed by the software profile and desirability option.The theory and concept of the desirability function were reviewed by Islam et al. (2009,2010).The best optimized conditions were found to be a mixing ratio of 2%(wood plastic mixing ratio 60:40), pressing time of 9 min and fire retardants at 0.This also optimized MOR and MOE at13.129 and 1,781.0 N mm-2, respectively(Fig.5).Finally,by using optimized levels of parameters(mixing ratio 60:40 without fire retardants, pressing time of 9 min,temperature of 150°C and press pressure of 4 MPa)a WPC board was produced and found the closestvalues similar to optimized MOR and MOE that indicated good fit to the predicted model.

Effects on physical properties

Density

Mixing ratio had a remarkable effect on density(Fig.6). Density increased with increasing content of plastic and reached a peak at a level of 2(ratio of sawdust:LDPE of 60:40),and then declined with further increase in LDPE content.Density increased with increasing pressing time, with highest density at approximately 9 min.Density declined with further increase in pressing time.Fire retardanthad an irregular effecton density.Highestdensity was achieved when fire retardant content was 0,and reduced sharply with increases in fire retardant.Density increased slightly when fire retardantcontentwas 5–6%.The highest WPC board density was observed 0.81 mg cm-3.

Fig.5 Desirability profile for optimization of process parameters for MOE and MOR

Fig.6 Effects of production parameters on density

Moisture content

The mixing ratio had a remarkable effect on moisture content(Fig.7).Moisture content declined with increasing plastic content and reached a peak at the level 1(the ratio of sawdust:LDPE was 70:30.Pressing time had no such effect on moisture content.Fire retardant

content had an irregular effect on moisture content;with highest moisture content when fire retardant was 6%.

Fig.7 Effects of production parameters on moisture content%

Thickness swelling

Thickness swelling(%)decreased with increasing mixing ratio(Fig.8).Swelling was highest at the mixing ratio level 1(70:30),and declined with increasing LDPE content.Pressing time had no remarkable effect on thickness swelling.It increased slightly and,after reaching a certain limit,decreased with further increase in pressing time. Content of fire retardant also had no remarkable effect on thickness swelling(Ayrilmis et al.2011).

Fig.8 Effects of production parameters on thickness swelling

Linear expansion

Linear expansion(%)decreased with increasing content of LDPE(Fig.9),and was highest at mixing ratio level 1 (70:30).Linear expansion was greatest at pressing time of 7 min,and then decreased with the increasing pressing time and after a certain time itdecreased with the further increase in pressing time.Again it has been observed that the linear expansion percentage decreased with the increasing content of fire retardant.

As atlevel1 which contains 70%ofsawdust,soak water more than level 3 which contains 50%of LDPE.So,the linear expansion is high at mixing ratio level 1.Pressing time has an irregular effect on linear expansion as it has a complex interaction with the other factors.The effectof fire retardant has got almost a similar trend with the water absorption.With the increase of fire retardant the linear expansion decreases gradually as the water absorption.

Water absorption

Fig.9 Effects of production parameters on linear expansion

Fig.10 Effects of production parameters on water absorption

Water absorption is very important parameters for the quality of WPCs in application.Several efforts have been made by different researchers to describe the water absorption behavior of the WPC(Najafiet al.2007;Tamrakar and Lopez-Anido 2011).The water absorption declined with increasing LDPE content(Fig.10).It was highest at the mixing ratio level of 1(70:30).Water absorption was highest at a pressing time of 11 min.It initially declined slightly with increasing pressing time and then sharply increased with increased pressing time.Water absorption declined with increasing contentof fire retardant.

Conclusion

Sawdust-incorporated WPC was manufactured in our study by applying a statistical optimization tool which helped to reduce the waste of raw materials.The maximum optimized MOE and MOR were achieved at a mixing ratio of 60:40, pressing time of 9 min and fire retardants of 0.The optimized MOR and MOE were 13.129 and 1,781.0 N mm-2, respectively.The pressing time and maxing ratio also had influenced on the all studied physical properties whereas fire retardanthas irregular effects on the physicalproperties of WPCs.Timber shortage is a growing global issue and scientific use of timber processing waste such as sawdustin composite board production can contribute greatly to meet demand for wood products.This study suggested that the mahogany sawdust incorporation with low density polyethylene(LDPE)is a very potential alternative to produce quality WPC.

ASTM D1037-12(2012)Standard test methods for evaluating properties of wood-base fiber and particle panel materials. ASTM International,West Conshohocken

ASTM D790-10(2010)Standard test methods for flexuralproperties of unreinforced and reinforced plastics and electrical insulating materials.ASTM International,West Conshohocken

Atuanya CU,Ibhadode AOA,Igboanugo AC(2011)Potential of using recycled low-density polyethylene in wood composites board.Afr J Environ Sci Technol 5(5):389–396

Ayrilmis N,Benthien JT,Thoemen H,White RH(2011)Effects of fire retardants on physical,mechanical,and fire properties offlatpressed WPCs.Eur J Wood Wood Prod 70:215–224

Ayrilmis N,Akbulut T,Dundar T,White RH,Mengeloglu F, Buyuksari U,Candan Z,Avci E(2012)Effect of boron and phosphate compounds on physical,mechanical,and fire properties of wood–polypropylene composite.Constr Build Mater 33:63–69

Azizi K,Tabarsa T,Ashori A(2011)Performance characterizations of particleboards made with wheat straw and waste veneer splinters.Compos B 42(7):2085–2089

Benthien JT,Thoemen H(2012)Effects of raw materials and process parameters on the physical and mechanical properties of flat pressed WPC panels.Compos A Appl SciManuf43(4):570–576

Box GP,Behnken DW(1960)Some new three level design for the study of quantitative variables.Technometrics 2(4):456–475

Chen JH,Zhong JC,Cai YH,Su WB,Yang YB(2007)Morphology and thermal properties in the binary blends of poly(propyleneco-ethylene)copolymer and isotactic polypropylene with polyethylene.Polymer 48(10):2946–2957

Islam MA,Sakkas V,Albanis TA(2009)Application of statistical design of experimentwith desirability functions for the removal of organophosphorus pesticide from aqueous solution by low cost material.J Hazard Mater 170(1):230–238

Islam MA,Nikoloutsou Z,Sakkas V,Papatheodorou M,Albanis TA (2010)Statistical optimization by combination of response surface methodology and desirability function for removal of azo dye from aqueous solution.Int J Environ Anal Chem 90:495–507

Islam MA,Alam MR,Hannan MO(2012)Multiresponse optimization based on statistical response surface methodology and desirability function for the production of particleboard.Compos B Eng 43(3):861–868

Kuo M,Adams D,Myers D,Curry D,Heemstra K,Smith J,Bian Y (1998)Properties of wood agricultural fiberboard bonden with soybean-based adhesives.For Prod J 48(2):71–75

Lee H,Kim HJ,Park HJ(2002)Performance of paper sludge/ polypropylene fiber/linocellulosic fiber composites.J Ind Eng Chem 8(1):50–56

Leu SY,Yang TH,Lo SF,Yang TH(2011)Optimized material composition to improve the physical and mechanical properties of extruded wood–plastic composites(WPCs.).Constr Build Mater 29:120–127

Myers RH,Montgomery DC(2001)Response surface methodology: process and product optimization using designed experiments, 2nd edn.Wiley,New York

NajafiSK,Kiaefar A,Hamidina E,Tajvidi M(2007)Water absorption behavior of composites from sawdust and recycled plastics.J Reinf Plast Compos 26(3):341–348

Nirdosha G,Setunge S,Jollands M(2011)Design of experiments (DOE)for investigating particleboard production.In:Fragomeni S,Venkatesan S,Lam NTK,Setunge S(eds)Incorporating sustainable practice in mechanics of structures and materials. Taylor&Francis Group,London

Obidiegwu MU,Nwosu CT(2012)Mechanical and flammability properties of low density polyethylene/ALSTONIA BOONEI wood fiber composite.Asian J Nat Appl Sci1(2):53–62

Panshin AJ,De Zeewe C(1980)Textbook of wood technology,4th edn.McGraw–Hill Book Company,Inc,New York,pp 1–402

Pramila R,Ramesh KV(2011)Biodegradation of low density polyethylene(LDPE)by fungi isolated from municipal landfill area.J Microbiol Biotechnol 1(4):131–136

Pritchard G(2004)Two technologies merge:wood plastic composites.Plast Addit Compd 6(4):18–21

Rahim S(2009)Wood waste utilization for wood composite industries in Malaysia.Paper presented at Asia Pacific Forest Products Workshop,14 Dec

Stark NM,White HR,Mueller SA,Osswald TA(2010)Evaluation of various fire retardants for use in wood flour-polyethylene composites.Polym Degrad Stab 95(9):1903–1910

Tabarsa T,Ashori A,Gholamzadeh M(2011)Evaluation of surface roughness and mechanical properties of particleboard panels made from bagasse.Compos B Eng 42(5):1330–1335

Tamrakar S,Lopez-Anido RA(2011)Water absorption of wood polypropylene composite sheet piles and its influence on mechanical properties.Constr Build Mater 25(10):3977–3988

Yamashita Y,Watanabe H,Takeda S(1999)Rubber wood characteristics of its supply and development of its utilization.For Resour 71:65–70

Yang WJ,Xie YJ,Wang HG,Liu BY,Wang QW(2013)Impacts of freezing and thermaltreatments on dimensionaland mechanical properties of wood flour-HDPE composite.J For Res 24(1):143–147

Youngquist JA,Krzysik AM,Chow P,Meimban R(1997)Properties of composite panels.In:Roger RM et al(eds)Paper and composites from agro-based resources.CRC Lewis Publisher, Boca Raton,pp 309–311

Yuan FP,Ou RX,Xie YJ,Wang QW(2013)Reinforcing effects of modified Kevlar?fiber on the mechanical properties of woodflour/polypropylene composites.J For Res 24(1):149–153

Zhang ZT,Sun ZB,Yang XH,Dong YH(2013)Manufacturing composite from regenerated wood fiber and polyethylene via paten-pressing process.Adv Mater Res 645:39–42

Zhao J,Wang XM,Chang JM,Zheng K(2008)Optimization of processing variables in wood–rubber composite panel manufacturing technology.Bioresour Technol 99(7):2384–2391

15 October 2013/Accepted:16 December 2013/Published online:27 January 2015

?Northeast Forestry University and Springer-Verlag Berlin Heidelberg 2015

The online version is available at http://www.springerlink.com

Corresponding editor:Yu Lei

Sofina-E-Arab·Md.A.Islam(?)

Forestry and Wood Technology Discipline,Khulna University, Khulna 9208,Bangladesh e-mail:iazharul@gmail.com