Department ofAgriculture

Service

Southeastern ForestExperiment Station

General TechnicalReport SE-53

Robert H. McAlister

March 1989

Southeastern Forest Experiment Stationp.0. Box 2680

Asheville, North Carolina 28802

The Research and Developmentof COM-PLY LumberRobert H. McAlisterResearch Forest Products TechnologistForestry Sciences LaboratoryAthens, Georgia

ABSTRACT

Between 1974 and 1986, a Southeastern Station Research WorkUnit developed standards for composite studs, joists, and trusslumber and manufactured and demonstrated the materials. Econom-ic feasibility was considered in every stage of research. Furtherdevelopment is left to industry.

KEYWORDS: Construction lumber, composite lumber, truss lumber,joists, studs, performance standards.

The characteristics of U.S. timber supplies havechanged markedly in recent decades, and furtherchanges are coming. Within 10 years, half of oursoftwood timber will be coming from trees that wereplanted (USDA Forest Service 1982). We have roughly470 million acres of forest land. Softwoods occupy200 million acres and hardwoods 270 million acres(USDA Forest Service 1974b) (fig. 1). Most of the

hardwoods are in the East, close to the majorconstruction markets, but these markets are almostexclusively softwood. Almost two-thirds of the annualtimber cut of 12 to 13 billion board feet (USDA ForestService 1982) goes into the construction, remodelingand repair of buildings (fig. 2). Most of the wood thatgoes into construction is softwood. That means thatthere is a lot of pressure on the softwood timberresource. A major problem is to find some means ofusing more hardwoods in construction.

A second problem is utilization. We do not useour timber efficiently. When we harvest a stand oftimber, we bring only about 70 percent of the woodvolume to the mill (Koenigshof 1977). We leave thesmall, crooked, or rotten trees and trees of undesirablespecies in the woods.

Figure 1. -Acreage of coniferous and hardwood forests in the continental United States.

1 7 %

FZA

NON-RESIDENTIALCONSTRUGTION

STRUCTURES ‘I II

RAIl!&ADS ~:I

M A N U F A C T U R E D

PRvTS ,”00

11% tS N I P P I N G

C O N T A I N E R S 1AND PALLETS ‘i

‘\ I,i.., ‘( NEW HOUSING AND

'I*.,''I*.,'

.^ :,’HOUSI~~~PKEEP

00/

‘, /i. . ..A

lR;!NG iII I’

MISGE~l;NEOUS “

Figure 2. -How wood is consumed in the United States; 1970 data.

Of the 70 percent brought to the sawmill, onlyabout 45 percent is recovered as lumber (Koenigshof1977) (fig. 3). The same factor applies to plywoodplants. Forty-five percent of 70 percent is a little over30 percent. We recover a lot of the waste in secondaryproducts and fuel, but, at best, we are using only 50to 60 percent of stand volume.

As early as 1972 many people could see thatthere was likely to be a serious softwood timbersupply problem in the United States by the year2000 (USDA Forest Service 1974b). The pressure on

the softwood resource would create shortages andrising prices. We needed to develop structuralproducts that utilized more than 30 percent of thestand volume and used both hardwood and softwoodtimber in structural lumber and panel products.

It appeared that some type of composite producthad the best chance of meeting the major goals.Most of the stress on both framing and structuralpanels is concentrated on the outermost fibers. Itmakes sense to put the strongest, stiffest wood onthe outside and put the lower quality material on theinside. One solution is to glue strong stiff veneerscut from the outer portions of the best trees to aparticleboard core made from ground-up particles ofwood and bark from the lower quality and lessdesirable trees. This concept was the basis forCOM-PLY (fig. 4) (Koenigshof 1977).

LESLES

Figure 3. -Conversion of logs into sawed lumber. Figure 4. -The composite structural lumber concept.

2

Originally, we calculated that 15 percent veneerwould be enough to give us the strength and stiffnessthat we needed for structural framing. What is more,it appeared that hardwoods would be suitable forboth the veneer facings and the particleboard core.

When a log is sawed into boards, much of thestrongest, stiffest, clearest, highest density woodfrom the outside of the tree is turned into chips,slabs, and edgings. Meanwhile, a high proportion ofthe wood from the center of the tree becomes lumber,even though the wood is knotty, less dense, andcontains juvenile wood, which may cause warp andtwist when the lumber is dried.

COM-PLY takes the strong, stiff, high-densitywood from the outer portion of the tree stem anduses it where it can add maximum strength andstiffness to the composite lumber or panel. The lowerquality wood from the center of the stem goes intothe particleboard core, and 90 percent of the logends up in the high-value primary product.

The concept is neat and elegant, but 15 yearsago no one was sure that it would work. In 1973, theDepartment of Housing and Urban Development(HUD) gave a research grant to the USDA ForestService to look into the COM-PLY concept. The workwas assigned to a small Research Work Unit (RWU)of the Southeastern Forest Experiment Station locatedin Athens, GA. The COM-PLY work was headedinitially by Richard F. Blomquist and, after Blomquist’sretirement, by Gerald A. Koenigshof. The membersof the research team and their areas of expertisewere: Dorothy H. Costa and Priscilla L. Floyd, Project

Secretaries; John E. Duff, wood/moisture relationsand fire; Louis I. Gaby, preservative treatments andfastener corrosion; Nolan Malcolm, Physical ScienceTechnician; Robert H. McAlister, mechanical proper-ties and forest resources; Charles B. Vick, adhesivesand gluing; Roy M. Walker, Physical Science Techni-cian; Dick C. Wiienberg, fasteners and mechanicaltesting; and Harold F. Zornig, engineering uses.

This report describes the research team’s ap-proaches and the findings. Information is provided inthree sections, The first briefly outlines the subject.The second describes research on COM-PLY studsthrough the pilot-plant production run and thedemonstration houses that used only the 2 by 4studs. The third describes the development ofCOM-PLY joists, COM-PLY truss lumber, and animproved flakeboard core.

Outline of Progress

We started by looking at residential framing andat the least critical structural member, the stud. A lotof the structural lumber in a house frame is in studsor other 2 by 4’s, and we thought we had a goodchance of producing a COM-PLY stud that wouldperform at least as well as the studs that were beingsold in the marketplace. The first question was ‘Whatshould a COM-PLY stud look like?’ We decided thatit should be solid, 1.5 by 3.5 inches, and fullycompatible with traditional construction practice forcutting and fastenings (fig. 5).

The next question was ‘How good does a studhave to be?’ That was a bigger problem because noperformance criteria existed for studs. Studs werestuds. The first suggestion, by HUD, was that theCOM-PLY stud had to be as good as a lodgepolepine stud. We looked at the grading rules and the

CONCENTRATED BENDING LOADS

REACTION Y---‘--l REACTION4 b

9 0 ’11 4,I- -1

Figure 6.-Schematic of loading method for combined bending test and end-loading.

published engineering design values for lodgepolepine (WWPA 1978) and figured that we could meetthe criteria with very little problem. In fact, we decidedto try to do a little better, since at that time very fewlodgepole pine studs were being sold. Our first job,then, was to develop the performance standards forcomposite studs.

The approach we used was to select a standardhouse, 30 feet wide and two stories high, and applyFederal Housing Authority (FHA) suggested designloads to the roof, floors, and side walls (Blomquistand others 1978). Studs were spaced 16 inches oncenter in the first-story walls and 24 inches on centerin the second-story walls. Loads for individual studswere then determined by standard engineeringprocedures. It turned out that the second-story studs24 inches on center were the most critical with anaxial load of 1 ,190 pounds and a bending load of 40

pounds per linear foot (due to wind). These loadswere used to develop strength performance criteria.Deflection criteria were developed for combinedloading of 1 ,190 pounds axial compression and 30pounds per linear foot bending (see fig. 6 for loadingdiagram). Maximum allowable deflection was l/240of the test span of 90 inches or 0.375 inch (Blomquistand others 1978).

In the performance standards for compositestuds, we developed a formula to account forvariability, repetitive member loading, a factor ofsafety, and the difference between the lo- to 15-minuteloading of test specimens and the long-term loadingof members in service. This expression becameknown as the HUD formula because key features ofthe variability adjustments were proposed by HowardC. Hilbrand, an engineer with FHA-HUD. The HUDformula is presented in figure 7.

HUD Allowable Design Load Formula

LOADt,,s - t

m.05

al low

where:x = average loadt.05.05

= t.,,for n-l d.f.S = standard deviationN = number of samples

F, = repetitive member (1.15)F, = factor of safety (1.5)

F, = time of test factor (1.18)

Figure 7.-The HUD formula for calculating allowable design loads for engineering properties of composite structural lumber.

Another criterion of performance is durability.The original thinking was that composite studs mightbe exposed to the weather for up to a year due toconstruction delays. Therefore the composite studsshould retain most of their strength and stiffnessafter 1 year of unprotected direct exposure to theenvironment. One-year tests were not practical forevaluation of alternative materials, but there were nowidely recognized accelerated tests to simulate 1year of outdoor weathering. However, the ASTMD1037 6-cycle accelerated aging test (ASTM 1979)was accepted as simulating several years of outdoorweathering for particleboards. One cycle of thisprocedure involves soaking in 200 “F water, freezingat 10 OF, drying at 210 OF, steaming at 210 OF, anddrying again at 210 OF. We decided that if thecomposite studs retained at least 50 percent of theirstrength and stiffness after the 6-cycle D1037 test,they would be sufficiently durable.

Dimensional stability was also part of the perform-ance standard for composite studs. The consensuswas that composite studs, after soaking and drying,should exhibit changes of dimension in width,thickness, length, and warp that did not exceed thecomparable values for sawed studs.

If we had done nothing more than developperformance standards for studs, the research wouldhave been worthwhile. But, we did much more. Allthrough the process we tested components anddifferent combinations of components to see howclose we could come to meeting our own performancestandards. We also analyzed softwood and hardwoodtimber resources to see if there was enough material,particularly high-quality veneer, to make manufactureof composite studs feasible, while utilizing all, ornearly all, of the timber in typical stands (McAlisterand Clark 1983; McAlister and Taras 1978).

Two problems were evident at the start. The firstproblem was a suitable particleboard core. No onewas producing a phenolic binder particleboard l-1/2inches thick, and the only phenolic particleboardsthat were being produced had a density of 55 to 65pounds per cubic foot (PCF). Boards of that densitysplit when nails are driven into the ends. We foundsome medium-density urea binder boards that wereabout right at 40 to 42 pounds PCF. It turned outthat a density of 38 to 42 PCF was a practical optimumfor all properties. Eventually we had some1-l/2-inch-thick phenolic board made to order in aparticleboard plant in southern Georgia. The secondproblem was with the percentage of veneer requiredin the composite stud. We had originally assumedthat two plies of l/8-inch veneer on each edge ofthe stud would be sufficient. In order to meet the

deflection criieria of 0.375 inch deflection for a loadof 40 pounds per linear foot over a span of 90 inches,we had to use at least two plies of l/&inch veneeron each edge. That meant that studs would be 20percent veneer and 80 percent particleboard. Theresource studies indicated that there was enoughsouthern pine grade C and better veneer in typicalstands to support the manufacture of compositestuds with enough left over to utilize the tops to a4-inch diameter.

By 1974, we had developed a composite studthat looked very promising. We were anxious toevaluate it on site in building trials. The first pilot-plantrun was made to produce some 3,000 studs (Vick1977a). This limited run and the demonstration housesbuilt with the studs were a success. The COM-PLYstuds were used and handled just like standardstuds. A demonstration house was constructed ateach of four locations: Columbia, MD; Aurora, IL (fig.8); Vancouver, WA; and Augusta, GA (Koenigshofand others 1977). Builders commented favorably.They noted, however, that the COM-PLY studs wereheavier (11 pounds vs. 9 pounds for spruce) andthat the 8-percent phenolic binder in the particleboarddulled ordinary steel saw blades very quickly.

Figure 8. - COM-PLY studs were used in demonstration housesusing standard construction practices.

5

Figure 9. - COM-PLY veneer-faced truss lumber and COM-PLY 2 by 6 studs and plates were used in the NAHB Research FoundationEER-II house in Damascus, MD, in 1982.

A lot more research remained to be done. Wewere convinced that we could develop satisfactoryCOM-PLY joists, and COM-PLY truss lumber seemeda good possibility. We hoped COM-PLY could beused for all the structural framing in residentialconstruction, but these uses had not yet been studied.In addition, building code officials, engineers, andarchitects were asking many questions that we couldnot answer. They asked about fire resistance,acoustical qualities, creep under long-term loading,and probable costs in relation to sawed lumber. Andwe still had not looked at the potential of the hardwoodresource.

These questions occupied us from 1975 through1986, when the Composite Research RWU wasterminated. We had some big milestones along theway. Our joist work was far enough along so that ademonstration house was built in Marietta, GA, in1978 that used COM-PLY joists, COM-PLY studs,and COM-PLY panels for subfloor and roof sheathing.Performance standards were developed and used toevaluate the composite joists in that home (Duff andothers 1978).

Our search for improved fire resistance inCOM-PLY floor joists led to the development of ahomogeneous, fully oriented strandboard that turnedout to be the key for the commercial development ofcomposite structural lumber. This development wasin cooperation with Tom M. Maloney and Roy Pellerinof Washington State University (Peters 1988). Thestiffness of the fully oriented strandboard was threetimes that of the randomly oriented flakeboards.Particle orientation increased the modulus of elasticity(MOE) of composite joists from 1 .l million poundsper square inch (PSI) to 1.6 million PSI, which iscomparable to the MOE of No. 2 southern pine.

Another accomplishment was the Energy EfficientResidence II (EER-II) (fig. 9) that was designed andbuilt by the National Association of Home BuildersResearch Foundation in Damascus, MD, in 1982. Wefabricated composite studs, composite floor joists,composite truss lumber, and special foundationgrade (FDN) 2 by 6 studs treated with chromatedcopper arsenate for the EER-II all-weather woodfoundation. Thus, we demonstrated that compositescould be used as framing anywhere in residentialconstruction.

6

We found that we did not need to limit the veneerused to the higher grades. Grade D veneer of bothpine and hardwoods was acceptable if certainguidelines were followed (McAlister 1983). Thisdiscovery made the full utilization of existing timberfor COM-PLY a definite possibility.

We determined that hardwoods could be used incomposite structural lumber almost interchangeablywith southern pine. In fact, yellow-poplar and sweet-gum veneers proved superior to southern pine veneersin some respects (McAlister 1979; McAlister 1982).The yield of veneer from the hardwoods is a littleless than from the pine, but use of hardwood veneermakes it possible to fully utilize the timber containingmixtures of pine and hardwoods (McAlister 1980).

We showed that composite lumber made with afully oriented strandboard core was suitable fortrusses. The loading properties for truss plates inoriented flakeboard are slightly lower than those fortruss plates in southern pine, but the loads are wellwithin the useful range of design values (McAlister1986).

The timber supply and utilization problems seenin 1972 stjll exist, and we are beginning to haveproblems with the fast-grown plantation trees (Megraw1985). The petroleum shortages have been delayed,but they seem inevitable. The cost of phenolic resins,which are petrochemicals, is an important determinantof the cost of composite structural products. Neverthe-less, composite structural materials hold muchpromise for the future. They could change forestmanagement in major ways.

Composite Studs: Concept to Reality

The research effort on composite framing materialsgrew out of a Housing RWU. The mission of this unitwas ‘More effective use of wood in housing.’ Workon innovative framing materials was underway in1968 when Zornig designed the FS-SE-5 house thatused a spaced Gothic-arch rafter made from 1 by2’s and 2 by 4 blocks (Anderson and Zornig 1972).Vick did some exploratory work on spaced columnsthat combined 1 by l’s or 1 by 2’s with hardboardwebs.

In early 1972 the Housing RWU was looking fornew areas of research. The unit had developed severaldesigns for low-cost housing which used traditionalwood-framing materials quite efficiently. The plans,which had been published by the U.S. GovernmentPrinting Office, were ‘best sellers.” We were told todevelop a new mission statement that would bemore concerned with the problems associated withthe use of southern woods in building construction.Gerald Koenigshof, from the Washington Office,USDA Forest Service, Division of Economics, cameto Athens to discuss a concept he had developedfor combining veneer with particleboard. Koenigshof’spresentation was illustrated with professionallyprepared drawings originally used for a TV programproduced and shown in the Washington, DC, area.These illustrated the problems of a shrinking resourcebase, smaller trees, inefficient conversion of trees tolumber and plywood by conventional processing,and the underutilization of hardwood species inconstruction.

Koenigshof had presented the composite conceptto Orville G. Lee and Howard C. Hildebrand ofFHA-HUD. They committed $400,000 to study theconcept if a satisfactory research unit was willing toundertake the work. Koenigshof had approached theForest Products Laboratory at Madison, WI, with theproposal, but its resources were already fully commit-ted to other high-priority research studies. Themembers of the Housing RWU team in Athens wereconvinced that the composite concept had merit andthat there was a good chance of solving most of theproblems within 5 years. The $400,000 from FHA-HUDwould provide a large share of the operating fundsneeded to study the concept. Also, the unit hadacquired from excess properly a120,000-pound-capacity universal test machine. Thislarge machine made it possible to do the full-scaletesting of structural members that would be required.

The final decision to begin work on the concept wasmade by Blomquist, the Project Leader, with theapproval of the Southeastern Forest ExperimentStation.

The scope of the RWU’s original mission wasquite limited. Our goal was to produce a verticalframing member-a stud- 92-5/8 inches long thatwould be capable of supporting both vertical buildingloads and bending loads due to wind in the samemanner as sawed studs made from species such asDouglas-fir, southern pine, spruce, or fir. There wereno restrictions on size or shape, but the new studwas supposed to be made from wood-based materials.A stud was chosen as the framing member becauseit was considered to be the least critical structuralmember in residential construction.

Research on structural composite panels, anintegral part of the total research effort, was theresponsibility of the American Plywood Association.It was given a research grant from FHA-HUD fundsreceived by the Forest Service. This research pro-ceeded concurrently with the work on studs.

Once the decision to undertake the researchhad been made, the problem was analyzed and aresearch direction developed in a series of meetingsbetween the Housing Research staff and Koenigshof.Personnel from FHA-HUD were also involved both inperson and in telephone conferences. It was decidedthat the new stud would be the same size (1.5 by3.5 inches) as existing sawed studs and that dimen-sional stability would have to be comparable to sawedstuds, at least in the 3.5-inch dimension.

The term ‘COM-PLY” was not coined until a yearlater. Initially, we referred to the stud as ‘the newstud’ or ‘the synthetic stud.’ These terms became aproblem because we did not want to develop a badimage for the product before it had a chance toprove itself. The term ‘COM-PLY” was developed in abrainstorming session with all of the research staffparticipating. The term ‘COM-PLY” was later copyright-ed by the American Plywood Association-with noobjection from the Forest Service-since we felt thatit could better regulate the use of the term.

Another decision made at this time was that themajor objective would be the development of theCOM-PLY stud and that results of the research wouldbe published in some son of manual rather than asindividual reports scattered through the literature.The major effect of this approach was that studies

were planned to answer specific questions related tothe eventual manufacture of the product. In manystudies, the number of samples and the statisticaldesign were inadequate for publication of the results.We did make rapid progress, but there was very littleoptimization and some questions were bypassed. .once a workable solution was found.

At the start of the research we made a list of allof the problems that we recognized and the approxi-mate order in which answers were needed. Thisexercise was valuable, but we were fortunate in findingas we went along that many of the anticipatedproblems either did not exist or had already beensolved. This phenomenon is sometimes referred toas ‘luck.’

From the start, we needed and got a team effort.We realized that all parts of the problem had to besolved at about the same time. The research assign-ments were very loose and were made principally onthe basis of previous work in a particular area. Workon the veneer resource, the particleboard core, thefastener and end use problems, and the gluing ofveneer to the particleboard core was conducted atthe same time. We got together regularly to discussthe progress of the work. Everyone knew whateveryone else was doing. This constant cross-fertilization of ideas and approaches to problemsmade the work go a lot faster. Plus, we could alwayscount on help.

This report presents some results of the unpub-lished studies to prevent their loss and to show thesequence of progress. An understanding of thesequence is important because a perfectly acceptablecomposite stud was developed, produced in apilot-plant operation, and extensively tested indemonstration houses with involvement of the majorbuilding codes about 2 years after the researchbegan.

At the outset, there were no real standards forthe strength and stiffness of sawed studs. Studswere being produced and sold from species as diverseas Douglas-fir with an average MOE of 1,700,OOO PSIand lodgepole pine with an average MOE of 950,000PSI. Grading rules for Stud grade 2 by 4’s werequite specific on such matters as size, moisturecontent, warp, crook, bow and wane; there was nota word about minimum strength or stiffness.

The ultimate marketability of the COM-PLY studwas a constant concern. We knew that particleboard

8

had a reputation for poor durability and quality. Byadopting the ASTM D-l 037 g-cycle, acceleratedaging test, we went out of our way to ensure thatthere was no question of the durability of the COM-PLYstud.

Initially, it appeared that there were five majorareas of research required.

1, The quantity and quality of veneer availablefrom trees in typical stands. As part of this problem,some decision needed to be made about the qualityand quantity of veneer required to produce a studwith the required strength and stiffness. We wantedto be sure that the veneer and the material forparticleboard were in reasonable balance so thatadditional residues would not be produced. Thisproblem was assigned to McAlister.

2. The physical characteristics of the particle-board core of the composite. In 1972, particleboardmade with phenolic binders was a rarity and no onewas producing a particleboard l-1/2 inches thick.The characteristics of the particleboard would deter-mine the quality and thickness of the veneer requiredand would affect the dimensional stability of thecomposite stud. Dimensional stability was deemedto be critical so that sawed studs and the new studscould be used interchangeably in the same wallsystem. This problem was assigned to Duff.

3. The fastening characteristics of the newstuds. We wanted the new studs to be completelycompatible with nails, screws, and other devices andwith traditional practices of construction. This problemwas assigned to Zornig.

4. The adhesive bond between the veneeredges and the particleboard core. This problemwas assigned to Vick.

5. The economics of the manufacture ofCOB&PLY studs. This area included resourceavailability, plant design and layout, cost of rawmaterials, labor costs, etc. This problem was assignedto Koenigshof.

Blomquist was responsible for the overall planningand direction of the research effort. Koenigshof, afterhis transfer from the Washington Office to Athens,was assigned as a team leader for structural consider-ations and testing procedures.

Resource StudiesSeveral simplifying assumptions were made at

the start of these studies. We assumed that grade Cand better veneer would be required for the edgesof the COM-PLY studs. The maximum open defectallowed for C grade veneer is l-1/2 inches, and anylarger defect seemed excessive for 1 -l/2-inch-widestructural lumber. We also assumed that three layersof veneer would be required to sufficiently randomizethe defects to produce a uniform strength and stiffnessof the COM-PLY studs. The first resource studieswere limited to southern pine because veneer yieldsby grade were available by l-inch diameter classesfor loblolly and slash pines (Clark and Schroeder1971; Phillips and others 1979; Schroeder and Clark1970). Data had been collected on veneer yields byblock and by tree. The objective of these studieshad been the development of yield equations relatedto tree diameter and height. It was necessary toanalyze the data from a different perspective to obtainthe information that we needed for the compositestud work. Clark, Phillips, Taras, and Saucier fromthe Utilization of Southern Timber RWU in Athens,GA, were quite generous in sharing the raw data,even to the extent of giving us a complete set ofdata cards.

These data were combined with data from theSoutheastern Station’s Forest Inventory and AnalysisRWU on the number of trees in each diameter classby stand type. The results showed that almost all ofthe C grade veneer in southern pine comes from thefirst three blocks in the stem. This analysis also showedthat there was enough C and better grade veneer inthe first three blocks of trees in average stands ofsouthern pine to completely utilize the entire stem(to a 2-inch-diameter top) in the production of theCOM-PLY studs. In fact, a small surplus was availableto provide veneer edges for particleboard made fromsome of the trees classified as ‘culls’ in typical stands.

The next questions were concerned with theMOE of veneer cut from southern pine trees of variousdiameters. MOE is a measure of stiffness of materialsand permits calculation of deflection of structuralmembers under load. We needed to know the averageMOE of the veneer and also whether the MOE wasnormally distributed. Measuring the MOE in bendingof veneer proved to be difficult. We adapted thetechniques developed by Koch and Woodson (1968)to estimate the dynamic MOE of veneer by measuringthe transit time of a stresswave through a veneerstrip. Koch and Woodson gave us a great deal ofveneer that had been stresswave tested for thelamination study and had not been used. This

9

pretested veneer speeded up our work considerablyin preparing studs for testing. A study was designedin cooperation with the Utilization of Southern TimberRWU in Athens to collect loblolly pine veneer from apreviously planned study of veneer yield by grade.Veneer was collected from each block just afterroundup, in the middle of the peel, and just beforethe core was dropped. Over 600 veneer strips 1.5 by100 by 0.167 inches thick were tested for dynamicMOE at the Forest Products Laboratory in Madison,WI (McAlister 1976). We found that the MOE of theloblolly pine veneer averaged about 1.8 million PSI,that the stiffest veneer came from the second blockin the tree, and that the distribution appeared to benormal. Normal distribution simplified subsequentprojections and analysis of the resource.

We did a short study to determine whether weneeded three plies of veneer or if two plies on eachedge were sufficient. We found that differences instrength and stiffness between studs made with twoplies or three plies of veneer were not of practicalsignificance. These results meant that we could usethicker veneer and fewer gluelines to produce thestuds.

The effects of using lower grade veneers and ofdifferent methods for laminating plies on the strengthand stiffness of the studs were studied intensively.Results indicated that at least half of the D gradeveneer from loblolly pine could be used if care wastaken that one of the plies was of C grade veneer.This finding meant that suitable veneer could betaken from more of the low-grade pine and hardwoodtrees in typical stands and that utilization of timberfor COM-PLY studs could approach 100 percent ofstand volume. It also became apparent that shortgrain in the veneer was the limiting factor in thestrength of the composite. Short grain was as likelyto occur in clear veneer as in veneer of the lowergrades.

The effects of variations in the MOE of theparticleboard core on the strength and stiffness ofCOM-PLY studs were also studied. The particleboardthat we were using as a core material at that timehad an MOE of between 250,000 and 450,000 PSI.Results indicated that variations in the MOE of theparticleboard did not significantly affect the MOE ofthe composite. The only significant factor was variationin the MOE of the veneer component.

Particleboard ResearchThere were several immediate problems with the

particleboard core. From the start, the COM-PLYstuds were intended to be fully weatherproof. One ofthe initial requirements was that the studs retain atleast 50 percent of their strength and stiffness.after 1year of outdoor exposure. We needed a particleboardwith a phenolic binder. The only phenolic particleboardavailable was a special II-B-2 mobile-home decking5/8 inch thick with a density of about 60 PCF. Tomatch the thickness of sawed studs, we needed aparticleboard that was 1.5 inches thick. Initial stepsinvolved gluing together three thicknesses of l/2-inchparticleboard or using two plies of 5/8 inch and oneply of 3/8 inch in the center. The thickness swellingof the laminated particleboard was close to 30 percent.We found one company in the East that was makingan extruded board 1.5 inches thick. This boarddisintegrated when soaked in water. We were toldthat a phenolic particleboard 1.5 inches thick wouldbe impossible to produce economically because thecure time would be at least 25 minutes.

One company on the west coast was makingparticleboard 1.5 inches thick for door core stockusing a urea binder. This material was produced intwo densities, 28 PCF and 42 PCF. Tests of thismaterial were encouraging. Thickness swell was onlyabout 20 percent-way too much but closer toacceptable than previously tested materials. Wemade up several studs with this core material andeven exposed some to outdoor weathering on a testfence. The COM-PLY studs made with the urea binderparticleboard cores lost only about 20 percent oftheir initial strength and stiffness after 1 year of outdoorexposure.

We continued to negotiate with particleboardplants to produce some 1.5-inch-thick phenolic binderboard. No plant was willing to guarantee the level ofinternal bond, MOE, modulus of rupture, or thicknessswell. Blomquist finally contracted with a majorparticleboard producer for a one-shift productionrun. The plant operators would try their best to meetour specifications with the phenolic binder, but wewould have to take what we got and the price was$10,000. It worked out very well in the end, but wewere pretty worried. The problem was that ureabinders have a lot of ‘tack” (stickiness) and phenolicbinders have very little tack. After the mat was formedand before it went to the prepress, it had to betransferred from one conveyor belt to another. Sincethere was so little tack and the mat for a 1.5-inch-thickboard was so thick and heavy, the mat cracked and

10

broke during transfer. For a couple of hours all themats were run back into the surge bins and therewere no boards going into the hot press. Eventuallythe percentage of phenolic resin binder was raisedto 8 or 9 percent rather than the 6-percent targetvalue. The mats held together and particleboardstarted coming out of the presses. The press timewas rather long -20 minutes- but it was expected.We ended up with two flatbed truckloads of particle-board. Thickness swell of this board was about 10percent; the MOE was about 450,000 PSI with aninternal bond of 80 to 100 PSI. Density of the boardwas about 42 PCF, which turned out to be ideal. Ourgamble had paid off handsomely; this core materialwas used in the pilot-plant production run. We knewfor the first time that we were likely to be successfulin producing a COM-PLY stud that would meet all ofthe standards that we had set. This milestone wasvery important.

Compatibility With Traditional ConstructionThe residential construction industry is highly

traditional. New products are accepted reluctantly,especially if they require new and special techniques,tools, or fastening procedures. We wanted theCOM-PLY stud to be as close to sawed studs aspossible in appearance, performance, and use. TheCOM-PLY stud should require no special fastenings.In practice, this meant that it had to be nailable withconventional nails.

We had predicted that a major fastening problemwould be splitting when nailing through the veneeredges and into the particleboard core with 6d and8d common nails. It turned out that the most seriousproblem was splitting of the ends when the studswere end-nailed to the plates with 16d common or16d sinker nails. The 60 PCF particleboard we hadlaminated together to make 1.5inch-thick corematerial split so severely that there was no nailwithdrawal resistance.

A great deal of ingenuity and effort went intodeveloping special fastener systems to attach thestuds to the plates. One method used a modifiedjoist hanger arrangement. Another used a crimped,folded, toothed plate (like a small truss plate) that fitinto a grooved plate. Later we noted that the28-PCF-density and 42-PCF-density urea door coreboards did not split when end-nailed with 16d commonnails. The 28-PCF-density board had very lowwithdrawal resistance, but withdrawal resistance forthe 42-PCF-density board was quite acceptable.Several other limited tests confirmed that particle-

boards in the 38 to 42-PCF-density range did notsplit when end-nailed and gave acceptable withdrawalvalues. Particleboards in this density range couldalso be toe-nailed into the plates with 8d commonnails.

The 38 to 42-PCF-density range for the particle-board core material is a critical specification so thatconventional fasteners can be used.

Gluing Veneer to the Particleboard CoreWe anticipated trouble in laminating the veneer

to the particleboard core because the two materialshave completely different dimensional stability charac-teristics. The joint must resist the differential swellingof the veneer and particleboard and still be able totransfer horizontal shear forces from the veneer edgesto the core. The cut edge of the particleboard isquite porous and not a good gluing surface. The flatgrain southern pine veneer has wide bands oflatewood, which is also a difficult gluing surface. Itlooked like an insoluble problem.

However, the first thing that we tried worked verywell. We used a 100 Ib/Mft* spread of a conventionalphenol-resorcinol adhesive that cures at roomtemperature. Best results were obtained with a5minute open assembly time and a 15 to 20-minuteclosed assembly time. Applied pressure of 125 PSIand a curing temperature of 70 “F were also important.The curing time could be reduced if the bondlineswere cured in a hot press at 300 “F.

The major disadvantage of the system was cost.Phenol-resorcinol adhesives are relatively expensive,compared with phenolic plywood adhesives andphenolic binders for particleboard. We used thisadhesive system for the veneer to core bond through-out all subsequent research on the COM-PLY studs,including the pilot-plant run that produced over 3,200studs (Vick 1977a).

We did experiment with an emulsified poly-isocyanate adhesive designed for a very rapid curingat room temperature. This adhesive was quite effectivebut more expensive and more difficult to use thanthe phenol-resorcinol adhesive. We extensively testedmelamine resins, melamine-urea resin blends, andphenolic resin molding adhesives that could be curedrapidly with radio-frequency energy. Nothing workedany better than the phenol-resorcinol adhesive thatwe started with. This adhesive system passed boththe ASTM D-1037 g-cycle accelerated aging test andthe 1 year of outdoor exposure with practically nodelamination or loss of strength.

11

Economics of Manufacture of COMA-PLY StudsResearch on the economics of manufacture

included the development of detailed cost analysesof the production processes for several alternativemethods of manufacture and plant layouts. Pricetrends for both raw materials and studs were studied.Detailed costs could not be estimated until themanufacturing process was known. The economicstudies, therefore, lagged behind COM-PLY studdevelopment. It appeared, however, that COM-PLYstuds could be produced and sold at a profit incompetition with conventional sawed studs. Sincestuds are one of the lowest priced commodities, itappeared desirable to be able to produce othertypes of structural lumber in the same plant to raisethe average price of products.

Summary of COM-PLY Stud ResearchThe original mission of the research was com-

pleted in June 1975 with the construction of the thirddemonstration house using COM-PLY studs andCOM-PLY panels. This demonstration house wasbuilt in Vancouver, WA, with the cooperation of theAmerican Plywood Association. It was completed intime for the 1975 meeting of the Forest ProductsResearch Society (FPRS) in Portland, OR. Our team’sreport on the development of the COM-PLY stud anda report of research on composite panels by theAmerican Plywood Association were presented atthe annual meeting (Blomquist and others 1975;Lyons and others 1975).

This achievement was remarkable. In less than 3years we had taken a complex idea from concept toreality. At this point, there were no publications onthe research. A great deal of what had been learnedin the 3 years of concentrated effort was containedin the summary report delivered by Blomquist at theFPRS annual meeting. The format for the definitivereport on COM-PLY was not determined until thefollowing year when it was decided to publish aseries of individual reports with a common format tobe known as the COM-PLY series.

The decision was made at this time to extendthe mission of the research to include joists andother structural members. The results of this researchare detailed in the following section of the report.

Joists

The decision to develop a COM-PLY joist was alogical outgrowth of the success of the stud program.In all development and analysis, we had consideredthe stud as an end-loaded and side-loaded beamrather than as a column with bending loads. Well-accepted standards for structural performance ofjoists existed, and we believed that we could developCOM-PLY members that met those standards. Aproblem analysis showed several areas that requiredresearch.

1. The MOE of a joist is critical because deflectionunder design load is limited to l/240 of the spanwith an absolute limit of one-half inch, no matterwhat the span. A target MOE of 1.4 million PSI wasset for COM-PLY (southern pine joists have an MOEof 1.6 million PSI) so that allowable span of COM-PLYjoists would be comparable to that of sawed joists.There was a problem connected with joist length.Since studs are less than 8 feet long, they can befabricated with full-length veneers on the edges.Joists are up to 32 feet long. This means that somemethod of joining sheets of veneer (scarfing orstaggered butt joints) had to be developed and theeffect on joist stiffness and strength determined. Thisarea of research was assigned to McAlister.

2. Utilization of hardwoods for both veneer andparticleboard components of the COM-PLY structuralmaterials. Even in pine stands, there is a considerablevolume of hardwood. The use of hardwoods was anintegral part of the COM-PLY concept. There wereno data on veneer yields by grade for hardwoods.Since most hardwoods other than oaks have lowerMOE than southern pine, there was concern overpossible performance problems of hardwood veneersin composites. This research area was assigned toMcAlister.

3. Creep is the continued increase in the deflectionof beams with no increase in load. This problem waspotentially serious because particleboard was knownto creep. Joists would be subject to long-term bendingdead loads. This investigation was assigned toWiienberg under the supervision of Koenigshof.

4. The fire resistance of joists. COM-PLY studsdid not perform as well as sawed studs in fire tests.That problem is more serious in joists than in studsbecause the load-bearing requirements for floors aremore critical than for walls. Fire performance for

12

walls is limited by burn-through and load carryingand has a 20-minute base for performance. Floorsare required to carry full design load (live and dead)for 10 minutes. This research area was assigned toDuff. After Duff left the project in 1981, the researchwas carried on by Wittenberg under the supervisionof Koenigshof.

5. Joists have different fastener requirementsthan studs. Many builders use joist hangers, othersuse ledger strips. Bearing, or compression resistanceperpendicular to the joist, was seen to be a possibleproblem. This research was assigned to Wittenberg.

6. The termite and decay resistance of untreatedCOM-PLY joists was a matter of great concern.Preservative treatment of joists is desirable or requiredin some situations. Treatment of a structural memberthat consisted of over 50-percent particleboard wasseen as a potential problem. This research wasassigned to Gaby.

7. Manufacturing COM-PLY joists economicallywas a potential problem because joists require moreveneer, probably thicker veneers and more gluelinesthan COM-PLY studs. A manufacturing process hadto be developed to bring particleboard, veneer cuttingand drying, and laminating technologies together toproduce a new product. This problem area wasassigned to Koenigshof.

8. The use of COM-PLY structural lumber forlight-frame wood trusses. This use was conceivedbecause of the large amount of structural lumberused annually in trusses. The general properties ofCOM-PLY structural lumber (dry and straight withuniform strength and stiffness) made it desirable foruse in a highly engineered product like a truss. Initially,the major problem was seen as the lateral-loadresistance of truss plates in the composite structurallumber. This problem area was assigned to McAlister.

As with stud development, the major objectivewas considered to be the development of theCOM-PLY joists and trusses rather than the productionof publications. The technology transfer effort was tobe concentrated on the commercialization of theCOM-PLY joist; that is, to persuade a manufacturerto build a plant to produce COM-PLY structural lumber.Some publications about the veneer resource studieswere in the works, but no decision had been reachedon a format for publishing the compendium of workon the COM-PLY project. Much of the research wasinterrelated. For example, decisions about the veneer

thickness required to increase joist MOE would affectfastener performance, creep, fire resistance, treatabil-ity, and economics. Therefore all research effortshad to be closely coordinated. Individual researchershad to exchange information frequently.

Joist StiffnessThe first development work was concentrated on

2 by 10 (1.5 by 9.25 inch) joists. A direct scale-upwith the same proportion of veneer as on the COM-PLYstud failed. The MOE of this joist was about 0.9 millionPSI, while the target MOE was 1.4 million PSI. Sincethe MOE of the particleboard core averaged about0.45 million PSI, the only solution was to increasethe amount of veneer on the edges of the joist.Calculations indicated that we needed between 1.25and 1.50 inches of veneer on each edge of a 2 by10 to meet the stiffness requirement. Veneer wouldmake up 27 to 32 percent of the total volume. Aquick check of the resource data on veneer gradeyields indicated that some D grade veneer wouldhave to be utilized. Another problem was with thenumber of gluelines. With l/8- to l/Sinch veneer, 8to 10 gluelines would be required for each edgelamination. These gluelines are quite expensive. Onesolution was to use l/4-inch-thick veneer. The use ofl/4-inch veneer on all composite lumber would resultin a nearly constant ratio of veneer to particleboardof 27 percent for everything from 2 by 4’s to 2 by12’s. However, the number of gluelines would beminimized, which would tend to minimize costs.

We could not find a veneer mill willing to peeland dry l/4-inch southern pine veneer. The consensuswas that peeling l/4-inch veneer would tear upequipment. It was a bit frustrating since for yearsveneer mills on the west coast had peeled 5/16-inchDouglas-fir veneer for core stock without tearing upany equipment. Finally we found a small veneer millin Murphy, NC, that agreed to peel and dry somesouthern pine l/s-inch veneer for us. We had to takemill-run veneer. They would clip the material to 12inches wide, and we would pay for the square feetof veneer actually cut. This veneer was the basis formost of the development work on the COM-PLYjoists,

We soon realized that the target value for MOEof 1.4 million PSI was a bit unrealistic with theparticleboard core that we were using. An MOE of1 .l to 1.2 million PSI was practical and within thesame stiffness range as spruce pine-fir joists thatresidential builders were using in many areas.Mechanical testing of many joists showed that if

13

joints in adjacent plies of veneer were staggered oroffset at least 12 inches, there was no effect on theMOE of the joists and the strength (MOR) was reducedvery little. We found short grain was the most seriousveneer defect, reducing both strength and stiffness.A great deal of otherwise clear veneer has extremelyshort grain. Much of the time, the severity of theshort grain is not evident from the surface appearanceof the veneer. The dynamic MOE of the veneerdetermined by stresswave techniques was an excel-lent predictor of joist MOE. However, even thestresswave technique would not detect areas ofshort grain in a long strip of veneer. We looked atmany alternative systems of grading veneer for usein COM-PLY products, but we never found any systemthat did a better job than the standard AmericanPlywood Association panel grading rules.

Many tests of COM-PLY joists indicated that, withthe number of plies of l/4-inch-thick veneer beingused in the composite beams, the grade of the veneerdid not have to be very good. We found that wecould use all of the D grade veneer if the outside plywas of C or better grade. Since all of the veneer thatcould be peeled would be usable, some source ofadditional particleboard would be needed to use allof the veneer. This development was of majorimportance to the economics of manufacture forCOM-PLY.

Utilization of HardwoodsMajor questions about the utilization of hardwoods

concerned the characteristics of the hardwoodveneers-their yield by grade and their strength andstiffness. We assumed that the particleboard furnishfor COM-PLY could be made of a mixture of softwoodand hardwood. (This was a trifle optimistic.) As nearlyas we could tell from the literature, very little researchhad been done on structural characteristics ofhardwood veneers. We initiated a hardwood veneerstudy in cooperation with the Utilization of SouthernTimber RWU in Athens as part of its hardwood biomassresearch. Veneer yields by grade and stiffnesscharacteristics were determined for sweetgum,yellow-poplar, and white oak from the GeorgiaPiedmont; for yellow-poplar, white oak, and soft maplefrom the North Carolina mountains; and for yellow-poplar, blackgum, and sweetgum from the SouthCarolina Coastal Plain.

Each study followed the same procedure. Siteswere selected in each area as being representativeof the timber. Three trees in each l-inch diameterclass from 10 to 22 inches d.b.h. were selected fromeach site. Only dominants and codominants with novisible decay were considered. The trees were. felled,measured, and bucked into l- or 2-block sections (ablock was considered to be 8.4 feet long). A l-inchdisk was taken at the top of each block. The treeswere weighed in the woods and then trucked to acooperating softwood plywood mill (the same millwas used for all three studies). One-sixth-inch veneerwas peeled from the blocks after conditioning. Acolor-code spray system was used so that the veneerfrom each block could be identified when tallied forsize and grade. Each block was measured at eachend and at the center immediately before beingchucked in the lathe. Each core was measured ateach end and at the center after it was dropped. Asimilar procedure had been used to determine theyield by grade for southern pine veneer.

Veneer was taken from each block to be usedfor strength and stiffness tests. The tests and theresults are fully described in published reports(McAlister 1980; McAlister 1982; McAlister and Clark1983).

Generally, the results of the studies were quitefavorable. The yield of veneer from the hardwoodswas somewhat less than from southern pine, but thequality of veneer was comparable. The strength andstiffness of the hardwood veneer were excellent.Yellow-poplar in particular was found to have anMOE equal to that of southern pine (McAlister 1982).Yellow-poplar and sweetgum were found to becompletely suitable for use in structural composites.This finding opened up the possibility of completestand utilization for production of COM-PLY by usingpine and mixed hardwood stands.

Other studies on the use of mixed hardwoodfurnish in the particleboard core of compositestructural members were done in cooperation withthe University of Georgia and Washington StateUniversity. The results of these studies were notpublished, but they were generally favorable. Particle-board made from mixed hardwood furnish hadadequate strength and stiffness for use in composites.

14

However, the thickness swell of the hardwood furnishparticleboards was somewhat greater than forparticleboards made from southern pine furnish. Thisproblem was left at this stage since we did not havefacilities for this type of particleboard research. Wefelt that our resources were better used on otherresearch.

Creep of Structural CompositesCreep-increasing deflection with no increase in

structural load-was seen as a possible problemwith COM-PLY joists, The problem is not unique tocomposites. Solid sawed joists are known to creepbecause of the visco-elastic nature of wood, andparticleboards used as flat panels have serious creepproblems. Creep in particleboard may be caused bythe density gradient that is a consequence of thereactions to temperature, moisture, and pressuregradients during the press cycle. Whatever themechanism, the perception was that particleboards(and products made from them) would creep.

This problem was discussed at great length.Two or three possible approaches to measuringcreep in COM-PLY joists resulted in study plans. In apreliminary study, joists under load were cycledthroughseveral cycles of high temperature and relativehumidity. The conclusion was that COM-PLY joistswere no more creepy than southern pine joists. Datafrom this preliminary study were not sufficient towarrant publication. No detailed study of creep incomposite joists was installed because (1) a finalspecification for the particleboard core material hadnot been developed, (2) sufficient temperature-humidity-controlled space for a large number ofspecimens was not available, and (3) both moneyand manpower were in short supply and had to beused for studies of higher priority.

One limited study was installed as the result of adeflection problem that surfaced at the EER-II testhouse in Damascus, MD, during 1982. Duringconstruction, it was noted that the outboard end of ajoist with a 2-foot cantilever had deflected almostthree-fourths of an inch after a heavy rain. The engineerin charge of the design of the EER-II house broughtup the question of creep. A study was initiated inhaste. Two floor sections were fabricated using 2 by10 joists spanning 14 feet and loaded to 125 percentof design load. Deflection measurements were takenalmost continually for the first couple of hours afterthe load was applied, then daily for several weeks,then weekly for a year, and finally monthly until thestudy was terminated some 5 years later. Both floor

sections survived the 125 percent of design load forover 2 years. Finally, the joists in one of the floorsections buckled and failed because lateral bracingwas lacking. Lateral bracing was added to the otherfloor section without removing the load and the testcontinued. When the load was removed, the floorsection was allowed to recover for 6 weeks. The finaldeflection, after full recovery, was about equal to theinitial deflection under load. This movement wastermed ‘irrecoverable creep.’ The joists were thenremoved from the floor section and tested understatic bending to failure. There was no change instiffness and, as nearly as could be determined, nochange in strength of the COM-PLY joists due to the5 years of loading at 125 percent of design load.

Due to the limited nature of the study, the resultswere not submitted for publication.

Fire Resistance of JoistsFire resistance of building components, especially

naturally combustible wood-building components, isvery important. Insurance rates and the danger tohuman life are major considerations. Tests of COM-PLY studs for fire resistance at the Forest ProductsLaboratory in Madison, WI, uncovered a problem.Generally, the composite materials were not as fireresistant as solid sawed materials. Most of the strengthof the composites came from the veneer on theedges of the members. When the veneer burnedthrough, the member failed. The best that could besaid for the composites was that they were almostas good as the sawed members.

This problem was especially acute for COM-PLYjoists. The fire resistance of joists is measured bybuilding a complete floor system over a large furnace.The floor is loaded to design load with vats filledwith water. When the fire is started in the furnace,the temperature rise must follow a standard curve.Failure is measured by collapse or excessive deflectionof the floor system. In initial ASTM and Underwritertests, COM-PLY joists failed in 7 minutes. The unofficialbenchmark failure time for joists was 10 minutes asestablished by FHA-HUD.

Obviously, these tests of complete floor systemsare quite expensive. Very few laboratories in theUnited States are capable of performing them, and agreat deal of material must be shipped. We needed

15

Figure 10. -Comparative fire performance of sawn lumber and many types of composite structural materials was evaluatedin the ‘coffin fire test.’

some method of evaluating the fire resistance ofstructural members that would be faster and lesscostly than the full-scale tests. This was the basis forthe development of the ‘coffin fire test.’

The ‘coffin’ was a small furnace made of firebrick. The heat source was a large two-arm pipeburner using propane gas. The test consisted ofmaking up a panel 32 inches wide and 96 incheslong from 2 by 4’s covered with l/Binch gypsumboard. This panel acted as the cover to the coffinfurnace. Design load was applied at third points with

lead weights set in place with a small derrick madefrom a length of pipe and a boat-trailer winch. Theinitial deflection was measured with a transit using areference point on the panel (fig. 10). The burnerwas ignited and the temperature rise was controlledby regulating the gas-flow according to output fromquick-acting thermocouples in the furnace. Failurewas determined from transit-based deflection read-ings. The burner was extinguished, the panel wasremoved, and the flames were doused with a gardenhose. The procedure was not a standard test, but itdid provide comparative values with known materialsand their performance in the standard test. Since thetest was quick and inexpensive, we were able to

16

determine the probable effects of many variables inthe construction of composite structural materials.

Some of the knowledge gained from the ‘coffin”test helped during design of composite joists thatdid meet the standard fire test value of 10 minutes.We found that (1) the stiffer the veneer used in thecomposite, the longer the time to failure; (2) thehigher the density of the core material, the longerthe time to failure; (3) the higher the MOE of thecore material, the longer the time to failure. It wasthis last conclusion, coupled with the developmentof the fully aligned particleboard core, that led to thedevelopment of a composite joist that exceeded thelo-minute benchmark time in a standard test.

None of the data developed from the fire tests(the ‘coffin’ test or the standard ASTM fire tests)were published.

Fastener InteractionsJoists have some unique fastener problems

compared with studs. End-nailing must support bothshear and moment loads (for example, around stairwellopenings) for the life of the structure. If joist hangersare used, several small nails near the end of the joistmust support large shear loads in a very small areaof the core. Some joists act as header beams whereledger strips nailed into the particleboard core supportheavy lateral loads.

These problems were studied by Wittenberg andWalker for several years. Results were reported inCOM-PLY Report 20 (Wittenberg 1981).

Decay and Termite ResistanceGenerally, framing materials are used in protected

environments where resistance to decay and termitesis not a major factor. However, some natural resistanceto decay and termites is desirable for any structuralmaterial, and preservative treatment of framing lumberis either desirable or required in some applications.Several studies were initiated in cooperation with theWood Products Insect RWU, Gulfport, MS, to checktermite resistance of untreated composite floor joists.Some natural termite resistance of the compositeswas found (Gaby and Carter 1981).

The treatability of composite structural lumberwas tested. Results indicated that composite structurallumber was easy to treat to any desired retentionwith waterborne preservatives. Full penetration of themembers was achieved in less than half the timerequired for southern pine lumber. Some of the

thickness swell that occurred during treatment ofcomposite lumber was irrecoverable on redrying, butthe swelling was not sufficient to be considered adefect. Decay and termite resistance of the treatedcomposites was outstanding. Composite structurallumber was treated with chromated copper._arsenateto a retention level of O.&pound PCF and used in anAll-Weather Wood Foundation System for the EER-IIresearch house in Damascus, MD. There were nopublications on this phase of the work.

Economics of ManufactureThe costs of manufacturing composite structural

materials were a constant concern. The overallresearch objective was to develop a product thatcould compete on a performance and price basisagainst No. 2 KD southern pine dimension lumber.This objective guided our research from the beginning,and there were some positive benefits to this ap-proach. Since hardwood stumpage prices aregenerally lower than softwood stumpage prices,economic considerations made research on hardwoodcomponents more attractive. Also, costs were consid-ered when we evaluated resin binders for patticle-board; times and temperatures for curing particle-board; the peeling, drying, and laminating of veneer;the amount and type of waste produced at eachprocessing step; the amount of electrical and heatenergy required for each processing step; and theamount of labor required for each step.

These factors, together with data on the basicforest resource and yield factors, were incorporatedinto a computer program that yielded estimates ofrequired inputs, amount and type of waste generated,fixed costs, variable costs, assumed cash-flows, andprofitability. The program was developed and refinedover a period of years by Koenigshof. Results of thedetailed analysis of joist manufacture under specificassumptions were published as COM-PLY Reports15, 17, and 26 (Koenigshof 1978, 1979, 1983).Seminars on the feasibility of manufacturing COM-PLYpanels in converted plywood plants were presentedin cooperation with the American Plywood Association.Koenigshof also prepared detailed economic analysesfor several major forest products manufacturers whowere considering the feasibility of producing compos-ite structural products.

Generally, the economic projections for compositestructural product manufacture were favorable. Thebasic assumption was that composite structuralproducts could be sold at a price at least equal tothe average price of No. 2 southern pine dimensionlumber.

17

An Intermediate ProductA workable composite joist was developed and

tested by early 1978. Enough joists and studs weremade in the laboratory to construct a demonstrationhouse in Atlanta, GA, in conjunction with the 1978Forest Products Research Society annual meeting(fig. 11). The demonstration house incorporatedcomposite floor joists that were 32 feet long andcontinuous over a center support. There were alsocomposite studs and composite panels that hadbeen commercially manufactured in a North Carolinaplant.

Even though a workable composite joist hadbeen developed, there was still a great deal of workto be done to check alternative methods of manufac-ture and the effects of many process variables onthe physical and mechanical properties of the material.The demonstration house was featured on anFPRS-sponsored tour at the annual meeting.

Composite Truss Lumber

While research continued on the composite floorjoists, the decision was made in 1979 to explore thedevelopment of composite truss lumber. There weretwo major factors considered in this decision: (1) thetruss fabrication industry consumes about 2 billionboard feet of dimension lumber every year, and (2)problems related to warped, twisted, and below-gradelumber could be reduced with composite structurallumber.

There were some immediate problems with theconcept. To begin with, the standard method offabricating light-frame trusses is with toothed metalplates made from 20-gauge or 16-gauge steel plate.The teeth are pressed into the wide face of thedimension lumber in the truss. With the standardconfiguration of composite structural lumber (i.e., theveneer on the narrow edges) the teeth are pressedinto the flakeboard. Preliminary tests showed thatthe lateral load resistance of plates installed into theparticleboard core did not yield useful design loadvalues. Another potential problem was with the columnbuckling characteristics of the composite structurallumber. Studs are loaded as columns, but they arecovered with sheathing or some type of panel materialthat braces the column. Truss components are

Figure 11. -Composite structural joists, 2 by 10 by 32 feet long, were used in a demonstration house in Atlanta, GA.

18

sometimes loaded in compression and, because oftruss design, their ratio of length to radius of gyrationfalls into the intermediate or long column range.Preliminary tests had indicated a possible bucklingproblem with composites in the intermediate columnrange.

The quick fix was to redesign the compositespecifically for use in trusses. Since the biggestperceived problem was with the lateral-load resistanceof the truss plates, we moved the veneer from theedges of the composite to the faces so that thetruss plates would be installed into the veneer. Wehoped that moving the veneer to the wide face ofthe composite would also eliminate the columnbuckling problem. Moving the veneer to the wideface would greatly increase veneer use, but the veneergrade did not have to be as high because the stresslevel would be lower. Resource analysis indicatedthat if all of the D grade veneer could be used, therewould still be enough veneer available to fully utilizethe trees in a stand.

Extensive research was conducted. The lateral-load resistance of truss plates installed in theveneer-faced composite was more than adequate. Infact, it was almost equal to the values for truss platesinstalled in southern pine dimension lumber. Thelower grade veneers did not lower the strength andstiffness of the composite if a few simple guidelineswere followed. The results of this research werereported in COM-PLY Report 25 (McAlister 1983).

Core Material for Composite Truss LumberOne of the requirements of the veneer-faced

composite was a different type of core material. Weneeded a core that was 7/8 inch thick. In combinationwith two plies of l/6-inch veneer on each face, thefinished thickness would be 1.5 inches. Also, thedimensional stability properties of the core neededto be unique. With the wide face of veneer, a certainamount of linear expansion across machine directionof the particleboard would be desirable to reducestresses induced by changes in moisture content.

A contract to produce a fully oriented particleboard7/8 inch thick was negotiated with a west coastproducer. It was assumed that the fully orientedparticleboard core for which we contracted wouldhave somewhat improved strength and stiffnessparallel to machine direction due to the orientation ofthe particles. It was also assumed that linear expansionacross the machine direction of the particleboardwould be 4 or 5 percent after 24 hours of water soaking.Since the veneer moves about 6 to 8 percent underthese conditions, such dimensional instability wouldbe desirable.

We had worked closely with FHA-HUD since thebeginning of the composite structural productsresearch. The demonstration houses using compos-ites had been insured under a special HUD programfor innovative use of materials. In 1981 we were askedto take on a real challenge. FHA-HUD and the NationalAssociation of Homebuilders Research Foundationwere cooperating in the design and construction ofa showcase Energy Efficient Residence, EER-II, nearWashington, DC. The construction was scheduledfor late spring 1982. What composite structuralmaterials could we supply? We agreed to producethe truss lumber, the joists, the studs, and thepreservative treated framing for an All-Weather WoodFoundation for the EER-II project.

This was quite an undertaking. Our testing labwas converted to a pilot-plant operation with a layupand clamp table 34 feet long and 4 feet wide. Everyonein the project was involved in furious activity twice aday as the layups were made and the pressure wasapplied to the gluelines. Vick had tested extensivelya special water-emulsion isocyanate laminatingadhesive and found that a 4-hour cure time underpressure was sufficient to produce strong durablebonds between the veneer and the particleboardcore. Special techniques were developed to laminatethe veneer and spread the glue. The compositejoists were made full length (32 feet). Each day’sproduction was checked off the bill of materials.

We had to postpone producing the truss lumberuntil last because the special oriented core was notavailable. On the very last day in what could becalled ‘good luck’ or ‘just-in-time inventory manage-ment,’ the special core material arrived, but therewas a problem. It was not what we had contractedfor: the material was a three-layer board with thecore oriented at right angles to the face orientation.A hurried test indicated that in other respects thecore was acceptable. There was no time to reorder ifwe were to meet the deadline for the EER-II framing.So we accepted the order and used it. This was awise decision.

Just as we started to fabricate the 2 by 4 trusslumber, we were asked to furnish some additionaltruss lumber for a two-car garage that had beenadded to the project. All the core material availablefrom the original order was earmarked for necessaryresearch projects. We ordered some special phenolic-bonded particleboard 7/8-inch thick with randomorientation that otherwise met our specifications forinternal bond and MOE. This material was used forthe garage truss lumber.

19

The deadline was met, and we loaded all of thecomposite framing on an open flat-bed trailer fordelivery to the truss fabricator and the building site.As the truck pulled away from the lab, it started torain, There were no tarps covering the load, but wewere not worried. This material had passed all of ourdurability tests. Murphy’s Law applied in thissituation-disaster. A week later we received a callfrom the truss fabricator. All of the garage trusseswere coming apart. The truss lumber was splittingright throught the middle. What were we going to doabout it?

We found that the garage trusses made with therandom core material were splitting and comingapart. The trusses made with the oriented core materialwere fine. We needed to find out why, so we carefullytested the physical and mechanical properties of thetwo core materials. Every test recommended in ASTMD-1037 was performed on both core materials plusthe I-l/Binch core used in the joists. There wasvery little difference in the test results between theparticleboards except for thickness swell after a24-hour water soak. The oriented core swelled about5 percent; the thick core swelled about 10 percent;and the random core swelled almost 30 percent. Wechecked some other particleboards for thicknessswell. In every case where the thickness swell after a24-hour water soak exceeded 15 percent, we wereable to reproduce the pattern of failure that had firstbeen observed in the garage trusses for the EER-IIproject. Thickness swell became a major factor inthe specifications for the core material for compositestructural lumber.

Oriented-Core PerformanceWe also checked the performance of truss plates

in the oriented particleboard. To our surprise, thelateral-load strength values for the particleboardwere high enough for effective use in trusses. Anothersurprise was the low linear expansion (less than 0.5percent) for the three-layer oriented particleboardcore. We assumed that low linear expansion wasdue to the cross-oriented core layer. We alsocontracted with the same producer for some fullyoriented flakeboard for core material; some 3/4-inchthick for lamination into l-l/e-inch stock and some7/8-inch stock for composite truss lumber.

The fully oriented core material was a real surprise.To begin with, the linear expansion across machinedirection was about 0.3 percent after a 24-hour watersoak. Thickness swelling was about 5 percent. MOE

of particleboard is usually measured on a panel withthe smooth surfaces top and bottom. The MOE ofthe flakeboard normal to panel surfaces (as a beamwith the cut edges top and bottom) averaged about1.2 million PSI with an MOR of over 6,000 PSI. Thetension and compression values were also quitehigh. There was about a 3 to 1 differential in MOEparallel to and across machine direction. Thus, theMOE of the oriented flakeboard across machinedirection was about the same as the MOE for randomoriented flakeboard. The lateral load resistance oftruss plates installed in the fully oriented core materialwas only 10 percent less than for the same platesinstalled in southern pine framing.

Actually, the fully oriented core material was afully structural product as it was, without any veneerinstalled either on the edges or on the faces. Thiswas a remarkable development. One problem sur-faced. Since the core material was so dimensionallystable, when veneer was installed on members widerthan 3.5 inches there was a serious problem withthe faces swelling so much that the member took ona barrel shape. The problem was severe enough tocause splitting of the core at the center of the memberafter a couple of cycles of wetting and drying.

No further development work was done on thefully oriented flakeboard as a structural materialbecause funding for the RWU was withdrawn andthe project was terminated in 1986.

Solid Accomplishments

Research on composite framing for studs startedin 1974. Standards for the evaluation of performancecharacteristics of stud framing were developed, anda fully satisfactory composite stud was developed by1975. Research on composite joists culminated in afully satisfactory joist in 1978. Satisfactory compositetruss lumber was produced by 1982. An experimentalcore material for composite truss lumber was foundto be a fully structural material in its own right in1985.

Economics of resource availability, market de-mand, and production feasibility were considered atevery stage of the research. The RWU achieved allof its research objectives and goals well within theprojected timeframes.

The RWU was terminated September 30, 1986. Itwas decided that further research and developmentof composite structural products would best beaccomplished by industry.

20

Bibliography for Historyof COM-PLY Project

Anderson, L.O.; Zornig, Harold F. 1972. Build your own low-costhome. New York: Dover Publications, Inc.; 204 pp.

American Plywood Association. 1974. U.S. Product Standard PS1-74 for construction and industrial plywood and typicalAPA-grade/trademark. Tacoma, WA: American Plywood Associa-tion; 35 pp.

American Society for Testing and Materials. 1979. Standard methodsof evaluating the properties of wood-base fiber and particlepanel materials. D-l 037-78. ASTM Stand., Part 22299341.

Batey, Thomas E.; Post, Paul W., Jr.; Matteson, DeForest; Ziegler,Gerald A. 1975. Flexural design procedure for the compositepanel. Forest Products Journal 25(9):49-55.

Blomquist, R.F.; Duff, J.E.; McAlister, R.H. [and others]. 1975. Com-Ply studs-a status report. Forest Products Journal 25(9):34-43.

Blomquist, Richard F.; Duff, John E.; Koenigshof, Gerald A. [andothers]. 1978. Performance standards for composite studs usedin exterior walls. Res. Pap. SE-155 (rev.). COM-PLY Rep. 2.Asheville, NC: U.S. Department of Agriculture, Forest Service,Southeastern Forest Experiment Station and Washington, DC:U.S. Department of Housing and Urban Development. 24 pp.

Bodig, Jozef; Jayne, Benjamin A. 1982. Mechanics of wood andwood composites. New York: Van Nostrand Reinhold Co., Inc.712 pp.

Carney, J.M., ed. 1977. Plywood composite panels for floors androofs: Summary Report. Res. Pap. SE-183. COM-PLY Rep. 3.Asheville, NC: U.S. Department of Agriculture, Forest Service,Southeastern Forest Experiment Station and Washington, DC:U.S. Department of Housing and Urban Development. 13 pp.

Clark, A., Ill; Schroeder, J.G. 1971. Veneer yields from turpentinedslash pine peeler blocks. Forest Products Journal 21(2):53-58.

Countryman, David. 1975. Research program to develop perform-ance specffications for the veneer-particleboard compositepanel. Forest Products Journal 25(9):44-48.

Duff, John E. 1977. Durability and dimensional stability of COMPLYstuds. Res. Pap. SE-172. COM-PLY Rep. 7. Asheville, NC: U.S.Department of Agricufture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 12 pp.

Duff, John E. 1980. Dimensional stability and durability of COM-PLYstuds exposed to soaking, high humidity, and weathering. Res.Pap. SE-207. COM-PLY Rep. 18. Asheville, NC: U.S. Departmentof Agriculture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 12 pp.

Duff, John E.; Koenigshof, Gerald A.; Wiienberg, Dick C. 1978.Performance standards for COM-PLY floor joists. Res. Pap.SE-192. COM-PLY Rep. 14. Asheville, NC: U.S. Department ofAgriculture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 14 pp.

Gaby, Louis I.; Carter, Fairie Lyn. 1981. Resistance of CCM-PLYlumber to subterranean termites. Res. Pap. SE-225. COM-PLYRep. 22. Asheville, NC: U.S. Department of Agriculture, ForestService, Southeastern Forest Experiment Station and Washing-ton, DC: U.S. Department of Housing and Urban Development.

8 PP.

Higdon, Archie; Ohlsen, Edward H; Stiles, William B; Weese, JohnA. 1988. Mechanics of materials. 2d. ad. New York: John Wiley8 Sons. 591 pp.

Koch, P.; Woodson, G.E. 1988. Laminating butt-jointed, log-runsouthern pine veneers into long beams of uniform high strength.Forest Products Journal 18(10) :45-51.

Koenigshof, Gerald A. 1977. The COM-PLY research project. Res.Pap. SE-188. COM-PLY Rep. 1. Asheville, NC: U.S. Departmentof Agriculture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 32 pp.

Koenigshof, Gerald A. 1978. Economic feasibility of manufacturingCOM-PLY studs in the South. Res. Pap. SE-197. COM-PLYRep. 15. Asheville, NC: U.S. Department of Agriculture, ForestService, Southeastern Forest Experiment Station and Washing-ton, DC: U.S. Department of Housing and Urban Development.28 PP.

Koenigshof, Gerald A. 1979. Economic feasibility of manufacturingCOM-PLY panels in the South. Res. Pap. SE-201. COM-PLYRep. 17. Asheville, NC: U.S. Department of Agriculture, ForestService, Southeastern Forest Experiment Station and Washing-ton, DC: U.S. Department of Housing and Urban Development.28 PP.

Koenigshof, Gerald A. 1983. Economic feasibility of manufacturingCOM-PLY floor joists. Res. Pap. SE-241. COM-PLY Rep. 28.Asheville, NC: U.S. Department of Agriculture, Forest Service,Southeastern Forest Experiment Station and Washington, DC:U.S. Department of Housing and Urban Development. 37 pp.

Koenigshof, Gerald A. 1985. Performance and quality-control stand-ards. Gen. Tech. Rep. SE-33. Asheville, NC: U.S. Department ofAgriculture, Forest Service, Southeastern Forest ExperimentStation. 20 pp.

Koenigshof, Gerald A.; McAlister, Robert H.; Lee, Orville G. 1977.Demonstration houses built with COM-PLY products. Res. Pap.SE-177. COM-PLY Rep. 10. Asheville, NC: U.S. Department ofAgricufture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 27 pp.

21

Koenigehof, Gerald A.; Wienberg, Dick C. 1977. Structural perform-ance and standards for COMPLY partition studs. Res. Pap.SE-161. COMPLY Rep. 13. Asheville, NC: U.S. Department ofAgriculture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 10 pp.

Lyons, Bruce E.; Rose, John D.; Tissell, John R. 1975.Performance of plywood and composite panels under concentrat-ed and impact loads. Forest Products Journal 25(9):56-66.

Mahoney, Leonard, Jr. 1975. Economic considerations for themanufacture of structural composite panels. Forest ProductsJournal 25(9) :61-63.

McAlister, Robert H. 1976. Modulus of elasticity distribution ofloblolly pine veneer as related to location within the stem andspecific gravity. Forest Products Journal 26(10):37-40.

McAlister, Robert H. 1977. Structural properties of COM-PLY studs.Res. Pap. SE-180. COM-PLY Rep. 12. Asheville, NC: U.S.Department of Agriculture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 15 pp.

McAlister, Robert H. 1979. Structural performance of COM-PLYstuds made with hardwood veneers. Res. Pap. SE-1 99. COM-PLYRep. 16. Asheville, NC: U.S. Department of Agriculture, ForestService, Southeastern Forest Experiment Station and Washing-ton, DC: U.S. Department of Housing and Urban Development.15 PP.

McAlister, Robert H. 1960. Evaluation of veneer yield and gradesfrom yellow-poplar, white oak, and sweetgum from the Southeast.Res. Pap. SE-226 COM-PLY Rep. 19. Asheville, NC: U.S.Department of Agriculture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 12 pp.

McAlister, Robert H. 1962. Modulus of elasticity and tensile strengthdistribution of veneer of four commercially important southeast-ern hardwoods. COM-PLY Rep. 23. Asheville, NC: U.S. Depart-ment of Agriculture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 6 pp.

McAlister, Robert H. 1963. Effect of veneer grade on bending andcompression of COM-PLY truss lumber. Res. Pap. SE-239.COM-PLY Rep. 25. Asheville, NC: U.S. Department of Agriculture,Forest Service, Southeastern Forest Experiment Station andWashington, DC: U.S. Department of Housing and UrbanDevelopment. 6 pp.

McAlister, Robert H. 1966. Performance of truss plate joints instructural flakeboard and southern pine dimension lumber.Forest Products Journal 36(3):41-44.

McAlister, Robert H.; Clark, Alexander, Ill. 1963. Veneer yields bygrade from three Coastal Plain hardwoods--blackgum, sweet-gum, and yellow-poplar. Res. Pap. SE-236. COM-PLY Rep. 24.Asheville, NC: U.S. Department of Agriculture, Forest Service,Southeastern Forest Experiment Station and Washington, DC:U.S. Department of Housing and Urban Development. 12 pp.

McAlister, Robert H.; Taras, Michael A. 1976. Yield of southernpine veneer suitable for composite lumber and panels. Res.Pap. SE-179. COM-PLY Rep. 9. Asheville, NC: U.S. Departmentof Agriculture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 15 pp.

McKean, Herbert B.; Snodgrass, James D.; Saunders, R.J. 1975.Commercial development of composite plywood. Forest Prod-ucts Journal 25(9):63-66.

Megraw, R.A. 1965. Wood quality factors in loblolly pine. Atlanta,GA: TAPPI Press. 66 pp.

National Forest Products Association. 1977. National design specifi-cation for wood construction. Washington, DC; 76 pp.

Peters, Thomas E. 1963. The use of electrostatic orientation in theproduction of fully aligned flakebaord for COM-PLY joists andtruss lumber. In: Maloney, Thomas M., ed. Proceedings of the17th Washington State University international particleboard/composite materials series symposium; 1963 March 29-31;Pullman, WA. Washington State University; 119-l 34.

Phillips, Douglas R.; Schroeder, James G.; Clark, Alexander, III.1979. Predicting veneer grade yields for loblolly and slash pinetrees. Forest Products Journal 29(12):41-47.

Public Land Law Review Commission. 1970. One-third of the nation’sland. Washington, DC: U.S. Government Printing Office.342 PP.

Saucier, Joseph R.; Holman, John A. 1975. Structural particleboardreinforced with glass fiber-progress in its development. ForestProducts Journal 25(9):69-72.

Schroeder, J.G.; Clark, A. Ill. 1970. Predicting veneer grade-yieldsfor loblolly pine. Forest Products Journal 20(2):37-39.

Southern Pine Inspection Bureau. 1977. Grading rules. Pensacola,FL. 214 pp.

Talbott, John W.; Maloney, Thomas M.; Huffaker, E. Max [and oth-ers]. 1961. Linear expansion design theory for COM-PLY panels.Res. Pap. SE-224. COM-PLY Rep. 21. Asheville, NC: U.S.Department of Agriculture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 56 pp.

U.S. Department of Agriculture, Forest Service. 1974a. Wood hand-book: wood as an engineering material. Agric. Handb. 72.Madison, WI: Forest Products Laboratory. [var. pp.].

U.S. Department of Agriculture, Forest Service. 197411. The outlookfortimber in the United States. For. Resour. Rep. 20. Washington,DC. 367 pp.

U.S. Department of Agriculture, Forest Service. 1960. An assessmentof the forest and range land situation in the United States. Rep.FS-345. Washington, DC. 631 pp.

U.S. Department of Agriculture, Forest Service. 1962. An analysisof the timber situation in the United States 1952-2630. For. Rep.23. Washington, DC; 499 pp.

22

Vick, Charles B. 1977a. Laminating COMPLY studs in a factory.Res. Pap. SE-167. COMPLY Rep. 4. Asheville, NC: U.S.Department of Agriculture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 15 pp.

Vick, Charles 8. 1977b. Adhesives for COM-Ply studs. Res. Pap.SE-1 71. COMPLY Rep. 6. Asheville, NC: U.S. Department ofAgriculture, Forest Service, Southeastern Forest ExperimentStation and Washington, DC: U.S. Department of Housing andUrban Development. 26 pp.

Walker, Roy M. 1977. Structural performance of nailed joints withCOMPLY studs. Res. Pap. SE-175. COMPLY Rep. 8. Asheville,NC: U.S. Department of Agriculture, Forest Service, SoutheasternForest Experiment Station and Washington, DC: U.S. Departmentof Housing and Urban Development. 18 pp.

Wittenberg, Dick C. 1978. Comparative racking strength of wallsframed with COMPLY and sawn timber studs. Res. Pap. SE-1 78.COM-PLY Rep. 11. Asheville, NC: U.S. Department of Agricutture,Forest Service, Southeastern Forest Experiment Station andWashington, DC: U.S. Department of Housing and UrbanDevelopment. 10 pp.

Wittenberg, Dick C. 1981. End bearing and joint strength of COM-PLY joists. Res. Pap. SE-222. COM-PLY Rep. 20. Asheville, NC:U.S. Department of Agriculture, Forest Service, SoutheasternForest Experiment Station and Washington, DC: U.S. Departmentof Housing and Urban Development. 10 pp.

Western Wood Products Association. 1978. Standard grading rulesfor western lumber. Portland, OR. 282 pp.

Wiienberg, Dick. C. 1977. Strength and stiffness of compositestuds. Res. Pap. SE-170. COM-PLY Rep. 5. Asheville, NC: U.S.Department of Agriculture, Forest Service, Southeastern ForestExperiment Station and Washington, DC: U.S. Department ofHousing and Urban Development. 9 pp.

23

McAlister, Robert H.

1999. The research and development of COM-PLY lumber. Gen.Tech. Rep. SE-53.Asheville, NC: U.S. Department of Agriculture, Forest Service, Southeastern ForestExperiment Station. 23 pp.

Between 1974 and 1986, a Southeastern Station Research Work Unit developed standardsfor composite studs, joists, and truss lumber and manufactured and demonstrated thematerials. Economic feasibility was considered in every stage of research. Furtherdevelopment is left to industry.

Keywords: Construction lumber, composite lumber, truss lumber, joists, studs, performancestandards.

McAlister, Robert H.

1989. The research and development of COM-PLY lumber. GenTech. Rep. SE-53.Asheville, NC: U.S. Department of Agriculture, Forest Service, Southeastern ForestExperiment Station. 23 pp.

Between 1974 and 1996, a Southeastern Station Research Work Unit developed standardsfor composite studs, joists, and truss lumber and manufactured and demonstrated thematerials. Economic feasibility was considered in every stage of research. Furtherdevelopment is leff to industry.

Keywords: Construction lumber, composite lumber, truss lumber, joists, studs, performancestandards.

The Forest Service, U.S. Department ofAgriculture, is dedicated to the principle of

multiple use management of the Nation’s forest resourcesfor sustained yields of wood, water, forage, wildlife, andrecreation. Through forestry research, cooperation with theStates and private forest owners, and management of theNational Forests and National Grasslands, it strives-asdirected by Congress-to provide increasingly greaterservice to a growing Nation.

USDA policy prohibits discrimination because of race,color, national origin, sex, age, religion, or handicappingcondition. Any person who believes he or she has beendiscriminated against in any USDA-related activity shouldimmediately contact the Secretary of Agriculture,Washington, DC 20250.