{HISIb This is to certify that the thesis entitled LABORATORY EVALUATION OF EUCALYPTUS GRANDIS AND EUCALYPTUS ROBUSTA FOR THE MANUFACTURE OF COMPOSITION BOARD presented by SIDON KEINERT, JR. has been accepted towards fulfillment of the requirements for hm Major professor Date 3/5/ /7f0 0-7639 OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MTERIALS: P'lace in book return to remove charge from circulation records LABORATORY EVALUATION OF EUCALYPTUS GRANDIS AND EUCALYPTUS ROBUSTA FOR THE MANUFACTURE OF COMPOSITION BOARD By Sidon Keinert Junior A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Forestry 1980 © COpyright by SIDON KEINERT JUNIOR 1980 ABSTRACT LABORATORY EVALUATION OF EUCALYPTUS GRANDIS AND EUCALYPTUS ROBUSTA FOR THE MANUFACTURE OF COMPOSITION BOARD By Sidon Keinert Junior Samples of plantation grown trees were obtained representing two species of the genus Eucalyptus, Eucalyptus robusta and Eucalyptus grandis. The differences between the species could be established in terms of physical and mechanical wood properties. E. robusta had the higher Specific gravity and correspondingly higher mechanical prOperties. Various types of resin bonded composition boards were manufactured in the laboratory from the same materials. These boards exhibited properties which compared favorably with specifications spelled out in the commercial standard for mat formed particleboard. Species characteristics were reflected in board prOperties only in the case of modulus of elasticity. Here the lower specific gravity species resulted in higher moduli at constant board density, confirming similar relationships reported in the literature. In most other cases, the relationships between raw material charac- teristics and board properties were obscured by the dominating effect of board density, a variable that is difficult to control in the laboratory. Linear expansion in particular is very difficult to relate to species characteristics and to any other single variable. To my parents Sidon Keinert and Astrid Keinert who are the light of my being. ii ACKNOWLEDGEMENTS I wish to express my sincere appreciation and admiration to my major professor, Dr. Otto M. Suchsland, whose experience, guidance, assistance and emotional support helped me get through the hardship of this investigation. I appreciate the keen interest and c00peration of the other committee members, Dr. Alan Sliker and Dr. Eldon A. Behr, both from the Forestry Department, Wood Science Section and Dr. Frank Mossmann from the Marketing and Transportation Administration Department. My thanks to Mr. Ivan G. Borton for his encouragement and assistance in fabricating samples and parts for composition boards manufacturing. I also appreciate the cooperation of Ms. Betty A. Briggs in typing this dissertation. I am particularly indebted to my country, Brazil, from which, through a scholarship, this graduate work was made possible. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . Chapter I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . Objective . . . . . . . . . . . . . . . II. CHARACTERISTICS AND IMPORTANCE OF EUCALYPTUS . . . . Overview . . . . . . . . . . . . . . . . . . . Eucalyptus Species in Brazil . . . . . . . . . . Some Characteristics of the Species Studied . . Plantation Background . . . . . . . . . . . . . III. WOOD COMPOSITION BOARD . . . . . . . . . . . . . Basic Processes . . . . . . . . . . . . . . . . Wood Composition Board Applications . . . . Wood Composition Board - PrOperty and Standards IV. EXPERIMENTAL DESIGN AND PROCEDURE . . . . . . Solid Wood Sampling and Testing Procedure. . . . Fiber Dimensions. . . . . . . . . . . Static Bending . . . . . . . . . . . . . . . Tension Perpendicular to Grain . . . Shear Parallel to Grain . . . . . . . . . . Swelling and Shrinkage . . . . . . . . . Specific Gravity and Moisture Content . . Composition Board Manufacturing, Sampling and Testing Procedure. . . . . . . . . . . . . . . Fiberboards Manufacturing . . . . . . . . . . Fiber Blending . . . . . . . . . . . . . . Mat Formation . . . . . . . . . . . . . . Pre-Pressing . . . . . . . . . . . . . . . Hot Pressing . . . . . . . . . . . . . . . iv Page . . vii . . viii 31 41 47 49 49 57 57 68 Chapter V. VI. Flakeboards, Sliverboards and Waferboards Manufacturing . . . . . . . . . . . . . Particle blending . . . . . . . . . . Mat Formation . . . . . . . . . . . . Hot Pressing. . . . . . . . . . . . . Composition Boards Sampling and Testing . . Modulus of Rupture and Modulus of Elasticity Internal Bond . . . . . . . . . . . . Thickness Swelling . . . . . . . . . Linear Expansion . . . . . Density Profile . . . . . . L ITERATURE REVIEW 0 I O O O O O O O O O O O PrOperties of the Wood Species . . . . . Fibers . . . . . . . . . . . . . . . . . Density . . . . . . . . . . . . . . . Other PrOperties . . . . . Composition Boards PrOperties . Modulus of Rupture . . . . . . . . . . . Modulus of Elasticity . . . . . . . . . Internal Bond . . . . . . . . . . . Dimensional Stability . . . . . . . . . Thickness Swelling . . . . . . . . . Linear Expansion . . . . . . . . . . RESULTS AND DISCUSSION OF RESULTS Statistical Procedure . . . . . . Regression Analysis . . . . . . . Covariance Analysis . . . . Solid Wood Physical Properties . . . . Specific Gravity . . . . . . . . . . . Modulus of Elasticity and Rupture Fiber Length . . . . . . . . . . . . . Tensile Strength Perpendicular to Grain Shear Strength Parallel to Grain . . Swelling and Shrinkage . . . . . . . . Composition Board PrOperties . . . . . . Modulus of Elasticity and Rupture. Actual Values . . . . . . . . . . . . Regression Analysis . . . . . . . . . Covariance Analysis . . . . . . . . Internal Bond . . . - . . . . . . . . Actual Values . . . . . . . . . . . . Regression Analysis . . . . . . . . Covariance Analysis . . . . . . . . Page 81 85 85 85 85 86 86 88 88 91 92 92 92 93 95 96 96 98 100 101 101 105 109 109 109 111 112 112 113 119 119 125 127 133 133 133 136 139 144 144 151 151 Chapter Linear Expansion . . . . Actual Values . . . . Regression Analysis Covariance Analysis . Thickness Swelling . . . Actual Values . . . Regression Analysis Covariance Analysis VII. CONCLUSIONS . . . . . . . . . . APPENDIX , A. Particle Geometry Nomenclature B. E. grandis and E. robusta Anatomical Description LITERATURE CITED. . . . . . . . . . . vi Page 153 153 154 157 159 159 165 168 171 174 175 178 10. 11. 12. 13. 14. 15. 16. 17. LIST OF TABLES Total Eucalyptus Planted Area (1977) . . . . . . . . . . . . 8 Eucalyptus grandis Planted Area (1977) . . . . . . . . . . . 10 Industrial Roundwood Production . . . . . . . . . . . 12 Cost Breakdown for Manufacture of Bureau Top with Particle- board Core and MDF Core. . . . . . . . . . . . . . . . . 43 PrOperty Requirements for Particleboard . . . . . . . . . . 45 PrOperty Requirements for Medium Density Fiberboard . . . . 46 Experimental Design and Manufacturing Data . . . . . . . . . 63 Bauer-McNett Fractions Obtained from E. grandis and E. robusta (without bark) Refined from Green Chips in a Bauer 418 I O O O O O O O O O O O O O O O O O O O O O O O 67 Furnish PrOperties . . . . . . . . . . . . . . . . . . . . . 84 Solid Wood PrOperties [Mean Values] . . . . . . . . . . . . 114 Solid Wood Properties Swelling and Shrinkage [Mean Values] . 128 Composition Board PrOperties-Modulus of Elasticity [Mean Values] 0 C C O O O O O C O C I C O Q C O O O O O O O O 137 Composition Board PrOperties-Modulus of Rupture [Mean valueS] O C O O O O O O O O O O O O O O O I O O I O O O 138 Composition Board PrOperties-Internal Bond [Mean Values] . . 150 Composition Board PrOperties-Linear Expansion [Mean Values]. 156 Composition Board Properties—Thickness Swelling [Mean Values]166 Composition Board Properties-Thickness Swelling [Mean Values]167 vii 10. 11. 12. 13. 14. 15. 16. 17. 18. LIST OF FIGURES States and Territory of Eucalyptus Occurrence . Population Density of Brazilian States . . Eucalyptus Plantation Experimental Plot . . Eucalyptus grandis — stem and leaves . . . . Eucalyptus robusta - stem and leaves . . The Watson Patent . . . . . . . . . . . . . Particleboard Processing . . . . . . . . . Typical Flow Chart of Particleboard Process Medium Density Fiberboard Plant Medium Density Fiberboard Plant Digesters and Refiners . . . . . . . . . Medium Density Fiberboard Plant Medium Density Fiberboard Plant — Former . Medium Density Fiberboard Plant Press . . . . . . . . . . . . . . . Lumber Core Furniture Panel . . . . . . . . Particleboard Core Furniture Panel . . . . . Raw Material Metering Bins Area Medium Density Fiberboard Core Furniture Panel . Design and Construction of Lumber Banded Bureau TOp Bolt Conversion . . . . . . . . . . . . . . viii Drier and Blender Press Feed Line and Page 18 20 22 25 27 30 32 33 34 35 36 38 39 4O 42 49 Figure 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. Boards from which specimens were taken . . . . . . . Solid Wood Static Bending Test . . . . . . . . . . . Solid Wood Tension Perpendicular to Grain Test . . . Solid Wood Shear Parallel to Grain Test . . . . . . Solid Wood Swelling and Shrinkage - Measuring Apparatus I O O O O O O I O I O O O O O 0 Forest Products Laboratory Standard Flakecutter and Hamermill I O O I O O O O O O O O O O O O 0 Double Disc Pressurized Refiner Model Bauer 418 Blending Operation — Medium Density Fiberboard . . Medium Density Fiberboard - Mat Formation. . . . . Mat Formation Equipment - Medium Density Fiberboard Medium Density Fiberboard - Pre-Pressing Operation . Medium Density Fiberboard - Pressing Operation . . . Illustration of Particle Geometry Range a) Flakes b We ers c 81 vers d Fibers o o o o o o o o o o o o o o o o o o o o Particles - Air Drying Operation . . . . . . . . Composition Boards - Sampling Procedure . . . . Internal Bond - Testing Procedure . . . . . . . . . Bending Specimens - Specific Gravity Distribution Tensile Specimens — Specific Gravity Distribution . Shear Specimens - Specific Gravity Distribution . . Swelling and Shrinkage - Specific Gravity Distribution Modulus of Elasticity Distribution . . . . . . . . . ix Page 50 53 55 58 6O 64 66 69 71 73 75 77 79 82 87 89 115 116 117 118 120 Figure 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. SO. 51. 52. 53. 54. 55. 56. 57. Solid Wood Modulus of Elasticity - Regression Lines . Fiber Lengths - Distribution . . . . . . . . . . . Tensile Strength Perpendicular to Grain Distribution Shear Strength Parallel to Grain Distribution . . . . Longitudinal Shrinkage Distribution . . . . . . . . . Tangential Shrinkage Distribution . . . . . . . . . Radial Shrinkage Distribution . . . . . . . . . . . . Volumetric Shrinkage Distribution . . . . . . . . . . MOE Mean Values - Lowest and Highest Average Board Densities O O O C O C C C O O O I O O I O O O 0 MOR Mean Values — Lowest and Highest Average Board De.nSitieS I O O O O O C O O I O O O O. O O O O O Modulus of Elasticity - Covariance Analysis (a) Adjusted Means (b) Tested for Differences due to species (c) Tested for Differences due to resin level (d) Tested for Differences due to particle geometry Relationship Between Board Density and Bending Strength of Particleboard Made From Various Species . . . Relationship Between Board Density and Modulus of Elasticity of Composition Board Made from E. grandis and E. robusta. . . . . . . . . . . . . . . . . . Fiberboards Density Profile . . . . . . . . . . . . Sliverboards Density Profile . . . . . . . . . . . . Flakeboards Density Profile . . . . . . . . . . Waferboards Density Profile . . . . . . . . . . . Internal Bond Mean Values - Lowest and Highest Average Board Densities . . . . . . . . . . . . . . . . . Page 121 122 124 126 129 130 131 132 134 135 140 142 143 145 146 147 148 149 Figure 58. 59. 60. 61. 62. 63. 64. 65. 66. Internal Bond Covariance Analysis (3) Adjusted Means (b) Tested for Differences due to Species (c) Tested for Differences due to resin level (d) Tested for Differences due to particle geometry Linear Expansion Mean Values - Lowest and Highest Average DenSitieS O O O O O O O O O O O O I O O O 0 Linear Expansion Covariance Analysis (a) Adjusted Means (b) Tested for Differences due to species (c) Tested for Differences due to resin level (d) Tested for Differences due to particle geometry Thickness Swelling Sorption Curves . . . . . . . . . . Thickness Swelling Sorption Curves . . . . . . . . . Thickness Swelling Sorption Curves . . . . . . . . . . Thickness Swelling Sorption Curves . . . . . . . . . . Thickness Swelling Mean Values - Lowest and Highest Average Board Densities . . . . . . . . . . . . . . Thickness Swelling Covariance Analysis (a) Adjusted Means (b) Tested for Differences due to species (c) Tested for Differences due to resin lexzel (d) Tested for Differences due to particle geometry. xi Page 152 155 158 160 . 161 162 163 . 164 . 170 CHAPTER I INTRODUCTION By world standards, Brazil is favored with a high rainfall and good soils. The inherited native forests were plentiful, but most of the coastal forests were heavily cut-over in the early centuries of European colonization. During the past half century the wonderful forests of Araucaria angustifolia on the great basalt flows of the south have been eaten into for internal consumption and for export, and the mixed rain forests of the central and upper Amazon region remain as a potential reserve of raw material. Brazil, as a developing country, provides an economical environment ideal for the establishment of a multitude of new investments in technology directed to the utilization of wood. Industries like the pulp, paper and particle board industries demand large quantities of raw material close to the manufacturing plant and markets to be served. Large reforestation programs have been underway due to government incentives. Exotic species like Eucalyptus species and pine species are the preferred ones. Fast growing, short rotation, high rate of return on investment are some of the reasons for their widespread use. Increased demand in Brazil for housing and furniture will encourage the ad0ption of efficient manufacturing techniques in these industries. In the furniture industry, these developments will parallel those that occurred in EurOpe and in the United States since the end of World War II, namely the introduction into furniture panel construction of the composition board core. These composition boards are made today in a variety of types and qualities depending on raw material availability and application. Whether or not the Brazilian housing industry will follow the North American example is much less certain. The preference in Brazil for masonry construction is possibly more a matter of tradition than the result of efforts to minimize construction costs. The large demand for structural wood panels as it exists in North America where it has resulted in the development of huge plywood capacity and more recently of structural composition board manufacture may not materialize in Brazil for some time to come. This does not rule out, however, the feasibility of a structural composition board industry for specific market applications. Both the furniture core and the structural composition board industries will be based on tropical and subtropical Pines and Eucalyptus species as the most logical raw materials. In contrast to the genus Pinus which is rather homogeneous, the genus Eucalyptus includes a large number of species with widely varying properties. Most laminated wood products reflect to a greater or lesser extent the properties of the species from which they are manufactured. These relationships between raw material properties and product properties have been the object of considerable research efforts. The development of the composition board industry in Brazil based on Eucalyptus species will greatly benefit by the investigation of these important technological relationships. Objective It is the objective of this study to make a contribution in this area by evaluating two Eucalyptus species as raw materials for the manufacturing of a variety of composition boards ranging from wafer- board to medium density fiberboard. The selection of the two species was limited by the availability of plantation grown Eucalyptus in the United States. The differences between the two species are relatively minor and do not represent the considerable variability of the genus as it may be found in Brazil. CHAPTER II CHARACTERISTICS AND IMPORTANCE OF EUCALYPTUS 1) Overview The Eucalyptus species' original habitat are the vast lands of Australia. They occur in the states of Queensland, New South Wales, Victoria, Tasmania, South Australia, Western Australia and the Northern Territory (Figure 1). The story of the cultivation of the Eucalyptus species and the early recognition of their economic potential commenced with the establishment of small plantations in Southern EurOpe and North Africa over one hundred years ago. Since then the ease with which Eucalyptus species can be cultivated, their rapid growth, and their adaptability, have led to their widespread introduction into many countries, especially in those which are poorly endowed with forest resources. They have become such an important factor in the economy of some countries that millions of trees are now planted each year throughout the world. *[48] Today, Eucalyptus species are planted in all five continents of the world confirming its importance as a raw material for manufactured products on a world wide scale. The area on which Eucalyptus species are grown outside Australia (original habitat) rose from 0.7 million ha in 1950 to 3.7 million in 1974 [2] and it continues to increase rapidly. The annual growth incre— ment of these new forests is estimated at 40 million m3, compared with *numbers in brackets indicate references. an estimated 9 million 1113 harvested annually from some 12 million (ha) of commercial Australian forests. [15] Many Eucalyptus species grow naturally on soils of low nutrient status but they have the capacity to respond with increased growth rates to more fertile conditions and especially to higher levels of nitrogen and phosphorus. Most Eucalyptus will not thrive on soils that are alkaline and have quantities of free calcium or sulphate in the profile. The effects of climate on the growth of Eucalyptus species are as important as soils effects and as a result they are planted in large quantities in trOpical and subtropical areas, like the states of Florida, California and Hawaii in the United States and almost all states in Brazil. (Carter, 1974) [10] reported yields ranging from 15.0 m3/ha/year on low quality sites to 31.9 m3/Ha/year on high quality sites for 10 year old E. grandis plantations in Southern Queensland. (Australia). (Rudolph et a1, 1978) [59] reported growth rates in Brazil averaging 40 m3 or more for 7 to 8 year old plantations. No data were available on growth rates for E. robusta. Some Eucalyptus species develOp very high levels of growth stress within the bole which may cause severe end splitting in logs, distortion during sawing, and severe shrinkage during drying. The causes of high stress levels are not well understood; the factors suspected include genotype, age, log size, growth rate, and lean. While it is clear that growth stresses can be high in very fast-grown trees, stress is generally less severe in larger logs than in smaller logs of the same age. Hillis and Brown [33]. Most of the "a. ‘.4.II. I .a .. . I. n z -1 A an .Hmma wnma czoum can maaaam "ouusom .mocopusuoo mnua%amosm mo >uouwuuow paw moumum .a shaman I I e ... ‘- III’I lull! 0 u .. J .1 Iu . . .\.I 5...... . . 2.. I. . it :- l. A u... .- ‘.I .3 a .I I .. I- - I o. III-Io .i. .I It I. I , I tn.I... fi .05 .I I II A . . . ........... . . . ........... [ill-II ll . i ’2 Out. u...o.ll||l!ai‘|.1lluUuJI’| TIL-"IiU. " filo—Ow: ham-n05 298300.; <3<¢pm2< N . . ., . . . V V - v. 1 .1". AI ’ .. Arr. ._ I. .1 ._ a . n. .s 41%.. “III! I. H , ..... .,.. .|.. I I .d ”1:: h I S l _. l. . 7H1). In) . w i ,J ”4 .I. lelI. III. obnl . n .., a. i .\. 4| 1 n i I . I.“ _ - "l > n I r1 my 4.. ;.,.iI 1. _ I. uv.n . I ., I. - I . p a PM! . x. .Ir M... j v I? . 1 . \‘I luv POI . II. J. i o I- , , g I ..| _ u WEIL 13.4 3%... It ., . .13! ill J11 F. .J 4 1...! .1 nl‘ I! a r. _ L. a . ‘5. / ll- :Slr .I I1..- |l Fj“J W4 s 0, II . v . a p , ' C 4 lo :I . 5 o ~ - L] ' I a. ........... «IPnlquLsnca: u I we. I. u .. a 1:. , «vol . x no a (i. .11! a 3.: It! i . . 1 , .1: A A -‘l . n \4 \fi. 9 I It | .‘(fin M 0m!- DI J \h \c . , j I. I L hall. ill a I~ . u a J...- t “It ' ‘1 1 , .I’ A . lo"... n!» .. ... u .l. x L .o . . -- ...z .. vq.. 1: a. ix. a, .ra.., ryri ... .x’ )‘I. 0‘ ulal .. l .J.., n . g Q n- .I! a. ,.a.: II: In, .I .. ... . Ill . . r :1 . . . , \s I! i x I [.61 iii... a u , \I. , .3. . I — F . .. I‘Du.c—v.r:6.~..1. I. T I. .i . ..I.. a. no. . a . a v :1. “‘3 J . research efforts dealing with silviculture, management, utilization and economics of Eucalyptus in Australia, Brazil, Africa and United States can be found in the Book Eucalypts for Wood Production by Hillis and Brown [33]. 2) Eucalyptus Species in Brazil Brazil is the leading Eucalyptus species planting nation totaling an area of about 1,500,000 ha +(IBDF, 1977), thanks to the pioneering efforts of Edmundo Navarro de Andrade, who in 1910 working for the Forest Service of the Paulista Railway Company in the state of Sao Paulo, took into his hands the task of turning Eucalyptus to an important species in the economic realm. He was responsible for the planting of more than 38 million trees throughout his life. Eucalyptus are planted mainly in the states of Minas Gerais, Espirito Santo, Rio de Janeiro, Sao Paulo, in the southern Maranhao and Bahia in the northeast, Parana, Sta. Catarina and Rio Grande do Sul in the south and Mato Grosso and Goias in the center part of the country (Fig. 2). Source: C¥IBDF - Brazilian Institute for Forestry Development. Table 1. Total Eucalyptus Planted Area (1977) Unit: (ha) §£a£g_ Period Total Minas Gerais 67 77 663,640 Esp. Santo 67 77 126,573 Rio de Janeiro 67 77 8,986 Sao Paulo 67 77 318,775 Parana 67 77 46,812 Sta. Catarina 67 77 13,211 R. G. do Sul 67 77 16,596 Maranhao 72 77 16,419 Bahia 72 77 16,409 Mato Grosso 7O 77 263,487 Goias 67 77 37,303 Total 1,511,782 Source: IBDF — Brazilian Institute for Forestry Development TERRITORY OF RORAIMA TERRITORY 0F A“APA 32°25' TERRITORY OF . - fi-FERNANDO DE CEARA RIO GRANDE ‘3.” DO NORTE -_-; PARA! BA ' ' PERNAHBUCO .? ALAGOAS SERGIPE LEGEND: MATo GROSSO . TERRITOR '“haf' 0F RONDONIA per (.7: E22 0" FEDERAL ::. 1-10 D'STR'CT ESPIRITO sguro f E23 '0'20 ,. RIO DE JANEIRC 1:] 20-50 PARAM-“r- GUANABARA 3;; 50-100 ....T El 100 200 —\SANTA CATARINA V as Over 200 / RIO GRANDE DO SUL I Source: Instituto Brasileiro de Geografia e Estatistica, Anuario Estatistico do Brasil (Rio de Janeiro, 1977). Figure 2.—-Popu1ation density of Brazilian states. 10 About 80% of the Eucalyptus grown in Brazil are Eucalyptus saligna and Eucalyptus urophyla (known as Eucalyptus alba) with Eucalyptus camal dulensis, Eucalyptus grandis, Eucalyptus globulus, Eucalyptus delega tensis, Eucalyptus tereticornis, Eucalyptus citriodora, Eucalyptus robusta, and Eucalyptus maideni making up the rest. Table 2 shows the planted area for Eucalyptus grandis that is of interest for this particular study. Table 2. Eucalyptus grandis planted area (1977) Unit: (ha) 84222 Mai Minas 95,147 Goias 5,347 Mato Grosso 28,610 R. G. do Sul --_ Parana 2,341 Bahia --- Sao Paulo 7,492 Sta. Catarina --- Esp. Santo 2,473 Overall 141,410 Source: IBDF - Brazilian Institute for Forestry Development. ll Eucalyptus plantation management includes complete site preparation, fertilization, and weed control. Emphasis is almost exclusively on maximizing the production of wood fiber per hectare per year, for pulpwood, fiberboard, charcoal wood and similar products. Cultural practices and rotations are becoming fairly well defined. Common practice is to plant by spacing 3 by 2 or 3 by 1.5 meters, and harvest the first crop by clearcutting at age 7 or 8 years, when mean annual increment in cubic meters per hectare culminates. Stump sprouts are profuse, and at about 10 months of age are reduced manually to the best two or three per stump. The second crop is harvested at 7 or 8 years, and the process is repeated a third time, so that three crOps are obtained over a 21 - 24 year period from one planting. The site is then cleared and a new planting established to repeat the three-crop sequence. Management for combinations of products on longer rotations is just beginning; where sawlogs are desired, thinning is part of the regimen. E. citriodora is managed for a combination of essential oils such as citronella, which is extracted from the leaves, and pulpwood, poles, and some sawlogs. Two storied stand management of this species is fairly common. Rudolph [59 ] made a theoretical comparison of a management alternative like the one cited before (three crops) in 21 years directed to one single product, like pulp— wood or charcoal against one based on one single crop in 21 years directed to multiple products, like lumber, plywood, etc., and came out with a net present worth in favor of this second one. As a result 12 of this, depending on future market behavior, management of Eucalyptus species should be directed to multiple products production without impairing the charcoal and pulp and paper industries. Eucalyptus growth rates are nothing short of phenomenal. In well managed plantations, even without the benefit of improved seed and planting stock, yields averaging more than 40 cubic meters per hectare per year are common. On good sites and with improved seed and planting stock, yields up to 62 cubic meters have been obtained on 7 and 8 year rotations. Industrial roundwood production in 1973 (Table 3) was Table 3. Industrial roundwood production, 1973 Million Product cubic meters Sawlogs 18.9 Logs for veneer and panel products 2.5 Pulpwood 4.5 Charcoal wood 20.0 Total 45.9 Source: IBDF - Brazilian Institute for Forestry Development. Charcoal is the leading wood product in terms of raw material required. Brazil has negligible fossil fuels, and the rather large steel industry that is centered close to the iron ore sources in the southeastern state of Minas Gerais relies heavily on charcoal. In that area, natural stands of broadleaved species have been all but eliminated for conversion to charcoal. The wood volumes involved are huge. In 13 1973, it took 20 million cubic meters of wood to produce 11.9 million cubic meters of charcoal. It takes 1.75 cubic meters of roundwood to produce 1 cubic meter of charcoal, and 3 cubic meters of charcoal are required to process a ton of steel. As the national forest disappears, the steel companies are recognizing the need to grow wood for charcoal by acquiring land and establishing their own plantations, and also encouraging other landowners to do so. National planning envisions tripling domestic steel output by 1985. [59 ] If that goal is realized, several million hectares of additional plantations will be needed to supply wood for charcoal. The lumber and panel products industries are concentrated in the southeast and south, close to the major markets. With the superb Araucaria forests disappearing, domestic softwood lumber will soon be scarce. Some lumber firms have already shifted location into the interior and to the Amazon Basin. In these regions, scarcity of capital, limited number of species, marketability, inadequate access to the timber, and long transport to major markets (Fig. 2) are major obstacles in the establishment of lumber and panel product industries. Nevertheless, deve10pment is inevitable because the natural softwood resource in the eastern and southern regions is practically gone. The increasing exotic pine and Eucalyptus plantations, primarily in the southeast and south, will, in time, make a major contribution to the hardwood and softwood lumber and panel products supply, but the projected demand through the year 2000 will not be met without immediate large-scale expansion of pine and Eucalyptus plantations. Pulp and paper production increased from 14 418,000 tons in 1963 to more than 1.3 million tons in 1974. Neverthe- less, consumption exceeds supply. Almost 300,000 tons of paper and paperboard were imported in 1974. The timber in the Amazon Basin is not likely to be deve10ped for pulp and paper products. Not only is access difficult, but the species are unsuitable using present manufacturing technology. Thus, the most logical approach to supplying adequate quantities of pulpwood for the near future is an immediate expansion of the planting program and the study of Eucalypts and pine species for an efficient utilization, primarily in the south and southeast. In these regions, there is an increasing number of fairly large pulp, paper, and other wood—fiber firms, many of them multinational companies Operating entirely on plantation—grown timber, much of it produced on land which they own. Thus increasing research needs to be aimed at efficient utilization of plantation—grown Eucalyptus and pines for multiple products use. 3) Some Characteristics of the Species Studied. Eucalyptus robusta Sm. - Swamp mahogany (syn E. multiflora poir). It is also known in Queensland as swamp messmate. "The size and strength of the tree, like that of the EurOpean Quercus robur, seem peculiarly to justify the name robusta." Thus wrote Sir J.E. Smith in his Specimen of the Botany of New Holland, published in London, 1793 [53]. The quotation concludes an ample description of the tree, and, if not the most fitting species to bear this fine title it must be remembered that it was applied when only half a dozen species of the whole genus had been encountered. Its original habitat being the coastal 15 areas from southern New South Wales to Southern Queensland (Australia), it came early under the observation of the first white settlers. As it occurs mostly on river beds and in swampy localities, it is commonly known as swampy mahogany. It grows to a fairly large tree, clad with a reddish, soft brittle bark. The foliage is coarse, the individual leaves being broadly lanceolate and firm, with many lateral veins. Von Mueller [21] states that they are "lighter colour above and more shining beneath," but this is clearly an error; the reverse being the fact. The buds have a distinct rostrate opercula, and seven are frequently arranged in stellate on a long, flattened peduncle. The fruits are elongate — ovoid, truncated, and when viewed from above, show a maltese—cross design that is a helpful guide to the preliminary identification of the species. Eucalyptus grandis Hill. Rose gum, is also known in New South Wales as flodded gum toolur (syn. E. saligna var. pallidivalvis). This is a magnificent and useful tree whose claim to specific rank has suffered some reverse. First described by W. Hill, director of the Brisbane Botanic Gardens in 1862,[52] it was not formally recognized as distinct from Eucalyptus saligna until Baker and Smith in 1902 described it under the title E. saligna var. pallidivalvis. In 1918 Maiden lifted it from it's similars under the fitting name quoted before. The evolutionary changes that are operating on the whole genus present, in the era of today, no types more confused with one another, than Eucalyptus grandis, Eucalyptus saligna, Eucalyptus botryoides. Examples of each may be found so opposed in their obvious characteristics that 16 any suggestions of kinship is unreasonable, on the other hand, the merging of one type to the other may be so gradual that two individuals close together in the scale may present no discoverable difference. The distinction between Eucalyptus saligna and Eucalyptus grandis is subtle and morphologically ill—defined. The oil of the former contains less Eucalyptol than that of the latter, and the timber of Eucalyptus saligna is darker in color and denser than the timber of Eucalyptus grandis, it is also less tough yet more durable. The young foliage and buds of Eucalyptus grandis are less often glaucous than its similars whilst its fruits are larger and generally coarser. The original habitat of the tree extends from about Goulburn, New South Wales in the south to northern Queensland. Its greatest concentration is in the northern rivers district of New South Wales, where on the coastal belt and in the gullies of the foothills magni- ficent specimens stand straight and clean of limb in the glory of a clear blue grey bark, towering over the surrounding trees and under— growth. The fruit is urceolate, either sessile or shortly pedicellate on an aneled peduncle, glaucous, with the valves prominently exserted. The valves are whitish from which feature the varietal name of the synonym suggested. [21] 4) Plantation Background. The Eucalyptus species used in this study came from a small experimental plot planted October 15, 1971 near Palmdale, State of Florida. 17 According to the U.S. Forest Service Experimental Station in Lehigh Acres, Florida, the seeds of Eucalyptus robusta were from Immokalee Seed Orchard, Florida, and Eucalyptus grandis from Biggar Site, Lee County, Florida. Some of the trees were machine planted, others hand planted on an area of 25 x 30 meters with a 5 x 5 feet spacing. At the time of cutting, the trees were approximately 7 years old. Average height was about 40 feet, average D.B.H. about 9.5 inches, and specimen trees were chosen according to uniformity of size (Figures 3, 4, 5). 18 Figure 3. Eucalyptus Plantation Experimental Plot. l9 20 Figure 4. Eucalyptus robusta - stem and leaves. 22 Figure 5. Eucalyptus grandis - stem and leaves. 23 «‘9 "5 {fig- CHAPTER III WOOD COMPOSITION BOARD Composition boards are a combination of the solid wood converted into a variety of comminuted forms and a binder that can be added or generated during the manufacturing process. The quality of these products is determined largely by the quality of the glue bond (its completeness and permanence), by the geometry of the particles, and the species of wood used. Composition boards can be grouped into these categories: Composition Board /\ Particleboard Fiberboard e Extruded PIatEn MDFI Hardboard Insulation ‘ pressed ; board Particleboard is made from wooden elements generally larger than the wooden cell. Fiberboard is made from fibers and fiber bundles having dimensions of the same order of magnitude as those of the wooden cell. Particleboard was invented 75 years ago by Henry Watson of Valparaiso, Indiana. A basic patent was issued by the U.S. Patent Office in 1905. This patent (Figure 6) shows clearly a flakeboard very similar to some types of board made today. Its industrial develOpment which started in EurOpe is a result of economic wood (raw material) scarcity and the necessity of utilizing large quantities of wood residues. In the 24 25 .moma .w umsma< .msm .oan .oz nausea .m.; . scones comma: aze .o shaman .3: .a i: 3::- ..2. .: .89 3.... 2:321: .amaom uh—womzoo .23th d .m .32 .c .094 nuhznzkm .0362. 62 26 beginning, technology was simply transferred from Europe to the United States, but as demand patterns and the raw material basis changed from roundwood to cheaper residues, different process modifications occurred. Fiberboard and the recent medium density fiberboard are U.S. developments. Production of particleboard in 1977 was 3,593 million square feet. Medium density fiberboard was 441 million square feet. This is based on 48 companies with 76 plants. (National Particleboard Association). 1) Basic Processes The particleboard process is a laminating process and can be distinguished from other laminating processes by the discontinuity of laminas and the extremely low glue spread. The laminas are small particles which are formed to a mat by gravity or other method of desposition. This loose mat of adhesive coated particles is then compressed in a hot press to a considerable degree of densification. Urea formaldehyde resin is most commonly used. A small part of the total particleboard production is manufactured with phenol-formaldehyde adhesive, a waterproof adhesive. Most particleboard today is manufactured by the so-called "platen" or "mat formed" process. Only a very small part of the total production is extruded. In the extrusion process adhesive coated particles are fed continuously through a vertical die (Figure 7). The major steps in the manufacture of "mat formed" composition board are: 27 Figure 7. Particleboard Processes. 28 P1. Pla { l/ Coated particles 1.5.3. in”; in... .. 1.. gr... .2; 2...?“ “1?... 3.5 m e. xgaaVV/a \ Oscil lating ram .1.‘ 29 1) Particle preparation 2) Particle classification 3) Particle drying 4) Blending (application of adhesive) 5) Mat formation 6) Pressing of mat between heated platens A general flow diagram is shown in Figure 8 (Suchsland, O.) 1968 [77]. Board properties of extruded particleboards are quite different from those manufactured by the platen system. In general, particle- board properties can be predetermined, controlled, and modified to suit certain applications. Medium density fiberboard (MDF) bridges the gap between fiberboard and particleboard technologies. MDF is based on a pulping process which reduces the wood raw material to fibers or fiber bundles. Yet, board formation and pressing are dry, and the final product competes with particleboard in the market place. Its properties are very similar to those of conventional particleboard. While conventional fiberboard products have densities of around 1.0 g/cm3 (wet or dry formed hardboard) or around 0.1 to 0.5 g/cm3 (insulation board) medium density fiberboard is manufactured at an average density of .75 g/cm3 (Suchsland, O.) 1978 [73]. The major steps in the manufacture of medium density fiberboards are: 30 .mmoooum canon mqufluuma we uumso 304m amoaaze .m muswfim ‘7'.-- Illlllllll I I — 1 5:22;. :..-.C guzczamaz :z< .mzcm .eulxmo uo4<> mwkwmwa w>4<> m>.._<> mwhmmoa >Kmm > 7F 54 [_ID a .0”. = AU Maw mmo>u>zoo p.103... .uopcoqn can Lumen I ucmam unmorpmawm zuwwcwc Esfivo: .HH shaman omEDONm mme._. “.0 m mo>w>zoo. wizom 19m; .104sz 5:10 25 xozmso , \zoEooa zaum / // T \ © D Y La JTHC J AUflv® ”.5340 mdmfiosz 35 poeuom I unmam Uumonuonwm zuflmsoa Esflpmz .NH muswam wZ.Io<2 ozimOu 1“ mommmmAZOouma RH TN mofi. .mwmum was mafia poow muons I ucmam pumcrucrflu muwmcwa Enwwmz .ma muswflm 34m 34m Baa: a to Bo 2.5 88 02;]: wgm U xx.“ Mu 4... mOFOMFwO 44.—ME ® ¢ . q f, % .523 n. m \ ® szimm3m4m2 mmwmm ¢ Q A , >F_m2w0 OOOI wmwma /.—.m 30 OOO¢ 25 ¢ mmew mhmdg 37 glue joint through the thin face veneers. Five—ply construction is then needed in these cases (Figures 14, 15). Particleboard has moved into different markets like the interior wall paneling business, store fixtures, kitchen cabinets, and other products, due to the development of techniques like direct printing methods, low and high pressure laminating, etc. In the mid-1960's particleboard entered the structural market in the mobile home field, where it supplanted plywood for floor under- layment and mobile home decking. It entered this market because of its lower cost, its smooth surface, and because it could be obtained in 10-12 and 14-foot lengths which eliminated the need for end joints when laying the panels on the mobile home floor joint system (Maloney, T.M.,)1977 [47]. There are strong indications that particleboard will make additional strong advances in the structural market. The development of the so—called "waferboard" is a good example. How successful structural particleboard will be, will depend on research efforts, economic advantages of the new products over plywood and lumber, availability of peeler logs, demand for housing, etc. MDF is a newcomer on the field, but has scored significant successes which are based on some of its distinct and unique proper— ties. The most significant property of MDF as far as its use in the furniture industry is concerned, is its uniformity of structure which is reflected in a smooth and tight edge that can be machined almost like wood (Figure 16). However, it is considerably more expensive than particleboard. In a panel with straight and plane edges the 38 5- PLY PLY 5 LUMBER CORE Figure 14. FURNITURE PANEL 39 Figure 15. PARTICLEBOARD CORE FURNITURE PA NEL PLY £3 .. 40 MEDIUM DENSITY FIBERBOARD CORE . Figure 16 FURNITURE PA NE L 41 additional cost would not be justified. Where edges are profiled, lumber edge bands and, therefore, cross band veneers can be eliminated. The panel edge can be machined and finished directly. The savings more than offset the higher board costs. This is demonstrated in Figure 17 and Table 4, showing a comparison of the cost of manufacturing a bureau top with particleboard and MDF (Suchsland, 0. 1978) [76]. Structural applications of medium density fiberboard are still undefined. Most of the research efforts in the structural board field are concentrated on achieving maximal mechanical and elastic board properties by using large, thin flakes in both random and oriented configuration. MDF, on the other hand, by virtue of its much more homogeneous structure may offer greater resistance to deterioration by swelling stresses. The medium density fiberboard segment of the industry is expected to have a strong impact on the board industry and markets. A number of plants have been built and others are in the process of being constructed. 3) Wood Composition Board - Property and Standards The standards for particleboard are formulated by the National Particleboard Association. A standard was first produced for mat formed wood particleboard in 1961. With more experience in the field and far more production developing in the 1960's this standard was rewritten in 1966 and is the present one in force, a new revised edition is expected to be published in 1980. It is designated Commercial Standard .Acmwflnofiz .mcflomm compo .%:mano ououflcuom nEoofiuoHZ snow xmouuoouv .aou Dawson vocamn ponaoa mo coauoouumcou pom cwwmoo .NH muowfim QOK qutbm 42 //.. / / wk / .. of _ _ k V _ - . L:V\n I fix i *1! n lug XV n!!.li.-il.;1u. |.:|| . T / .. L _ _ 3.36:3 v _ _ _ _ _ _ _ ..n.mm omqommdttcm " u _ _ _T Ii; " Sud ..§-~\_ _ L . T _ _ 3.2. 43 was 8.2 58 383.1 .58 EN 5m zmfism m momfi .52 mm. lvmul ©2258: .8sz 21 I «I .33 1.83 mm .m on .n 232.”: I on. 5E 93. 9.5 zmemzm m momfi 86 mod .3522 .58 2. MW! was I mm .1 82% 396 mm. mm. meow ozq xoqm mmmzms .8. km. 8E mmm2m> mvm 8.. 938 I 3.1 Emssu 43qu SEER o: 8.523 33 .Qmmmmt .38 .8: 3336135 .muoo mm: paw ouoo oumonmaofiuuma :uw3 ecu nomads mo mHSuommoawE now QBvamoun umou .q mHan 44 CS 236-66 "Mat-formed Wood Particleboard." Table 5 presents the property requirements for particleboard specified in this standard. The standard distinguishes two types of particleboard: Type 1 generally made with urea formaldehyde resin binders for interior applications and Type 2 generally made with phenolic type binders suitable for interior and certain exterior type applications when so labelled. Within each type products are further differentiated by density grade with an A, B, and C level for the interior applications and A and B density grade for exterior applications. As a further differentiation there are two quality classes in each of the density grades with each class having a separate set of physical properties established for modulus of rupture, modulus of elasticity, internal bond, linear expansion and face and edge screw holding. The main innovation in the new standard now undergoing revision is the estab- lishment of minimum property requirements for the so-called waferboards. In August 1973, the National Particleboard Association published NPA 4-73 "Standard for Medium Density Fiberboard," Table 6 presents the property requirements of this board, which are quite similar to particleboard. Properties of the different experimental particleboards in this study were compared to use classification Type 1 and density grade B, and Class 2 which is the standard specification for industrial core stock. Amv colcmm mo "mousom .uco.uou.—aaa noosoauo taco-cu ac. seduces. has o~au~.:; .ocdoou u.~c=0&o a—_ouocuou 0.00:.5 vanes—no. you: can easy-sol >_uou: 0c. can-.39 nu.) tool v.aono«u.uuon valuouuuqx .n uhhhu .oco.uou.~nao no..oa:_ no. oqnduaao .unonc‘n cauou ovsgovqoluOuuoou: sun: oval unucuocoo_ euooaouu.uu¢s 00tuououul .— Ohxha 45 . . . a so \- -..-----------------------------------------.i a: on 2 o 3 2:. 3. can a . ~ . . eaten-nu usavuo uO-uouuu 00. n- «n.o no 000.oa~ 00..“ ~ .au.ocom $5.008. .u acuu0)o cane—log Una-nous ‘0‘: no. aqua. 0.00 can 00a ma.0 00v 000.00o co..n ~ a. awkwnL-vum n - - - -- - 36 2. 8062 8.; ~ .35.: .32: n- u- -v -- -n------ -----a--------a-.-u--c- .-o no. 4 nonooau-----o-u--u----------------u-uuu---u-o -au as“ 0n.0 0n 000.0nn 00... u .uou¢= at. a. aux-An ha 2o... 32. is: --- w: 2.: 2 So a: 8. ~ audio :3. nn.o 00 000.000 00..~ ~ .0. so .04. on .xu.-:00 (5.06.. duca- ohou oo~ r- 0:. a. resins; .— 2251325 .8: 2: a: $6 2 8°62 22.. ~ .325” 5.3.: . I 'AYI' I j nunuua-u-u..o.-uuc-u..o:a..----..-anuu...os...--o --- - -- («.0 Own 000.cmn .00.." m :9}... 7:. .. atxw.. on --------------------- ..... -1- ...... --------- --- .9. .....o 2: 25...: soc; _ .35.... cos: a a an: 5: ..r.<. 0.... .u; :1. I... .III] III O- '91.. illl; '11 LY! [4 ll. 5 .I.II.I.TII . 11.]: .I- 4.! ..o>o .:.I. ..o>o .xcs. ..o>o .c.£. filfl.7>o .c.l. ..u>o .caI. ..9>o .c... asuhdunqaid a‘uut: orfiu ULzu zo—v....<..xu 3230 >h—u:.v.< o? o? 93 o? o? o? o? o? 2.2: 3583.. any unannoun o m m m.o a w m a 0» nodua acou90o ousuouos one owouu>< o2 o2 o2 o2 o2 o2 o2 o2 at 238828 couwuo oooum Na 0 Na 0 Nu m NA NA 0 Na 0 NH 0 Nu 0 NH 0 NH 0 NA - Nu a ma Nu 0 NH m§- w Anv acou:0o anon: . . oouuqocuv 2. oo. 2. 8. 2. 8. 2. 8. 2. 8. 2. oo. 2. oo. 2 7... $53 33.. 35.82 . N acuuoHHOu one m.n c m m.n n n m.n m.m 0» news: acouuoo assuage! ouuuo>< ououoz ooxoam ouo>uam ouonfim momma: noxomu ouo>uaw ouonuu auuoaooo odouuuom ouvcouw .m oumsaou .m ooquoam .111 .oqu ucuusuuousco: 00m summon Aoucosuuummm .u munch 64 Figure 24. Forest Products Laboratory Standard Flakecutter and Hammermill. 65 66 Exhaust C_\ c hne " Clup Um t; A I . ‘ ‘ \ N HrLes-el ‘ -. \ I Dfleuor _o-. ‘ .. . ‘ - Instrument “It Lu-va-l Detmtor 4%] Rotary mm.- 458 0|:QSICI RJdIO Achvr Lcsel Drlectur (tom. .\ Irv. “a Helmet Figure 25 . Double Disc Pressurized Refiner Model Bauer 418. 67 Green chips of E. grandis and E. robusta entered the refiner at a bulk density of 12 and 16 lbs/ft3 and moisture contents of 62.6 and 65.4 percent (0.D. basis). Wet fibers emerged at a bulk density of 1.5 lbs/ft3 for both species with moisture contents of 52.3 and 60.6 percent respectively. Fibers were dried in a small rotating drum dryer capable of drying 100 pounds of wet fibers per load. Hot air (about 2403F) was introduced through the tumbling fibers from the center; wet fibers were dried to less than 5 percent moisture content. Table 8 shows the distribution of particle sizes from the dry furnish used to make the fiberboards for the study. Data were collected from a Bauer—McNett Model 203-A Classifier. Table 8. Bauer-McNett fractions obtained from E. grandis and E. robusta (without bark) refined from green chips in a Bauer 418. Mesh designationl- —8/+14 -l4/+28 —28/+48 -48/+100 -100 Percent E. grandis 37 16.9 15.8 10.9 19.4 E. robusta 33.6 15.8 16.7 13.3 20.6 1Tyler Standard Sieves. All material passing a given mesh is indica— ted by a minus sign (-). Material retained on the mesh is indicated by a (+) plus sign. Looking at the two distributions it can be seen that the two species are very similar, and that a high percentage of fine material was present, approximately 20 percent in both cases. 68 2.1.1 - Fiber Blending. A treatment of 8 percent and 12 percent Urea Formaldehyde resin solids (Allied Chemical Two-Component Fiberbond Binder) was accomplished in a rotating wooden drum (see Figure 26). Fibers were tumbled through spray from a center mounted spray gun. Treated fibers were removed with a vacuum system mounted on a barrel. 2.1.2 - Mat Formation Appropriate quantities of treated fibers were run through a pilot forming machine and formed into a forming box 18 x 20 inches (Figures 27 and 28). 2.1.3 — Pre-Pressing. Mats were pre-pressed in a Riehle Testing Machine equipped with a floating load head parallel to the base of the machine (Figure 29). All mats were pre-pressed at a pressure of 330 psi. 2.1.4 - Hot Pressing. A11 boards were hot pressed by the "platen" or "mat formed" process (Figure 30). Press cycle was as follows: Platen temperature 2F 330 Pressure (psi) 450 Pressing Time (min) 9 Board Thickness (in) 5/8 For average closing times of different fiberboards refer to Table 7: experimental design and manufacturing data. Closing time was recorded for each board as the time period between reaching full pressure and reaching the stops. Thickness stops were used to control the thickness 69 Figure 26. Blending operation - medium density fiberboard. l I .‘ ‘1!“ w r i“ o 'U. 1w" 2 A ‘3; \' . .4‘99i-Wfi'3?‘ . g A! 71 Figure 27. Medium density fiberboard - mat formation. 73 Figure 28. Mat formation equipment - medium density fiberboard. 75 Figure 29. Medium density fiberboard - pre-pressing operation. 77 Figure 30. Medium density fiberboard - pressing Operation. 78 79 Figure 31 a, b, c, d. Illustration of particle geometry range. a. Flakes b. Wafers c. Slivers d. Fibers 80 81 variation between boards. 2.2 - FlakeboardsigSliverboards and Waferboards Manufacturing. The raw material left over from the air dried solid wood and most of the remaining bolts were converted into small blocks, soaked in water, and after saturation fed radially into a standard Forest Products Laboratory disc flaking machine. Flakes were produced with a nominal .020 inch knife projection controlling the thickness of the particles, and a distance of 1/4 inch in between scoring knives controlling the length of the particles along the grain. After this operation the flakes were hammermilled without screen. Slivers were produced with a nominal .040 inch knife projection controlling the thickness, a distance of 1/4 inch in between scoring knives controlling the length of the particles along the grain. After this operation slivers were hammermilled through a screen with 1/2 inch circular openings. Wafers were produced with a nominal .030 inch knife projection controlling the thickness, a distance of 1 1/4 inch in between scoring knives controlling the length of the particles along the grain. For the different particle geometries used in this study refer to Figures 31 a, b, c, d. Particles were air dried to a moisture content below 5 percent (Figure 32). Average values of flake dimensions, sliver dimensions and wafer dimensions measured on a random sample of about 70 particles per species are shown in Table 9. 82 Figure 32. Particles - air drying operation. 84 mo.H ~.n ~.m A3\Av ofiumm mmocumam mm 2.mH on Ae\av ofiumm mmmcumocmfim mmo. mmo. mac. A.cfiv mmochALH mm. mo. 5H. A.cHV nucfiz mN.H ms. as. 2.:Hv :uwcma mummmz muo>HHm mmxmam huumEoow mHoaupmm mfiwcmuw moua%amosm m©.m m.© 0.0 A3\AV OHumm mmmcumfim w.2m m.mH 2.wm Aa\ov ofiumm mmmcumocmam mmo. Hmo. «Ho. A.cHV wmmconLH on. mo. ma. A.:Hv nuoflz 0N.H so. so. A.:HV zuwcma muwmm3 muo>aam mmxmfim huuoEoow oaoauumm Amosam> mwmum>ov vumvcmuw mum momwzucouma CH moonE:z « HM Auoo.v Ammo.v Ammo.v Aa.~n~v Am.~mmv An.ooNMAw.HOHV A~.0ov Am.nm AN.~0mHv Am.~anv Am.mMH~V Amc.omHVANH.Hc~V Amm.ooHv o.HH me. mm. on. nqu «OBH omMH own One Hun 00H.HH cum.0H qo<.HH mNNH OcNH NHmH umoLm .mcoh .0cmm HHmpo>c .mcmh .0wm AHouo>o .wcwh .mma HHmum>c .wcme .0mm HHmoo>c .mcwe .vmm 2N0 Auamcm>muo onHommw Lowdwuuw umocm camcouum oHHmcmH ousumsm we msHsvoz huHUHummHm mo msstoz HHmuo>o .mchmuw mzumNHmozm II Amoo.v AH~0.V Amoo.v AH.0w~v no.mwwv Am.moHv Aq.onv Am.moHV Am.¢ou AoooHv A¢.0amHv Am.0ova Aw.0-0 A-.No~v Aac.m-v o.HH um. H0. on. wcoH 0HwH mHmH moo 0mm mom coo.~H 05¢.NH HNm.NH ¢~oH mQMH aocH umwsm .mcoH .0cmm HHmuo>o .wcmh .0mx HHmuo>o .mcmh .vmx Hkuw>o .wcmh .vmm HHmuo>c .wcoh .mom ANV nommEc>H~I, Afimav Asmav clone bu H afimav ac: Afima oooAV mo: .u.z NMH>ouo oHLHoomw :uwcwuum ummnw camcouum oHmeMH opsumsm mo msHsvoz NuHoHummHm mo wSHsvoz HHmpo>o HmosH m> cmoz_ .mmsutmmota coo: ofifiom .oH OHecH mumsnou wsummeusm 115 o—- E. grandis --" E. robusta 7 d I E grandis n = 50 - ' ’ = .50 [I \. := 6 1 I I "/|\ 5. i I X l\ t calc = 7.08 ° I I. [ ‘\E. robusta n = 50 C(X) 4'] I|I\| ;=,59 ' II \. u I II N 0‘ 0‘ ‘I\ M I...” on.” ’ (D II Specific Gravity (12% M.C.) Figure 35. Bending Specimens - Specific Gravity Distribution 116 ’—"' E . grandis --’ E. robusta 7 .. 1"“ E. grandis 2 : 1:3 6 . i : ‘. s = .055 5. I’ l V8 C(X) ' I I“: \E. robusta n = 19 4 I I I” ll \‘ 1.: = .61 ‘ \ s = .071 3 - I I - \ - \ 2' 'I " A \\ 1 ‘ I], |\. ‘\ '// I ! \. \ . - x 11 2' 3’4? 5' 6' .7 .8 .9 Specific Gravity (12% M.C.) Figure 36. Tensile Specimens - Specific Gravity Distribution 117 0—0 B. grandis "u E. robusta 8 1 7 4 [\E. randis n = 50 6 < I . [R E = .49 s = .062 I; (X) I 5 q . \ ‘I 4‘ .l {X ‘E.robusta n=49 I I, \i ‘K x j .57 3.. I i ‘ s-.065 2. i I I ‘ I'I'] '\ \ l d . " I I \\~ ‘\\ / / I \ 1 v #f I. '1 1“)?— 1 2 .3 .4 .5 6 7 .8 .9 Specific Gravity (12% M.C.) Figure 37. Shear Specimens - Specific Gravity Distribution C (X) 118 E. grandis E. robusta 8- 7w 61 E. grandis n = 24 f". a i = .51 I \ s = .066 5' 1%. 4- ' I I I \ I I, \ \E. robusta n = 22 3. I I J \ I = .60 . I k \\ S = .073 2‘ I o \ 1. I I \ / I - \ /// I \ \\ T‘ I 4‘1 TI 1 1"fiA—I .l 2 .3 .4 .5 6 .7 .8 .9 Specific Gravity (12% M.C.) Figure 38. Swelling and Shrinkage - Specific Gravity Distribution. 119 difference, which means the two species are different as far as average MOE is concerned. Linear regression equations were developed for MOE over specific gravity for both species showing a "good fit" (r- .77 for E. robusta and r = .765 for E.4grandis) (Figure 40). Distribution of MOE's is shown in (Figure 39). A covariance analysis was conducted adjusting MOE means over Specific gravity. After this adjustment means were compared at the 1 percent significance level, showing a significant difference. It can be concluded that the significant difference in between MOE's is not only due to specific gravity, but other variables that were not possible to exercise control over in this study also have a definite influence over the MOE of the species. 2.3 - Fiber Length. Fiber length was obtained for both species. Measurements were made on about 150 fibers after maceration by the Jeffrey Process and a t-test comparing the two overall means conducted at the 1 percent level of significance. No difference in this respect was verified, which means that fiber length does not contribute to the difference in MOE. (Figure 41). 2.4 - Tensile Strength Perpendicular to Grain. This test was usually carried out as an optional test because the stress is not evenly distributed over the minimum cross section. But it looks to be a good indication of the internal wood matrix resistance or the internal bond of the original solid wood matrix. 120 coausnfiuumfia hufiowuwmam mo moasvo: .mm muswfim Aama ooofiv mo: ooqm ooom coca OONH com 006 [I’llr‘o b - r b p 11 _ / / II III _ _ \n\\ x ./ n _ \\ w.oa~ u m I, . _ _ \\ a m a: u m ’ / _ x . on n c mumsnou .m” . u _ \\ / . . ll. ._ _.\\ \l . OH x _ . 38.2: / . . \ CC u I \\“ \ 8.~ u 23.3.5 . / . 8a a £86 2 mo.aaa a m . _ mama u m I, _ . cm H c mfivcmuw .m ./_\ I 0N «Jonson .m Ill-l 3?»me .m 6'... Modulus of Elasticity (1000 psi) 1168 700 121 “~4> E. robusta MOE = -Sl9.75 + 3269 SG r = .77** MOE = - 89.64 + 2755 SG r = .765 ** E. Covariance F(1, 97, .01) = 6.85 F = l4.51** calc 1 1 .2 .4 .6 .8 1 1.2 Specific Gravity in. E. grandis oo- 13. robusta Figure 40. Solid wood Modulus of Elasticity - Regression Lines 122 mama. mm. omH Asev numcma umnflm coausnfipumwa I mzumcog umnwm .Hs mpawfim osH. wNH. awfl. .ww. mo. we. Awm. “HI/II”; d_ " \x.\\\....\...r / m A, _ _ w o :. mavcmu m a, _ _ .~ 1OH // u _ «a ’V _ _ o \\ I _ \ m.N u Awam am.vu ON _ _ . x .52. n m /_\\\ 9N n 38“. as. n m omH u a mumsnou .mé\ tom .oq mum—#90“ o m "| 333» .m .l. Axv u 123 Tensile strength was measured loading the specimens, for both species, in the radial and tangential directions. Tensile strength was higher in the tangential direction for both species; in E. robusta about 33 percent and E. grandis about 40 percent higher (Table 10). According to Killmann, F. 1968 [41] the test carried out in this experiment is not true tensile strength perpendicular to grain but the so called "double cleavage" test. Results are 50 percent lower than the true tensile strength perpendicular to grain. Results obtained in this study are comparable to results reported in his analysis. Looking at the overall mean, E. robusta was about 20 percent higher than E. grandis. Compared by a t-test at the 1 percent level the two overall means are different. It can be concluded that E. robusta tensile strength is different than E. grandis. The distribution of tensile strength values around the means is shown in Figure 42. Linear regression equations of tensile strength over specific gravity were developed showing no significant relationship. The equations are as follows: TSpr = 28.5 + 716.7 SGr r = .42 f = 3.7 not significant TSpg = 360.1 + 43.2 SGg r = .02 F = .008 not significant where TSpr = tensile strength perpendicular for E. robusta TSpg = tensile strength perpendicular for E.figrandis SGr = specific gravity for E. robusta SG = specific gravity for E. grandis 8 124 o—o E . grandis 501 --- E. robusta 40¢ tealc 33.36 ’\E. grandis n = 38 t(.99,78)=2.66 / ° N i .. 386 I [.I \\ s = 101.8 3. . L' . 1 / \ , \ w” / ”I I I (x 10,000) I l I I \ \ 204 I . \ I I, I \ \ E. robusta n = 39 I I ' \ a? = 465 I I I \ \\ s = 108.4 1°‘ - ’ I \ \ ’ \ // I I . . \ LA’? I‘ I 1 \.\§— ‘ 200 400 600 800 1000 Tensile Strength I (psi) Figure 42. Tensile Strength Perpendicular to Grain Distribution. 125 No equations for tensile strength perpendicular to grain over specific gravity were found in the literature. These very low correlation coefficients could be well due to irregular stress distribution, difficult orientation of specimens, etc. 2.5 — Shear Strength Parallel to Grain. The ultimate shearing strength parallel to grain is related to the strength in tension, but the shear test is problematic due to superimposed, mostly bending stresses. Compressive stresses, stress concentrations and internal checks are other factors which may mask a clear picture of the shear phenomenon. Shear strength was measured in both species in the radial and tangential planes. Shear was higher for both species in the tangential plane, about 20 percent higher for E. robusta and about 23 percent higher for E. grandis. Looking at the overall mean for both species, E. robusta was about 8 percent higher in shear than E. grandis (Table 10). A t-test comparing these two means at the 1 percent significance level was conducted, and due to the large variability in results, no significant difference was verified. This indicates that a larger sample should have been used. The distribution of shear strength values around the mean is shown in (Figure 43). Linear regression equations of shear strength over specific gravity were calculated showing a significant relationship at the 1 percent level of significance. The equations are as follows: SSr = 16.7 + 2910 SGr r = .65** f 34.06 significant SS = 408.8 + 2320 SG r .52** f g g 18.05 significant 126 .COfiusnfiuumHo :Hmuo Ou HmHHmumm :uwcmpum ummzm .mq mpswwm Afimmv Hmaampmm :uwcmuum ummzw 83 BEN 88 83 83 com .8.» - I - O P D D \ .0 : II I/ / \\.\ / . \\ I . \ . 63 u .6. II . _ \ £2 u m I \ \ rm own : mumsnou .m/ . \ \ I x . / . x I x . 9.5 n 23.8.: I . \ \ .3 I / \ . 2a u 380 EN u m I . . 33 u m II \ x. on u c mawsmuw .m umH .8 mumfi—QOH .m 'I' WHUGNHw .m 0'. Aooo.oa xv Cc u 127 where SSr = shear strength parallel to grain for E. robusta. SSg = shear strength parallel to grain for E. grandis. 2.6 - Swelling and Shrinkage. Wood is dimensionally stable when the moisture content is above the fiber saturation point. wood changes dimension as it gains or loses moisture below that point. It shrinks when losing moisture from the cell walls and swells when gaining moisture in the cell walls. This shrinking and swelling can result in defects or performance problems that affect its use. It is therefore important to have a clear picture from where to depart in defining these variables for wood. Longitudinal, tangential, radial and volumetric swelling and shrinkage were determined for both species. Swelling and shrinkage were higher in every determination for E. robusta over E. grandis (Table 11). A t-test comparing the overall means for shrinkage longitudinal, tangential and radial of the two species at the l per— cent significance level showed a significant difference in all cases. The distribution of shrinkage values around the mean for both species are shown in (Figures 44, 45, 46, 47). Linear regression equations were developed for shrinkage over specific gravity, showing no significant correlation. In this case, no equations will be presented. In general, most properties, with the exception of shear strength parallel to grain, showed a significant difference in between species. 128 «a cm u .o.: quxucuuaolm so unassuugm "adaluom usu an vounqau~ou sudau_udasu who) sun =o>o cu :ooum mead—03a on. ououcuugm «cc .ucuuuuao no owuxcuugo «outs» on «cud—03m so amaze—uzo undocuucqu mo o_uI¢ I anxhn cc .a:o_uau>oo assoc-u. a». ooausucvuqn cu canals: a Aooo.v Ao.cv Am.~v Ac.nv An0.~w Ac.~v Aco.v An.qv Anm.~v nnc.v An.nv Ae.~v an.~v Amn.v fimso.v Ano.v an. c.n~ $.0u ~.m~ a.“ oe.s N.n m.n~ N.o~ c.~ co ad ~.o~ ~.s N.“ ~.n no “a nN. coo. suu>ouo huvso>o huvco>c anaco>o aueco>c huvco>o >uv=o>c uuvco>o xuvco>o Any any >uvco>c xuvco>o >chu>o xuvco>o Auv Anvsuvco>o auvso>o ouuuuoam 0» Ga On On On On Du On N a Ga Ga Du 0» N a Du 0» ouc~o>< cause .u.: RNA cacao .u.x NN- cacao .u.: NNu cause .u.: N- u: u: noose .u.: N- guano .u.x N- u: u: cacao .u.x "~— .u.m aquuo-!~o> unuucomcoh unuvox ouuuoI=~o> anxhn ~o—ucoucuh unavox nccuvauuucoa at any acuumoam Any ocnxcuuzm .auvcouu nammudouam Ansc.v Ac.vv Aco.uv Ac.ov Ac.av An.~v Aso.v as.nv Ano.~v Rom.v An.~v A-.nv Aoc.~v Aow.v Aoco.v A~no.v co. n.cn c.n~ ~.n~ «d n.0q c.o ~.- ~.- n.~ co «a o.- w.a n.o ~.c so an o~. «a. nuu>auo xuv=o>o subco>o auvco>9 auvco>n >u¢¢u>¢ auvco>o Auvco>c xuvco>o any ARV xuvco>c zuvco>o uuvco>o >uv¢o>o Any Auv auvco>é >uvco>o uuuuuoam on on ou ou cu cu ou cu cu o. o» ou N A ou o» ouuuu>< coupe .u.z "NH cause .u.: nag coouo .u.: u- coouu .u.: NNq Nu; ~UI couuu .u.x "Nu cacao .u.: N- u: coupe .u.z and .u.m 330632, 33:23:. :33. ouuuoguo> wrath a: 2.055... 12.3. nagging Any wcuuuuam ANV ounxcuunm .ouaaaou naumhauosm .au=~a> cam:— MMchuuzm vcthcuadoJm .a.m....a.,..s.a....=nafla 31.1.3 C (X) 129 °—° E . grandis -—— E. robusta 15. E. grandis n = 29 /\"|‘\\ i = .09 tcalc = 2'8 S/l’ / s - .03 10- t(.99,58) = 2. 66/ \‘\\ l/ 5. //// \\ E. robusta n = 33 I, \\ \\ a? = .11 /’ \\\ s = .032 /a’ kg, '10 ' \:;_:r>_-_' .02 .04 .06 .08 . .12 .14 .16 .18 .20 .22 Longitudinal Shrinkage (%) Figure 44. Longitudinal Shrinkage Distribution. 130 ~-—- E. grandis 30 q -—- E. robusta A , ‘ E. grandis n= 24 25 ‘ . . §= 7.2 I I \ s= 1.43 I \. tcalc = 3.71 I I I C (X) c(.9950)=2.704 . I I (x100) I I 15- . I I, I 1-, ' I 4 l \ 10. I, ‘\ , 0 ° I \ E. robusta n = 22 4, \1 \s i = 9.8 I; I, I 'I ‘\\ s = 3.22 5- , / I ’ I \ . ” \\ xf’ | l \ \ W i 1 ’q 1 \‘q 4 8 12 16 20 Tangential Shrinkage (Z) Figure 45. Tangential Shrinkage Distribution. 130 -—- E. grandis 30 q -—- E. robusta i‘ E. grandis n= 24 25 ‘ . §= 7.2 [I \ s= 1.43 20.I I \ _ ' I tcalc 3. 71 I I C (x) c(. 9950)= 2. 704 . l lSq I \ Il . \ I 4 l \ 10d / \I \\ E. robusta n = 22 I? § = 9.8 1’ l \I \‘\\ s = 3.22 5. II / I/ | \ ’f’ ' \\ / I l \ \ filL___‘-t" I 443-; ‘ “!I' 4 12 16 20 Tangential Shrinkage (2) Figure 45. Tangential Shrinkage Distribution. C(X) (x10) 81 7d 61 51 tcalc =53 t(.99,50)=2. 131 E. grandis E. robusta 704f\ I I\\ E. grandis 2: 2:1 [I l \f\ s = .55 I I l \ . i I], I l \ E robusta n = 22 ' I, II \\ i = 4 2 Figure 46. Radial Shrinkage (Z) Radial Shrinkage Distribution. C (x) (x100) 25 I 20 a 15 u 10 . 132 Figure 4 8 12 16 20 Volumetric shrinkage (%) 47. Volumetric Shrinkage Distribution. 133 shear strength not being different could be best explained by the inherent variability of specific gravities within the wood structure. Cell wall thickness and cell cross section dimensions are directly related to specific gravity of the wood and, together with ring widths and earlywood-latewood proportions, define specific gravity variations. In this study, specific gravity was determined not looking deep into the specific gravity differences within the wood structure. Practically the significant difference in between the average specific gravities of E. robusta and E. grandis in this particular study was not sufficient to explain all the differences in properties. In conclusion, when looking at differences in properties, where the species variability in respect to specific gravity, is somewhat large, a larger and much more careful selected sample and a more detailed look at the distribution of specific gravities within the structure should be taken. 3) Composition Board Properties 3.1 - Modulus of Elasticity and Rupture. 3.1.1 - Actual Values. Modulus of elasticity (MOE) and modulus of rupture (MOR) are two important board properties particularly for structural applications. It is well documented in the literature that both prOperties increase with board density. In this study MOE and MOR increased as board density increased for fiberboards, flakeboards, Sliverboards and wafer— boards at 8 and 12 percent resin levels, for both species E. robusta and E. grandis (Figures 48, 49). MOE average values range from 338,000 134 1.: :37... 121.: 1..“ nT—>( 7227:: 1.... 7.3.2... I a...:___,.> 25;: :32 .x.\ .CZLT. u ; 1t:cctu>__m a m n T. spacer—9.2;: u 3 Amsoxuv .3?ch undo: Amsiuv >395: 32.: o. x. N. o. m. o. x. N. c. m . P, D b b b F n b b \P ween ween ‘23 v 03 m :5... v . 0% \U 0 m nu 129» 1 Doc .2: . o2 —.J>U._ Cd m0“ Nm _..>:._ :_muz NN_ I wow—:55: .H._ 5950» ...._ (15d 0001) 30K 1:35 .mo*u~mcoc choc: cmuuo>< umo;u_: cc: umczc; u mc2~w> cam: xcz .oc ¢L:u_u vunc;uc£fiu u L vcwccuc>_—m n m mucosoxc~u u an vuaoaptbcz I 3 AnEu\uv xuumcwa vumoc AmEu\mv zuamcwn cums: Pa. L”. b“. Fm. Plo- xhwo ho a». mo :58 a 82 1 K i coo: “0a cooe mw S U 1o00m wOOOm accoo . coco 32: 1 82 u n... muvcmum .m H02: .593. Nw [III #264 so“; NNH I mumssou .m (18d) now 136 psi to 728,000 psi for boards made out of E. robusta and from 400,000 psi to 704,000 psi for boards made out of E. grandis. (Table 12). MOR average values range from 2,775 psi to 6,960 psi for boards made out of E. robusta and from 2,711 psi to 7,599 psi for boards made out of E. grandis. (Table 13). These values are in accordance to standard CS 236-66 1B2 for mat formed particleboard and NPA 4—73 for medium density fiberboard. It can be seen from (Figures 48, 49) that average MOE and MOR values for every kind of board made are very close to each other for the two different resin levels 8 and 12 percent. 3.1.2 - Regression Analysis. Linear regression equations were developed only for modulus of elasticity over board density for all composition boards and both species. Equations are as follows: E.ggrandis. Fiberboards - MOE = -242.6 + 1116.3 D R .83** - significant. Flakeboards - MOE = —215.4 + 1204.5 D R = .86** - significant. Sliverboards— MOE = -210.6 + 994.8 U 75 ll .94** - significant. Waferboards - MOE = 111.6 + 741.2 D R .46 - not significant. E. robusta. Fiberboards - MOE = -309.4 + 1091.1 D R = .87** - significant. Flakeboards - MOE = —382.3 + 1369-0 D R = .97** — significant. Sliverboards- MOE = -23l.8 + 1001.5 D R = .91** — significant. Waferboards - MOE - -159.8 + 1075.3 D R = .55** - significant. All coefficients were tested by means of an F-test at the 1 percent significance level. Most of the equations show a "good fit" in 137 10.! U. I --uvol¥.o-o--9i--¢J .......... -- alluulmrcla-ll.llnt--a-.us sang“ choonuouuz unaccuo>udm vucoaoxu~u vuaonuua«m on>h mecca .fl.uod§.&m Noe New .ms «"4 soc Nan cos ~14 Adan oooEV acne: vouaaav< mo: a a a a o a a a a o s a o - _~ o~ any ocuucoo otauu_oz Aao.~hv A~.eoo Aca.¢nc Aoo._eo Ana.~cc An~.eev “no.0no .No._~v An~.oev .hn.~o. An.nov Aom.~ov a~m.a~c1oq.~ev Aeo.~ov A~e.¢~c Nae sec ago o_n Non one nae can QNN cam -o has sen ads doc Non 11.; coonv mo: Ammo.v Aoao.v Acuo.v An~o.v Aano.c Ammo.v Am~o.o Aaco.v Ammo.v Ammo.v Awno.v Ammo.v Aneo.c Am~o.v Aono.v Ao_o.o ameu\mv as. as. as. a». a1. mo. ~a. on. an. no. as. an. o“. «o. «a. No. suancoo _.:ou< - - a a - Na a a Na N. a o - - o a ARV ~o>~s c..o¢ VbflOAhUums; vhco.tt>‘—m thOLUxuum flungtvnr— anxh flung _mo:~a> cm»:— iti I [111.11 $111.11 ¢l|i|3 N._c_umm~m ho m:—:13: .Wo_mctmcp: ot€¢: coqu‘momfims .N— o—c1+ .mumanOuKmsumNmmmmw 138 n m a a a a a o a a a o 3 2 Z .2 A5 2.3.8; 3330: Am.o~myaa.omn1Ao.cqov Ao.oo¢v A-ov Ao.~eNVAa.s~oVAe.~<~v A~.ncsv Au.omev Acme—v A~.ncmV Am~o~yA~.mcmv Ao-~v an.ooo_v oo~c suom anon omen "can e~a~ Noam -n~ Oman ance «do: amsm oomn once came «an: Aunmv so: Auuo.v Auno.#An¢¢.v Anno.v Aocc.v Annc.v a~mc.:A0mo.v Am~c.v Asnc.v Aaqc.v Ao~c.v noeo.ya~mo.v Aoqo.v Anco.v nauxw nu. ac. an. we. as. no. «N. go. as. no. as. no. us. no. Nu. ~0. >uuocua dosuu< - Nu a a Na - o 0 Nu Na o a Na Nu m a ARV ~u>ua canon vunonuouaz vuwonuo>uum vuaonuxoam vHQOAuonum oa>h vane: .auvceud‘azume-oom a a m a a a a a a a n o a Na nu ed ARV acoucou wuzuauo: as.~em_ Acmnv Au.nn5v Ao.-mv an.cnnv Ao.~moy Achv AN.anov an.m~ev A~.nqcv Aa.monv A~.Necv Auwnv Ac.~oov Am.wn-v An.conv «nae o¢an nnOn cc~m osco oaou ween cmnw comm nnmn cuec mm- oooo Q~On nonn anon Admacaot anno.x Aadc.uao~o.v Ammo.v Auno.v Ammo.v Anne.“awao.v An~0.v Ammo.v Awno.v Acno.v Aneo.uno~o.v Acno.v Acuo.v neo\w as. as. as. no. as. me. NR. an. an. no. as. an. cs. co. as. «o. muumcoo unsuu< NH «a o a Nu Nu m a - Nu m c - Nu m a Any ~o>oa canon vuaonuuuuz unaccuw>dum uuaoaoxn~m unnoauunuh wash venom .aumanou nauqumoam _ao:—c> coo:— Mmmwmsz uo ma—svmm .mwuutmeEm whooa :ouuuaomeoo .m— o~snh 139 relationship to board density. Composition boards at constant average density possess higher MOE values as the wood density decreases or the compaction ratio increases. Due to large variation in process variables, it is impossible to manufacture two equal boards to the same constant average density. 3.1.3 - Covariance Analysis. In order to make a valid comparison of the two species composition boards, all prOperties had to be adjusted to a constant average density. This was possible through a very valuable tool called covariance analysis. This adjustment was made at three different stages to allow comparisons of the effect of Variables like species, resin level and particle geometry over the different board properties. At this point only the adjustment of modulus of elasticity over board density is going to be discussed. At first, modulus of elasticity was adjusted over board density to allow comparisons of the species effects on this prOperty. The first conclusion can be drawn from Figure 50: as the density of the species decreases modulus of elasticity for all composition boards increases. When these two adjusted modulus of elasticity means were compared to verify if the species effects were significant at 1 percent significance level, only the fiberboards at the 8 and 12 percent resin levels and flakeboards at the 8 percent resin level were significantly different, this means that as the species specific gravity decreases for fiber- boards (8 and 12 percent resin levels) and flakeboards (8 percent resin level), modulus of elasticity increases significantly (Figure 50). .mum>~wc< oucmwum>ou c >u.0.umc_: mo m:~::cz .v .u .; .ch oczuwu .auumsoox wuuuuumo 0» wow .~m>m# cameo ea 0:: moucoumuuuv new voumwk I a mmocwuouuus u0u ccumck a u .mouuonm od 0:: moocoswuuuv nob woumck I m memo: roamspc< I < >ua>-0 ouuuomam mwuuwam mum: o» .m om. >u~>muu ouufiooam mo.umnm n cmcu . mum: on . m cccu .. «v on. m Mm. m «c on. m zuu>mpc ouufiowom mw_ocom sud>cuu o_c*omam mwuooam «an: o; . m cccu .. mum—AOL .. m cmcu . mm. m «u Om. L m . u «v cm. m woos woos . aces .ooe 214C) fiOOm fioOm woom seem ill (18d 0001) 30m pansnrpv «I (159 000I) 30w Pansnrpv (Isd 0001) 30K P9350§PV (18d 0001) 30K pansnrpv wooo 78 WV.” 58 38 § \ \S foam fican fiooh co“ .muumwuo.mwuowam cob wocmouuucmdm on come mwcwfi uxuumuum .muquomm wfiouuuma mam H~>o~ :«mmu am mmucouwuuvu Haw acuumou umuum ocean vofioom .Ilzll H0>mws~ CHQUK NC 0"- etmoruwsaa u a vhfiOn—Nv—Nfik I H...— vuwosuo>u~m a m vuuocswumz n 3 1:6.— cwmmx NNH III 141 From this important observation if we were supposed to draw a graph like the one developed by Klauditz [39] (Figure 51) for modulus of elasticity at the 12 percent resin level (Figure 52), it could be concluded that the effect due to species gravity over bending strength, MOE or any other property is not necessarily true for all composition boards. In this study for example, Klauditz' relationship is only true in the case of fiberboards. This could mean that fiberboards are more responsive to anatomical differences within and between species as far as influence over its physical properties are concerned. In the second stage MOE values were adjusted over board density to observe differences in between the 8 and 12 percent resin levels for every kind of composition board. Higher adhesive contents normally increase modulus of elasticity. This is the case for this study: as the resin level increased from 8 to 12 percent modulus of elasticity increased for all composition boards. When these two adjusted modulus of elasticity means were compared to verify if the difference in resin level significantly increased this property at the 1 percent significance level, only the fiberboards and the waferboards increased significantly due to the 4 percent difference in resin level (Figure 50). After the second stage many results were pooled. In the third stage, MOE values were adjusted over board density to observe differences due to the four different particle geometries. In general, as the length of the particles increased, modulus of 142 IOOOO 8000 6000 4000 2000' BENDING STRENGTH — PSI Figure 51. 0.5 0.6 0.7 0.8 BOARD DENSITY - G/CM’ Relationship between board density and bending strength of particleboard made from various species (Klauditz and Stegmann, 1957). 143 mumsnou mumanou mumonou .m sum .m new .m UGG Eoum mvme mwumon cowufimoasoo mo xufiowuwmam mo .mumaaou .m can wwwcmuw .m msaswoz mam hufimaww wumon amwzuon afiLwGOHumHom .Nm muowwm AmEo\wv muamcwa vumom a. m. pm. bw. Ln. .ooq numonumnfim fl. rbwumonum>wam .oom wumonuwnfim \. q. mfiwcmuw .m? \\ mumanou .m? \ \lkumoamxmam rooo \. mfivcmuw .mnu \ \( mfiwcmuw .Mo1.\| wumoauommz r con mwvcmuw .mmox\ (18d 0001) son 144 elasticity increased. When these adjusted modulus of elasticity means were compared to verify the difference mentioned above, at the l per- cent significance level, a real difference was obtained among all composition boards. The analysis for particle geometry in some of the cases was made with the pooled lines; waferboards having the highest MOE and Sliverboards the lowest (Figure 50). The low MOE of the Sliverboards could be explained by the breakdown in width after hammermilling and somewhat because of the high thickness of the particles. Density profile was determined in all boards (Figures 53, 54, 55, 56). Not very much difference can be observed, this fact could be expected because the process parameters were kept quite uniform. 3.2 — Internal Bond. 3.2.1 — Actual Values. The internal bond (IB) or tensile strength perpendicular to the board surface, is a widely determined property and a very controversial one in terms of the analysis of results as is well documented in the literature. In this particular study, looking at the overall results, in general there is a tendency of internal bond to increase with increasing board density for all composition boards at 8 and 12 percent resin levels for both species E. robusta and E. grandis (Figure 57). Although this tendency is not a very clear one in all cases, internal bond average values range from 137.3 psi to 219.09 psi for boards made out of E. robusta and 118.5 psi to 223.7 psi for boards made out of E.#grandis (Table 14). These values exceed the minimum values of 1.27 Y g... O H Relative Densit H O 145 82 Resin Level robusta (high nominal density) robusta (low nominal density grandis (high nominal density) I new!“ grandis (low nominal density) od——————— Figure 53. .2 .3 .4 .5 fi 1 1 1 Relative Board Thickness Fiberboards Density Profile Relative Density 1. 2 146 8% Resin Level __- E. robUSta (high nominal density) -—m-—_ E. robusta (low nominal density) ~---' E. grandis (high nominal density) E. grandis (low nominal density) Figure 54. Sliverboards Density Profile 1 .3 .4 Relative Board Thickness Relative Density 1.2 1.1 1.0 147 d . robusta (high nominal density) . robusta (low nominal density) . grandis (high nominal density) l'. |-!. P! n: tn E. grandis (low nominal density) 4 J .1 .2 .3 .4 .5 Relative Board Thickness Figure 55. Flakeboards Density Profile. Relative Density 148 ----—- E. robusta (high nominal density) ._-_. E. robusta (low nominal density) -—- E. grandis (high nominal density) -——-- E. grandis (low nominal density) Figure 56. 1 1 T .1 .2 .3 .7. .§ Relative Board Thickness Waferboards Density Profile 149 .mowuwncmc cumo; cumuo>m unocuus use umozo~ I mc:_=> cue: tcc: .ccumucu .mm cu:u_m cumcsue£«u I k vpmogum>_~m u m vsmoxwxmuh u an up¢¢£u05m3 I 3 AmEo\uv auwncoc venom Ansoxuv nuance: vumoc a. m. N. be. m. a. w. h. N0 m. am so... .2: 1 32 m. a J m \ I 32 a \\ nah 1.2 m x m p ., f... m b 28 0 .SN ~96; Emma um .llu' ~96.— cummm NNH l nuvcwuw .m mumssou .m (1sd) puog TEUJBJUI 150 assay nos so" has «a. nos ass «as Nsa use»: cou.=.v< .— o a o a m o a o o o o a s o s B Any ueuscou message: A«.—~v “~.o~v Aa.c~v a~.mqv Ao.nnv Ao.on A~.wnv Am.-v An.cv Ao.~v Aw.~_v A~.ov Am.anv .o.m~v Ao.nuv afi.a~. Aqua. ~.~c~ ~.~o~ ~.oo~ o.qo~ c.o- ~.~w~ ~.ec~ o.oo~ ~.n- o.mo~ n.n- o.no~ m.o- ~.~¢~ o.oo~ m.m- neon ~1¢u0u¢~ Anno.. Aoo.s Aso.. Aamo.s loo.s Aoo.e A-.. Aamc.. -- Ammo.c Anno.. Ammo.v Asso. a~mo. Ano.s Amo.s Annu\u. co. co. KN. no. ~N. No. as. on. ~m. no. «s. «a. as. «0. ms. ea. sud-coo unsuu< - - n Q ~_ Nu m a «A mu m a nu - a o .uv ~o>oa suave vunonuounz theonuu>qdm vuwoaoxmdu vuaonumnum umxh vu-On .muvcasd,n:umh~moam ~o~ ms A_mav a a a o a a a a o a a a h o 5 c ANV acoscoo 9.:unao: Ada.sv Ao.~ev Ao.onv Ao.nm. Ao.-v An.a~v Au.ov Am.n~v A~.c~v an.av A~.c~. Ao.n~v Aw.m~v Am.n~v an.-v Ao.n_v aqua. e.-~ ~.~o~ n.nm_ c.wo~ ~.-~ ~.~a_ a.-~ ~.om_ a.o- c.eo~ n.o- e.no~ ac.o- o.an~ ~.~a~ n.~n~ econ unseen:— Ano.v A-o.v Acco.v A-o.v Anno.v Adso.v Anuo.v A-o.v Amno.v Amc.v Aoo.. Acmo.v noeo.v Annc.v Aneo.v Ahno.v Asu\x. co. co. am. No. gm. no. as. on. «a. ~o. as. on. as. do. «n. do. asuncoo ~n=Uu< «a «A m o - N_ a a N_ Na o a - Nu m a Anv «0:04 swam: vueonuouuz memocco>_~m vuaonuxaum unconsoA«m verb mecca .mo_uczummw memo: :o_uunomsou _n0:~m> caut— (I'll!!! when assuage— .em o~;ck .mumsa0u naumh~muam 151 CS 236-66, 1-3-2, for mat formed particleboard and NPA 4-73 for medium density fiberboard. 3.2.2 - Regression Analysis. Linear regression equations were deve10ped for internal bond over board density for all composition boards and both species. Equations are as follows: E. grandis. Fiberboards - IB = -l39.6 + 427.5 D R = .64** significant Flakeboards - IB = - 47.9 + 344.7 D R - .86** significant Sliverboards - IB = 45.2 + 224 D R = .62** significant Waferboards — IB = 159.3 + 49.1 D R = .18 not significant E. robusta. Fiberboards — IB = -88.3 + 391.2 D R = .85** significant Flakeboards - IB = 56.4 + 201.3 D R = .78** significant Sliverboards - IB = 7.3 + 271.2 D R = .80** significant Waferboards - IB = 65.9 + 169.9 D R = .37 not significant All coefficients were tested by means of an f-test at the 1 percent significance level. Most of the equations show a "good fit" in relation- ship to board density with the exception of the waferboards. 3.2.3 - Covariance Analysis. A covariance analysis was deve10ped for internal bond in the same fashion as for modulus of elasticity. At first internal bond was ad— justed over board density to allow comparisons of the species effects on this prOperty. The first conclusion can be drawn from Figure 58 £1 = 152 noosuuouuav you nuance I a .auuoeoow oquuuuwm Ou mac auw>auu uuuuoumm moquomm monsoon .m mavens» am. On .m j ""'I|l .\ .III.III.II.IIL 1.. k" H. (rsd) puog teuzanuI paisnfpv .~o>u~ sumo» cu use nuocououufiv sou assume I u >u~>uuc ouuuummm mwuumam nunsnou .m savanna .m mm. on. .zuumaomo mauuuumm van Hw>wa suaosuossc vuaonmxmdu .m«n>~m=< coco—um>oc meow uncumucu .v .u .; .nmm shaman .nmmooam cu mam moucmuwuuuv so» wouwob I a sumo: vmumanv< I < >uq>euo uuuvommm muuowmm >uq>auo ouuuooom mmfiuoem nunanou .m mavens» .m numasou .m mavens» .m on. em. on. an. n n m w n D. W. W. 1.2 m .2: m 32 m 32 n n ., n u m w m M... u m. u m. a» w a w .nIII w v-|l'||ll \.I v||"'ll|l) \\\ .m h\ .d 1.\\ gm MW 5 S o .v a n u up I Va. alol-lol .I' fi SN b3.” 58 52 1 SN _\ annex cu mmocmuouuav uom wcuumou umuum mun“ wouoom .llzlll d0>01~ afiuum N” '|' m mumocuo>uum u m an uumonumumz u 3 Ho>ma :«nmz NNH nIlIlII 153 no clear relationship exists as far as the specific gravity of the species are concerned for the composition boards at 8 and 12 percent resin levels. When these two adjusted internal bond means were compared to verify influence of specific gravity of the species at the 1 percent significance level, only the fiberboards at the 12 percent resin level were significantly different. This means that as the specific gravity of the species decreases, internal bond decreases for fiberboards (12 percent resin level) significantly (Figure 58 b)- In the second stage 13 values were adjusted over board density to observe differences between the 8 percent and 12 percent resin levels for every kind of composition board. When tested at the l per- cent significance level, only the fiberboards increased significantly due to the 4 percent difference in resin level (Figure 58c). After the second stage of testing, the non-significant lines were pooled together. In the third stage 13 values were adjusted over board density to observe differences due to the four different: particle geometries. When the adjusted IB means were compared at the 1 percent significance level, no real difference was obtained among any composition boards, this means that in this study particle geometry did not significantly affect the prOperty internal bond (Figure 58 d). So only the fiberboards were affected by changing variables as far as internal bond is concerned. 3.3 - Linear Expansion. 3.3.1 - Actual Values. Linear expansion like MOE and MOR is a very important property when panels are used for structural purposes. Some scientists have 154 found linear expansion to increase along with increasing board density, some have found no clear relationship. In this study, no clear relationship between linear expansion and density exists for all composition boards at 8 and 12 percent resin level for both species E. robusta and E. grandis (Figure 59). Linear expansion average values range from .124 percent to .397 percent for boards made out of E. robusta and from .145 percent to .344 percent for boards made out of E. grandis. Most of the composition boards are in accordance with standard CS 236-66 and NPA 4—73 with the exception of the Sliverboards which exceeded the maximum average allowed (Table 15). 3.3.2 - Regression Analysis. Linear regression equations were deve10ped for linear expansion over board density, for all composition boards and both species. Equations are as follows: E. grandis. Fiberboards - LE = -.014 + .276 D R = .75** significant Flakeboards - LE = -.047 + .290 D R = .63 not significant Sliverboards - LE = .368 - .051 D R = .05 not significant Waferboards — LE = .028 + .234 D R = .19 not significant E. robusta. Fiberboards - LE = -.079 + .343 D R = .73** significant Flakeboards - LE =.00009 + .321 D R = .43 not significant Sliverboards - LE = .199 + .163 D R = .19 not significant Waferboards - LE = .123 + .164 D R = .19 not significant IL55 .mmwuvucoc cacuc>m umoxu—: ccc um03c~ I mo:~e> cmoz cc_mccaxu Lccc.; AmEu\wv humane: vunom armcELogsa n e etacsmx¢~u a .1 AmEo\uv >u*mcm: memo: .om cuzxfim TLwCLLc>__m u m tcecscoue3 u 3 o. m. N. o. m. o. m. N. c. m. a. woos. 1.8. h 3 «V .1 \I\\ m. mVnIJ lthHI % J \W\\ \\ 15w. 3 $8. LKn. \\ e .\‘\ w t .. gs . con. u .ocn. \\lllI mu no III ”AK . ‘ O s cos cos mavens» .m ->us gamma Ne 7:5.— =~m~¢ NNH ll mum:Lou .m (z) norsuedxg Jeauxj .1556 «3. of... nun. 3m. . . . . . 3: 5.2.: 253:! msu nm~ om~ «Nu couuceexm saved; c.o~ o.- Nu ~.d~ ~.- n.- m.- e.- ~.c~ o.n~ o.o~ m.o~ _.n~ o.o~ a.n~ c.¢~ N n~v acousou assuage: o.o o.» o.o a.» ~.a ~.m 0.x ~.¢ «.0 a.o a.» n.m ~.w o.» ~.n a.n u adv usuucou ossuoqoz Ado.v Ameo.v Ao~o.v Aano.v Ammo.VAoco.v Anu.v Aao.v ANmo.v Asuo.v Am~c.v Anqo.v Aeco.v Aouo.v A-O.V Aemo.v co annex. have «as. -_. «RN. cos. ~Nn. m~n. «en. “an. as". use. Nes. _n~. mes. ma_. wON. mes. 1". a 1 as Anwo.v Aooo.v Awo.v Aooc.v Ammo.v Anoc.v Aooo.y Aoec.v Acoo.v Anno.. Amec.v Ao~c.v Ammo.%qnc.v Anc.v Aco.v Amsu\wv ms. 05. s“. 00. ms. no. as. no. «a. no. on. «e. co. co. um. no. undocun ~I5uu< «a «q a m «a «u a a «a «a m o - Nu m a ARV ~o>od sumo: unaccu0uu3 vhaonuo>u~m vuuoaoxuam unaccuonuu cash venom .nuvcmnd usumawaunm o-. cum. now. can. «mm. mam. «Nd. and. Auv acne: vuuaaav< sounsoaxm have“; s.c~ m.- n.- ~.- ~.- o.- Q.~N o.- a.o~ e.- n.o~ m.- ~.o~ o.o~ o.o~ o.o~ NA~V useucou wean-do: o.m o.o ~.m 0.x ¢.o c.a o.w m.o 0.0 u.o ~.o o.o ~.m n.o o.~ ~.~ NAnv usuucoo assuage: Acmo.v Aomo.v Aaac.v Am~o.v Anno.v And.v Awmo.v Aomo.v Acn.v Aonc.v Auuo.v Aumo.v A-o.v Acac.v Acno.v Aqnc.v ANV :oqnsemxm nous“; ~n~. Now. «RN. m-. mom. oo~. non. nn~. o~n. man. oHN. cum. nod. ova. on“. c-. a~mo.v Awmc.uadmo.v Ao~o.v Aac.v Acme.v nomc.v Acco.g ANo.V Aomc.v Asno.v Audo.v Amno.v Ammo.VAeco.v Awno.v anEu\mv am. as. an. no. no. co. KN. uo. mm. ac. an. do. as. on. cs. no. auuocoo nasuu< «a Nu a a mu - a a Na - a o N— - a a ARV ~u>oa canon uuaonuuum: unaccum>umm vuaonoxaqm vsmonuonqb uo>h venom .mun250u onumfiwwunm .n0=~a> snot. codenammw news—4 .myuuummmwm venom couuumqmsou .n~ o—csh 157 All coefficients were tested by means of an f-test at the 1 percent significance level. Most of the equations show a "bad fit" with the exception of the fiberboards for both species showing a good correlation to board density. Even where correlation is high, practical signifi- cance is very low. 3.3.3 - Covariance Analysis. A covariance analysis was deveIOped for linear expansion looking at the same three variables effects, species specific gravity, resin level and particle geometry. At first linear expansion was adjusted over board density to allow comparisons of the species effects on this property. The first conclusion can be drawn from Figure 60 a: no clear relationship exists as far as the specific gravities of the species are concerned for the composition boards at 8 and 12 percent resin levels, When these two adjusted linear expansion means were compared to verify influence of specific gravity of the species at the 1 percent significance level, only the flakeboards at the 8 percent resin level were significantly different. This means that as the specific gravity of the species decreases, linear expansion in this single case decreases significantly (Figure 60 b). In the second stage LE values were adjusted over board density to observe differences between the 8 percent and 12 percent resin levels for every kind of composition board. When tested at the 1 percent significance level the flakeboards increased linear expansion signifi- cantly due to the 4 percent difference in resin level (Figure 60 c). 2158 .m«n>~mn< mocc3cm>ou cofimcmexm came“; .v .u .2 .s cc ounwum .huuosouw «defiance cu 03v ~o>m~ same» 0» 02v .mmuomam On one nousouuuuuv new pounce I a nwucououufin com pounds I u mousoumuuvn you pounce I a menu: counsaa< I < >uu>auu uuuuuoum nwuoomm aua>nac ouuuomam mouoomm >uu>muo ouuuuoam moquwew auq>wco ouuuusmm mmuoonm unannou .m mavens» .m nunsaou .m nqpsecw .m uncanny .m numCMuw .m «unseen .m mavens» .u m. on. on. an. an. on. an. an. 53. 52. v2. v2. \1 \\\\ rI:II:MMHuuuVAn:III r OCN . \‘AKI ‘3 Q Illllllllllnlnlnl \: \\ \\ runuuummmw .\\Q "l||l \‘ rI.....lII.|||L u, 28. \I s. 1.... can . FIIIIIIJ \‘ r can . o2 . hi .7 .IIIIH'IIII (z) notsuedxg xeaurq paisnfpv (z) uotsuedxg Jeaurq pansnfpv (z) uorsundxg Jeau11 pansntpv (z) norsuedxg Jeaurq pansnfpv . 8s . Joe. . :3. cos. .nuuouuu «nausea new cosmouuuswuw on some nucufi unwaauum .xuuoaoow vacuumed new Ho>o~ swoon ca moucououuuv uou mauunou umuum wosua vonoom Iluzll 32.3 52:. «a III. numoAuonfiu a m vuaonuu>«_c I m cunonoxodu I Am vasoauoun3 I 3 Hm>oa sumo: NNH _lIIlII .m«m>~m:< woes—um>oo couwcmaxm cmm:*4 .v .u .2 .m 0c munwum .nmuuoam cu one mwucououuvn you manner I = .>uuoaoow odouuumm cu 03v nousououuuv you wounoh I o ~m>o~ Games 0» o:v neocououufiv you pounds I u mean: voumsem< I < huu>uuo cauuuomm nmuuoew >ua>ouu ouuaumnm mwuooem auu>muu aquauuem mouovnm >ua>oco uuuuosnm mouooam 1138 7"- "||l \‘ [ol-l.|ol u: (z) notsuedxg Jeaurq pansnrpv .con. 3 ace. '|'I""l \. III.III.III.III:II, av (z) uorsuedxa JBGUIj paisnrpv can. \I face. my luuuusIIIJ (z) norsundxg 123011 pansntpv 1.... as (z) uorsuedxa Jeaurq pazsntpv .ouomuuo wwaoomm ecu cosmoauucwun on some mocam usmfiauum .xuuoeoom oaowuuma can a0>oH swoon nu nuusououuqv sou mewunou nouun nosufi vomoom Ilzill eumoncussc I m anmosuxaam I 1m vHQOAuw>udo I m vuaonuouaa I 3 30>»; :aamx Na sunspou .m nausea» .m oumnaou .m aumcecw .m sunsnou .m mwvsauw .m mumscou .m aqvcmuu .m m. an. on. on. an. em. on. cm. .ooq. recs. you. van. Ln \1 ‘1 .0 ‘1 \ \ """ ‘1 ,rl:ll!nll.ldu1lrall. nu \\m . c2 \\\.4 a. 38 \\\i 3. as as. r .1, mm .M\ 8m . ace. Hm>Oa~ CHWOZ NNH I 159 After the second stage of testing, the non—significant lines were pooled together. In the third stage LE values were adjusted over board density to observe differences due to the four different particle geometries. When the adjusted LE means were compared at the 1 percent significance level, only the Sliverboards had a very high and significant linear expansion. This very high linear expansion could be well explained by the breakdown in width after hammermilling and somewhat because of the high thickness of the particles (Figure 60 d). This is not in complete agreement with Bryan [7] who found that as the length of the particle increases, LE decreases. 3.4 - Thickness Swelling. 3.4.1 - Actual Values. Thickness swelling is another very important property when considering most of the uses of composition boards. It is well documented in the literature that there is no clear relationship of this prOperty to board density. Sorption curves for thickness swelling for both resin levels, densities and species are shown in Figures 61, 62, 63, 64 . From the figures we can see that the increasing size of the particles increased thickness swelling, and that in general as resin level increased, thickness swelling decreased for both species. In this study looking at the overall means, there is no clear relationship between thickness swelling and board density for all composition boards at 8 and 12 percent resin level for both species E. robusta and E.vgrandis (Figure 65). 16f) .Aam>m~ :«mmu Rm I mnemos cauuamoeeoo mvvcmum .wv .mo>u:u scuunuom I acquaoam mmocxo—sb .ae musmwm vuuonuwnuu I m vuaonuo>uum I m vumosoxmau I an vasonuoumz I 3 Any uncucou musumuoz ANV ucoucou eununqoz ON ea NA a c on ca Nu m e I e we mes # w an Is u. T. r- m. sud m .uvnIx n m. n \8 1 3 F 1.: TL 7? m x m ob sou {a n. weu sea ram rmN unmucoU summu Rm xu ment ~m:_Eo: :23: ucmucou Canon Na >uwmcwv Amcch: 30a manna» .m mwmcmuu .m (z) 8n111ans ssauaatqi .A->0~ cums» NNH .mmumos mafia—MOLEOU m_=:mcu my .mm>u:u cc.ueuom I uc_-ozm mmocxofish .Nc munumu lfil vumCLumcau u u accescc>‘_m I m eu60£mxc~u n ~u cuccccebmz I 3 Auv acousoo wusumwoz ANV unsucco venue‘s: ON ca NH m c Cw c~ Ne w e wnw we we 1 m” 1 w 3 w w w a a u m 12“.. #2 I I .m «H w oarx 4 ed ch fioN a «N w an r r ucoucoo canon Nmfi I zofimcov stfieoc sum; mN acoucoo Canon de I >uqmcmv Hes—so: 3oH w~ muvcmuw .m mavens» .u (2) 80111905 IsaUIDIQI .A_;>;_ C—mvu N3 I wanna; :o_u_moasou num:50u .xv .w;>L:u :c_.;c:m I u:_—_03m xx;:x;_;+ ..: Lu:x_; 1152 tczcscm;_m u a mc332s0>_~m I m accosmxe~m n ~m mucosuemea I 3 Auv ucmucoo zunumuoz ANV ucmucou munumuoz ON c~ Nn m c ON c~ Na a e b P * Ir F b h D b o o .q fie r» in l u. T... 9 .NH m. .NH 3 s s c. n a T: n 1: I . u .\ 8 P mu m. I. 3 ram cu 1cm .eu va va ucmucoo :«mmu N@ I Auwmcwv Handsoc 5w“: ucmucoo swam» Rm I auwmcuv Hacueo: 3o~ mumnso» .2 mumnsou .m (z) auTttaflS ssauaatqi 1133 A o~ .A~m>o~ sysop NNH I mvumo; couuumomscu eun2£o» .mv m0>c=o cc_uauOm I acumfiozm mmocxo_;k .ec change NV ucwucou ousumuoz ca Nu acoucou cine» NNH I hufimcov accuse: xx“; numnaou .w w a 1N~ veN ram (2) 8n111305 ssaunatqi accostw;_m I t etmonmxuac I 3e any acoucou ousuwfioz om en N— w vuwccto>uum I m vuwoscouna I 3 'e occucoo sumo» Nm. I >u~wccv Headsoc aofi mumspou .m Va ‘v Q wag j 0 —I (z) Surttans ssaunarqi fl 0 N wew 2164 3 A 2, \(i '>. I, \ (z) SuIIIans BSBURDIQI \Q 0 MS J \5‘ .mmuuumcov vumon mumum>m unocuaz vac unkaH I mw=~m> :ccz uc¢dficam mmoCJomzh .mc scam—m utmo;»0:_m I a vcmcctc>¢fin I m vswocoxmpm a fig vnccztemma I 3 mnEo\mV auqnswa venom AnEU\uV >u«mcmn memo: p m. m. m. m $0 ”0 N. c ALI. L‘ATJ I m \nIJ K3 1 CH \¢¢ \ \ k \\ \ \ v \\ _c\ d. AI \ ..cN m. 31:: 5mm» Nw III H02: 5mm» N: I savanna .m mumnnou .m T m V c> H ssaunorql j m ... (z) 8n1118ns if C N 165 Thickness swelling average values range from 6.65 percent to 18.52 percent for boards made of E. robusta and 5.77 percent to 20.42 per- cent for boards made of E. grandis. Thickness swelling is a prOperty that is not covered by the standard CS 236-66 or NPA 4-73 (Tables 16, 17). Compared with commercial particleboards in the study made by Suchsland, O. [71] some of these average values, like for the sliver- boards, waferboards and flakeboards are relatively high. 3.4.2 - Regression Analysis. Linear regression equations were developed for thickness swelling over board density for all composition boards and both species. Equations are as follows: E. grandis. Fiberboards - TS = 8.5 - 2.4 D R = .13 not significant Flakeboards — TS =21.0 - 9.2 D R = .38 not significant Sliverboards - TS = 8.7 + 9.00 D R = .34 not significant Waferboards — TS =12.4 + 4.5 D R = .08 not significant E. robusta. Fiberboards - TS = 9.8 - 2.4 D R = .12 not significant Flakeboards - TS 316,0 — 2.8 D R = .18 not significant Sliverboards - TS - 5.5 + 16.3 D R = .74**significant Waferboards - TS =12.7 + 5.4 D R = .12 not significant All coefficients were tested by means of an f-test at the 1 percent significance level. Most of the equations show a "bad fit" in relation to board density what could be expected. .1656 .mcoddnd>ou canvases mum managucscea causdsz :d muucenz c o od 0 ad a a o a ed a 0d a c o a a dell/[III od dd dd dd dd od od dd IIMH cd dd Tied 0 od e od co I .u.: dd dd dd dd Nd dd dd dd dd dd dd dd 0 ad ad ad no III/II ad ad ad ad ad ad ad od Nd ed wd sd nd ed nd ad on I .u.: dd dd dd od dd dd dd dd ad ad cd ed a od 0 0 he III/II d~ d~ d~ d~ om d~ dN d~ ow dN ow du md ed od ed use I .u.: a a a o a a a a a m a a d o a o ”as I o x dqo.v Ann.d qu.v do~.d doe.v dod.v doc.d dn~.v II dmoo.v II doe.. d-.. dad.d dad.v dmd.v as ac. do. so. no. on. an. on. mm. ed. mu. ed. we. on. «N. me. an. III/III]! dme.v dw~.d dmc.v dqe.v d~c.dv dad.v dom.v ddn.v II dne.v II dwc.v ddn.v den.d dnd.v dod.v co c~.~ dn.d am.d os.d Na.d Nd.d dn.d od.d od. nu. ma. mo.d co.d cm. Nd.d mo.d doo.dd doc.v doo.v doo.dv doo.dv ddo.v Ace.dd dow.v doo.dVdac.dV dda.. do~.dv dad.v doo.v dan.d deq.v dc no.n -.c d~.~ w~.c mo.m c~.m od.c c~.m d¢.~ d~.e od.c o~.m o~.~ me.~ dd.n mm.n IIII/IIII dod.v ddd.v dn~.dv dac.dv doo.dv doo.v dam.dv don.v den.v zoo.v doo.d.d~o.v dno.v dec.vddno.v dcc.v em.o ed.~ n.od o~.a n~.m ma.~ oo.od co.a N.n od.s ad.~ Nd.o mo.n Nc.e dc.m d~.m om dmc.mv don.md dm¢.~v de~.~v dnd.dv dao.~v deo.dv dn~.~v d~e.dv de~.dv dmn.v deo.~v dnd.ddd~o.d deo.dVAdo.v me mm.w e~.od mm.dd we.cd -.dd we.» mm.~d on.o Nd.o eo.~ dxmm dn.m no.u em.~ dd.m dw.¢ .IIIIIIII doo.ov dm.av dce.nv dmo.nv dao.v dcd.~d ddo.v ddo.nv dnn.dv d~5.dv dmo.v ddc.~. dad.dvded.v doo.dddo~.v me I do no.md o.od m~nmd cN.od oo.nd w~.cd ~m.md nd.md oq.~d -.cd waned m<.cd mo.o dw.o ed.o nn.o 5 Ana .=.¢ o. dus.:.¢ vocodudvcouox vocodudvcou dud wadddoam unacxudze dnc.d dddo.v deco.. d~do.# dsmo.v dddo.v dm~0.v ddmo.¥ Ammc.v dnc. doo.v dONc.Vroeo.Vdmdo.jdmeO.v dsno.v dn su\uv cm. «a. on. do. do. no. «s. on. an. ~o. on. an. as. do. nu. do. Andaman dwnuu< Nd ~d c a Nd ~d x w ~d ~d a a ~d Nd a a dud do>oa :daox vcwoscodx3 atmosuo>ddm memosmxndm speedumadu 0a>h venom I I .mumanod mnudxdmunm _mcnda> sec:— w.:. dId-d....flw.. .mdmflIv—IUI _ ...)... . wwwufiwdamm. I...._.e.cI=II:~..d-nIdmass. u. . de ...I 31......” .1(57 d o d o d cd o o e pWITd I -blIIrIWI.IId d d dsI/ ed o. ad ad dd dd dd dd od dd ed ddII.-IId. d o d dd .3.: ad ad ad dd dd dd dd dd od dd od -dd d od o d dell, dd dd dd dd dd dd dd dd dd dd dd dd dd dd dd dd dd u z dd dd dd od dd dd ad ad od dd ad ad d d o d dsl/II ed dd dd dd dd dd dd dd ad dd ad ad ad dd dd dd do .u.: d d d o d d d d d d d d d d d d dds dd .o.: dod.dv dod.v dad.v dam.v do~.V de¢.v dwq.v doo.v II dnn.v do~.v dm~.v dom.v ddd.d ddoo.v ddd.d dc dd.d dd. dd. dd. dd. od. es. dd. dd. do. dd. dd. dd. dd. dd. dd. III ddd.dd ddd.d dod.d ddd.d ddd.d ddd.d dod.d ddd.dd I- ddd.d ddd.d ddd.d ddd.d dad.d ddd.d ddd.d Ill/1111 dd.d dd. do. dd.d dd. dd.d do. dd.d.Ixss. dd.d do.d do. dd. co. dc.d do. dd dod.dd ddd.dd dod.d ddd.dd ddc.dd dad.d dds.dd ddd.dd dad.dd dsd.d dsd.dd ddo.dd dad.d dod.d ddd.d dod.d dd dd.d dd.d dd.d do.d dd.d dd.d dd.d dd.d dd.d dd.d dd.d dd.d oc.d do.d dd.d dd.d III/11111:, ddd.dd dsd.d ddd.dd ddd.dd ddd.dd ddd.d ddd.dd ddd.dd ddd.dd dod.d ddd.dd ddd.d ddo.d ddd.d ddd.d ddd.d dd dd.d dd.d o.od dd.od dd.d do.d dd.d dd.d dd.d dd.d dd.d dd.d dd.d dd.d dd.d dd.d dod.dd dsd.dd ddd.dd ddd.dd dsd.dd ddd.dd ddd.dd ddd._d ddd.dd dad.d dod.dd ddd.d ddd.d ddd.d ddo.dd ddo.dd dd do.d dd.d d.dd ds.od dd.d dd.d dd.dd dd.d dd.d dm.d dd.d ddmm. dd.d dd.d od.d dd.d [III/11111 dad.dd ddd.dd ddd.dd ddd.dd ddd.dd dds.dd ddd.dd ddd.dd dod.dd dsd.d ddd.dd ddd.dd ddd.d ddd.d ddo.dv ddd.dd dd dd.dd dd.dd ed dd dd.dd dd.dd dd.dd dd.od dd.dd dd.dd do.dd dd.dd dd.dd Idd.d dd.d do.d dd.d d > d5 ....d. S S. .24. vocodddnCOUnz nocodudvcou ddd dd5:25. dmmcdudgd dddc.d doo.d ddo.d dddo.d doc.d doo.v ddd.d ddsc.d -I dddo.d dddo.d dddo.d dddo.mdddo.d ddo.d ddo.d d ddsuddd on. dd. dd. dd. dd. dd. dd. dd. dd. dd. dd. dd. dd._ dd dd. do. ddmcoc daadu< dd dd d d dd dd d d dd dd d d dd _ dd d d ddd du>ad added 2.84;}: Emo.,t.._...> dId..d 3253.5: - I -- Enofmdd: 2:; 38d 1---: --I.-IIII- -I-.-IIl-|II.I- rd! . m....I_ ......d..,d.dd3._._ .1.dd.5d_ cod u ..m.c.—_=au . warms—2w I -74:qu N d .mwmddq... _ma:da> amaz— fidl _ dudlsaImIIWJmIdIc x.,d.d.__._. .dIdI ...d.-._..,..._. 168 3.4.3 — Covariance Analysis. A covariance analysis was conducted to observe the prOperty thickness swelling in relation to the three factors mentioned in other sections before. At first thickness swelling was adjusted over board density to allow comparisons of the two species effects over this property. The first conclusion can be drawn from Figure 66 a 2 no clear relationship exists as far as the specific gravity of the species are concerned for the composition boards at 8 and 12 percent resin levels. When these two adjusted thickness swelling means were compared to verify influence of specific gravity of the species at the 1 percent significance level only the fiberboards at the 8 percent resin level were different. This means that as the specific gravity of the species decreases thickness swelling for the fiberboards at this 8 percent resin level decreases significantly (Figure 66 b). In the second stage TS values were adjusted over board density to observe differences in between the 8 percent and 12 percent resin levels for every kind of composition board. In general, thickness swelling decreases as the resin level increases. This is the case in this study for all composition boards. When tested at the 1 percent significance level only the fiberboards decreased thickness swelling significantly as the resin level increased (Figure 66 c). After the second stage of testing, the non significant lines were pooled together. In the third stage TS values were adjusted over board density to observe differences due to four different particle geometries. When 169 the adjusted TS means were compared at the 1 percent significance level only the fiberboards had a very low and significant thickness swelling, which could be well explained by the uniformity of raw material, the fibers being the ultimate form of wood element making the springback behavior of the board matrix smaller in relationship to the other increasing size of particles (Figure 66 d). 17C) had>ndu oduduodm mmdooam mdvcmdw .m sunspou .m an . .Aduoeoow mdodudod cu man moocououddv you voumob I a Om. (Z) SUIIIans ssauaatui pansnfpv .od Ind d od .ddquomm ododudme new .do>0d admod od one nmocwdodudv sou vmumwh I u Add>uuo ududumaw moduomm .mdmddcc< ousmddm>ou wadddm3m wmocxodze .v .o .g .m we mdswdm .modomdm Ou van moosmuouudv sod voumob I a memo: vmumnnv< I < >Ud>nuu ududoudm moduoem dud>mdc ududuoem modowdm mucosdmndu I m eddOEQdddc I dd am. On. on. Co. on. Om. m r m m 1 m rm. fin dun m \\ n I“ 3 I“ 3 u dd» 1 m I w L w \\\ \\\ \\\ I.‘ M «u. \I‘I I‘ Mr \I‘ s“ I. X. I OH 2. v OH m. v Cd u u a a a s s s I! S S s S ,Jull . I. I! S S 1% w L\ a. u m IIIIIOIQI I w a I I h. T. A“ A. T. .l [cl-lohol on“. V n.” l'l'I'N' m rnfi .m. WmH lololallol 8 I'I'u'I'II'IS 8 I; .w d. I, I, _\! 9» wk us ( "I|'\|||l ( ( .od od dod .muuoudo moduomm nod cosmodddcwdm o: anus mocdd uzwdndum do>wd :dmod :d moocodouudv dew wcdumou undue mocdd vodoom .II.III dm>md :dmmd Nw vdmosdv>ddm u m dd.dmodemdmdd u 3 do>od admmd de l l) 2) 3) CHAPTER VII CONCLUSIONS The two Eucalyptus species Eucalyptus robusta and Eucalyptus grandis show significant differences in several of their physical and mechani- cal prOperties. The most important one is the difference in specific gravity. This difference is reflected in most of the other tested solid wood properties. The relationship between specific gravity and physical and mechanical properties are similar to those found in other species. Sufficiently high mechanical prOperty levels and adequate physical prOperties can be developed in a wide range of composition boards manufactured from the two species within reasonable limits of board density and binder addition. Compliance with particleboard specifications such as those in the National Particleboard Association Commercial Standard CS 236-66 can readily be achieved with all particle configurations. The dominating variable as far as most board prOperties are concerned is clearly the board density. It is also a variable which is most difficult to control under laboratory conditions, both between boards and within boards. In order to study the effects of species specific gravity and resin level on board prOperties, these board prOperties must be adjusted for density variations by means of a covariance analysis, the board density being the covariant. The 171 5) 172 results of the covariance analysis indicate that variables like species specific gravity are of secondary significance at least within the variation given by the two species used. The most responsive to species specific of the four particle configurations is the fiber. This is in contradiction to other findings which indicate that fiberboard is less sensitive to species specific gravity than particleboard. In fact, this is one of the important attributes of medium density fiberboard which allows the utili- zation of heavier hardwoods without undue increases in board density. In the case of this study, the greater sensitivity of fiber— board to species specific gravity may be due to the fact that it might have been possible to form the fiberboard mats with much greater uniformity, thus reducing the variability of the board density. Linear expansion of particleboard and fiberboard cannot readily be related to the major raw material and process variables. While it must, at least theoretically, derive from the swelling and shrinkage characteristics of the solid wood, there are probably too many interactions obscuring the first order relationships. Complicating the matter is the severity of exposure conditions. The high humidity condition and the long term of exposure cause relaxation of stresses, and deterioration of glue lines. Thickness swelling is less complex. In this study, fiberboards had the lowest thickness swelling values due to uniformity of structure. 6) 173 With regard to future work in this important area of the relation- ships between raw material characteristics and board pr0perties, this final conclusion is offered: while most mechanical properties of composition board can easily be adjusted to the required level by changing the compression ratio of the particles, some physical prOperties like linear expansion cannot be so adjusted. Only very careful study of all the interactions and possibly modification of measuring technique will be successful here. APPENDICES APPENDIX A Particle Geometry Nomenclature. - Definitions of the various types of particles have been developed by the American Society for Testing and Materials as a part of the "Standard Definition of Terms Relating to Wood - Based Fiber and Particle Panel Materials" ASTM Designation D 1554. The following definitions are important in defining the geometries used in this research. Fibers. - The slender threadlike elements or groups of wood fibers or similar cellulosic material resulting from chemical or mechanical fiberization, or both, and sometimes referred to as fiber bundles. Flake. - A small wood particle of predetermined dimensions specifically produced as a primary function of Specialized equipment of various types with a cutting action across the grain (either radially, tangen— tially or at an angle between). The action being such as to produce a particle of uniform thickness, essentially flat and having the fiber direction essentially in the plane of the flakes, in overall character resembling a piece of veneer. Slivers. - Particles of nearly square or rectangular cross section with a length parallel to grain of the wood of at least four times the thickness. Wafers - There is no standard definition for this geometry. For the purpose of this study wafers will be defined as a longer thicker flake used in composition boards for structural purposes. 174 APPENDIX B "E. grandis Anatomical Description - pores large, varying in number from 129-165 per area of approxi- mately 20 square mm.‘ rays large, 17-30 cells high, numbering 57-89 per square mm., in majority of samples near 80, biseriate rays common, average 40 percent, some triseriate rays present in half samples examined; parenchyma not abundant and mostly paratracheal; practically no resin in paren- chyma cells, ray cells only half filled with resin, cells generally open, lumina of wood fibres free from resin, giving a more open appearance distinc- tive from such woods as E. marginata, E. resinifera, E. tereticornis, and others —— wood practically identical in most respects with that from E. saligna, and no attempt has been made to separate these two species. For general cell structure see photo- micrographs of E. saligna, Plate No. l. Dadswell and Burnell (1932) [12]. E. robusta Anatomical Description - pores medium to small in size, not numerous, averaging 150 per area of approximately 20 square mm.; rays large, broad and up to 24 cells high, averaging 70 per square mm., biseriate and triseriate rays common, ray cells mostly filled with resin; parenchyma abundant, mostly diffuse but some paratracheal, cells containing some resin; lumina of wood fibres and vessels generally free from resin, vellels ty- losed. For typical cell structure see Plant No. 2." Dadswell and Burnell (1932) [12]. 175 176 Plate No. l E. saligna (Sydney bluogum). ‘ -""—".7- Va... ‘ ' _ ... dd. . L . & Q . 1 m fig“ ‘x-mfl? “ ... ...—1"”-7I.‘ V ‘ 4°C" ‘- I'.’ .n’f' FIG. 1 (top).—Trnnsverso Section. x 95. FIG. 2 (bottom).—T:Ingential Section; X 95. Nora's—(a) urge tylosed vessels. (4) Long biseriate rays, the cell. of which are onlv partly resin filled. E. grgndI's has II similar structure. Compare with E2 haemastoma and E. bdrymdes. 176 Plate No. l E. saligna (Sydney bluegum). _. «vqsmn ' '_‘ ‘16" ’ . _ . u... . . j a” ‘ r—Turm‘x’i’fiu‘ ‘ 9k . _; j 1 angr» a {fiat-no‘ ‘ ‘ “_..1-\‘L.d-r'\r FIG. I (t0p).—Tmnsverso Section. x 95. FIG. 2 (bottom).—T:mgentiul Section; X 95. Norms. —(a) Large tylosed vessels. (‘) Long label-lute rays, the cells of which are only partly resin filled. E. grandix has s similar structure. Comps with E. .‘Iaenmloma and E. bdrym'du. 177 Plate No. 2 E. robusla (swamp mahogany). FIG. 1 (top).—Tmnsverse Section. X 95. FIG. 2 mummy—Tangential Section. x 95. Norris—(a) Abundance of parenchyma cells, some of which are resin-filled. (b) Pores tylosed and containing resin, some of the fibres containing resin in their lumim. (0) Presence of broad rays, some of which are triseriate. LITERATURE CITED 10. 11. 12. 13. LITERATURE CITED American Society for Testing and Materials. 1978. ASTM Standards, Wood; Adhesives, Philadelphia, Pa. Anon. 1976. Eucalypts for Planting (draft). Fo:Misc./76/lO (Rao: Rome). Bamber, R.K., Floyd, A.G. and Humphreys, F.R. 1969. Wood PrOper- ties of Flooded Gum. Aust. For. 33 (1) (3-12). Bamber, R.K. and Humphreys, F.R. 1963. A Preliminary Study of Some Wood Properties of E. Grandis (Hill) Maiden J. Inst. Wood Science II (63—70). Banks, C.H. 1954. The Mechanical Properties of Timbers with Parti- cular Reference to those Grown in the Union of South Africa. J.S. Afr. For. Assoc. No. 24. (44-64). Brumbaugh, J.I. 1960. Effect of Flake Dimensions on PrOperties of Particleboard. For. Prod. J. 10(5) (243-246). Bryan, L.B. 1962. Dimensional Stability of Particleboard. For. Prod. J. 12(12)(572-576). Burley, J., Posner, T., and Waters, P. 1970. Sampling Techniques for Measurement of Fibre Length in Eucalyptus Species. Wood Science and Technology, Vol. 4 (240-245). Carter, W.G. 1974. Growing and Harvesting Eucalypts on Short Rotations for Pulping. Aust. For. 36 (214-225). Commercial Standard CS 236-66 - Mat Formed Wood Particleboard 1968. Childs, M.R. 1956. The Effects of Density, Resin Content and Chip Width on Springback and Certain Other Properties of Dry Formed Flat Pressed Particleboard. M.S. Thesis. Dept. of Wood and Paper Science, North Carolina State University, Raleigh. Dadswell, H.E. and Burnell, M. 1932. Identification of the Colored Woods of the Genus Eucalyptus. Bull. Coun. for Scient. and Ind. Res. 67. (50 pages). DeVilliers, A.M. 1973. Utilization Problems with Some Eucalypts in South Africa. Proc. Meet. Div. 5 - Int. Un. For. Res. Ore. South Africa. Vol. 2 (256-270). 178 14. 15. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 179 Edwards, D.W. 1973. Defects of Fast Grown Eucalyptus in New South wales. Proc. Meet. Div. 5 - Int. Un. For. Res. Ore. South Africa. Vol. 2 (256-70). Forwood. 1975. Report of the Forestry and Wood Based Industries Development Conference (Aust. Govt. Publ. Serv.:Canberra). Franklin, E.C. 1977. Yield and Properties of Pulp from Eucalypt Wood Grown in Florida. Forest Engineering - Reprint from Tappi Conference Papers. 1977. Gatchell, C.J., Heebink, B.G. and Hefty, F.J. 1966. Influence of Component Variables on PrOperties of Particleboard for Exterior Use. For. Prod. J. V. 16 (4) (46-53). Geimer, R.L., Montrey, H.M., and Lehmann, W.F. 1975 Effects of Layer Characteristics on the Properties of Three-Layer Particleboards. For. Prod. J. 25(3) (19-29). Gertjenjansen, R., Hyvarinen, M., Haygreen, J., and French, D. 1973. Physical Properties of Phenolic Bonded Wafer-Type Particleboard from Mixtures of Aspen Paper Birch, Tamarack. For. Prod. J. 23 (6) (24-28). Gertjejansen, R. and Haygreen, J.G. 1973. Effect on Aspen Bark from Butt and Upper Logs on the Physical PrOperties of Wafer- Type and Flake-Type Particleboards. For. Prod. J. 23 (9) (66-71). Grimwade, R. 1930. An Autobiography of the Eucalyptus. Augus S. Robertson, Ltd., Sydney-Australia. Gueneau, P. 1969. The Characteristics and Utilization of Eucalyptus robusta in Madagascar. Bois For. Trap. 124 (53-65). Guha, S.R.D., Sharma, Y.K., and Jadhau, A.G. 1965. Chemical Pulps for Writing and Printing Papers from Eucalyptus robusta. Indian For. 91(5) (294-296). Guha, S.R.D. et.al. 1967. Chemical, Semi-Chemical and Mechanical Pulps from Eucalyptus grandis. Indian. For. 93(6) (360-372). Halligan, A.F. 1970. A Review of Thickness Swelling in Particleboard. Wood Science and Technology 4(4) (301-312) [73 ref]. Halligan, A.F. and Schniewind, A.P. 1972. Effect of Moisture on Physical and Creed PrOperties of Particleboard. For. Prod. J. 22(4) (41—48). 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 180 Hann, R.A., Black, J.M. and Blomquist, R.F. 1962. How Durable is Particleboard? For Prod. J. 12(12) (577-584). Hans, A.S., Burley, J. and Willianson, P. 1972. Wood Quality in Eucalyptus grandis (Hill) Maiden Grown in Zambia. Holzforschung, 26(138-141). Hart, C.A. and Rice, J.T. 1963. Some Observations on the DevelOp- ment of a Laboratory Flakeboard Process. For. Prod. J. 13 (11) (483-488). Heebink, B.G., Hann, R.A., and Haskell, H.H. 1964. Particleboard Quality as Affected by Planer Shaving Geometry. For. Prod. J. 14(10) (486-494). Heebink, B.G. and Hann, R.A. 1959. How Wax and Particle Shape Affect Stability and Strength of Oak Particleboards. For.Prod. J. 9(7) (197-203). Heebink, B.G. 1974. Particleboards from Lodgepole Pine Forest Residue. U.S.D.A. For. Serv. Res. Paper FPL 221. For. Prod. Lab. Madison, Wisconsin. Hillis, W.E. and Brown, A.G. 1978. Eucalypts for Wood Production. Csiro, Australia. HSE, C-Y. 1974. Reaction Catalysts of Urea-Formaldehyde Resins as Related to Strength Properties of Southern Pine Particleboard. J. Jap. Wood. Res. Soc. 20(11) (538-540). HSE, C-Y. 1975. Preperties of Flakeboards from Hardwoods.Growing on Southern Pine Sites. For. Prod. J. 25(3) (48-53). Johnson, D.R. 1973. Panel Discussion on Experiences with Pressure Refined Fiber. Seventh Particleboard Proc. Wash. State Univ., Pages l9-42. Kelly, J. 1969. Eucalypts. Nelson, Plates. Kelly, M.W. 1977, Critical Literature Review of Relationships Between Processing Parameters and Physical Properties of Particleboard. U.S.D.A. For. Serv. General Tech. Rep. FPL-10 For. Prod. Lab. Madison, Wisconsin. Klauditz, W. and Stegmann, G. 1957. Uber die Eignung von pappelholz zur hersetellung von holzspanplatten. Holzforschung 11 (174-179). Kimoto, R., Ishimoir, E., Sasaki, H. and Maku, T. 1964. Studies on Particleboards VI. Effects of Resin Content and Particle Dimension on the Physical and Mechanical Properties of Low Density Particleboards. Wood Res. Kyoto U. 32(1—14). 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 181 Kollmann, F.P., Kuenzi, E.W. and Stamm, A.J. 1975. Principles of Wood Science and Technology. Vol. I and II. Springer-Verlag, New York, pp. 703. Kusian, R. 1968. Model Investigations about the Influence of Particle Size on Structural and Strength PrOperties of Particle Materials II. Experimental Investigations. Holztechnologie 9(4) (131-134). Lehmann, W.F. 1974. PrOperties of Structural Particleboard. For. Prod. J. 24(1) (19-26). Lehmann, W.F. and Hefty, F.J. 1973. Resin Efficiency and Dimensional Stability of Flakeboards. U.S.D.A. For. Serv. Res. Pap. FPL 207 For. Prod. Lab. Madison, Wisconsin. Lehmann, W.F. 1970. Resin Efficiency in Particleboard as Influenced by Density, Atomization, and Resin Content. For. Prod. J. 20 (11) (48-54). Maloney, T.M. 1975. Use of Short-Retention—Time Blenders with Large-Flake Furnishes. For. Prod. J. 25(5) 21-29. Maloney, T.M. 1977. Modern Particleboard and Dry Process Fiberboard Manufacturing. Miller Freeman Publications. (672 pages.) Penfold, A.R. and Willis. J.L. 1961. The Eucalyptus. World Crop Books. London. Plath, L. and Schnitzler, E. 1974. The Density Profile A Criterion for Evaluating Particleboard. Holz Roh Werkst. 32(11) (443-449). Post, P.W. 1958. The Effect of Particle Geometry and Resin Content on Bending Strength of Oak Particleboard. For. Prod. J. 8(10) (317-332). Post, P.W. 1961. Relationship of Flake Size and Resin Content to Mechanical and Dimensional ProPerties of Flakeboard. For. Prod. J. 11(1) (34-37). PrOperties of Australian Timbers: 1959 Rose Gum (Eucalyptus grandis) For. Prod. Newsletter Csiro, Aust. N2 259(4). Properties of Australian Timbers: 1960 Swamp Mahogany (Eucalyptus robusta) For. Prod. Newsletter. Csiro, Aust. N3 262(4). Rackwitz, G. 1963. Influence of Chip Dimensions on some PrOperties of Wood Particleboard. Holz Roh Werskt. 21(6) (ZOO-209). 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 182 Ranatunga, M.S. 1967. A Study of the Fibre Lengths of E. grandis Grown in Ceylon. Ceylon. Fore. 6(101-112). Rice, J.T., Snyder, J.L. and Hart, C.A. 1967. Influence of Selected Resin and Bonding Factors on Flakeboard PrOperties. Roffael, E. and Ranch, W. 1972. Influence of Density on the Swelling Behavior of Phenolic-Resin Bonded Particleboards. Holz Roh Werkst. 30(5) (178-181). Rudolph, J.J., Simoes, J.W. and James, L.M. 1978. Forestry in Brazil: An Awakening Giant. Journal of Forestry, Vol. 76 No. 12. (784-786). Rudolph, J.J. 1978. M.S.U. Brazil-Mac Project Report N2 91. Skolmen, R.G. 1972. Specific Gravity Variation in Eucalyptus robusta Grown in Hawaii. U.S. Fore. Serv. Racific-Southwest For. Range. Exp. Sta. Res. Paper PSN-78. Srivastava, J.S. and Mathur, G.M. 1967. Chemical Pulps for Writing and Printing Papers from Eucalyptus grandis. Indian Pulp and Paper Calcutta. 19(3) (215-216). Stewart, H.A. and Lehmann, W.F. 1973. High Quality Particleboard from Cross Grain Knife Planed Hardwood Flakes. For. Prod. J. 23(8) (52-60). Strickler, M.D. 1959. Effect of Press Cycle and Moisture Content on Properties of Douglas-Fir Flakeboard. For. Prod. J. 9(7) (203-205). Suchsland, O. and Woodson, E.G. 1974. Effects of Press Cycle Variables on Density Gradient of Medium Density Fiberboard. Proc. Eight Partic. Symposium - Wash. State U. Pullmann, Wash. (375-396). Suchsland, O. 1959. An Analysis of the Particleboard Process. Q. Bull., Mich. Agr. Exp. Sta. Mich. State Univ. 42(2) (350- 372). Suchsland, O. 1967. Behavior of a Particleboard Mat During the Press Cycle. For. Prod. J. 17(2) (51-57). Suchsland, 0. 1960. An Analysis of a Two-species Three-Layer Wood Flakeboard. Mich. Agr. Exp. Sta. Q. Bull. 43 (2) (373-393). 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 183 Suchsland, 0., Lyon, D.E., and Short, P.E. 1978. Selected PrOperties of Commercial Medium Density Fiberboards. For. Prod. J. 28(9) (45—48). Suchsland, O. 1962. The Density Distribution in Flakeboard. 0. Bull. Mich. Agr. Exp. Sta. Mich. State Univ. 45(1) (104-121). Suchsland, 0. 1972. Linear Hygrosc0pic Expansion of Selected Commercial Particleboards. For. Prod. J. 22(11) (28-32). Suchsland, O. 1973. HygrOSCOpic Thickness Swelling and Related PrOperties of Selected Commercial Particleboards. For.Prod. J. 23(7) (20-30). Suchsland, O. and Woodson, E.G. 1976. PrOperties of Medium Density Fiberboard Produced in an Oil Heated Laboratory Press. South. For. Exp. Sta. Res. Pap. SO 116. Suchsland, O. 1977. Compression Shear Test for Determination of Internal Bond Strength in Particleboard. For. Prod. J. (27) (1) (32-36). Suchsland, O. 1970. Optical Determination of Linear Expansion and Shrinkage of Wood. For. Prod. J. (20) (6) (26—29). Suchsland, O. 1978. Medium Density Fiberboard Production Possi- bilities in the Tennessee Valley Region. Prepared for Division of Forestry, Fisheries, and Wildlife Deve10pment. 173 pages. Suchsland, O. 1978. Markets and Applications of Medium Density Fiberboard Proceedings Complete Tree Utilization of Southern Pine. Forest Products Research Society. (473-481). Suchsland, O. 1968. Particleboard From Southern Pine. Southern Lumberman. December. Talbot, J.W. and Maloney, T.M. 1957. Effect of Several Production Variables on the Modulus of Rupture and Internal Bond Strength of Boards Made of Green Douglas-Fir Planer Shavings. For. Prod. J. 7(10) (395-398). Taylor, F.W. 1972. Anatomical Wood PrOperties of South African Grown E. Grandis. S. Afr. For. Journal. 80(20-24). Taylor, F.W. 1973. Variations in the Anatomical Properties of South African Grown E. Grandis. Appita 27 (191-178). Turner, H.D. 1954. Effect of Particle Size and Shape on Strength and Dimensional Stability of Resin Bonded Wood-Particle Panels. For. Prod. J. 4(5) (210-222). 82. 83. 84. 85. 184 Vital, B.R., Lehmann, W.F. and Boone, R.J. 1974. How Species and Board Densities Affect the Properties of Exotic Hardwood Particleboards. For. Prod. J. 24 (12) (37—45). Wasserman, W. and Neter, J. 1974. Applied Linear Statistical Models. Richard D. Irwin, Inc. (842 pages). Woodson, E.G. 1976. Effects of Bark, Density Profile, and Resin Content on MDF from Southern Hardwoods. For. Prod. J. (26) (2) (39-42). Woodson, E.G. 1977. Medium—Density-Fiberboard from Mixed Southern Hardwoods. Proc. Eleventh Part. Symp. Wash. State Univ. Pullmann, Washington, (131-140). TA ER ”'IIIIIIIIII‘IIIIIIIIIIIIIII’IIII“