Relationship between fiber furnish properties and the structural performance of MDF

Relationship between FiberFurnish Properties andthe Structural Performanceof MDF

Les GroomU.S. Forest Service, Pineville, Louisiana

Laurence MottPerstorp AB, Perstorp, Sweden

Stephen ShalerUniversity of Maine, Orono, Maine

Ab&actThe structural performance of medium density

fiberboard (MDF) is attributable to three primaryvariables which are: physical and mechanical prop-erties of individual wood fibers; fiber-to-fiber stresstransfer; and fiber orientation. These origins of fiberproperties and stress transfer can be traced to thefiber generation method wherein fiber orientation isassociated with mat formation. This paper is part ofan on-going study to determine the mechanismsgoverning the stiffness and strength of fiber-basedcomposites. Preliminary data are presented in thispaper focusing on the effect of juvenility and fibergeneration on the mechanical properties of individ-ual wood fibers and the subsequent properties ofMDF panels. Development of panel stiffness andstrength is also discussed with regards to fiber pack-

Groom Mott

Shaler

ing and stress transfer as determined by testingoriented and un-oriented panels as well as directobservation with microtomography.

IntroductionOver the past few decades, the need for improved

use of this nation’s natural resources, particularlyforest products, has developed. One of the best waysto address this challenge is through an expansion ofthe reconstituted wood fiber philosophy. This phi-losophy began with rectangular dimension lumbercut from cylindrical logs and has evolved to the pointof fiberboard. Breaking down wood into its constit-uent elements is not exceedingly difficult. The chal-lenge lies in reassembling the elements to meet de-sired end specifications. The manner in which theseelements are reassembled dictates the performance

Groom, Mott , and Shaler l 89

of the final product, which is the overall premise ofengineering wood products.

This axiom is applicable to the production of me-dium density fiberboard (MDF) with improved per-formance. The stiffness and strength of MDF, likeany other wood composite is dependent upon theproperties of the individual wood components andthe manner in which these components are com-bined. Most research efforts to improve the mechan-ical performance of MDF are proprietary in natureand have focused primarily on the binding system.Other variables commonly studied include particleorientation, board density, additives, sizing agents,species, and pressing variables (Maloney 1986). Dueto technological limitations, studies correlating fiberproperties and panel properties have only recentlybeen possible.

The long-term goal of this on-going work is to as-certain the mechanisms that govern the physical andmechanical properties of structural fiberboard.These goals are achieved by studying the three fac-tors governing MDF: physical and mechanical prop-erties of individual wood fibers; fiber-to-fiber stresstransfer; and fiber orientation. Data presented in thispaper are the compilation of several studies that ad-dress these factors. Specifically, the objectives of thispaper are to ascertain the relationship between fiberproperties and the structural performance of MDF;determine the effect of refining on the mechanicalproperties of individual fibers, and subsequently theproperties of MDF; and ascertain the in-situ physicalcharacteristics of fibers in MDF.

Experimental ProceduresThe experimental procedures can be divided into

four main categories based on objectives and antici-pated outcomes. In all cases, the raw material for theconstruction of MDF panels and mini panels wasloblolly pine (Pinus tueda) harvested from a conven-tional plantation in southern Arkansas. The juvenileand mature portions of the bole were separatedmanually, The term “juvenile” in this study refers towood up to the tenth growth ring whereas “mature”refers to wood beyond the 30th growth ring. Thetransition wood between these zones was discarded.

Effect of Fiber PropertiesFull-Size Panels. -Thirty 16- by 16- by 3/8-in.

MDF panels were manufactured at the University ofMaine to investigate the relationship between fiberfurnish mechanical properties and MDF perfor-

90 l Groom, Mott, and Shaler

mance. The panels were comprised from furnisllc.3with varying degrees of juvenile and mature fibersFibers were generated with a pressurized disc refirlcrat feed pressures of 10 and 40 psig. Fibers were ;lit

dried and bagged for panel manufacture. Panclbwere manufactured at a target density of 48 Ibs./li, ’and with a panel resin content of 8 percent urc;Iformaldehyde (solid content .= 65%). Panels wc’r~pressed for 3 minutes at 270”F, conditioned (70”1:,50% RH), and subsequently evaluated for mechanical properties, linear expansion (LE), density, andinternal bond (IB) strength.

The mechanical properties of individual fiberswere ascertained on the four fiber types (juvenile OImature, 10 psig or 40 psig). In addition, unrefinctlportions of juvenile and mature chips WCI’Cmacerated in acetic acid and hydrogen peroxide. Vibers were then tested in tension to determine thrmodulus of elasticity (MOE) and ultimate tensilestress (UTS) of the fiber furnish. A detailed explan;ltion of the maceration technique can be found inMott (1996), and a detailed explanation of the rn(chanical property determination can be found iI1Groom et al. (1996).

Mini Panels.-Miniature MDF panels measurin}:4- by 5- by l/8-in. were constructed to verify full-sizespanels and, subsequently, to investigate the effects OIfines and fiber orientation. Mini panels were construtted with a press schedule and adhesive that were*identical to full-size panels. Mini panel and fiberproperties were determined similarly to the full-sizepanels.

Effect of Fiber OrientationThere exist two primary methods to achieve fiber

alignment: electric field (Woodson 1977; Talbot t1974) and physical manipulation. This study usedphysical fiber alignment, consisting of a series of vi-brating fins located approximately 1 mm above theforming mat. The degree of fiber orientation wascontrolled by the gaps between the steel fins. Gallspacings were chosen to be 2,4, and 8 mm. Fiberspassed through the vibrating fins and settled downonto a 4- by S-in. mat with fiber orientation locatetlin the S-in. direction. Mini panels were constructedwith varying degrees of fiber orientation and varyin!:degrees of juvenile and mature fibers. Panels weresubsequently tested for mechanical properties, LE,density, and IB.

IXcct of Fines LoadingMini panels were constructed with varying pro-

~l~~rtions of long fibers and fines as well as fibers in-I(.rmediate in length. Approximately 3 kg each of ju-wnile and mature wood fibers were placed on awries of shakers in which various fragment lengths\vcre segregated. The shaker trays were chosen suchI 1~11 the largest to smallest segments consisted of:

Tray ,lD-5% Shives and bundles Discarded.

I ‘ray 220-30% Mostly fibers, with some Designated

bundles and broken fibers. fibers.‘I’lny 3

30-50% Mixture of fibers and fines. Designatedintermediate.

Tray 420-30% Comprised mostly of fines. Designated

fones.

A total of 24 oriented mini panels were con-\I rutted to investigate the effect of fiber packing and\I ress transfer on the mechanical properties of MDF.‘I’he 24 mini panels were composed of 12 juvenile.Ind 12 mature mini panels, with three replicatesc*;\ch of:

l 100 percent fibers:0 percent fines.l 50 percent fiber:50 percent fines.l 0 percent fibers: 100 percent fines.l 100 percent intermediates.

‘I’here were 24 corresponding mini panels con-51 tucted with random orientation.

I’ffect of Refining LevelsIn addition to the full-size MDF panels described

llreviously, mini panels were constructed to evaluateI he effect of refining levels on fiber and MDF panelIlroperties. Juvenile and mature wood fractionsI’rom a loblolly pine tree were chipped and refined atteed pressures of 48, and 12 bar. These pressures arer-cspectively, below, at, and above the glass transitionlcmperature of lignin. Fibers from these six fur-nishes, as well as juvenile and mature chemicallymacerated fibers, were analyzed for mechanical pro-perties. Mini panels were also constructed to evalu-,Ite the effect of refining on MDF mechanical prop-crties, LE, and IB.

Microtomography of MDF panelsThe in situ physical characteristics of fibers in

MDF, including a three-dimensional fiber network;Ind void structure, were investigated using microto-

mography and image analysis techniques. The state ofindividual fibers within the panel including lumencollapse (if present), fiber length, fiber curl, and ori-entation in three-dimensional space were evaluated.

Microtomography is a technique for nondestruc-tively evaluating the internal three-dimensionalstructure of a material. The measurement principles(Deckman et al. 1991) and its application to cellulosicstructures including solid wood and paper were pre-viously described (Shaler 19%‘; Shaler et al. 1998).

The specific results reported are for an MDF spec-imen from 4 bar refined mature loblolly pine. Thevolume of the specimen scanned was 1.28 by 1.09 by0.37 mm, with the 0.37-mm dimension correspond-ing to the panel z-direction. The specimen was ex-cised approximately 5 mm from the top surface ofthe panel. A lOxlens was used in the microtomo-graphy setup so that the individual voxel resolutionwas 2.4 p. The x-ray energy was 8.5 keV. Informationwas collected at the X2B beam line built and run byExxon Research &Engineering at the Department ofEnergy managed National Synchrotron LightSource (NSLS). Visualizations of the internal struc-ture were developed through volumetric clipping,pseudo coloring, and opacity change techniques.

Results and DiscussionPanel density was determined on IB specimens

prior to test and averaged 45.9 Ib./ft.3 on full sizespecimens and 39.3 lb./f?.3 for mini panels. Averagemoisture content at time of testing was 7.5 percent.The average test duration for bending tests was sixminutes. For mini panels, four bending specimenswere tested for each panel. However, only the mostinterior parallel and perpendicular specimens Wereused in the analysis due to a significant edge effect.

Mean mini panel properties were comparable tothe properties of full-size specimens. The coefficientof variation (COV) of the mini panel mechanicalproperties were generally less than 15 percent, espe-cially for tests measuring in-plane properties. Thelowest variability existed for the linear expansionspecimens with an average COV of 12 percent.Bending properties‘ had a similar variability, withCOV values for MOE and modulus of rupture(MOR) of approximately 14 percent. Internal bondstress values were extremely variable; average COV’swere 36 percent with some samples having COV’sover 80 percent. It appears as though the mini panelsmay be too thin to gather reliable IB data, with re-

Groom, Mott, and Shaler l 91

sults being altered by the process of specimen attach-ment to the IS blocks.

Effect of Fiber PropertiesFull-Size Panels.-The most significant finding

of this study is that there exists an inverse relation-ship between fiber properties and MDF mechanicalproperties (Figs. 1 and 2). The refined, juvenilewood fibers used to make the un-oriented MDFpanels shown in Figures 1 and 2, had a MOE and ul-timate tensile stress (UTS) of 500,000 psi and 29,700psi, respectively. The corresponding mature fibershad MOE and UTS values of 970,000 psi and 62,800psi. This seemingly paradoxical inverse relationshipof stronger panels from weaker fibers must be ex-plained based on something other than fiber me-chanical properties. All panels in the study were ran-domly oriented, so orientation did not play a role.

If weaker fibers make stronger MDF panels, thenit is probable that the governing strength mecha-nisms originate with fiber-to-fiber stress transfer.Stress transfer is encompassed by many peripheralvariables such as adhesion, panel density, and fiberpacking. Adhesive application and amount used inall panels were kept constant, although adhesive dis-tributions may have been different due to the sizeand morphology differences between juvenile andmature fibers (Zobel 1961; Larson 1962). Panel den-sity and density profiles were similar among panels.The juvenile wood fiber components in Figures 1and 2 were shorter, and contained a higher percent-age of fibers less than 0.5 mm, and a lower percent-age of fibers greater than 2.5 mm.

One important note to consider is that the me-chanical properties of MDF may not be governed so

PJUVenlle 7bJ:2bM bOJ:bOM 25Jz75M Malure

Panel Composltion

Figure l.- Full-sized MDF panel MOE shown as afunction of juvenile: mature wood ratios. Verticalbars represent one standard deviation.

much by fiber orientation but rather by fibril orien-tation. Wood fibers are themselves cylindricalmulti-laminate composites that derive their me-chanical properties from the orientation of micro-fibrils. A wood fiber with a microfibril angle of 45”such as a juvenile fiber has in theory MOE values,which are equivalent parallel’and perpendicular tothe long axis of the fiber. The same condition doesnot exist for mature fibers with fibril angles of ap-proximately 5 to lo”, where the longitudinal MOE ismuch greater than the transverse MOE. Thus, al-though the longitudinal modulus of mature fibers isgreater in magnitude than juvenile fibers, the trans-verse modulus of juvenile fibers far exceeds the cor-responding mature fibers. The transverse modulusof mature fibers provides a weak linkage in compos-ite panels and thus may produce MDF panels withdiminished mechanical properties.

Mini Panels.-Mini panels were constructed toconfirm full size panels as well as to evaluate the ef-fectiveness of orientation. Stiffness and strength ofmini panels are summarized in Figures 3 and 4, re-spectively. MOE and MOR show identical trends aswas seen in full-size panels; panel stiffness andstrength increase with increasing percentages of ju-venile fibers. As expected, parallel and perpendicularproperties were identical for un-oriented panels.There is a significant separation of parallel and per-pendicular mechanical properties for the orientedpanels, but both properties follow a similar trend ofincreased mechanical properties with increasingpercentages of juvenile fibers.

6.000

P1.000’

JlJVWlll@ 75Jt25M boJ:M)M 255:75M t&hue

Panel CompositionFigure Z.-Full-sized MDF panel MOR shown as afunction of juvenile: mature wood ratios. Verticalbars represent one standard deviation.

9 2 l Groom, Mott, and Shaler

Effect of OrientationResults for the mechanical property testing of

these varying levels of orientations are shown in Fig-ures 5 and 6. It was anticipated that, according to tra-ditional mechanics theory, improved orientation ofthe individual elements, i.e., fibers would increasemechanical properties in the alignment direction. Asthis orientation improved, the mechanical propertiesof the composites with the strongest elements wouldincrease at a faster rate and surpass correspondingcomposites made of the weakest elements. Indeed, in-.creased orientation did improve the mechanicalproperties of both juvenile and mature panels. How-

ever, the data show some trends that were unantici-pated. The parallel specimens for the juvenile panelsincreased at a faster rate than the mature panels. Al-though the mechanisms for this result have not yetbeen identified in this study, the answer most likelylies either individually or in concert, with alteredbonding areas and the relative differences in stiffness

_

between the mature and juvenile fibers. Mature fibersare generally much stiffer and stronger than their ju-venile counterparts longitudinally. Although it mayseem ideal, this longitudinal stiffness and strengthmake the mature fiber less flexible in a fiber mat. Thisdiminished flexibility decreases the fiber-fiber con-

OJ0 % 25% .Fm 75% lm%

Parcmd Juwnlle Flbm

25% 50%

Pment Juvonlia Fiben

Figure 3.-MOE of un-orientedand oriented MDF mini-panels inrelation to percent of juvenilefiber content.

Figure 4.-MOR of un-orientedand oriented MDF mini-panels inrelation to percent of juvenilefiber content.


tact area and thus the number of sites in whichstresses can be transferred among the fibers.

Effect of FinesShaker trays for separation of fiber fractions var-

ied depending on fiber type.Juvenile fibers were separated as follows:l Long fibers passed through a 20 mesh (0.0394

in.),butwereretainedona45mesh(O.O139in.)tray.

l Fines were the fraction that passed through a140 mesh (0.0041 in.) tray.

Juvenile long and fine fiber fractions, respectively,comprised 23 and 26 percent of the total.

Mature fibers were separated as follows:l Long fibers passed through a 10 mesh (0.0787

in.), but were retained on an 18 mesh (0.0394in.) tray.

l Fines were the fraction that passed through a 50mesh (0.0117 in.) tray.

Mature long and fine fiber fractions, respectively,comprised 20 and 36 percent of the total.

The effect of tines loading on the stiffness andstrength of MDF mini panels is shown in Figures 7and 8, respectively. The mechanical properties areinversely correlated to fines loading, regardless oforientation or fiber type. The compliance and weak-

350,000 -.---.- ._____ ~-- -..-- - - _..._.--- - - -

300.000 ’

260.000 -

EB

1 200,000 .

i‘sg 150,000 -

i

100,000 .

Figure S.-Effect of orientationof MDF mini-panel bending stiff-ness. The smaller fin spacingsequate to increased fiber orien-tation.

3.500

Figure 6.-Effect of orientationof MDF mini-panel bendingstrength. The smaller fin spac-

1,~

WIings equate to increased fiberorientation.


2

Fin Spacing (mm)

ness of the panels comprised of fines was due pri-marily to the lack of fiber length, resulting in lowfiber aspect ratios and ultimately poor physical in-terlocking and fiber-fiber contact. It should be notedthat the mini-panel specimens made up of juvenilefibers outperformed the panels comprised entirelyof mature fibers. In addition, the stiffness ofjuvenilemini panels was less susceptible to fiber loading ef-fects as compared to mature mini panels due to theshorter initial fiber length.

The effect of fines loading on LE is shown in Fig-ure 9, which shows that the ideal furnish for minimalI,E would be some combination of long fibers and

fines. The data are insufficient to specify the optimalratio, but it appears to exist between 0 and 50 percentfines loading. Moss and Retulainen (1995) foundthat addition of fines to kraft paper increasesstrength properties by filling in voids and thus pro-viding a greater degree of bonding. A similar effectmay be occurring in the MDF samples. Figure 9 alsoshows that perpendicular specimens exhibitgreater LE than parallel specimens; juvenile speci-mens generally have a greater LE than their maturecounterparts; and in all cases, mini panels com-prised of only intermediate fiber lengths have thelowest LE values.

1 Figure ‘I.-Effect of fines load-100% ing on the MOE of oriented and

un-oriented MDF mini-panels.

I _ _ - - . . . , _ _ _ _ _ . _ _ _ _ _ - . - _ . - - _ . . _ _ - _ _ _ _ _ _ _ _

‘zcor : htst : PU- o- 0r:Mal:Perp-Or : Juv : P.,- A- 0r:Juv:Perp-o-IJIIor : Ma: PM

s o n

Pwcmt Finer

Figure S.-Effect of fines load-ing on the MOR of oriented andun-oriented MDF mini-panels.


.

The IB data for the fines loading mini panels aresummarized in Figure 10. Fines are detrimental tothe IB strength of MDF mini panels. The deleteriouseffect of decreased IB strength as a function of finesloading is equally applicable to all panel types, re-gardless of orientation or juvenile fiber content.

MicrotomographyMicrotomography successfully imaged the struc-

ture of the MDF material. A perspective view of thethree-dimensional volume (1.28 by 1.09 by 0.37 mm)is given in Figure 11. The grid lines superimposedupon the volume are spaced 0.12 mm apart. Addi-tional detail is visible from observing a representa-tive side of the specimen (Fig. 12), which represents a

digitally obtained 2.4 p thick slice in the interior ofthe specimen, 391 u from one end. The identifiablefibers clearly show a lack of collapse in the lumens ofthick walled cells, which is in contrast to a fully col-lapsed fiber structure that has been observed in com-mercial kraft liner board material prepared from lob-lolly pine. Some fibers are oriented parallel with thefield ofview and exhibit a ribbon-like snake structure.This can also be observed from a top-view of the samespecimen (Fig. 13). The white regions represent voidstructure with the dark regions corresponding to thecell walls. The hazy/milky regions may representfines, which absorb the x-ray energy but whose struc-ture is undefined or size is so low it is not discernible.

Figure 9.-Effect of fines load-ing on the linear expansion ofMDF mini-panels. Six maintraces represent fines loadingfor 0 percent to 100 percent. Sixpoints to the right of the main

10

graph represent values of linearexpansion for mini-panels com-prised solely of intermediate fi-bers.

40% 60%Fines (percent)

D

0 or: Juv, Perp.

A Unor: Juwnile0 Q: Mat, Perp.

A Unor: Melure

. a: Jw, Par.0 Or: Mat Par.

Figure lO.-Effect of fines load-ing on the internal bond strengthof MDF mini-panels.

5 0Fines (percent)


Although difficult to communicate through1 wo-dimensional images, the three-dimensionaloiientation and length of a sample fiber was ob-lained from the volume. This was accomplished us-ing manual digital image interpretation of the vol-ume. The ends of a fiber within the volume wereidentified and their coordinates (x, y, and z) re-corded. The length of a straight line between theliber ends was designated as the secant length. Thecurved nature of fibers was evaluated through defin-ing the perimeter of fiber cross sections at 12 to 24 pintervals in a principal direction (X or v) of the MDFvolume. The length of each of the individual fiberwas then calculated from the distance between the(x, y, and z) coordinate of adjacent slice centroids.

Figure Il.-Isometric view of 1.28 by 1.09 by0.37 mm MDF specimen imaged with microto-mography.

The length of the individual segment was thensummed and defined as the total fiber segmentlength. This fiber segment length is a truer measureof fiber length than the Secant length. While manydescriptions of fiber curvatures have been made inthe literature (Nguyen and Jordan 1994; Trepanier1998), one particular measure was selected for de-scription of the MDF fiber due to its simplicity andease of calculation. The curl index (%) is defined as:

Fiber secant length x 1oOFiber segment length

A perfectly straight fiber would have an index of100 while a fiber bent in half would have an index of50. The curl index was measured on 7 1 fibers. A sum-mary of the fiber secant and segment lengths as wellas curl index is given in Table 1.

The distribution of fiber lengths was not normalalthough free fibers have been identified to havesuch a distribution (Wang and Shaler 1998). Thereason for this distribution is that in most cases theentire fiber length was not contained within the sam-ple volume measured. This phenomenon leads to afiber length distribution that is skewed to shortlengths.

While the fibers measured by either the segmentor secant method were underestiniated, the curl in-dex as defined as Equation [l] should still be valid.The mean and median curl index was 94.7 and 96.0percent, respectively, with a range of 70.6 to 100.0

Figure 12.-Edge view (0.92 by 0.81 mm) of in-ternal fiber structure 391 p from a specimen end.

Figure 13.-Plain view of internal fiber structure(0.2 by 0.81 mm).


i

.

Figure 14.-Fiber curl index vs. fiber secantlength of 71 fibers from interior of MDF specimen.

Table l.-Summary of length and orientation of71 fiber segments from within a 1.28 by 1.09 by0.37 mm volume of MDF.

M e a n M e d i a n Range

Fiber segmentlength (pm)

Fiber secantlength (pm)

Curl index (%)dyldx

dz/dx

dzldv

314.2 262.6 30.1 to 961.3

296.0 247.8 29.7 to 847.6

94.7 96.0 70.6 to 100.0-0.3 3.5 -87.8 to 89.8

7.7 7.5 -73.8 to 82.9

0.2 -2.9 -77.0 to 74.1

percent (Tab. 1). Curl index as a function of fiber se-cant length (Figs. 13 and 14) indicates that theamount of fiber curl was independent of fiberlength. It is important to note that the observed fibercurl is impacted by the fiber latency of the pulp priorto consolidation as well as additional fiber distor-tions introduced by the hot pressing operation.

The fiber orientation in three-space was mea-sured from the ends of 71 fibers (Tab. 1). Three an-gles were identified including:

l dy/dx = the angle of the projection of fiberlength in the x-y plane relative to the x-axis;

l dz/dx = the angle of the projection of fiberlength in the x-zplane relative to the y-axis; and

l dzldy = the angle of the projection of fiberlength in the y-z plane relative to the x-axis.

Figure 15.-Fiber orientation vs. curl index of 71fibers measured in the xy, xz, and yz specimenplanes.


The @/&angle corresponds to in-plane fiber orien-tation. The degree of alignment has been tradition-ally described in oriented strandboard (OSB) usingeither percent flake alignment or the von Mises dis-tribution (Shaler 1991). The mean angle was nearzero with a range of angles from + 90” (Table 1). Theexpected random nature of fiber orientation is ap-Parent from Figure 15. The vertical component ofliber orientation as defined by the dz/dx and dz/dyangles indicate that most fibers have an orientationwith + 25” of the x-y plane, but that an appreciablepercentage of fibers exhibit orientations up to 85”out of plane. This significant vertical orientation offibers is not typically described in MDF material.The presence of out-of-plane fiber alignment (Her-manson et al. 1997) impacts the mechanical proper-I ies of MDF and merits further investigation. The in-lluence of fiber handling and forming operations onthe three-dimensional fiber orientation within MDFoffers potential to control and improve panel prop-erties.

ConclusionsMiniature MDF panels are indicative of full-size

panel stiffness and strength. Specific inferences thatcan be extracted from this study are:I . MDF stiffness and strength increase with increas-

ing percentages of juvenile wood fibers.2. LE values are greater perpendicular to the direc-

tion of fiber orientation.3. Orientation increases MDF stiffness and

strength.a. MDF mini panels comprised of juvenile fi-

bers have a greater benefit from fiber orien-tation than do the mature counterparts.

b. Increase in stiffness and strength was in-creased both parallel and perpendicular tothe direction of fiber orientation, which ismost likely due to improved efficiency infiber packing.

Inclusion of fines in the fiber furnish decrease thestiffness, strength, and IB strength of MDF panels.Fines do appear to restrict LE between packing levelsof 0 to 50 percent. However, insufficient numbers ofsamples were manufactured to pinpoint the optimallevel of fines loading with regards to LE.

AcknowledgementsThe authors gratefully acknowledge financial

support of this study by the Composites Panel Asso-ciation. The authors are also indebted to The Bio-

Composites Centre for fiber refining and analysis,Dr. Dennis Keane for acquisition of the microto-mography data, John Alexander and Matt Brown fordetermination of MDF panel properties, Eric Phil-lips for image interpretation, and Mike Roessler forMDF panel production. Part of this research wasconducted at the National Synchrotron LightSource, Brookhaven National Laboratory, which issupported by the U. S. Department,of Energy, Divi-sion of Materials Sciences and Division of ChemicalSciences (DOE contract number DE-AC02-76CH00016).

L i t e r a t u r e C i t e dDeckman, H. W., J. H. Dunsmuir, K. L. D’Amico, S. R. Fer-

guson, and B. P. Flannery. 1991. Development of quan-titative x-ray microtomography. In: Proc. Matl. Res.Sot. Symp. Materials Research Society. Pittsburgh,Pennsylvania. 217:97-l 10.

.

Groom, L. H., S. M. Shaler, and L. Mott. 1996. The mechani-cal properties of individual lignocellulosic fibers. In:Proc. Woodfiber-Plastics Composites Conf. Madison,Wisconsin. pp. 33-40.

.

Hermanson, J., D. Stahl, S. Cramer, and S. M. Shaler. 1997.Transformation of elastic properties for lumber withcross grain. ASCE J. of Structural Engineering.123(10):1402-1408.

Larson, P 1962. A biological approach to wood quality.TAPPI 45:443-448.

Maloney, Thomas M. 1986. Juvenile wood - problems incomposition board products. In: Proc. of a CooperativeTechnical Workshop, Juvenile Wood: What does itmean to forest management and forest products?ForestProducts Research Society. Portland, Oregon. pp.72-74.

Moss, P. A., and E. Retulainen. 1995. The effect of fines on fi-bre bonding: cross-sectional dimensions of TMP fibresat potential bonding sites. In: Proc. of the Int’l. PaperPhysics Conf. Niagara-on-the-Lake, Ontario.TAPPI:97-101.

Mott, L. 1996. Micromechanical properties and fracturemechanisms of single wood pulp fibers. Ph. D. disserta-tion, Univ. of Maine, Orono, Maine.

Nguyen, N. , and B. Jordan. 1994. Curvature measurementsof crossing fibers by image analysis. J. Pulp Paper Sci.20(8):J226-J230.

Shaler, S. M. 1991. Comparing two measures of flake align-ment. Wood Sci. and Technology. 26( 1):53-61.

Shaler, S. M. 1997. Microtomography of wood compositemicrostructure. In: Proc. of the First European PanelProducts Symp. Llandudno. Wales, UK. pp. 28.

Shaler,S. M., D. T. Keane; H. Wang, L. Mott, E. Landis, and L.Holzman. 1998. Microtomography of cellulosic struc-


tures. In; 1998 TAPPI Proc. on Product and ProcessQuality. Milwaukee, Wisconsin. pp. 89-96.

Talbott, J. W. 1974. Electrically aligned particleboard andfiberboard. In: Proc. 8th Int’l. Particleboard/CompositeMaterials Symp. T. M. Maloney, ed. Washington StateUniv., Pullman, Washington. pp. 153- 182.

Trepanier, R. J. 1998. Automatic fiber length and shapemeasurement by image analysis. Tappi 81(6):152-154.

Wang, H., and S. M. Shaler. 1998. Computer simulated threedimensional microstructure of wood fiber compositematerials. Journal of Pulp and Paper Science. 24( 10):380-386.

Woodson, G. E. 1977. Density profile and fiber alignment infiberboard from three southern hardwoods. Forest

Products J. 27(8): 29-34.Zobel, B. J. 1961. Juvenility in Wood Production. Problems

in Forest Tree Breeding. Recent Advances in Botany.Univ. of Toronto Press. Toronto, Ontario, Canada.


5 Washington State University

33rd International Particleboard/Composite MaterialsSymposium Proceedings

April 13 - 15,1999

EditorsMichael I! Wolcott

Robert J. TichyDonald A. Bender

Associate EditorLinda C. Miklosko

Sponsored by

Washington State UniversityWood Materials and Engineering Laboratory

College of Engineering and Architecture

Pullman, Washington, USA1999

Relationship between fiber furnish properties and the structural performance of MDF

Documents