Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Improved method for evaluating the quality of phenolic resin bonds of Douglas fir Northcott, Philip Lachlan 1954

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1954_A7 N63 I6.pdf [ 8.59MB ]
Metadata
JSON: 831-1.0106855.json
JSON-LD: 831-1.0106855-ld.json
RDF/XML (Pretty): 831-1.0106855-rdf.xml
RDF/JSON: 831-1.0106855-rdf.json
Turtle: 831-1.0106855-turtle.txt
N-Triples: 831-1.0106855-rdf-ntriples.txt
Original Record: 831-1.0106855-source.json
Full Text
831-1.0106855-fulltext.txt
Citation
831-1.0106855.ris

Full Text

IMPROVED METHOD FOR EVALUATING THE QUALITY OF PHENOLIC RESIN BONDS OF DOUGLAS FIR  by PHILIP IACHIAN NGRTHCOTT  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE i n the Department of Forestry Faculty of Applied  Science  We aooept t h i s thesis as conforming t o the standard required from candidates f o r the degree of MASTER OF APPLIED SCIENCE  Member of the Department of Forestry  Member of the Department o f Mechanical Engineering  THE UNIVERSITY OF BRITISH COLUMBIA  A p r i l , 1954-  - 1 ABSTRACT The purpose and h i s t o r y of glue bond t e s t i n g have been reviewed, .Certain d e f i c i e n c i e s of standard test methods have been elaborated, and the need f o r a more aoourate t e s t procedure stated*  The objective of t h i s par-  t i c u l a r research was to develop a t e s t that would meet such a need.  Exper-  imentation was limited to Douglas f i r veneers bonded with hot-press phenolic r e s i n adhesives. The following requirements of an i d e a l method of estimating glue bond q u a l i t y were used as a guide i n selecting new designs. requirement i s r e p r o d u c i b i l i t y of t e s t r e s u l t s .  The foremost  Other e s s e n t i a l features  include a u n i v e r s a l l y acceptable unit of measurement and a t e s t speoimen which i s simple and economical t o prepare.  Other desirable features are  that a maximum number of specimens be obtainable from a given sample, and that the method be adaptable t o both research and production t e s t i n g . The Glueline-Cleavage Test developed through t h i s research meets a l l of the above requirements.  The p r i n c i p l e i s t o measure the force r e -  quired to s p l i t or cleave a one i n c h square plywood specimen along the glue l i n e by means of a "knife" or wedge.  The aotion i a s i m i l a r t o the s p l i t t i n g  of any wood with a wedge except that the knife i s p a r t i c u l a r l y positioned along the glue l i n e . or laminated.  Specimens for t h i s purpose may be either cross-banded  I f the data are t o be d i r e c t l y comparable every d e t a i l of the  specimen, manufacture, t e s t i n g machine, and test procedure must be standardized.  This requirement i s common to a l l methods of t e s t .  For q u a l i t y oon-  t r o l purposes t e s t specimens would be out from the plywood at an angle of f o r t y - f i v e degrees to the g r a i n . tested.  Every glue l i n e of the speoimen may be  Special two-ply specimens have been designed f o r research purposes  where aocuracy i s of the utmost importance.  In t h i s type, the material i s  - 2 cut so that the c e l l s interseot the veneer surfaoe a t an angle of t e n degrees.  This small angle, plus the r e l a t i v e weakness of wood i n tension  perpendioular t o the grain, tends t o concentrate the stress i n the glue line.  This insures a glue l i n e f a i l u r e when the knife i s applied i n the  correct d i r e c t i o n .  I f the knife i s not applied i n the correct d i r e c t i o n  the specimen tends to s p l i t along the grain of the wood away from the glue l i n e rather than toward i t as intended.  These research specimens are pre-  pared from two edge-grain veneers glued with the springwood-summerwood bands crossed.  Any cross-banding angle up t o ninety degrees  (common f o r  oommeroial plywood) may be used. The strengths of matched phenolic-bonded Douglas f i r glue l i n e s were compared by f i v e methods, four being mechanical and one r e l y i n g on wood f a i l u r e .  The mechanical methods were the Blook Shear, Glueline-  Cleavage, Tension Normal t o the Glue Line, and Tension Shear Tests. The Per Cent Wood Failure Method employed wood f a i l u r e readings from the Tension Shear specimens as estimates of bond q u a l i t y .  The Glueline-Cleavage  and Tension Shear Methods inoluded several t e s t specimen designs. The above-mentioned comparisons y i e l d e d the following information. (1) mechanical methods proved more acourate than those based upon wood f a i l u r e estimations, (2) f o r q u a l i t y - c o n t r o l purposes the Glueline-Cleavage Test was shown t o be equal or superior t o the other meohanical methods, and (3) the Glueline-Cleavage Test, when used with specimens designed for research purposes, proved of superior accuracy to a l l others t e s t e d . Add i t i o n a l advantages of the Glueline-Cleavage Method include:  (1) the  simplest possible test specimen shape and therefore simple and inexpensive manufacture, (2) a maximum y i e l d of specimens from a given plywood sample (a valuable feature with experimental designs r e q u i r i n g large numbers of  - 3matohed specimens), (3) a lower time requirement per t e s t , (4) much less expensive machinery i s required t o perform the t e s t , and (5) the exposure of every glue l i n e f o r inspection purposes when research-type specimens are used* Although the Glueline-Cleavage Test i s believed t o be the most aocurate method yet developed, imperfections remain and several further methods have been proposed f o r increasing i t s aoouraoy.  TABLE OF CONTENTS PREFACE Chapters  Page  I - INTRODUCTION  1  I I - THE MECHANICAL TEST VS. PER CENT WOOD FAILURE FOR ESTIMATING BOND QUALITY  5  I I I - MECHANICAL TESTS OF PLYWOOD GLUE BONDS 1.  The Most Desirable Stress D i s t r i b u t i o n i n a Test Speoimen 2. Methods of Reduoing V a r i a b i l i t y 3» Theories Regarding Sources of Excessive V a r i a b i l i t y i n Test Results 4« Theories Regarding the Design of a Better Test Method 5 • Summary  16 20 24 27 30  IV - THE GLUELINE-CLSAVAGE TEST 1* 2. 3. 4. 5»  6.  Introduction D e f i n i t i o n of the Glueline-Cleavage Test Exploratory T r i a l EGxlO° Speoimen Comparison of Glueline-Cleavage, (EGxlO ), with Tension Shear, Tension Normal, and Block Shear Tests Summary  32 32 33 35  0  40 43  V - ELABORATION OF THE GLUELINE-CLEAVAGE METHOD 1* 2. 3» 4. 5.  6.  Preamble A u x i l i a r y Tools Experimental Design Working Plan Results (a) Per Cent Wood Failure Vs. Mechanical Test (b) S e n s i t i v i t y of Designs (o) Glue Strength vs. Wood Strength Summary  VI - CONCLUSIONS AND BIBLIOGRAPHY  RECOMMENDATIONS FOR FUTURE RESEARCH  45 46 49 50 55 55 57 62 63 64  71  iii APPENDIX A Table IA - Tension Shear Tests, Springwood bonded to Springwood and Summerwood t o Summerwood Table IB - Blook (Compression) Shear Tests, Springwood bonded to Springwood and Summerwood t o Summerwood Table 1C - Tension Normal to the Glue Line Plywood Tests, Springwood bonded t o Springwood and Summerwood to Summerwood Table 2  - Coefficients of V a r i a t i o n of Breaking Loads Exploratory Comparisons  Table 3  - Comparison of Breaking Loads, e t c . , Tension Normal t o the Glue Line, Tension Shear, Block Shear, and Glueline-Cleavage Tests  Table 4  - S t a t i s t i c s of Table 5 Data  Figures 1 and l a - The Glueline-Cleavage Maohine Figure 2 - I l l u s t r a t i n g the Cutting Plan f o r "Veneers" Figure 3 - Cutting and Numbering Plan I l l u s t r a t i n g the Layup of Glue Blanks f o r Glueline-Cleavage EGxlO Specimens 0  Figure 4 - Cutting and Numbering Plans f o r Speoimens Figures 5 and 5A - Tension Normal Specimens Figures 6 and 6A - Tension Shear Speoimens Figures 7 and 7A - Block Shear Speoimens Figures 8 and 8A - Glueline-Cleavage EGxlO Specimens 0  APPENDIX B Table 1 - Randomization  of Pressloads  Table 2 - O r i g i n a l Data Table 3 - Percentage Reductions i n Strength (or $WF) Table 4A- Comparisons of the Differences i n Predicted Percentage Reduction ( i n breaking load or per cent wood f a i l u r e ) Design 11 - Design 11% WF  iv APPENDIX B (oont'd)  Page  Table 4B - Comparisons of the Differences i n Predicted Percentage Reductions ( i n breaking load) Design 11 - Design 4  93  Table 5A - Summary of Slope Ratios  94  Table 5B - Analysis of Variance of Slope Ratios  94  Table 6  95  - Single Degree Comparisons of Slope Ratios  Figure 1 - Cutting Plan f o r Boards  96  Figure 2 - Designs of Test Specimens  97  Figure 3 - Marking System for Glue Blanks  98  Figure 4 - Marking and Cutting Plans  99  Figure 5 — Logarithmic Transformation; Typical Plots  10G  Figure 6 - Representative Plots of Data of each Design for each Treatment  101,  102  Figure 7 - Patterns of Curves f o r Weathering Treatments I to VI f o r each Design  103,  104  V  PREFACE The researoh forming the basis for t h i s t h e s i s was 1950,  initiated i n  Chapter IV, The Glueline-Cleavage Test, was the subject of a paper  presented before the Sixth National Annual Meeting of the Forest  Products  Researoh Society at Milwaukee, Wisconsin, i n June, 1952^^^.-^ The author wishes to thank C o l . J.H. Jenkins, Mr. R.M. Mr. E.G. Fensom, Mr. J.B. Alexander, and Mr. W.J.  Brown,  Smith, a l l of the Forest  Products laboratories of Canada, f o r t h e i r co-operation i n authorizing the Project Work included i n Chapters TV and V, and f o r other a i d that made possible the oonduot of t h i s researoh.  He wishes to thank the members of  the Faculty of Forestry, e s p e c i a l l y Professors R.W. J.H.G. Smith, and J.W.  Ker,  Wilson f o r t h e i r valuable assistance.  Professor S.W. W.O.  Wellwood, J.W.  Nash of the Mathematios Department and Professor  Richmond of the Meohanical Engineering Department of the University  of B r i t i s h Columbia, and Mr. D.G.  M i l l e r of the Forest Products  of Canada contributed valued advice. ohecking was  The stenographic work, compiling and  g r e a t l y f a c i l i t a t e d by the help of Mrs. P.D.  H.G.M. Colbeok, Mrs. E. Cutforth, Miss W.G. the Vancouver Forest Products  Laboratories  Birrell,  Dyer, and Miss M.L.  Mr.  Wells of  Laboratory.  The work involved i n planning, conducting, and analysing the researoh reported, and e s p e c i a l l y the preparation of t h i s t h e s i s , has been an invaluable experienoe to the author both personally and p r o f e s s i o n a l l y i n h i s work at the Forest Produots Laboratory.  I t i s hoped that further  opportunity w i l l be provided f o r a d d i t i o n a l researoh suggested by t h i s study.  The figures i n parentheses r e f e r to the bibliography a t the end of the text.  CHAPTER I - INTRODTJCTIOS 1.  The general purpose of t e s t i n g glue joints i s to measure the  strength of the bond*  This i s necessary i f step "by step improvements are  to be made i n the " e f f i o i e n o y " ^ / of glues and the economy of gluing processes* The a r t of veneering i t s e l f i s an ancient one, having been pract i s e d by the craftsmen of twenty or t h i r t y centuries ago.  In spite of the  antiquity of i t s o r i g i n no r e a l e f f o r t was made t o investigate the more s c i e n t i f i c aspeots of plywood manufacture u n t i l very r e c e n t l y .  Although  s u f f i c i e n t p r a c t i c a l knowledge has been accumulated to allow reasonably s a t i s f a c t o r y mass production, manufacturers  are oontinually faced with  problems about which they have l i t t l e , i f any, fundamental knowledge.  The  broad reason for plywood research i s to provide some of t h i s needed fundamental knowledge, i n t h i s oase plywood bonding information. One method of appraising bond q u a l i t y i s to make use of the percentage of wood f a i l u r e developed as a r e s u l t of rupturing the bond. broken glue l i n e can be used.  Any  To use the method t o advantage i t i s neces-  sary to make the observations on specimens which have been prepared by a  ^An  " e f f i c i e n t " adhesive i s defined by K n i g h t ^ ' as* 0  One that maintains an adequate bond between the wooden elements under conditions of exposure that the j o i n t has to withstand. The term "adequate bond" has a wide i n t e r p r e t a t i o n ; i n the small, highly stressed j o i n t s of a i r c r a f t or high-powered boats, only the maximum possible strength i s adequate, but for very many purposes a comparatively weak j o i n t gives everything necessary. Whether the j o i n t has high or low strength, i t must be permanent for so long as i t s use demands. Here again, there are degrees of permanence; high-grade furniture i s expeoted t o l a s t at least a generation, but i t i s s u f f i c i e n t i f non-returnable packaging containers survive the one journey. This u s e f u l d e f i n i t i o n embraoes the multitudinous d e t a i l s of the gluing prooess as w e l l as the adhesive.  - 2 standard!eed procedure.  Broken Plywood Glue Shear speoimens^*  are one common source of such observations.  ^»  ^  The Knife Test(®» ^4) method  i s another standardized system f o r estimating percentage wood f a i l u r e * A seoond type of estimate of bond quality which has found some favour i s the v i s u a l examination or de lamination method^*  2  4)  #  Here  the prooedure i s to expose the speoimen t o the elements, or t o some form of accelerated weathering treatment calculated to set up stresses i n the bond*  The amount of delamination which develops i s used as the o r i t e r i o n  of bond quality*  This i s r e a l l y a mechanical t e s t with the forces s e l f -  applied by the hygrosoopioity of wood but laoking any measurement of these forces* Neither the v i s u a l method nor the percentage wood f a i l u r e method provides a u n i v e r s a l l y acceptable estimate of the strength of a bond*  Eaoh  f a i l s , therefore, t o meet a fundamental requirement of the perfect bond test* The most universally accepted method of estimating bond quality i s by means of mechanical t e s t s *  These may be grouped into four types  which have been designed to apply the following stresses to the bondt (1)  Tensile (Tension Perpendioular to the Grain Plywood Test(45), R u d k i n ( ) , G l u e l i n e - M e t h o d ( \ Platow and 58  (2)  Tftetz( \  27  Shear (e*g* Blook Shear( * 40» 41) 1  f  Double  2  eompression  Shear( 4), Double Tension S h e a r ^ \ 1  (3)  Combined t e n s i l e and shear (Simple-lap Test P i e o e ^ * Plywood Glue ShearU* 40, 4l))  (4)  #  Combined t e n s i l e and compressive (Spandau(^»  •^)*  In the oour3e of routine mechanical t e s t i n g of glues i t has frequently been necessary t o repeat t e s t s because the specimens broke  ^ \  through the wood (100$ wood f a i l u r e ) but at loads less than the allowable w r f T r i - i r — T h e s e conditions developed i n t e s t i n g f o r Tension Shear i n accordance with R.C.A.F. S p e c i f i c a t i o n C-22-2^7) (using either sugar maple or yellow b i r o h ) .  The v a r i a b i l i t y i n t e s t r e s u l t s appeared t o be exoessive.  and a t t r i b u t a b l e t o the wood or method o f t e s t rather than t o the adhesive-/ under t e s t .  When studies were undertaken regarding the gluing properties  of Douglas f i r veneers, using either the Plywood Glue Shear T e s t ^ "Tension normal t o the Glue Line Plywood T e s t " ^ ) ,  or the  the problem appeared t o  be even more acute. D i s s a t i s f a c t i o n with the r e s u l t s obtained when using standard t e s t methods, and a continuing demand f o r a means of appraising the q u a l i t y o f glued j o i n t s , made i t advisable t o investigate the problems pertaining t o the t e s t i n g of plywood bonding. Most of the plywood produced i n B r i t i s h Columbia i s manufactured from Douglas f i r .  A l l of the Douglas f i r and western softwood plywood  manufactured i n t h i s provinoe i s bonded with hot-press phenolic r e s i n adhesives and i s marketed as "weatherproof" grade.  In spite o f the faot that  Douglas f i r plywood has been successfully manufactured f o r more than twenty years there are s t i l l many problems requiring i n v e s t i g a t i o n .  One of these  and one whioh i t has been the author's p r i v i l e g e to study, i s the e f f e c t of d i f f e r e n t dryer temperatures upon the bond q u a l i t y .  Because c e r t a i n  problems of t e s t i n g are d i f f e r e n t when researoh i s limited t o one speoies of wood rather than a ohoioe of woods and for the reasons discussed above i t was decided t o l i m i t a l l t e s t i n g t o Douglas f i r and t o hot-press pheno l i c r e s i n adhesives.  In t h i s t h e s i s the words adhesive and glue are used synonymously.  - 4 This chapter has been devoted t o a b r i e f explanation of the purpose of testing i n general and of t e s t i n g plywood bonds i n p a r t i c u l a r . A short description of methods of estimating bond q u a l i t y and some of t h e i r deficiencies has been included.  Reasons have been presented f o r  l i m i t i n g the research t o phenolic r e s i n bonds of Douglas f i r veneers. The primary objeotive has been s e t t o develop a better method f o r estimating the q u a l i t y of plywood bonds.  - 5 CHAPTER I I - THE MECHANICAL TEST VS. PER CENT WOOD FAILURE FOR ESTIMATING BOND QUALITY For the purpose of t h i s thesis bond q u a l i t y w i l l be defined as the maximum stress, which may be imposed upon the bond without causing i t s f a i l u r e or rupture of the adjoining wood.  I t i a the e f f e c t i v e strength of  the bond at any ohosen i n s t a n t , i . e . , the maximum stress which the bond was  capable of withstanding, minus i n t e r n a l stresses, fatigue stresses,  and reductions i n the strength properties of the wood.  The q u a l i t y of the  bond between two pieces of wood w i l l vary with any treatment to which i t i s subjeoted, such as ohanges i n moisture content, or i n temperature, or by ohendcal or fungal attack, A mechanical t e s t (of a wood-glue bond) i s one which measures the bond q u a l i t y by means of a "breaking load" that i s r e l a t e d to the ultimate stress which the bond or the wood withstood occurred,  just before  failure  Acoording t o the l i t e r a t u r e mechanical t e s t i n g of adhesives i s  of very recent o r i g i n although gluing i t s e l f dates back many oenturies. The e a r l i e s t reoord of mechanical tests n o t e d i s  reproduced, i n part,  because of i t s reference to e a r l i e r work and i t s generally applicable background.  This was the f i r s t of three very comprehensive reports issued by  the ( B r i t i s h ) Adhesives Research Committee between 1922 and  1932.  10, An examination of the scattered l i t e r a t u r e of the subj e c t quickly shows the unsatisfactory state of a f f a i r s as regards the t e s t i n g of adhesives by means of mechanical t e s t s of cemented wood j o i n t s . There i s no standardised method of carrying out such t e s t s , and i t i s evident that muoh i n v e s t i g a t i o n must be undertaken before there can be adopted any generally acceptable standardised prooedure. Thus, among the v a r i a b l e faotors i n such t e s t s are the nature of the t e s t s themselves (whether tension, shear or impact), the nature and oondition of the wood used, the method of preparing the adhesive and the wood for the t e s t j o i n t s , and the form and size of j o i n t employed i n making up the t e s t pieoe.  - 6 11. No r e a l l y s a t i s f a c t o r y methods of t e s t i n g glues, e t c . , mechanically, otherwise than by means of wood j o i n t s , appear to have been devised. Weidenbusoh (1859) employed rods made up of plaster of Paris and glue, but the r e s u l t s appear to "be uncertain. G i l l (Journ. Ind. Eng. Chem., 1915, 7# 102) used briquettes made up of fuller»s earth, diatomaoeous earth, quartz sand or sawdust with various glues, but obtained s i m i l a r l y unsatisfactory r e s u l t s . The same investigator employed a modification of Setterberg s method (Sohwed. Teoh. T i d s k r i f t , 28, 52) i n which the strength of glued paper i s measured; the results of t h i s t e s t are of q u a l i t a t i v e , rather than quantitative, s i g n i f i c a n c e . G i l l ( l o o . c i t . ) has a l s o used p o r c e l a i n , glass and t i l i n g i n the making of glued t e s t j o i n t s , but has found them a l l unsatisfactory. Hopp (Journ. Indust. Eng. Chem., 1920, 12, 356) used s t r i p s of dried glue ground to standard s i z e , and tested them i n a Sohopper machine. He states that the r e s u l t s of h i s t e s t s were oondordant, but i t i s very doubtful whether the t e n s i l e strength of a dried sample of glue i s s t r i c t l y i n d i c a t i v e of the behaviour of that sample of glue when used i n making a j o i n t between two pieces of wood In the ordinary way. I t would seem to be generally agreed that i n order to t e s t the value of an adhesive f o r use i n j o i n t i n g timber, the most u s e f u l information i s to be obtained from wood joints made up with that adhesive. ,  12. At the outset of t h e i r work i t seemed to the o r i g i n a l Committee that while the e x i s t i n g meohanical tests of adheslves for timber were s u f f i c i e n t l y good t o enable high-quality adhesives to be d i f f e r e n t i a t e d from poor ones, i t would be d i f f i c u l t to make step-by-step improvements i n good or bad glues unless mechanical t e s t s were devised which would make i t poss i b l e to d i s t i n g u i s h with oertainty between samples showing moderate differences of merit. With t h i s objeot i n view i n v e s t igations were undertaken by Major A. Robertson, R.A.F., at the Royal A i r o r a f t Establishment, and a report was submitted by him to the Committee e a r l y In 19l9. This report has sinoe been issued* by the Aeronautical Research Committee, and i t w i l l here suffioe b r i e f l y to indicate i t s nature. 13. The various tests whioh, up to that time, had been employed f o r the determination of the strength of oemented joints i n timber were examined and were each found t o be i n adequate i n some p a r t i c u l a r d i r e o t i o n . Tension tests are of two classes:- (a) Those of j o i n t s made along the grain; (b) those of end grain j o i n t s . In connection with the former  "Report on the Materials of Construction used i n A i r o r a f t " a n d A i r o r a f t Engines", Chap. I I , page 132. Published by H.M. Stationery O f f i c e , price 21s. net.  - 7 c l a s s , two types of t e s t piece formerly used at the Royal A i r c r a f t Establishment were investigated and found to be unsatisfactory on account of the unequal d i s t r i b u t i o n of stress produoed by the bending of the specimen when loaded* Test pieces of two other designs (described i n the report) were examined and found t o be s a t i s f a c t o r y i n securing a . reasonably good stress d i s t r i b u t i o n . I t i s i n t e r e s t i n g here to note that from some tension t e s t s on a very small piece of glue (turned up from a pieoe of cake glue) the tension stress of the glue i t s e l f appeared to be a t least 3,000 l b s . per square inoh, i . e . , more than twioe what i s obtainable normally between glue and wood. As regards the second class of j o i n t s — e n d grain j o i n t s — a t convenient form of t e s t piece i s described i n the r e p o r t . Preliminary t e s t s on t h i s type gave promising r e s u l t s , and i t was found that good r e s u l t s could be obtained with soft timbers as w e l l as with hard. Reference i s a l s o made i n the report t o the Spandau t e s t (used i n Germany), which i s a modification of the d i r e c t tension t e s t ; i n t h i s t e s t the pieces are glued end g r a i n to end g r a i n .and the speoimen i s broken by bending. An advantage of the Spandau t e s t piece i s that t h i s oan be used f o r an impaot t e s t which i t may ultimately be desirable t o introduce. 14. Major Robertson's report further deals with shear t e s t s , and an examination i s made of the various types of t e s t pieces which have been used by d i f f e r e n t experimenters. It i s shown that the majority of these are unsatisfactory f o r i n none i s the stress on the j o i n t a simple shear. I t i s suggested that the easiest metnod of carrying out a shear t e s t i s to make the j o i n t i n c l i n e d a t about 15 or 20 degress t o the axis of a tension test specimen (figured i n the r e p o r t ) . The stresses i n the j o i n t are then a shear and a tension and, i f the angle i s between the l i m i t s given above, the stress causing f a i l u r e w i l l generally be the shear s t r e s s . 15. The report concludes with the outline of a procedure for t e s t i n g adhesives, based on the experience gained i n the work mentioned. The above gives the barest synopsis of the report, which should be oonsulted i n i t s o r i g i n a l form by those interested i n strength tests of adhesives f o r timber. 16. Since the issue of Major Robertson's report, the a t t e n t i o n of the Committee has been directed to other i n v e s t i gations, and the report consequently represents the stage t o which the enquiry has been brought; i t i s hoped to r e t u r n to the subjeot at a l a t e r date. From various r e f e r e n c e s 1 8 , 42) ^  a  pp  e a r s  serious t e s t i n g of adhesives was done during World War  that the f i r s t I t o meet the war-  time demands of the (wooden) a i r c r a f t industry. During t h i s period meoh-  - 8 a n i o a l t e s t s of glued j o i n t s appear t o have been developed  independently  by the Royal A i r o r a f t Establishment a t Farriborough, and the United States Forest Products Laboratory, Madison, Wisoonsin, eaoh having developed t e s t speoimens of d i f f e r e n t design* The two designs developed by the U.S. Forest Products Laboratory, known as the "Block-Shear" and the "Plywood-Shear" t e s t s , have been described by several a u t h o r s ^ * 40, 4 l )  #  Some eight or more designs of t e s t  specimen were investigated by the ( B r i t i s h ) Adhesives Research Coinmitfcee between 1920 and 1932^ '» 1-  ^ )  #  o  r  m  o  r  e  0  f these had been used  during World War I f o r purposes of glue s p e c i f i c a t i o n *  A f t e r an extensive  i n v e s t i g a t i o n two designs were recommended as the best of those considered* One, the "Simple Lap Test-Pieoe" was designed f o r shear t e s t 3 , and the other, the "Spandau Test Piece" f o r t e s t i n g adhesives i n t e n s i o n ^ * ^ )  #  Under the pressure of the F i r s t World War great strides were made i n both p r a c t i o a l and t h e o r e t i o a l f i e l d s of adhesion i n America and Britain*  The researches carried out at t h i s time by such men as Anderson,  Browne, Brouse, Hopkins, Lee, McBain, Robertson and many others were thorough*  These workers were f u l l y aware of the imperfeotions, as w e l l as the  merits, of t h e i r (mechanical) t e s t methods.  They r e a l i z e d that breaking  loads are not stresses, t h a t every change i n the dimensions  of a test  specimen a l t e r e d the breaking load, and a f t e r taking every precaution t o standardize conditions and procedures a very large v a r i a t i o n would remain between their test r e s u l t s .  I n faot t h e i r work was so sound that t h e i r  recommendations have formed the basis of most consequent adhesive t e s t i n g * In 1945 Platow and D i e t z ^ ) presented another specimen design 2  for t e s t i n g bonds i n tension. A new approach was set f o r t h , that of d e l i b e r a t e l y introducing stress concentrations i n the glue l i n e instead of  attempting to eliminate them* The wood-to-wood speoimen has been designed to i s o l a t e glue f a i l u r e s from wood .failures by incorporation of a notch to localize stresses.^) In recent years several new mechanical t e s t methods have been developed.  Wakefield^)  desoribed a "Tension Normal to the Glue Line  Plywood Test" designed "bo give a positive and quantitative value t o the adhesion between p l i e s that can be duplicated within normal experimental error".  Rudkin^^) presented "A Simple Method of Testing Glue Lines i n  Tension", whereas Elmendorf(17) used the "Torsion Shear Test" f o r plywood. More reoently Laoey and Howe^ ?) have desoribed a "Glueline Method", employ2  ing a wedge to oleave a speoimen along the glue l i n e , while Elmendorf has introduced a v a r i a t i o n of the "Cantilever" t y p e ^ ^ ) developed by Robertson f o r the Adhesives Researoh Committee. The work of the Adhesives Researoh Committee^'*  15)  &  i\  based upon breaking loads obtained from meohanioal t e s t s of glued j o i n t s . Wood f a i l u r e was a nuisance which they t r i e d to minimize i n t h e i r  attempts  to measure the strength of (strong) glues by means of wooden t e s t pieoes. I t was quite clear t o these workers that t h e i r i n t e r e s t lay i n the strength of the bond and t h e i r objeotive was to devise means of measuring t h i s bond strength.  One of the deciding factors i n t h e i r choice of a standard t e s t  piece was as followst I t was found that i f the simple-lap t e s t pieoe were assembled with speoifio a t t e n t i o n to g r a i n direotion, gross timber (wood) f a i l u r e could be eliminated. To obtain a minimum of timber (wood) f a i l u r e , the timber should be out so that the grain makes a small angle with the t e s t pieoe. In t h e i r researches no reoord has been presented of the per cent wood f a i l ure ooourring i n the broken specimens.  - 10 The e a r l i e s t U.S. Forest Products Laboratory r e p o r t ^ ) whioh i t has been the author's p r i v i l e g e t o read, when r e f e r r i n g t o the Plywood Shear Test r e s u l t s , states:  "The glued surface must not f a i l a t a load of  less than 150 pounds per square inch".  No mention i s made of per oent  wood f a i l u r e . . Another Forest Produots Laboratory r e p o r t ( ^ l ) states: For eaoh (block shear) specimen tested, notation i s made of the breaking load and the estimated percentage of the glue l i n e area i n whioh the wood s p l i n t e r s . Two or more duplicate j o i n t s are usually prepared, each one giving 10 speoimens f o r t e s t . The average and the minimum breaking load and the average percentage of wood f a i l u r e are generally taken as the f i n a l reoord of the t e s t . No mention i s made of how muoh weight i s given t o each i n evaluating the q u a l i t y o f the bond.  Truax (40) has discussed the f a l l a o y of using wood  f a i l u r e as a means of evaluating the strength of glued j o i n t s . It appears that early researchers accepted measurement of bond strength as t h e i r objective. The i n t e r p r e t a t i o n of r e s u l t s was complicated by a laok of knowledge regarding the stress d i s t r i b u t i o n s e x i s t i n g i n speoimens a t the time of f a i l u r e , and by the great number of souroes of variabi l i t y with whioh they had t o oontend. It appears that by 1938 serious consideration was being given t o the use of per oent wood f a i l u r e as a o r i t e r i o n of bond quality, a t least for  plywood s p e c i f i c a t i o n purposes i f not f o r researoh purposes, as witness  the following quotation from Perkins  (36), Fortunately, the Forest Produots Laboratory d i d have, (experience with t e s t s f o r E x t e r i o r Type Plywood) from panels of various kinds of plywood whioh had been weathering f o r nearly f i v e years. I t was t h e i r conclusions that i f standard plywood shear speoimens revealed 50 per oent wood f a i l u r e a f t e r having been subjected either t o three and a h a l f cyoles of cold soaking and drying or to two cycles of b o i l i n g and 145°F. drying, that the panel oould be expeoted t o have a long l i f e i n s e r v i c e . Just how long, no one was w i l l i n g or able t o prediotj ... One of the f i r s t impressions gained i n t h i s study  -  11  -  was the importance of the amount of wood f a i l u r e developed when t e s t i n g the j o i n t s . It i s not clear whether or not t h i s use  of per cent wood f a i l u r e as a sole  estimate of bond quality developed from a oonviotion that i t was method, or whether i t was  a superior  put forward as a convenient substitute for break-  i n g loads with a l l t h e i r vagaries.  At any r a t e , by 1950 the Douglas F i r  Plywood Association had b u i l t up data whioh tended t o confirm t h e i r " b e l i e f i n a close r e l a t i o n s h i p between wood f a i l u r e and d u r a b i l i t y "  The  breaking loads of t h e i r specimens used f o r per oent wood f a i l u r e determinat i o n do not show appreciable possible explanations  c o r r e l a t i o n with d u r a b i l i t y . There are  for t h i s laok of c o r r e l a t i o n .  a change i n the thickness  Brouse^) has  several  shown that  of any veneer from that of the ohosen standard  ohanges the anticipated breaking load for that speoimen.  These a n t i c i p a t e d  breaking loads for yellow b i r c h 3 - p l y plywood varied from 120 pounds f o r 1/64" face and baok veneers and  l / 8 " oore veneer t o 7 4 0 pounds for l / 8 " faoe  and back veneers and l / 6 4 " oore veneer.  I t appears that t h i s faotor was  not  kept under control and that plywoods of various thicknesses were included i n the Douglas F i r Plywood Association t e s t s .  Bethel and Huffman^ ) have shown 6  that the o r i e n t a t i o n of lathe checks with respect to the saw  cuts can make  highly s i g n i f i c a n t differences i n the values of breaking loads and per cent wood f a i l u r e s obtained.  There i s no i n d i c a t i o n that t h i s faotor was  con-  t r o l l e d i n the Douglas F i r Plywood A s s o c i a t i o n work. The Douglas F i r Plywood A s s o c i a t i o n ^ ' ^ was  s a t i s f i e d that per oent A  wood f a i l u r e , as read from stanflard Plywood Shear specimens subjected to 4 hours b o i l i n g , 20 hours drying, and 4 hours b o i l i n g , provides an estimate of the bond quality superior t o that obtainable nitude of the breaking loads.  from considering the mag-  - 12 In 1949 Laoey and Howe^ "^, i n discussing the ohoioe of a measure 2  of bond q u a l i t y f o r researoh purposes, stated: The choice of methods l a y i n the f i r s t plaoe between a meohanioal strength test and the v i s u a l examination of glue lines that have been s p l i t open* Experience at the Laboratory with veneer-base joints has favoured the v i s u a l t e s t as being more sensitive t o i n i t i a l degrade, quicker to carry out and independent of s p e c i a l t e s t i n g f a c i l i t i e s , and more convenient f o r examining large areas and p a r t i c u l a r l y for ready-made j o i n t s . No s o l i d wood t e s t piece has, however, been developed f o r examination i n t h i s way. I t seems clear that had they been studying veneer bonding problems they would have used either the v i s u a l or the per cent wood f a i l u r e method i n preference to the r e s u l t s of mechanical t e s t s . According t o K n i g h t ^ ^ ) , " i n B r i t i s h s p e c i f i c a t i o n s , wood f a i l u r e 2  has no plaoe, although most t e s t e r s agree that adhering f i b r e i s desirable". I t appears from N e w a l l ^ ) that a move i s afoot t o introduce per cent wood 2  f a i l u r e as an a l t e r n a t i v e t o the meohanioal t e s t .  The existenoe of the  following statement i n B.S. 1455 (1948)^/, and B.S. 1088 (1951)2/, "by means of a knife test on plywood of any species or thickness, or pieces of any convenient size being tested dry or a f t e r appropriate water treatment", would seem t o indicate that per oent wood f a i l u r e , or something olosely akin to i t , d e f i n i t e l y has a plaoe I n B r i t i s h s p e o i f i o a t i o n s . On the basis of the author's experience t h i s trend, while i t may be j u s t i f i a b l e from an adhesive c e r t i f i c a t i o n or production t e s t i n g standpoint, i s t o be deplored from a researoh standpoint.  I t i s not possible t o  t e l l i f a bond i s adequate when no attempt i s made t o measure i t s strength. I t i s conceivable that t h i s strength may be approximated by using other than meohanioal t e s t s , but only a f t e r the bond strength has been estimated by the use of these t e s t s .  I f per cent wood f a i l u r e i s t o be used as an estimate  4/ — B r i t i s h Standards I n s t i t u t e S p e c i f i c a t i o n .  - 13 of bond strength a c o r r e l a t i o n must be established between per oent wood f a i l u r e and meohanioal t e s t data*  A f t e r such a r e l a t i o n s h i p has been  established bond strength could then be predicted from per oent wood f a i l ure*  I t i s axiomatic therefore that adequate mechanical t e s t s are fund-  amental to a l l are  (wood) glue and gluing research*  These (meohanioal methods)  the basic tools used t o evaluate the r e s u l t s of a l l other methods of  t e s t , be they physioal, chemical, pathological or combinations of these and others* It i s not intended to imply that t o be useful the r e s u l t s of meoha n i o a l tests must be reduoed t o glue l i n e stresses, although t h i s would be desirable. are  Breaking loads as measured by mechanical t e s t s , provided they  p o s i t i v e l y correlated with the a c t u a l stresses i n the glue l i n e , may be  equally useful f o r many purposes* A fundamental consideration i s that the strength of the bond i s of major i n t e r e s t , not that of the adhesive i t s e l f .  An adhesive whioh  develops the f u l l strength of the wood under the required service conditions f u l f i l l s a l l requirements except possibly that i t might be less expensive or more t r a c t a b l e .  No more oan be expected of the wood, and attempts t o  estimate the strength of a strong adhesive through the medium of weak wood (and many of them have been made) have always been doomed t o f a i l u r e * A drawback t o the use of per oent wood f a i l u r e as an absolute measure of bond quality i s the great difference i n the appearance  of j o i n t s  which may be induced by a l t e r i n g the angle, of g r a i n with reference t o the d i r e c t i o n of a p p l i c a t i o n of the load or saw outs. Test Pieoe(*4» 15)  The ( B r i t i s h ) Simple Lap  purposely designed t o minimize per cent wood f a i l u r e  so that any representative estimate provided by these specimens would be muoh lower than those provided by the (United States) Block Shear Test(l» 42) ^  - 14 or the (Canadian) Tension Shear Test P i e o e ^ * ^ .  These objections are min-  imized f o r s p e c i f i c a t i o n t e s t i n g of one speoies of wood, one class o f adhesive (e.g. phenolic r e s i n adhesives), and one standardized design of t e s t piece.  This i s the s i t u a t i o n with the Douglas F i r Plywood Association,  and the use of per oent wood f a i l u r e as a measure of bond q u a l i t y may w e l l be j u s t i f i e d on the grounds of expediency. The Douglas F i r Plywood Association, working under conditions that favour the use of per oent wood f a i l u r e , found i t necessary t o teach estimators how to "read" per cent wood f a i l u r e so that data colleoted by d i f f e r e n t persons, or the same person a t d i f f e r e n t times, w i l l agree w i t h i n acceptable limits.  Their method i s t o c i r c u l a t e standardized sets of specimens from  estimator to estimator and require that each check his readings against the standard.  This standard, although adopted only a f t e r c a r e f u l study by ex-  perts i n the wood gluing research f i e l d , i s s t i l l an a r b i t r a r y one. Another group of experts would quite l i k e l y have chosen a d i f f e r e n t standard with the same specimens, and would almost c e r t a i n l y have a r r i v e d a t a d i f f e r e n t standard had they used a d i f f e r e n t design of t e s t piece or d i f f e r e n t species. I f the Douglas F i r Plywood A s s o c i a t i o n has found i t neoessary t o take these precautions t o coordinate the estimates made by d i f f e r e n t "readers", how much more elaborate a system i s required i f per oent wood f a i l u r e i s t o be used under less suitable conditions. Different test pieces, species, p l y wood and laminated wood, d i f f e r e n t adhesive classes, to say nothing of estimators from laboratories spread across the world greatly increase the hazards associated with using per cent wood f a i l u r e as a universal estimate of bond q u a l i t y .  I f t h i s i s t o be used as a measure of bond q u a l i t y a world-  wide exchange of standardized specimens i s required.  Photographio keys or  written descriptions have not proven s a t i s f a c t o r y substitutes.  Other  - 15 references(4» 24» 26, 31) have been h e l p f u l i n dealing with per cent wood failure. In spite of the defects of mechanical test r e s u l t s , such as excessive v a r i a b i l i t y and inaoourate knowledge of how the breaking loads should be interpreted (due t o inacourate knowledge of the stress d i s t r i butions a t the moment of f a i l u r e ) , they s t i l l oome closer t o a c t u a l l y measuring the q u a l i t y of a bond than do other so-called "measures"• At best, other estimates such as per oent wood f a i l u r e , can only be used as substitutes f o r the mechanioal t e s t a f t e r the two have been shown t o be correlated.  Only then can the mechanical strength of a bond be predicted  from per cent wood f a i l u r e readings.  Furthermore, the defects of the mech-  a n i c a l t e s t s appear to be more amenable to s o l u t i o n through further research than do those associated with per cent wood f a i l u r e ; c e r t a i n l y t h i s i s one of the impressions gained from the research herein reported.  When a l l things  are taken i n t o consideration i t appears that the r e s u l t s of meohanical t e s t s , i n spite of their imperfections, provide a more u n i v e r s a l l y applicable standard of glue l i n e q u a l i t y than does per cent wood f a i l u r e .  - 16 CHAPTER I I I - MECHANICAL TESTS OF PLYWOOD GLUE BONDS 1.  THE MOST DESIRABLE STRESS DISTRIBUTION IN A TEST SPECIMEN I t i s i n t e r e s t i n g t o note that when meohanioal t e s t i n g of ad-  hesives was undertaken by the Adhesives Researoh Committee (-^) the aims appear t o have been t o develop test speoimens, and methods of t e s t i n g them, whioh would apply uniform shear or uniform t e n s i l e stresses over the e n t i r e area of the bond i n order t o obtain a true measure of the shear or t e n s i l e strength of the bond. The authors of the F i r s t R e p o r t ^ o f that Committee had t h i s t o say i n reference t o tension tests made along the graint Two types of t e s t pieoe formerly used a t the Royal A i r o r a f t Establishment were investigated and found t o be unsatisfactory on account of the unequal d i s t r i b u t i o n of stress produced by the bending of the specimen when loaded. In reference t o shear t e s t s i Major Robertson's report further deals with shear t e s t s , and an examination i s made of the various types of t e s t pieoes whioh have been used by d i f f e r e n t experimenters. I t i s shown that the majority of these are unsatisfactory f o r i n none i s the stress on the j o i n t a simple shear. I t i s clear that t h e i r objective was t o obtain uniformly d i s t r i b u t e d (pure shear or pure t e n s i l e ) stresses over the t e s t area.  It i s  equally clear that they were not s a t i s f i e d with the r e s u l t s obtained. approach i s i n contrast t o the views of some of today's i n v e s t i g a t o r s . When discussing d i f f e r e n t t e s t p i e o e s ^ 4 ) the Committee says» I t has a t various times been suggested that glued j o i n t s should be tested under stress conditions approaching as close as possible t o pure shear i n the plane of the j o i n t . In a l l attempts t o obtain a pure shear stress i n p r a c t i c e , bending or d i s t o r t i o n of the parts has introduced secondary stresses whioh have probably influenced the actual f a i l u r e s . Sinoe, however, suoh an effeot usually occurs i n the f a i l u r e of actual glue j o i n t s i n servioe, there does not appear t o  This  be any l o g i c a l reason why adhesives should be tested by any close approximation of pure shear. Both test pieoes which were eventually recommended by the  Com-  m i t t e e ^ S ) , the Simple Lap Test Pieoe (Tension-shear) and the Spandau Test Pieoe (end-grain j o i n t tested i n bending), were types which d i d not give even a close approximation t o pure shear or pure t e n s i l e s t r e s s e s . Data are g i v e n ^ 5 ) to show that breaking loads of simple lap (tension-shear) specimens of d i f f e r e n t areas are not constant when reduoed to pounds per square inch of glue l i n e t e s t e d , but vary from 1,252  p . s . i . for a  one-inoh  lap to 2,334 p . s . i . f o r a one-quarter-inoh l a p . With t h i s type of speoimen,  therefore, d i v i d i n g the breaking load by the oross section of the glue  l i n e does not give a r e a l i s t i c measure of the magnitude of the stresses to which the glue l i n e has been subjeoted.  The r e s u l t s of t h i s type of t e s t  should be recorded as breaking loads i n pounds rather than as pounds per square inch, a t l e a s t u n t i l an acceptable stress analysis i s performed. Only then may the r e s u l t s be reported as maximum stress on a unit b a s i s . What constitutes the most desirable stress d i s t r i b u t i o n i n a t e s t speoimen?  The following quotations from Platow and D i e t z ^ indicates that  t h i s problem was  s t i l l not resolved as reoently as  1945.  Strength t e s t i n g of adhesives i s a peculiar problem because i n any u s e f u l t e s t specimen the adhesive being tested i s only a minor part of an assemblage, and the only way to exert any foroe upon the adhesive i s through the m a t e r i a l bonded, or substrate. Ho f e a s i b l e way of grasping the adhesive d i r e c t l y has been found. Complicating the problem i s l i m i t e d knowledge of the nature of adhesion. .Although i t appears to be generally accepted that adhesion i s primarily an a c t u a l adherence of the adhesive to the surface of the material bonded and not some mechanical keying or anohoring phenomenon, no matter whether the surfaces be porous or p e r f e c t l y smooth, l i t t l e i s known about the forces, moleoular or otherwise, which cause the adhesive and the substrate to c l i n g t o each other. Consequently, i n a l l strength tests a considerable degree of  - 18 uncertainty e x i s t s as to what i s being tested* A further complication i s that i n any strength t e s t of adhesives a t least two strength factors are being t e s t e d simultaneouslyt the strength of adhesion of adhesive t o the adjacent surface, and the strength of oohesion of the adhesive i t s e l f . Since the only f e a s i b l e way t o exert foroe upon the adhesive i s through the substrate, i t follows that the substrate i s being tested at the same time as the strengths of adhesion and cohesion* The r e s u l t i s three simultaneous strength t e s t s , with the ohanoes often more than even that f a i l u r e w i l l not ocour i n the adhesive at a l l * Keeping these complications i n mind, there are two p r i n c i p a l approaches t o the problem of devising strength t e s t s for adhesives. One, whioh may be c a l l e d the 'purist' or aoademio approach, i s to devise t e s t s i n whioh only pure stress of one k i n d — f o r example, tension or s h e a r — i s exerted on the adhesive, and a t the same time care i s taken to assure f a i l u r e i n the adhesive, i n either adhesion or oohesion, and not i n the subs t r a t e . A second approaoh.is the more pragmatic one of using a t e s t specimen r e l a t i v e l y easy to make; t e s t i n g i t i n as simple a manner as possible consistent with obtaining more or less pure stresses; and taking the viewpoint that i f f a i l u r e a c t u a l l y oocurs i n the bonded material, rather than i n the adhesive, i t proves that the adhesive i s the stronger, therefore has successfully passed the t e s t , and no more need be asked. I f f a i l u r e occurs i n the adhesive, so much the b e t t e r , since numbers oan be reported representing strength values under the conditions of t e s t . There are objections to both, procedures. While perhaps fundamentally sound, the p u r i s t ' s approach i s d i f f i c u l t and may e a s i l y require rather elaborate equipment as w e l l as complicated t e s t speoimens to assure pure s t r e s s . Furthermore, with strong adhesive and oohesive forces and weak substrates, i t may be almost impossible t o obtain f a i l u r e s i n the adhesive i t s e l f without so a l t e r i n g the conditions as to destroy t h e i r p r a o t i c a l significance. The more pragmatio approaoh i s open to the c r i t i c i s m that i f combinations of stress are present i n a t e s t , there i s no sure way of determining whioh stress was mainly responsible for the f a i l u r e and therefore there i s no good way of t r a n s l a t i n g the r e s u l t s i n t o pure strength values f o r the use of designers. The oonsequenoe i s to force the employment of a great many d i f f e r e n t kinds of test specimens to r e f l e c t the d i f f e r e n t types of assemblage that might be found i n actual use. A point i n favour of suoh a procedure i s , of course, the faot that i n any complex struoture the designing engineer seldom i s c e r t a i n of the exact d i s t r i b u t i o n of various kinds of stress i n a j o i n t and that i t may for that reason be necessary to t e s t any given j o i n t t o determine i t s strength under  - 19 oauditions simulating a c t u a l use* On the other hand, the designer could approach his design with more confidence i f he were sure that the strength values he was using a c t u a l l y represented those strengths uncomplicated "by other factors™/. These arguments help to foous the reader's a t t e n t i o n upon the problems involved but seem to confuse rather than o l a r i f y the issue* far  As  as a designer i s oonoerned, a l l he requires i s that the bond "develop  the f u l l strength of the wood"^/ under the required conditions of servioe* Once assured of t h i s he has no further i n t e r e s t i n the adhesive but uses the strength properties of the wood for his purposes*  Inasmuoh as there  are many adhesives of t h i s c a l i b e r a v a i l a b l e there seems to be l i t t l e need for  pure stress values. Platow and Dietz (2) appear to have resorted to the "more prag-  matio" approaoh because they designed a t e s t speoimen to d e l i b e r a t e l y concentrate the maximum stresses i n the glue l i n e . ... The wood-to-wood speoimen has been designed to i s o l a t e glue f a i l u r e s from wood f a i l u r e s by the incorporation of a notoh t o l o c a l i z e s t r e s s e s ^ ' . 2  It i s i n t e r e s t i n g to note that photoelastio a n a l y s e s ^ * 2  ^»  ^9)  indicate that the stress d i s t r i b u t i o n over a seotion subjected to a "pure" stress i s not uniform across the seotion.  The p u r i s t ' s dream of achieving  a uniformly d i s t r i b u t e d pure t e n s i l e , shear, or compressive stress was, i t appears, foredoomed to f a i l u r e . The aotual d i s t r i b u t i o n of stresses i n a s p e c i f i c wood-to-wood bond w i l l probably never be known due to the number of contributing faotors among which may be mentionedt —  —  —  — Phrases suoh as t h i s or ( r e f e r r i n g to the strength of a glue) "weaker than the wood", "stronger than the wood", etc., w i l l be found i n the t e x t . I t i s realized.that these are inexact statements of the e x i s t i n g oondition but they express the idea more c l e a r l y than a more involved statement would.  - 20 (a)  the s t r e s s - s t r a i n c h a r a c t e r i s t i c s of the wood,  (b)  stress concentrations e x i s t i n g i n the wood before gluing,  (0)  the s t r e s s - s t r a i n c h a r a c t e r i s t i c s of the adhesive,  (d)  stresses induced i n the adhesive by shrinking or swelling when during,  (e)  stresses induced i n the j o i n t by swelling of the wood due to moisture absorbed from the adhesive,  (f)  stresses induced i n the j o i n t by swelling or shrinking due t o moisture picked up or l o s t by the wood a f t e r being bonded and before being tested, the most severe stresses probably being induced by soaking or b o i l i n g the specimens p r i o r t o t e s t ,  (g)  stress concentrations induced i n the wood due t o d i f f e r e n t i a l swelling or shrinking of springwood and summerwood with ohanges i n moisture content of the wood,  (h)  the pattern of stresses induced during t e s t i n g of the specimen; each method of t e s t , size of specimen, speoies of wood, e t c , w i l l induce a d i f f e r e n t pattern, and  (1)  the e f f e c t of the moisture content of the wood upon i t s s t r e s s strain characteristics. In spite of the apparent i m p o s s i b i l i t y of obtaining aoourate  knowledge of the stress d i s t r i b u t i o n i n i n d i v i d u a l specimens a t the moment of f a i l u r e there i s , nevertheless, some hope of reducing the v a r i a b i l i t y i n t e s t r e s u l t s through stress analyses.  The accuracy of estimates based on  these r e s u l t s can be improved by more intensive study of t h i s stress d i s t r i b u t i o n problem. 2.  METHODS OF REDUCING VARIABILITY Our p r i n c i p a l objects were f i r s t to discover i f possible the causes of the wide variations whioh occur between the r e s u l t s  - 21 of i n d i v i d u a l t e s t s of the same glue when any of the usual form of t e s t pieoe i s employed and then to develop a form of test pieoe from which more concordant results might be obtained* This quotation, from the Second Report of the Adhesive Researoh Committee(H), could well have been used i n the foreword of t h i s t h e s i s . The problems s t i l l remain.  The sources of v a r i a t i o n were investigated  rather exhaustively under the Committee's auspices, and complete i n s t r u c t i o n s were prepared f o r t h e i r minimization under the following headingsj I.  Timber (a) (b) (c) (d) (e) (f)  II.  Glue (g) (h) (i) (j) (k)  III.  Moisture content. Structure, e t c . Mechanical preparation of the surface. F i t of glued surfaoes. I n c l i n a t i o n of grain. Temperature at -time of gluing.  I n i t i a l variations i n glue. Previous h i s t o r y . Proportions of constituents, Temperature of heating. Duration of heating.  Operation of gluing (1) Method of a p p l i c a t i o n . (m) Room temperature and humidity. (n) Delay i n bringing surfaoes i n contact. (o) Clamp pressure. (p) Time under pressure. (q) Aoouracy of pressure j i g . (r) Rate of cooling. (s) Glue f i l l e t . (t) Temperature and humidity of .conditioning chamber, (u) Time of conditioning.  IV.  Testing (v) Acouraoy of alignment of specimen. (w) Loading g r i p s . (x) Temperature. (y) Humidity. (z) Rate of loading.  - 22 This work was overlooked.  so w e l l done that few, i f any, sources of v a r i a t i o n were Other investigations have oonfirmed the conclusions regarding  these sources of v a r i a b i l i t y 4 2 )  #  Having done t h e i r utmost to minimize the v a r i a b i l i t y a t t r i b u t a b l e to these sources the a u t h o r s ^ ^ ) s t i l l lament that, One of the most unsatisfactory features associated with a l l present methods of t e s t i s the large v a r i a t i o n which occurs, the cause of whioh has not yet been traoed. In any one group of t e s t s the difference between the maximum and minimum values i s usually of the order of 20 to 30 per oent of the arithmetic mean. Claims for smaller variations should not be accepted without i n v e s t i g a t i o n , as the anomalous p o s i t i o n e x i s t s that i f the t e s t s are suoh as to allow timber f a i l u r e , the apparent v a r i a t i o n i s frequently reduced, the strength of timber being usually more constant than the apparent strength of the glued joint. Variations appear t o e x i s t , not only among joints from any one group, but a l s o among groups which have been prepared under apparently similar conditions. This increases the a d v i s a b i l i t y of using a large number of t e s t pieces i n each group to give greater accuracy i n the mean r e s u l t , but owing to experimental d i f f i c u l t i e s (provision of j i g s , etc.) the number composing each group i s very l i m i t e d . It should therefore be remembered that i n many of the experiments here described, the probable aoouraoy of the mean r e s u l t s i s i n s u f f i c i e n t t o allow of d e f i n i t e conclusions being formed, and although the t e s t s are useful i n i n d i c a t i n g the probable trend, i n p a r t i c u l a r oases the i n d i c a t i o n may be erroneous. To a s s i s t the estimation of the amount of r e l i a n c e whioh may placed on the r e s u l t s , the mean errors have been calculated using Peter's approximation. These have been given i n two forms: (a)  Plus and minus l i m i t s on the arithmetic mean. (These give the l i m i t s within which the true r e s u l t i s as l i k e l y to l i e as outside).  (b)  The probable percentage error of a single observation whioh gives some guide to the r e l a t i v e degree of v a r i a t i o n which i s ooourring among groups containing similar numbers of r e s u l t s .  The s i t u a t i o n i s not greatly ohanged except i n one  be  important  - 23 respectj the science of s t a t i s t i c s has been developed t o a s s i s t i n analyzi n g variable data suoh as the test r e s u l t s of glued j o i n t s .  To t h i s can  be added the inoreased knowledge regarding d i s t r i b u t i o n of stresses i n stressed structures. The theory of s t a t i s t i c s , especially C o r r e l a t i o n Analysis and Analysis of Variance techniques, provides one of the most, powerful tools a t the disposal of the researcher.  S t a t i s t i c a l methods  offer a means of segregating controlled v a r i a t i o n induced by treatments from the uncontrolled or "error" varianoe on a mathematical b a s i s . Early workers laoked these methods except i n the very primitive form mentioned i n the above quotation. The d i s t r i b u t i o n of stresses i n a specimen i s obtainable through the use of either photoelastic or complicated mathematical analyses. Certain of these analyses have dispelled any suggestion that stresses a c t uniformly over the t e s t area.  The influenoe of lathe check orientation on  Plywood Shear Test Results (6) i s another source o f v a r i a t i o n for whioh a method of reducing v a r i a b i l i t y has been introduced only reoently. Another a i d i n the campaign t o reduce v a r i a b i l i t y i s the use of the r a t i o breaking load of glue l i n e breaking load of the wood  instead of a breaking load.  used a method applicable with the Blook Shear Test.  Orth(35)  Marra and W i l s o n ^ ^ )  have used the same p r i n c i p l e f o r what they c a l l a " g l u a b i l i t y index". Selbo and Olson(39) employed the same p r i n c i p l e for presenting data on a v a r i e t y of joint types.  This p r i n c i p l e , applied t o plywood, has been used  i n the research recorded i n t h i s t h e s i s . A further method of increasing the aocuraoy of mechanical  test  i n t e r p r e t a t i o n concerns the f a l l a c y of attempting t o t e s t a strong adhesive through the medium o f weak wood.  Much research e f f o r t has undoubtedly been  wasted because the worker attempted t o measure the strength of strong glues  - 24 by means of weak wood.  In order t o minimize recurrence of t h i s d i f f i c u l t y  the following method i s proposed i n contrast t o the normal praotioe of using an adhesive only as recommended by the manufacturer. that several formulae of t h i s adhesive be used.  I t i s proposed,  These a d d i t i o n a l mixtures  d e l i b e r a t e l y include some "weaker than the wood" and, when used i n conjunct i o n with the method suggested i n the next paragraph, the pattern of t e s t r e s u l t s w i l l indicate which of the adhesive formulae, i f any, form bonds "as strong as the wood".  This i s the information required t o avoid the  f r u i t l e s s task of t r y i n g t o draw conclusions regarding the strength of an adhesive whioh i s "stronger than the wood". The following i s another method whioh should add t o the accuracy of predictions based on the results of mechanical t e s t s .  The p r i n o i p l e i s  to base the judgment of bond q u a l i t y upon the rate of strength reduction induced by a series of "accelerated weathering" oycles of increasing severi t y , i . e . , upon the rate at which the strength of the bond i s reduced rather than upon the absolute strength. 3.  THEORIES REGARDING SOURCES OP EXCESSIVE VARIABILITY IN TEST RESULTS A review of the inherent defeots of mechanical methods was r e -  s t r i c t e d t o a few commonly accepted t e s t s : to the Glue Line, and Block Shear. reasons:  Tension Shear, Tension Normal  This r e s t r i c t i o n was imposed for two  f i r s t l y , the author has had some experience with these p a r t i c u l a r  t e s t s , and secondly, most other types already have been discarded by competent a u t h o r i t i e s . The f i r s t step was t o consider i n some d e t a i l a l l the factors whioh might be responsible f o r the great v a r i a b i l i t y of the t e s t r e s u l t s , only the more important  of which w i l l be discussed.  - 25 In s t a t i s t i c a l theory, the Total V a r i a t i o n occurring i n a populat i o n of t e s t r e s u l t s i s the sum of a l l the i n d i v i d u a l sources of V a r i a t i o n . In t h i s instance the Total V a r i a t i o n i s the sum of those inherent i n the wood, the design of the t e s t speoimen, the adhesive, the t e s t i n g machine, the workmanship, and f i n a l l y that d e l i b e r a t e l y introduced f o r the purpose of study.  These are commonly segregated i n t o two classes, one a t t r i b u t e d  t o "treatments" and the other to " e r r o r " . The i d e a l t e s t specimen and method of t e s t i n g should be such that a l l v a r i a t i o n , except that purposely introduced f o r study, w i l l consequently be a t t r i b u t a b l e to the wood and the adhesive and none t o the method of t e s t , workmanship, or other oontrollable f a c t o r s .  Many broken t e s t specimens were  examined f o r the purpose of locating sources of v a r i a t i o n which would lend themselves t o corrective a c t i o n .  The more important of the theories con-  sidered w i l l be described b r i e f l y . (A)  Tension Normal to the Glue Line Plywood Test(45)  In t h i s method of test the load i s applied perpendicular to the glue l i n e so that a l l sections p a r a l l e l to the glue l i n e are, t h e o r e t i c a l l y , equally stressed.  I f wood were a homogeneous material and workmanship  perfect t h i s would probably be true.  There are, however, suoh factors as  differences i n the physioal and mechanical properties of springwood and summerwood, imperfections  suoh as minute seasoning cheoks, manufacturing  defects, and the addition of two extra glue l i n e s .  I t seems i n e v i t a b l e  that a stress concentration w i l l be created at some section i n the t e s t pieoe other than the bond under study, and that f a i l u r e at that point w i l l r e s u l t except when the bond i s very weak.  This was  i l l u s t r a t e d by the  broken speoimens, only a minority of which had broken i n the glue l i n e or i n i t s immediate v i o i n i t y .  - 26 (B)  Tension Shear T e s t ^ * 9  1 0  »  5 7  »  42)  In t h i s t e s t method the specimen i s designed so that stress concentrations are set up at c e r t a i n seotions, notably at the base of the cuts.  saw-  Depending upon the q u a l i t y of the workmanship this,may or may not be  at the glue l i n e where a stress concentration would be of value.  I t was  reasoned that t h i s causes the wood of many speoimens to f a i l f i r s t i n tension, followed by oleavage, whereas the objective i s to t e s t the joints i n shear.  I t was reasoned further that, apart from the design of the t e s t  specimen, two factors would tend to introduce undesirable v a r i a b i l i t y i n t o the t e s t r e s u l t s .  These factors are s l i g h t differences i n the depth of  the saw outs and the pattern of springwood and summerwood at the gluing faoe. For glue aooeptanoe tests these factors have been minimized by using the most suitable speoies; for production t e s t i n g they have been accepted as inevitable conditions of the t e s t ; but for researoh into the gluing properties of a p a r t i c u l a r species the issue oannot be evaded by substituting another species, neither i s i t desirable to aooept these conditions as i n e v i t a b l e . (C)  Block Shear T e s t ^ ' 1  4 0  *  4 2  )  In t h i s t e s t method the foroe i s applied so as to produce a (compressive)  shearing stress i n the plane of the bond.  This method as  described^ * 40, 42) probably comes as close as any to producing a uniformly 1  d i s t r i b u t e d (shear) stress over the j o i n t area, the i d e a l sought by e a r l y workers (14). When an attempt was made t o use the method with specimens modified to suit plywood the wood f a i l e d i n compression p a r a l l e l to the grain before the bond was ruptured.  This made i t neoessary to glue blocks  of wood to the specimen (see Figures 7 and 7A. Appendix A) before the j o i n t  - 27 oould be sheared, 4.  THEORIES REGARDING THE DESIGN OF A BETTER TEST METHOD The major requirement of a good t e s t method i s that the r e s u l t s  s h a l l be reproducible.  With bonds of uniform q u a l i t y the method of t e s t  with l e a s t v a r i a b i l i t y should be the best.  With p e r f e c t l y uniform materials  the i d e a l test would always give the same breaking  load.  I t i s axiomatic  that the r e s u l t s of a t e s t ishould be expressed i n u n i v e r s a l l y acceptable and unequivooal terms.  Unit stresses or breaking  loads i n pounds for a  standardized t e s t specimen are reasonable examples.  Probably the next most  important requirement i s the s i m p l i c i t y of the t e s t speoimen. whioh i s d i f f i c u l t to prepare f a l l s short of p e r f e c t i o n .  A specimen  In addition, the  cost of the equipment required t o prepare and test the specimens should be excessive.  not  F i n a l l y , i t i s desirable, though not e s s e n t i a l , that a t e s t  method be adaptable for either researoh or production t e s t i n g . To meet the requirement that t e s t s of uniform q u a l i t y bonds s h a l l exhibit minimum v a r i a b i l i t y the method would be expeoted t o incorporate as many as possible of the following features: 1.  The  load should be applied i n suoh a manner that the  stress w i l l be concentrated imated.  critical  i n the glue l i n e whose q u a l i t y i s to be est-  The region of maximum stress should be known. A stress system  which i s subjeot to mathematical analysis would be an asset. l e a s t two methods whioh may be used to achieve t h i s objeotive.  There are at The  first  has to do with the shape and method of loading the speoimen, i . e . , by means of a t e n s i l e , compressive, or shearing force, or wedging a c t i o n .  The  latter  u t i l i z e s the differences i n strength properties of wood i n the l o n g i t u d i n a l , r a d i a l , and tangential directions to concentrate portion of the bond.  the stresses i n the desired  - 28 2,  In the oase of softwoods the normal v a r i a t i o n of springwood  and summerwood pattern at a glue l i n e must introduce some unnecessaryv a r i a b i l i t y into the t e s t r e s u l t s .  I t i s reasoned that  oross-banding  edge-grain veneers would y i e l d specimens with s u b s t a n t i a l l y the same percentages of springwood and summerwood i n the glue l i n e of each speoimen and thus tend t o minimize v a r i a b i l i t y from t h i s source. 3,  When veneers are lathe-out from a l o g , i r r e g u l a r i t i e s i n the  g r a i n of the wood r e s u l t i n o e r t a i n c e l l s being cut l o n g i t u d i n a l l y whereas others are out at an angle.  Those out at an angle a f f o r d an opportunity  for glue to be drawn i n t o the lumen by c a p i l l a r y a c t i o n , or to be forced i n by pressure; the others do not.  I t seems l o g i c a l that by cutting  veneers at an angle to the grain t o expose the lumen of most c e l l s , vari a b i l i t y from t h i s souroe would be reduoed. 4.  The surfaces of veneers, whether s l i c e d , peeled, or sawed,  suffer damage t o some extent during manufacture.  The surfaoe damage, con-  s i s t i n g of oheoks, uneven thiokness, torn s p l i n t e r s , or t o r n f i b r e s , must add to the v a r i a b i l i t y of t e s t r e s u l t s . from this source, i t was  In order to minimize v a r i a b i l i t y  considered desirable to saw and plane  veneers  whioh were t o be U3ed i n studying the gluing properties of a species. 5.  The f i n a l method of providing f o r minimum v a r i a b i l i t y between  t e s t r e s u l t s i s standardization. Standardization i n t h i s connection means that the dimensions of the p a r t i c u l a r source are defined exactly, so vari a b i l i t y from that souroe w i l l be minimized.  I t Is known f o r instanoe,  that any change i n the dimensions of a test speoimen, or of the veneers of whioh i t i s composed, a l t e r s the magnitude of the breaking load.  This  being so i t i s necessary t o specify i n d e t a i l the veneer and speoimen dimensions.  Another example would be s p e c i f i c a t i o n of the exaot moisture  - 29 content at whioh speoimens s h a l l he tested.  The same p r i n c i p l e applies  throughout the gamut of sources of v a r i a b i l i t y .  Probably there i s not a  single feature conoerning a method of t e s t which oan be varied from the chosen standard without introducing some consequent change i n the breaking load.  Standards must be set and adhered to f o r suoh widely divergent  features as the shape and dimensions  of the t e s t speoimen, the d e t a i l s of  the machines and accessories used to t e s t the specimen, the d e t a i l s of wood and glue preparation, the d e t a i l s of the bonding process and the moisture content of the wood. A rather complete l i s t of these factors i s included at the beginning of Chapter I I I , Part 2, "Methods of Reducing 6. ship.  Variability".  Every precaution should be exercised to assure f i n e workman-  The t e s t specimen should be as simple as possible to manufacture.  In hot-press bonding with veneers, single speoimens are more d i f f i c u l t to prepare than sheets from whioh t e s t specimens are out.  Sawcuts whioh must  touch the glue l i n e are d i f f i o u l t to make accurately. The more complicated the specimen shape the more s k i l l e d the labour required t o prepare specimens to the required high standard of acouracy.  A rectangular s o l i d (box) i s  probably the specimen shape most simple t o prepare to a high standard of dimensional accuracy. Certain types of specimen, such as the Tension Shear, require three or more times the area of wood a c t u a l l y tested, i n order t o apply the load.  Where the entire speoimen i s tested, three times as many speoimens  oan be prepared from the same quantity of wood.  This can be a worthwhile  feature when large numbers of matohed specimens are required f o r a part i c u l a r design. The cost of a t e s t i n g machine may  not be of major consequence for  the large researoh organization but f o r many smaller establishments t h i s  - 30 feature alone presents a c r i t i c a l r e s t r i c t i o n of methods*  An aocurate  method suitable f o r production t e s t i n g and employing an inexpensive apparatus would be of value t o industry*  I t would generate more i n t e r e s t  i n quality control of the gluing process which would lead t o higher standards with smaller losses from substandard 5.  produots.  SUMMARY In summary, the evidence provided by photoelastic a n a l y s e s ^ * ^ » ^ \ 2  stress a n a l y s e s ^ ' JS) and experimental evidence^ * 9  14, 15, 2 1 ) con-  2  $  t  firms the theory that the most desirable stress d i s t r i b u t i o n i s one which i s concentrated i n the glue l i n e .  The most desirable t e s t specimen, there-  fore, would be one i n which the maximum stresses are concentrated i n the glue l i n e . Methods of reducing v a r i a b i l i t y i n meohanioal t e s t r e s u l t s , and the i n t e r p r e t a t i o n of these r e s u l t s , include the f o l l o w i n g : the methods stressed by the Adhesives Research C o m m i t t e e ^  1 4  , 15) h e United States ft  Forest Produots Laboratory^* 42), lathe-check o r i e n t a t i o n ^ ) , reducing breaking loads to percentages  of wood strength(30, 35, 39), using a range  of adhesive strengths from "weaker than the wood" t o "stronger than the wood" i n combination with the rate of reduotion i n breaking load induced by a series of weathering oyoles o f increasing severity, and s t a t i s t i c a l techniques such as c o r r e l a t i o n analysis, slope r a t i o assays., and analysis of varianoe. The features to be kept i n mind when designing a method of t e s t i n g bond quality f o r research purposes include the following: (A)  TO minimize v a r i a b i l i t y among t e s t r e s u l t s : (1)  Concentrate the stress i n the glue l i n e (a)  by means of the shape of the specimen and method of  - 31 applying the load, (b) (2)  by oontrol of the f i b r e orientation of the wood.  Use edge-grain veneers and oross the springwood-summerwood bands when gluing.  (3)  Prepare veneers so that the wood oells are out a t a s l i g h t angle•  (4)  Use planed rather than peeled or s l i o e d veneers.  (5)  Standardize  the dimensions of the t e s t specimen and the manu-  facturing and t e s t i n g procedure to be followed, using the l i s t s at the beginning of seotion 2 of t h i s chapter as a guide. (6) Demand the highest q u a l i t y of workmanship. (B)  Use the simplest shape of speoimen consistent with loading requirements.  (6)  Keep the cost of equipment as low as i s consistent with the neoess a r i l y high standard  (D)  of aocuraoy required.  A desirable but not e s s e n t i a l feature would be that the method be suitable for production t e s t i n g as w e l l as for researoh purposes.  - 32 CHAPTER IV - THE GLUELIHE-CLEAVAGE TEST 1.  INTRODUCTION The unsatisfactory state of fundamental knowledge regarding the  gluing prooess has been presented, followed by a comparison of v i s u a l and mechanical methods of estimating glue l i n e q u a l i t y .  The theory of the  superiority of the l a t t e r technique was then propounded.  Deficiencies of  c e r t a i n meohanical methods of evaluating bond q u a l i t y have been tabulated, leading to the conclusion that whereas a meohanioal method i s d e s i r a b l e , none e x i s t i n g approaches the desired standard of accuracy.  Requirements of  the i d e a l method have been reviewed. 2.  DEFINITION OF THE GLUELINE-CLEAVAGE TEST The Glueline-Cleavage Test has been designed t o meet t h i s need  for a more accurate method of appraising the strength of wood to wood bonds. I t i s made by placing a knife-edge along the glue l i n e of the speoimen and measuring the force required for cleavage.  Figures IA, Appendix A. show  the knife and specimen i n p o s i t i o n f o r t e s t i n g .  Figure 1 of the same  appendix shows the hydraulio jack which has been adapted t o apply the load, and the gauge which reoords the foroe applied. Any a l t e r n a t i v e method of aoourately measuring the applied load would be s u i t a b l e .  I n i t i a l l y tests  were made with a 3,000 pound Universal t e s t i n g maohine, but t h i s prooedure was  discontinued due to pressure of work f o r that machine.  apparatus  The hydraulio  (calibrated against the Universal t e s t i n g maohine) has given r e -  produoible r e s u l t s .  I t has the added advantages of being simple, inexpensive,  and much faster i n operation. Three knife angles were compared} angle was  30°, 60°, and 90°.  The  90°  chosen f o r further work, although i t s superiority was not estab-  lished conclusively.  The knives were o a r e f u l l y ground, then f i n i s h e d on an  - 33 o i l stone.  Later work on stress analysis proved t h i s ohoiee to be a con-  venient one*  I t was found neoessary t o grease the knife-edge before t e s t i n g  each specimen to keep p i t c h from c o l l e c t i n g on the knife and a l t e r i n g the f r i o t i o n a l component of the applied force* The standard chosen for speoimen dimensions was 1*00" x  1.00"  x 0.400", the l a t t e r measurement being the thickness of "plywood" obtained when two 0.200" veneers are glued together*  This two-ply construction has  been adopted as standard for oertain research work but f o r other purposes suoh as quality control t e s t i n g i t has been necessary t o use three or more ply oonstruotions* 3.  EXPLORATORY TRIAL The Glueline-Cleavage Method was invented t o y i e l d more reproduc-  i b l e r e s u l t s than those mechanical methods with which d i s s a t i s f a c t i o n has been expressed. iment was  In order to t e s t the v a l i d i t y of t h i s assumption an exper-  set up t o compare the r e s u l t s obtained from three test methodst  Tension Shear, Tension Normal, and Glueline-Cleavage. This comparison was e f f e c t of drying temperature  oonduoted i n conjunction with a study of the on the strength of phenolic r e s i n adhesive  bonds with Douglas f i r veneers.  Glue blanks 4  W  x 4" x 0.200" t h i c k were  prepared with rotary out veneer from each of 26 Douglas f i r trees with a wide range of density and rings per inch.  Twelve pairs of these side-  matched glue blanks were prepared from each t r e e .  Corresponding pairs from  eaoh of the twenty-six trees were dried using a d i f f e r e n t sohedule f o r each of the twelve sets.  A f t e r having been dried, a l l speoimens were allowed t o  reaoh equilibrium moisture content i n an atmosphere of approximately 50 per oent r e l a t i v e humidity and 70 degrees Fahrenheit i n preparation for bonding. Complementary pairs of veneers were then glued with the grain of the p l i e s  - 34 parallel* treatment*  Care was taken to ensure that a l l speoimens reoeived l d e n t i o a l One Tension Shear, one Tension Normal, and one Glueline-Cleav-  age speoimen was out and tested from eaoh of these 4" x 4" laminates*  This  -work y i e l d e d twelve populations of laminates, a l l members of eaoh populat i o n having been treated l d e n t i o a l l y and then t e s t e d by eaoh method. Table 2,  Appendix A* presents the r e s u l t s of these t e s t s expressed  as Coefficients of V a r i a t i o n , bases, of comparison*  These are considered t o be the most r e l i a b l e  With t h i s system the minimum C o e f f i c i e n t of V a r i a t i o n  indioates the most r e l i a b l e method of t e s t *  The Glueline-Cleavage Test  yielded the lowest Coefficient of V a r i a t i o n i n every case*  This evidenoe,  based on 936 t e s t s of speoimens matohed as nearly as i t i s possible t o matoh wood speoimens, indicated that the Glueline-Cleavage Test was e r i o r and worthy of further intensive i n v e s t i g a t i o n .  sup-  At t h i s time i t  appeared that the method had made use of the following p r i n c i p l e s which have been set down as p r e r e q u i s i t e to an i d e a l t e s t s 1,  the test r e s u l t s are recorded i n pounds, whioh i s a widely aooeptable unit of measure,  2*  the speoimen i s of the simplest possible shape, and therefore economical t o prepare,  3.  the load i s applied i n a manner such that the maximum stress i s developed i n the portion of the glue l i n e just below the knife edge,  4.  the equipment required i s very inexpensive,  5.  the method of t e s t i s r e a d i l y adaptable f o r production t e s t i n g as w e l l as f o r researoh  6.  purposes,  the maximum possible number of speoimens oan be from a given sheet of plywood.  prepared  This desirable feature i s  - 35 made possible by the entire speoimen entering i n t o the t e s t . No material i s wasted for gripping purposes. 4.  EGxlO SPECIMEN^/ 0  The exploratory tests proved so encouraging that further attent i o n was turned toward the design of a test specimen f o r research purposes. The f i r s t step when commencing suoh a design was t o cheok theories regarding the requirements  of the i d e a l method of test and t e s t speoimen.  Some  of these could not be r e a d i l y v e r i f i e d ; e.g., the one regarding more uniform bond strength being obtainable with wood c e l l s out at a s l i g h t angle. One theory whioh d i d lend i t s e l f t o experimental confirmation was that one souroe of the differences between t e s t r e s u l t s i s the varying percentage o f springwood and summerwood i n the glue lines of i n d i v i d u a l specimens.  Tests were conducted  on Douglas f i r specimens prepared with spring-  wood bonded t o springwood and summerwood t o summerwood t o v e r i f y t h i s theory (see Tables IA, IB, and IC, Appendix A ) .  :  •  I t should be noted that the very  An EGxlO Speoimen i s defined as one which incorporates the following p r i n c i p l e s i n i t s design and manufacturej 0  1. I t i s prepared from sawed and planed edge-grain veneers i n whioh the c e l l s of the wood intersect the surface a t an angle of 10° (see Figure 2, Appendix A ) . 2. The veneers are cross-banded, either a t r i g h t angles or at some smaller angle, t o provide for a more uniform d i s t r i b u t i o n of springwood t o springwood, summerwood t o summerwood, and springwood t o summerwood bonding than i s otherwise obtainable. For i l l u s t r a t i o n s of 90° and 10° oross-banded veneers see Designs 1 and 3 of Figure 2, Appendix B. 3. The specimens are marked and sawn i n suoh a way that when the oleaving foroe i s applied by the knife the stress w i l l be concentrated i n that portion of the glue l i n e immediately below the knife edge. This stress concentration Is aohieved, apart from the use of the knife edge, by a comb i n a t i o n of the angle at which the wood c e l l s interseot the glue l i n e and the fact that wood i s r e l a t i v e l y weak i n tension perpendicular to the g r a i n .  - 36 l i m i t e d numbers of speoimens were prepared from only one t r e e , consequently r e s u l t s should not be interpreted at t h i s stage as i n d i c a t i n g other than a probable trend. From the nature of the f a i l u r e s i n Tension Shear speoimens i t appeared l o g i c a l t o conclude t h a t , i n springwood to springwood bonding, the wood broke i n tension at the base of the saw-out.  This i n i t i a l tension  f a i l u r e was followed by a peeling or clearing a c t i o n r e s u l t i n g i n complete rupture.  In those speoimens bonded summerwood t o summerwood, the wood was  s u f f i c i e n t l y strong i n tension t o withstand f a i l u r e , and most of the bonds appeared to have been tested i n (tension) shear. The appearance of the broken Block Shear specimens indioated that the springwood t o springwood specimens f a i l e d i n compression p a r a l l e l t o the grain combined with shear through the springwood, whereas the summerwood t o summerwood speoimens tested the bond i n (compression) shear.  I t appeared,  therefore, that the type of f a i l u r e obtained with springwood bonded t o springwood was d i f f e r e n t i n appearance from that obtained with summerwood glued t o summerwood. This difference i n appearance was a l s o r e f l e c t e d i n the induoed stresses, the mean breaking load of summerwood bonded t o summerwood being 90 per oent greater than that of springwood bonded t o springwood i n the case of Tension Shear Tests (Table IA, Appendix A ) .  I t was 42 per cent greater  i n the case of Block Shear Tests (Table IB, Appendix A ) .  A "t  t t  t e s t f o r the  s t a t i s t i c a l significance of the differences between the means indicates that these are both highly s i g n i f i c a n t (beyond the P - .01 l e v e l ) .  These large  differences i n breaking loads of springwood bonded t o springwood and summerwood bonded t o summerwood may explain, i n part, the great v a r i a b i l i t y enoountered i n t e s t r e s u l t s obtained by standard methods.  - 37 The C o e f f i c i e n t of V a r i a t i o n of the breaking loads of springwood bonded t o springwood calculated from Tension Shear speoimens was 15.4 per cent, and that of summerwood bonded t o summerwood 14.4 per oent.  When wood  i s glued i n praotioe, some of the area i s bonded springwood to springwood, some summerwood to summerwood, and the balance springwood t o summerwood. I f these two samples are combined i n t o one composite sample i t approaches (though exaggerated) that of normal t e s t specimens.  The C o e f f i c i e n t of  "Variation calculated f o r t h i s composite sample i s 35.3 per oent.  The Co-  e f f i c i e n t of V a r i a t i o n has thus been inoreased from 14.4 and 15.4 per cent to 35.3 per cent. summarized  Data for both Tension Shear and Block Shear methods are  below. Coefficients of V a r i a t i o n , per cent!/  Test Method  Sp. t o Sp.  Su. to Su.  Tension Shear  15.4  14.4  35.3  Block Shear  14.8  16.9  23.9  Sp. t o Sp. • Su. to Su.  These r e s u l t s demonstrate mathematioally how i t i s possible for the pattern of springwood and summerwood i n a bond to influenoe the v a r i a b i l i t y of t e s t r e s u l t s and how cross-banded edge-grain veneers may be used t o reduoe vari a b i l i t y f o r research purposes where t h i s i s e s s e n t i a l . From the Tension Normal data presented i n Table 1C, Appendix A, and the nature of the breaks, i t was observed that only one specimen broke i n the v i c i n i t y of the (Douglas f i r t o Douglas f i r ) bond.  The remainder  f a i l e d either through the springwood or through the two a d d i t i o n a l glue  7/ -* See Tables IA and IB, Appendix A» for the basio data. Sp. and Su. are used as abbreviations f o r springwood and summerwood.  - 38 l i n e s required t o prepare the speoimen for t e s t , therefore the bond under examination was not t e s t e d .  This was due t o imperfections i n the method  of t e s t , workmanship, or weakness of the wood.  Theories.concerning t h i s  behavior have been outlined previously. As might be expeoted, the d i f ference between the means of the breaking loads of the two samples d i d not prove s i g n i f i c a n t .  These data i l l u s t r a t e how rupture at some undesired  place i n the specimen can be a t t r i b u t e d to f a i l u r e t o ensure that the maximum stress i s developed i n the glue l i n e . Although the experimental work upon which the following two pieoes of information are based w i l l not be presented u n t i l l a t e r i t seems appropr i a t e t o use portions of the data a t t h i s time for i l l u s t r a t i v e  purposes.  The f i r s t i s the effect of concentrating the stress i n the glue l i n e by cont r o l l i n g the angle at which the wood o e l l s are oriented with respect t o the glue l i n e .  from 8.012  when f i b r e s were  p a r a l l e l t o the glue l i n e , to 10.978 when f i b r e s were oriented at an angle of 10° to the glue l i n e and when the knife was applied i n suoh a d i r e o t i o n as to concentrate the stress i n the bonded area. r a t i o the more sensitive the t e s t ) .  (The larger the slope  The second item concerns the e f f e c t of  greasing the knife i n reducing v a r i a b i l i t y .  The C o e f f i c i e n t of V a r i a t i o n  was reduoed from 13.4 to 9.6 per cent^/. A l l r e s u l t s of the various comparisons seemed t o indicate that the general approaoh wa3  correct and that the v a r i a b i l i t y w i t h i n t e s t r e s u l t s  could be reduoed by making use of the p r i n c i p l e s already outlined i n / — See Table 5A, Appendix B, for data; see (7) f o r d e f i n i t i o n of Slope R a t i o .  8  See Table 4. Appendix A.  - 39 Chapter I I I . The EGxlO Speoimen design^/ therefore incorporated as many 0  of these p r i n c i p l e s as possible.  This one inch square speoimen i s the  simplest possible shape to manufacture, the only t o o l s required being those normally a v a i l a b l e i n the average woodworking shop.  Aoouraoy requires, of  course, that these be i n good meohanioal condition.  Suoh a specimen shape  makes i t possible t o cut the maximum number of test pieces from a sheet of plywood because no gripping area i s required and the whole speoimen enters i n t o the t e s t .  These are worthwhile features when i t i s neoessary to pre-  pare large numbers of speoimens from matched veneers.  Another feature i s  that maximum stress i s oonoentrated i n the glue l i n e immediately below the knife edge.  Some evidence i s a t hand that a stress analysis of t h i s type  of speoimen i s possible (33). Use i s made of the d i f f e r e n t strength prope r t i e s of wood i n i t s various d i r e c t i o n s , notably i t s r e l a t i v e weakness i n tension perpendicular to the grain, t o concentrate the stress i n that portion of the glue l i n e immediately below the knife edge.  The v a r i a b i l i t y i n t r o -  duced by d i f f e r e n t speoimens having widely d i f f e r e n t percentages of springwood (or summerwood) i n t h e i r bonds has been minimized by oross-banding edge-grain veneers.  The bond uniformity has been enhanced by using planed  instead of peeled, s l i c e d , or sawed veneers.  The governing p r i n c i p l e that  the many d e t a i l s of gluing research must be performed according t o prearranged standards has been used throughout.  Broadly these inolude every  d e t a i l of wood s e l e c t i o n and treatment, glue s e l e c t i o n and gluing, machinery, speoimen size and moisture content.  - 40 -  5.  COMPARISON OF GLUELINE-CLEAVAGE (EGxlO ) WITH TENSION SHEAR, TENSION NORMAL, AND BLOCK SHEAR TESTS 0  (A)  Experimental  Procedure  In order to study the effeot of the above speoimen design i n reducing the v a r i a b i l i t y of t e s t r e s u l t s , and i n order to more f u l l y examine the preliminary conclusions, a further i n v e s t i g a t i o n was undertaken.  This was  intended to determine whioh of Tension Shear, Tension  Normal, Block Shear, or Glueline-Cleavage with EGxlO specimens was 0  most suitable f o r evaluating the q u a l i t y of Douglas f i r veneer  the  joints  bonded with a hot-press phenolio r e s i n adhesive. It was e s s e n t i a l that a l l test specimens be olosely matched i n every respect t o permit a d i r e c t comparison of r e s u l t s .  A l l specimens  therefore were out from a single clear, straight-grained pieoe of Douglas f i r heartwood 4" x 4  n  x 8* with the annual rings p a r a l l e l to one face.  The selected pieoe of wood was as nearly flawless as oould be by v i s u a l inspection. sawed from the 4 Appendix A. weeks.  W  determined  Boards one-half inoh thick by four inches wide were  x 4" x 8' timber according to the pattern of Figure 2,  These boards were end-ooated and allowed to a i r - d r y f o r two  They were then dried f o r one week a t 122° F., to seven per oent  moisture content, followed by two days at 144° F., whioh reduoed t h e i r moisture oontent to f i v e per cent.  This drying schedule produoed check-  free lumber of uniform moisture oontent.  The boards were transferred to a  humidity chamber and conditioned to equilibrium moisture oontent i n an atmosphere of 32 per oent r e l a t i v e humidity and 70°F.  They were then  planed to produce veneers 0.200" j 0.005" t h i c k , and f i n a l l y they were sawed to glue blanks 3.50"  x 3.50".  The thickness of each blank was  measured a t the four corners with a micrometer t o a tolerance of jO.001".  - 41 This information was used when matching glue blanks to obtain layups of as uniform thickness as possible* Blanks f o r the preparation of EGxlO Specimens were matched, 0  numbered and marked i n readiness for gluing, using the veneers out at an angle of 10° to the wood c e l l s .  From the balanoe of the veneers, blanks  for the preparation of Tension Normal, Tension Shear, and Block Shear specimens were matched, numbered and marked preparatory to gluing.  Both  Tension Normal and Tension Shear blanks were cross-banded while Block Shear blanks were laminated* The adhesive consisted of 2000 grams of PF512i a phenolic r e s i n manufactured by Monsanto (Canada) Limited, 400 grams walnut s h e l l flour and 400 grams water*  The walnut s h e l l f l o u r was added to the r e s i n and the  mixture s t i r r e d for ten minutes.  Following t h i s the water was added and  s t i r r i n g continued for f i v e minutes at which time the adhesive was for use.  The i n i t i a l temperature of the r e s i n was 19*5° C ,  temperature  of the adhesive was 24*0° C.  ready  and the f i n a l  This was spread on one face at  39 (S s 4)i2/ pounds per 1000 square feet of glue l i n e with a mechanical glue spreader.  Five minutes open-assembly time and f i v e minutes olosed-  assembly time were allowed, and pressure of 200 pounds per square inch pressure was applied a t a platen temperature Fahrenheit f o r ten minutes.  of 280 (S a 7)22/  degrees  The glued blanks were then removed from the  press and hot-stacked i n a olosed wooden box f o r 18 hours.  Following t h i s  they were returned t o the humidity chamber f o r reoonditioning (32 per cent r e l a t i v e humidity and 70° F . ) . A f t e r reaching equilibrium moisture content, the glued blanks  S i s the standard deviation of the population as estimated from a sample.  n  n  - 42 were out i n t o t e s t specimens which were then returned t o the humidity chamber f o r f i n a l reconditioning.  The test specimens were cut and numbered  according t o the plan of Figure 4, Appendix A.  Specimens were tested by  t h e i r respective methods, immediately upon removal from the humidity ohamber.  The following information was reoorded f o r eaoh specimen t e s t e d :  (a)  breaking load, i n pounds,  (b)  wood f a i l u r e , i n per cent,  (o)  weight of the speoimen a t t e s t , i n grams,  (d)  oven-dry weight of the speoimen, i n grams, and  (e)  volume of the oven-dry specimen, i n oubio centimeters.  (B)  Analysis of Test Results See Table 3, Appendix A, f o r the data.  I t was anticipated that  there would be l i t t l e difference i n the s p e c i f i c gravity of i n d i v i d u a l speoimens since a l l had oome from a single timber of uniform q u a l i t y .  This  proved t o be the oase so i t was unnecessary t o o u l l any speoimens, or to apply a correction for s p e c i f i o gravity differences.  Likewise, moisture  contents of the speoimens were s u f f i c i e n t l y uniform so that no c o r r e c t i o n was necessary. Analysis of the breaking loads y i e l d e d the data presented i n Table 4» Appendix A.  The Glueline-Cleavage Test (EGxlO  0  Speoimens), with  the knife greased, gave the minimum Coefficient of Variation, 9 . 5 per cent. Greasing the knife had reduced the C o e f f i c i e n t of V a r i a t i o n from 13.7 t o 9 . 5 per cent, a very worthwhile reduction emphasizing  the importance of  greasing the knifo before each t e s t . In addition to y i e l d i n g the least v a r i a t i o n between t e s t r e s u l t s , the Glueline-Cleavage Test with EGxlO oross-banded 0  speoimen has another  - 43 advantage; the glue l i n e i s opened f o r i n s p e c t i o n even i n a case where the wood i s "weaker than the glue".  This feature may be observed by oomparing  the photographs of broken Glueline-Cleavage specimens with those of the Tension Shear, Tension Normal, and Block Shear methods (Figures 5 to 8 inolusive, Appendix A ) . 6.  SUMMARY The Glueline-Cleavage Test oonsists of measuring the foroe r e -  quired t o cleave a one inch square test specimen with a 90° knife (wedge) plaoed along the glue l i n e of the specimen.  Almost any machine capable of  applying and measuring the necessary load could be adapted t o t h i s method. The small hydraulic machine i l l u s t r a t e d i n Figure 1, Appendix A, b u i l t a t nominal cost, was used i n a l l of the Glueline-Cleavage tests herein reported. The EGxlO  0  specimen has been designed to minimize v a r i a b i l i t y i n t e s t r e -  sults for research purposes. The advantages of the Glueline-Cleavage Method (as separate from the a d d i t i o n a l advantages of the EGxlO Specimen) w i l l be summarized f i r s t . 0  Undoubtedly the most important advantage of the method i s that i t has been shown t o give more reproducible r e s u l t s ( i . e . , to be more accurate) than the other methods against whioh i t was compared.  Second, the r e s u l t s are  reoorded i n pounds, a u n i v e r s a l l y acceptable unit of measure.  Third, the  specimens are r e l a t i v e l y easy to prepare, requiring only the t o o l s normally available i n any woodworking shop. may  Fourth, a maximum number of specimens  be out from a unit area of plywood, whioh f a c i l i t a t e s the use of experi-  mental designs requiring large numbers of matched specimens.  F i f t h , the  region of maximum stress, and there i s only one, i s more d e a r l y defined than i n most other test methods.  This i s advantageous when inspecting  specimens f o r the causes of low f a i l i n g loads.  Sixth, the method lends  i t s e l f r e a d i l y to production t e s t i n g or "trouble-shooting".  Seventh, a l l  of the glue l i n e s i n a partioular plywood speoimen oan be tested, regardless of the number. The EGxlO  Eighth, only inexpensive equipment i s required. 0  Specimen has two additional advantages.  F i r s t and  foremost, the v a r i a t i o n i n t e s t results i s less than i n any other style of speoimen t e s t e d .  Second, a l l glue l i n e s are opened f o r inspection even  though the adhesive i s "stronger than the wood".  CHAPTER 7 - ELABORATION OF THE GLUELINE-CLEA7AGE METHOD 1.  PREAMBLE The Glueline-Cleavage Test, when used i n conjunction with EGxlO  Speoimens oross-banded at 90°,  0  has been demonstrated t o be a superior  method for estimating the adhesive bond q u a l i t y of plywood.  The speoimens  upon whioh t h i s statement i s based were prepared so that they would have as uniform bond strength as p o s s i b l e . The question arose as t o whether the method would maintain i t s superiority with a wider range of adhesives and weathering treatments.  Apart from t h i s general problem there were three  speoiflo investigations whioh i t was desired t o make. The f a l l a c y of attempting t o measure the strength of strong glue with weak wood has been considered.  I t may happen that a researoh worker  does not know the r e l a t i v e strengths of the wood and adhesive which he proposes t o use, but without t h i s information i t i s quite possible to draw erroneous conclusions from an i n v e s t i g a t i o n . may be avoided would be of value.  A system whereby t h i s  pitfall  The development of suoh a method i s the  f i r s t of the three speoial investigations noted i n the previous  paragraph.  The second point requiring examination was whether or not the design of the EGxlO Speoimen used i n Chapter 17 was the best. 0  How muoh  improvement, i f any, oould be effected by orossing the springwood-summerwood bands at an angle less than 90 degrees?  EGx0°12/ or F G x O  0  What would be the effect of using  ^ / veneers instead of EGxlO veneers? 0  These questions  appeared t o require answers. i^EGxOO i s an abbreviation for edge-grain veneer (with the wood c e l l s p a r a l l e l to the surfaoe). 12/ — FGxO° i s the corresponding abbreviation for f l a t - g r a i n veneers. 1  - 46 The t h i r d problem was t o demonstrate mathematioally that the r e s u l t s of mechanical tests were more r e l i a b l e as an estimate of bond quality than were those based upon per cent wood f a i l u r e , 2.  AUXILIARY TOOLS Consideration was given to methods of s e t t i n g up, conducting,  and analysing the r e s u l t s of an experiment designed to investigate the problems outlined above.  I f an adhesive i s suspeoted of providing bonds  "stronger than the wood", and i f i t i s desirable to v e r i f y t h i s suspicion, then i t appears that some method must be found f o r preparing a glue that i s "weaker than the wood".  This should not ohange the glue's nature so  d r a s t i o a l l y that conclusions are invalidated.  To achieve suoh a purpose  i n t h i s researoh i t was considered advisable t o add a standard mixture of extenders (walnut-shell f l o u r , caustio, and water) to d i f f e r e n t percentages of the phenolic r e s i n .  The range of mixtures varied from that required by  the manufacturer's formula (27&6 r e s i n s o l i d s ) to one known t o be defi n i t e l y "weaker than the wood" (12§jS r e s i n s o l i d s ) .  Any resultant design,  then, must include more than one adhesive strength. The comparison of several possible speoimen designs would include a s a t i s f a c t o r y measure of t h e i r r e l a t i v e e f f i c a o i e s .  The objective of suoh  a comparison would be to determine which of several t e s t speoimen designs and methods of t e s t would be the most sensitive to small changes i n glue line quality.  The design for a measure of s e n s i t i v i t y * ' / i  following premisej  s  based upon the  f o r an adhesive, suoh as the phenolic r e s i n PF512 ex-  tended with walnut-shell f l o u r and water, there must be some combination of r e s i n and extender whioh w i l l be only just as strong i n cohesion and/or — - "Measures of s e n s i t i v i t y " for the purpose of t h i s thesis w i l l be defined as the a b i l i t y of the "measure" t o d i s t i n g u i s h between small changes i n the strength (quality).of a glue bond.  - 47 adhesion as the wood f i b r e s which i t i s to bond.  Successively greater  extensions w i l l produce bonds whioh w i l l f a i l at progressively smaller loads.  Smaller extensions, however, w i l l not produce bonds which f a i l at  increasing loads because the f a i l u r e w i l l always ocour i n the wood and not i n the adhesive.  In other words, the bonds formed by successively smaller  extensions w i l l a l l break at the same load, that determined by the.strength of the wood. The f i r s t method investigated as a measure of s e n s i t i v i t y  was  the degree of c o r r e l a t i o n existing between breaking loads and per cent resin solids.  A f t e r careful consideration i t was  o'oncluded that t h i s  degree of correlation i s not the best available measure of test method sensitivity.  A c o r r e l a t i o n curve represents the r e l a t i o n s h i p between the  variables, i n t h i s oase breaking load as recorded by the chosen method of test and per oent r e s i n s o l i d s .  This, however, i s only of i n d i r e c t i n t e r e s t .  The primary i n t e r e s t i s direoted toward a r e l i a b l e i n d i c a t o r of the f i r s t sign of reduoed bond strength as the r e s i n s o l i d s content of the adhesive i s reduoed.  This i s estimated by the rate at whioh the breaking load f a l l s  o f f with decreased r e s i n s o l i d s .  More broadly, i t i s desired to be able to  say that the t e s t r e s u l t s t r u l y represent the strength of the adhesive bond just before i t ruptures. were of i n t e r e s t *  It appeared, therefore, that the following points  f i r s t l y , the reductions i n breaking load rather than the  absolute breaking loads, seoondly, the transformation of these figures to percentages  so that comparisons oould be made between methods with accuracy,  t h i r d l y , the rate at which breaking loads decreased with reduced r e s i n s o l i d s ( i . e . , the steeper the slope the more sensitive the t e s t , other things being equal), and f o u r t h l y , the aoouracy of slopes, ( i . e . , the r e p r o d u c i b i l i t y of r e s u l t s as measured by the Standard Error of Estimate).  -  48 -  A s t a t i s t i c a l technique known as "Slope Ratio A s s a y s " w h i c h  is  used by medical men t o determine whioh of several drugs i s most e f f e c t i v e seemed t o be applicable and has been used with modifications.  In essenoe,  the procedure i s to p l o t treatment effectiveness over dosages (the analogy would be breaking load over per cent r e s i n s o l i d s ) using a transformation such that the data w i l l p l o t as a straight l i n e . largest value of the r a t i o b/Se  =  The medicine giving the  Slope of l i n e i Standard Error of Estimate  s  declared  to be the most e f f e c t i v e f o r each increase of dosage (for t h i s purpose the largest b/Se would be declared the most effeotive for each reduction i n per cent r e s i n s o l i d s ) . A comparison of accuracy between meohanioal and wood f a i l u r e methods posed a problem, the data of the former being recorded i n pound units, and that of the l a t t e r i n per oent.  A l l data must be reduoed t o a  percentage basis before d i r e c t comparisons may be made.  In the oase of the  mechanical test r e s u l t s , t h i s was percentage of the maximum breaking load whereas the wood f a i l u r e r e s u l t s are already expressed as per cent.  Ifa  v a l i d c o r r e l a t i o n exists between bond strength and wood f a i l u r e then per cent of maximum breaking load and per cent of wood f a i l u r e should be d i r e c t l y comparable.  I t was decided t o use t h i s p r i n o i p l e i n analysing the data.  The method used t o compare the accuracy of bond strength predictions based upon mechanical tests and wood f a i l u r e i s described below. I f specimens matched for bond strength are tested by two mechanical methods (e.g., Glueline-Cleavage and Tension Shear) a t least four independent  est-  imates of bond q u a l i t y may be made from them; two breaking loads and two •fowr per oent wood f a i l u r e s .  A l l o f these designs, i f they are t o be u s e f u l , A  must record p a r a l l e l trends when the bond q u a l i t y i s a l t e r e d . -* I f the strength of the bond has been reduoed by 25%, then each design should reoord  - 49 this.  The meohanioal tests should show a 25% reduction i n breaking load  whereas the Per Cent Wood F a i l u r e Designs should show a 25% reduction i n per oent wood f a i l u r e .  The difference (per oent reduction i n breaking  load minus per oent reduction i n per cent wood f a i l u r e ) i s a measure of the incapacity of one (or both) to accurately portray t h i s reduction i n bond strength. 3.  EXPERIMEMTAL DESIGN (STATISTICAL) The experiment was not l a i d out as a single s t a t i s t i c a l design.  This would have been too complicated t o be p r a c t i c a l , but c e r t a i n of the necessary p r i n c i p l e s were incorporated so that portions of the data oould be analysed by s t a t i s t i c a l procedures.  For instance, care was taken that  specimens from eaoh r e p l i c a t i o n were included i n every treatment  combin-  ation. Nine designs of Glueline-Cleavage Specimen and four designs of Tension Shear Specimen were included. i n Figure 2, Appendix B.  The details of these are i l l u s t r a t e d  Thirteen meohanioal estimates of the bond q u a l i t y  are provided, one from eaoh design.  In a d d i t i o n , an independent estimate  of the bond q u a l i t y i s provided by eaoh design through the medium of the per oent wood f a i l u r e s read from the ruptured specimens.  A t o t a l of  twenty-six estimates of bond strength therefore were made a v a i l a b l e by t h i s experiment.  Not a l l of these have been used i n t h i s t h e s i s .  Four  strengths of phenolio r e s i n adhesive, ranging from "stronger than the wood" t o "weaker than the wood" were inoluded.  These have been designated as  Adhesives A, B, C, and D containing, respectively, 27g-, 22^, 17^, and 12-| per oent r e s i n s o l i d s .  I t was decided to t e s t speoimens a f t e r each of s i x  increasing s e v e r i t i e s of the "aooelerated weathering b o i l i n g i n water and drying).  oyole" (alternate  Eight t r e e r e p l i c a t i o n s were included, one  - 50 specimen from eaoh tree being a l l o t t e d t o each of the treatment combinations. 4.  WORKING PLAN Having f i r s t s e t t l e d upon the treatment combinations the next  step was t o prepare a working plan i n d e t a i l .  Materials were selected and  the d e t a i l s of wood preparation inoluding sawing, seasoning, planing, and conditioning were performed, with one exception, aeoording to the standards set f o r t h i n Chapter IV.  Cutting t o the pattern of Figure 1, Appendix B,  instead of Figure 2, Appendix A, was the single exception.  From the 0.200"  veneers prepared with the f i b r e s a t an angle of 10° to the surface, twentyfour glue blanks 3.80" x 3.80" were sawed, to give eight f o r eaoh of the three Glueline-Cleavage designs using t h i s m a t e r i a l . Using the edge-grain veneers prepared with the f i b r e s p a r a l l e l t o the surfaoes, 64 glue blanks 3.80" x 3.80" were sawed, 24 f o r the three Glueline-Cleavage designs, 16 for the Tension Shear, and 24 f o r the Tension Shear Plywood designs.  From the  f l a t - g r a i n veneers prepared with the f i b r e s p a r a l l e l t o the surfaoes, 64 glue blanks 3*80" x 3.80" were sawed, 24 f o r the three Glueline-Cleavage designs, 16 for the Tension Shear, and 24 f o r the Tension Shear Plywood designs. The glue blank thicknesses were measured and reoorded on each oorner of the specimens, t o *0.001".  Glue blanks were matched on a t h i c k -  ness basis, f o r each speoimen design, so as to obtain "layups" with as nearly as possible the same thiokness.  Each layup was marked d i s t i n c t l y t o  f a c i l i t a t e matohing of the speoimens with a minimum of confusion when gluing.  For uniformity, a l l specimens were marked i n the upper l e f t  corner, as shown on Figure 3, Appendix B.  The number, representing the  specimen design, glue mix, and tree, was placed on eaoh veneer of the layup.  Specimen designs and names are i l l u s t r a t e d on Figure 2,  - 51 Appendix B. Due to the volume of work, four days were required t o complete the gluing.  Because there were four grades of glue i t was convenient t o  a l l o t one day to each glue type (glues and days were randomized). For the f i r s t two t r e s s tested, the order of gluing the layups was chosen by a random sampling procedure the results of which are set f o r t h i n Table 1, Appendix B.  Replications were treated i n a similar manner.  The four formulae below provided s u f f i c i e n t adhesive t o glue a l l of the blanks whioh were bonded with that p a r t i c u l a r mix ( i . e . , for eaoh day's gluing).  In a d d i t i o n each provided s u f f i c i e n t for waste, as i t was  neoessary to clean the r o l l s a f t e r every second or t h i r d press load. Forty pounds of adhesive per 1000 square feet of single glue l i n e (1.82 grams-per 3.80" x 3.80" speoimen) was applied by a power-operated glue spreader. ADHESIVE "A"; 27^5 r e s i n s o l i d s . This, the standard mix, contained 65.0# PF512 l i q u i d r e s i n or 27.6% r e s i n s o l i d s .  To 32 grams of oaustio soda dissolved i n 40 grams of  water were added 520 grams of water at 200°F.  These were s t i r r e d f o r one  minute a f t e r which s t i r r i n g was continued while 320 grams of walnut-3hell f l o u r and 80 grams of soda ash were added.  The mixture was then s t i r r e d  continuously f o r 20 minutes, keeping the temperature above 180°F. a t i o n was oontinued u n t i l the temperature was reduced t o 140°F.  AgitFinally  2000 grams of PF512 r e s i n , 2 grams of stove o i l , and 80 grams of walnuts h e l l f l o u r were added and s t i r r i n g continued f o r 20 minutes during whioh time the temperature was further reduced t o 70°F.  The v i s c o s i t y was then  measured with a McMiohal Visoosimeter, and the adhesive was ready for use.  - 52 ADHESIVE "B"; 2 2 ^ r e s i n s o l i d s This formula contained 22.5% r e s i n s o l i d s .  To 46 grams of  oaustio soda dissolved i n 58 grams of water were added 759 grams of water at 200°F.  These were s t i r r e d f o r one minute a f t e r whioh s t i r r i n g was con-  tinued while 468 grams of walnut-shell f l o u r and 117 grams of soda ash were added.  The mixture was then s t i r r e d continuously for 20 minutes, keeping  the temperature above 180°F. was reduoed t o 140°F.  S t i r r i n g was oontinued u n t i l the temperature  F i n a l l y , 1626 grams of PF512 r e s i n and 2 grams of  stove o i l were added and s t i r r i n g oontinued f o r 20 minutes during which time the temperature was further reduoed t o 70°F.  During the l a s t 5 min-  utes' s t i r r i n g , the v i s c o s i t y was adjusted (by the addition of a small quantity of water, or walnut-shell f l o u r ) t o that of the standard mix, i . e . , Adhesive "A".  The adhesive was then ready f o r use.  ADHESIVE "C ; 1 7 ^ r e s i n s o l i d s tt  The procedure followed i n preparing Adhesive "C was the same as n  that f o r Adhesive "B" with the exoeption that the proportions of ingredi e n t s were adjusted so that the r e s u l t i n g mixture contained 17§?S r e s i n solids. ADHESIVE "D ; 12^S r e s i n s o l i d s n  The prooedure followed i n preparing Adhesive D M  n  was the same as  that for Adhesive "B" with the exoeption that the proportions of ingredients were adjusted so that the r e s u l t i n g mixture contained 12-^2 r e s i n solids. Five minutes open-assembly time was allowed, followed by f i v e minutes closed-assembly time. 285°F. and 200 p . s . i .  Pressure was applied f o r ten minutes at  The glued blanks were allowed t o oool and then were  conditioned f o r 96 hours i n an atmosphere of 32 per oent r e l a t i v e humidity  - 53 and at a temperature of 70°F. Figure 4 , Appendix B, shows the marking and cutting plans for the d i f f e r e n t designs.  With EGxlO  0  Glueline-Cleavage specimens i t was  essential that the d i r e c t i o n of knife a p p l i c a t i o n be marked on the specimens before they were out. An e f f o r t was made to saw Tension Shear specimens i n such a manner that the lathe checks "pulled closed" during testing^ ). 6  One t e s t specimen from each tree, and from every design and adhesive, was tested a f t e r eaoh of the following "aocelerated weathering" treatments} I. II.  dry} a f t e r 4 hours b o i l i n g i n water;  I I I . a f t e r 4 hours b o i l i n g , 20 hours drying a t 145°F. and 4 hours b o i l i n g , or, abbreviating, a f t e r 4,20,4; IV. V. VI.  a f t e r 4,20,4; 20,4; a f t e r 4,20,4; 20,4; 20,4; 20,4; a f t e r 4,20,4; 20,4; 20,4; 20,4; 20,4; 20,4.  With the exception of weathering treatment I above, a l l speoimens were tested wet a f t e r having been cooled to room temperature. For each specimen tested the following data, with the exceptions noted, were reoordedi (a)  breaking load, pounds,  (b)  wood f a i l u r e , per cent,  (o)  weight of the specimen at t e s t , grams,  (d)  oven-dry weight of the speoimen, grams, and  (e)  volume, of specimens at t e s t (by micrometer measurements), oubio oentimeters.  Specifio gravity measurements obtained from speoimens whioh have been b o i l e d i n water are of doubtful accuracy.  The s p e c i f i c g r a v i t i e s of those  specimens tested dry were, therefore, aooepted as being representative of the s p e c i f i c gravity of the glue blank from whioh each was cut.  The same  procedure was applied as a basis for c a l c u l a t i n g the percentage moisture content.  In order to estimate the s p e c i f i o gravity and percentage  moisture content of eaoh glue blank, the specimen chosen f o r t e s t i n g dry from that blank was weighed, i t s dimensions measured (*0.001 ) immediately n  before t e s t i n g and i t s volume oomputed. pieces were oven-dried and re-weighed.  A f t e r having been tested, the The i n d i v i d u a l speoimen for any  given t e s t was chosen at random from among those cut from that p a r t i c u l a r glue blank.  Those specimens from each glue blank which were to be b o i l e d  were kept together.  This f a c i l i t a t e d the seleotion of one speoimen f o r  t e s t i n g from each glue blank a f t e r eaoh treatment.  The broken speoimens  were allowed to s i t faoe up on a table i n the laboratory f o r approximately 24 hours before the per oent wood f a i l u r e estimation was made.  The pieces  of each test specimen and eaoh glue blank were f i l e d together, as were the trimmings, so that t h i s material could be used f o r further study.  The g l u -  ing and t e s t i n g prooedure was repeated three times as outlined. Altogether eight trees were tested, two i n the f i r s t series and three with each repetition. Each design of test specimen yielded two independent estimates of the quality of that glue l i n e .  The f i r s t was the breaking load and the  second the per cent wood f a i l u r e .  In t h i s thesis the terms Design 1,  Design 10, Design 13, e t c . , r e f e r to breaking loads, percentages of maximum breaking loads, or percentage reduction i n breaking loads. When estimates based upon per oent wood f a i l u r e are being considered they w i l l  - 55 be referred t o as Design 1$WF, Design 10$WF, Design 13?2WF, e t c .  Estimates  of per oent wood f a i l u r e are based upon the standard used by the Douglas F i r Plywood A s s o c i a t i o n . 5. RESULTS The f i r s t step i n analysing the data was reduction of a l l ures t o a common basis of comparison.  fig-  The one chosen was the peroentage  reduction i n breaking load (or per oent wood f a i l u r e ) whioh had been induced by each combination of treatment f o r eaoh tree and eaoh design tested.  The reasons f o r t h i s ohoice of transformation have been developed  i n the section of t h i s chapter e n t i t l e d " A u x i l i a r y Tools".  In Table 2,  Appendix B, i s presented the o r i g i n a l data and i n Table 3, Appendix B, these data transformed t o percentage reductions i n breaking loads (or per cent wood f a i l u r e ) . (a)  Per Cent Wood Failure Vs. Meohanical Test For the purpose of t h i s comparison wood f a i l u r e s obtained from  Tension Shear (Plywood Shear) specimens w i l l be used.  This i s a widely  employed t e s t procedure for wood f a i l u r e estimates, and the system for estimating per cent wood f a i l u r e was developed for t h i s p a r t i c u l a r speoimen, wood, and glue.  Thus conditions are optimum for the use of per cent a  wood f a i l u r e as an estimate of bond q u a l i t y .  110F  w i l l be used for i l l u s t r a t i v e  Data from Designs 4, 11, and  purposes.  Design 11 vs. Design 11$WF Table 4A» Appendix B, l i s t the differences (per cent reduotion) ( i n breaking load ) minus (of Design 11 ) for  (per cent reduction ) ( i n per cent wood f a i l u r e ) (of Design ll^WF )  each treatment combination and f o r eaoh t r e e tested.  differences i s 5*5%, i . e . , using the averages  The mean of the  of 196 specimens per design  - 56 i t appears that the two designs gave approximately the same r e s u l t s . Design lljSWF indicated 5.5% greater reduction i n bond strength than did Design 11.  Upon closer inspection of the data i t i s found that the  Standard Deviation of these differences i s j37.4$.  Although the estimate  based upon the mean difference appears reasonably accurate, any i n d i v i d u a l difference i s very unreliable (68 out of 100 of the differences being equal to or smaller than +37*4% and, of course, the other 32 out of 100 being larger than £.37.4$) • Design 11 v s . Design 4 By comparison, when the same technique i s used w i t h the d i f ferences (per cent reduotion) ( i n breaking load ) (of Design 11 )  minus  (per oent reduction) ( i n breaking load ) (of Design 4 )  for matched speoimens (matched for treatments and t r e e ) the mean of the differences i s 6.5%. •15.4$.  The Standard Deviation of these differences i s  See Table 4B., Appendix B, f o r the data. The average difference i n the estimates has changed from 5.5%  t o 6.5%.  This differenoe i s probably not s i g n i f i c a n t .  The Standard Dev-  i a t i o n , however, has been reduced from +37*4% t o ±15»4-% reduotion.  t  a very notable  Design 11 and Design 4 agree much more olosely than do Design  11 and Design ll^WF i n spite of the fact that the l a t t e r were made from the same specimen and the former ware made from d i f f e r e n t specimens. Suppose that two specimens were prepared from the same material (wood and glue), treated i d e n t i c a l l y by a treatment calculated t o reduce the bond strength, and then one tested by Design 11 and the other by Design 4.  The  expectation would be that the difference between the two meohanioal estimates (of the percentage reduction i n bond strength) would not exceed  ±15.5%  with  - 57 a p r o b a b i l i t y of .63 or •30.8# with a p r o b a b i l i t y of .95. Had two e s t imates been made using the single speoimen tested by Design 11, one using Design 11 and the other Design 11#WF, the expectation i s that the d i f farenoe (between the two estimates of the percentage reduction i n bond strength) would now not exoeed *37.4^ with a p r o b a b i l i t y of .68 or +74.8^ with a p r o b a b i l i t y o f .95. The inescapable conclusion i s that per oent wood f a i l u r e  (Design  ll^WF) i s not a r e l i a b l e estimate of the bond q u a l i t y of any single s p e c i men (although i t may have merit when an average i s taken of a great many specimens).  The d i f f i c u l t y i s that only one speoimen i s examined a t a time  and the researcher attempts, oonsoiously or otherwise, t o say that the estimate of per oent wood f a i l u r e measures the q u a l i t y of the bond i n that p a r t i c u l a r speoimen.  Then, too, there i s an ever present pressure t o  reduoe the number of specimens f o r economic and p r a o t i o a l reasons.  This  oould lead to incorrect conclusions, wasted researoh, or faulty' production oontrol.  Per cent wood f a i l u r e should, for researoh purposes, be used  only i n conjunction with mechanical t e s t s , not as a substitute f o r them. Should i t be necessary t o use per oent wood f a i l u r e as an estimate of bond quality t h i s should be done only a f t e r the required c o r r e l a t i o n with meohanioal methods has been established, (b)  S e n s i t i v i t y of Designs A comparison w i l l next be made of the s e n s i t i v i t y of i n d i v i d u a l  designs, one of the major objectives of t h i s research being t o e s t a b l i s h whioh of several designs of t e s t speoimen i s the most s e n s i t i v e .  This  comparison w i l l be r e s t r i c t e d t o meohanical methods for reasons whioh have been established. When the analysis was commended the Slope Ratio Assays method  - 58 was applied t o only part of the data, namely t o Adhesives A and B (27i# and 22^> r e s i n s o l i d s ) *  These data provided l i n e a r c o r r e l a t i o n without  transformation and were i n that portion of the curve of most i n t e r e s t , namely, the region where a loss of strength was f i r s t notioeable* This approach appeared t o he favorable; a f t e r further consideration, however, i t was deoided that a l l of the data (Adhesives A, B, C, and X>) should be inoluded*  Transformation of the data was necessary t o meet the l i n e a r  c o r r e l a t i o n requirement of Slope Ratio Assays* w i l l be explained l a t e r *  The transformation ohosen  This deoision t o use a l l of the data was prompted  by an i r r e g u l a r i t y i n the data, namely, that Tension Shear r e s u l t s i n d i c a t ed a tendency f o r the breaking loads of Adhesive A t o be less than those of Adhesive B*  This was oontrary t o expectation based on preliminary t e s t s ,  and contradictory t o the r e s u l t s given by most, but not a l l , of the Glueline-Cleavage t e s t s * Shear Tests*  Furthermore, i t was not true i n a l l oases of Tension  I t should be noted that although the trend was a3 desorlbed  above, the differences between the means was i n no case s i g n i f i c a n t *  It i s  possible, therefore, that t h i s apparent trend i s a r e s u l t of sampling vari a t i o n s and oould be mathematically discarded on that basis* This question of whether i t was l i k e l y that Adhesive B (22j^S r e s i n s o l i d s ) would give greater bond strength than Adhesive A (2lj0t r e s i n s o l i d s ) , was disoussed with the ohemists of the r e s i n manufacturer*  They  reported that an adhesive formula containing 2 2 ^ r e s i n s o l i d s had been subjected t o faotory t e s t s and they oonoluded that " i t w i l l not quite come up t o weatherproof standards"•  The inference i s that t h i s adhesive d i d not  produoe as good a bond as the standard mix. Adhesive A* There i s no apparent reason why the glue lines (of Tension Shear and Glueline-Cleavage Speoimens) prepared from the same sheet of plywood  - 59 should not have i d e n t i c a l strengths* loads should follow the same pattern*  This being the oase t h e i r breaking There i s no l o g i c a l explanation  why Tension Shear and Glueline-Cleavage breaking loads should show opposite trends i n t h i s o r i t i o a l region*  I t was decided t o proceed therefore on  the assumption that there i s an aotual decrease i n the strength of the bond when r e s i n solids are reduced from 27^S t o 2 2 ^ and that the differences noted are the r e s u l t of sampling "errors"• A logarithmic transformation of the per cent r e s i n solids soale was  chosen t o allow the use of l i n e a r c o r r e l a t i o n analysis as required f o r  Slope Ratio Assays* i n d i v i d u a l curves*  Figure 5, Appendix B, i l l u s t r a t e s t y p i c a l p l o t s of I t may be observed that a straight l i n e plot i s a  compromise between what should be a s l i g h t l y oonvex downward curve f o r the least severe accelerated weathering treatment, t o a s t r a i g h t l i n s f o r the medium treatment, and a s l i g h t l y oonvex aooelerated weathering  upward' curve f o r the most severe  treatment*  Description of the Method Used In order to obtain the greatest accuracy when comparing t e s t speoimen designs, comparisons should be made w i t h i n eaoh treatment combination*  In order to present a v i s u a l picture of the data, percentage  reduc-  tions i n breaking load were plotted over per oent r e s i n s o l i d s f o r eaoh of Designs 1 to 1 3 , and f o r eaoh accelerated weathering treatment*  In spite  of the faot that a logarithmic soale was used f o r per cent r e s i n solids i t was  decided t o plot the data on a n a t u r a l scale (see Figures 6 , Appendix B  for the data of speoimens t e s t e d a f t e r accelerated weathering treatment I ) . This necessitated p l o t t i n g the (straight) regression l i n e as a curve* Examination w i l l show that the use of logarithmic transformation f o r the l i n e a r c o r r e l a t i o n analysis and subsequent p l o t t i n g of the data on natural  - 60 co-ordinates serves two useful purposest  (1) i t meets the linear c o r r e l a -  t i o n demands of Slope Ratio Assays, and (2) i t gives a true picture of the shape of the f i t t e d curve and the spread of the data about i t . Linear regression analyses were made between peroentage reduction i n breaking load and the logarithm of per oent r e s i n solids f o r each of the above.  The regression l i n e s were plotted on the oorrespending graphs of  Figures 6, Appendix B.  The slope r a t i o , b/Se s Slope , Standard error of estimate  was next oaloulated f o r eaoh of the above mentioned regressions and t h i s figure added t o eaoh graph of Figures 6, Appendix B. The next step was t o graph the families of ourves for each design; these are Figures 7, Appendix B.  Upon inspection of these ourve  patterns f o r d i f f e r e n t designs i t i s noticeable that c e r t a i n designs, notably 1, 2, 4» 5. 6 and 11 have produced f a i r l y regular patterns'subh.as oould be expected from a consideration of the probable effects of the t r e a t ments.  Others, notably 8, 9, 10 and 12, have the symmetry of t h e i r pattern  broken by certain curves crossing others.  This i s interpreted as an i n -  d i c a t i o n that r e s u l t s obtained from these designs are l e s s r e l i a b l e than those obtained from the other designs. A d d i t i o n a l study has been l i m i t e d therefore t o those designs which produced regular and l o g i c a l curve patterns. Designs 3, 7 and 13, having r e l a t i v e l y minor d i s t o r t i o n i n t h i s respeot, have been included f o r further a n a l y s i s . Analysis of Varianoe of Slope Ratios Next, the slope r a t i o s were arranged i n the form of Table 5A, Appendix B, and an analysis of variance (Table 5B, Appendix B) was made t o determine whether or not the differences between the means were s i g n i f i c a n t . When t h i s was found t o be so, a further comparison was undertaken between i n d i v i d u a l means or groups of means (Single Degree Comparisons).  Table 6,  - 61 Appendix B, l i s t s the Single Degree Comparisons  of most i n t e r e s t *  The oomparison between Designs 1 and 13 (means 10.978 and 9*173 from Table 5A. Appendix B) f a i l e d t o show a s i g n i f i c a n t difference the means).  (between  Because Designs 1, 2 and 3 have much smaller differences there  i s no s i g n i f i c a n t difference between t h e i r means. Study of the comparisons l i s t e d i n Table 6, Appendix B, prompts the following conclusions} (1)  the three Glueline-Cleavage Designs u t i l i z i n g EGxlO veneers are 0  s i g n i f i c a n t l y more sensitive than both the Tension Shear Designs and those Glueline-Cleavage Designs using either edge-grain or f l a t - g r a i n material (comparisons o f Designs 1 • 2 • 3 v s . 4 • 5 . * 6, 1 v s . 11, 3 v s . 11,  I v s . 13, 3 v s . 13, and 1 • 2 • 3 v s . 1 3 ) i ^ / , (2)  oross-banded specimens prepared from either f l a t - g r a i n or edge-  g r a i n veneers exhibit no s i g n i f i c a n t difference between the means of Tension Shear or Glueline-Cleavage methods (comparison of Designs 4 • 7 v s .  I I • 13), (3)  the data f a i l t o y i e l d s i g n i f i c a n t differences i n s e n s i t i v i t y  between designs 1, 2 , and 3 , the three EGxlO Designs, a t the accepted 0  l e v e l of s i g n i f i c a n c e .  Inspection of the Slope Ratios tabulated i n Table  5A, Appendix B, w i l l show that i n every oase the Slope Ratio f o r Design was greater than that f o r Design 2 .  I t i s highly probable that a  larger sample would indicate Design 3 t o be s i g n i f i c a n t l y more sensitive than Design 2 .  Comparison of the Slope Ratios of Design 3 with Design 1,  on the other hand, w i l l show that i n one-third of the comparisons the Slope Ratios of Design 1 exceed those of Design. 3 whereas i n the remaining two_ See Figure 2 , Appendix B, f o r speoimen designs.  - 62 t h i r d s the opposite was true*  I t seems l i k e l y that further sampling would  substantiate the theory that no s i g n i f i c a n t difference i n s e n s i t i v i t y exists between -these designs* use of EGxlO  0  Glueline-Cleavage Designs 1 and 3 (speoimens making  cross-banded veneers), have been shown t o be the most sens-  i t i v e of the twenty-six designs inoluded i n t h i s study. i l i t y that Design 2 should be inoluded on an equal b a s i s *  There i s a possibDesign 1 would,  therefore, be the ohoioe when (oross-banded) plywood adhesive j o i n t s are under study.  Design 3 (or Design 2 when more applicable t o the work i n  hand) would be the ohoioe when laminated adhesive j o i n t s are under study* (o)  Glue Strength vs* Wood Strength The f a l l a c y of attempting t o measure the strength of a strong  adhesive through the medium of weak wood has been discussed, and ene o f the objectives o f t h i s work was t o f i n d a method of avoiding t h i s .  A simple  method i s proposed f o r e s t a b l i s h i n g whether a given adhesive i s stronger or weaker than the wood whioh i t i s t o bond*  The p r i n o i p l e i s t o use several  formulations o f the adhesive ingredients instead of the single mixture recommended by the glue manufacturer.  Some of these are purposely made  "weaker than the wood". A few t r i a l s w i l l soon e s t a b l i s h which formulae are "weaker than the wood" and whioh are not. The i n t e r p r e t a t i o n t o be taken from such data depends t o some extent upon the r e l a t i v e strengths of the wood and of the glue bond or the glue.  When the glue i s "weaker than the wood", and treatments have been  suoh as not t o a l t e r the strength of the wood, any reduction i n strength would be a t t r i b u t a b l e t o the e f f e c t of the treatment upon the adhesion or oohesion of the glue*  Different conclusions may be required f o r hot-press  phenolic r e s i n adhesives, whioh are known t o be very i n e r t chemical substanoes when completely polymerized*  These adhesives, when used t o bond  - 65 Douglas f i r aeoordlng to the manufacturer's  Instructions, form bonds whioh  are "stronger than the wood". Any reduction i n strength induoed by a t r e a t ment may now be subjeot t o two or more i n t e r p r e t a t i o n s . between the wood and the glue has decreased.  One i s that adhesion  Another i s that the strength  of the wood i n the v i c i n i t y of the bond has been reduoed.  Should i t be  necessary t o d i f f e r e n t i a t e between these, a determination of whether the adhesive i s or i s not "weaker than the wood" may provide a key t o the s o l ution. 6.  SUMMARY Estimates of plywood bond q u a l i t y based upon the breaking loads  of meohanioal t e s t s have been shown t o be superior i n aoouraoy t o those based upon per oent wood f a i l u r e . The importance  of determining whether the adhesive i s "weaker  than the wood" or "stronger than the wood" has been stressed and a method for making t h i s determination i n d i c a t e d . When used f o r "trouble shooting", "production t e s t i n g " , or simi l a r work where the EGxlO Speoimen oannot be employed, the i n d i c a t i o n i s 0  that the Glueline-Cleavage and Tension Shear methods are of equal aoouraoy. The Glueline-Cleavage Method, when used with EGxlO Specimens, 0  designed f o r maximum aoouraoy, has been confirmed as the most acourate method subjected t o t e s t .  - 64 CHAPTER 71.  CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH  The purpose of t e s t i n g wood-to-wood bonds in. general and plywood bonds i n p a r t i c u l a r has been reviewed. the history of glue bond t e s t i n g . has been summarized.  A b r i e f summary has been made of  The excellent researoh of early w o r k e r s ^  F i n a l l y , reasons have been given f o r a continuing need  of methods t o estimate bond q u a l i t y .  Since established t e s t s f a i l to meet  t h i s need within the required l i m i t s of aoouraoy there i s a neoessity t o develop more d e f i n i t i v e methods.  The scope of t h i s researoh has been  limited to hot-press phenolic r e s i n bonds of Douglas f i r veneers. I n i t i a l l y adhesive researoh workers used mechanical t e s t s t o measure bond q u a l i t y .  More recently, estimates based upon the percentage  of wood f a i l u r e i n a ruptured bond have become popular, e s p e c i a l l y f o r glue s p e c i f i c a t i o n or quality control purposes.  The mechanical t e s t has been  shown t o be b a s i c , t e s t s based upon estimates of per oent wood f a i l u r e being of value only a f t e r a r e l i a b l e c o r r e l a t i o n has been established between the two methods.  D i f f i c u l t i e s of i n t e r p r e t i n g wood f a i l u r e have been reviewed  and the neoessity f o r a world-wide exchange of standardized per oent wood f a i l u r e specimens suggested.  The f a l l a c y of attempting t o measure the  strength of strong glues through the medium of weak wood has been pointed out.  Some of the d e f i c i e n c i e s of e x i s t i n g meohanical tests have been r e -  viewed.  The continuing need f o r meohanioal methods of measuring the  strength of adhesive bonds has been established.  In a d d i t i o n , i t has been  shown that the defects of mechanical methods may be corrected more e a s i l y than those of other methods. Mention has been made of the f u t i l i t y of attempting t o obtain uniform pure shear, t e n s i l e , or compressive stresses across glue l i n e t e s t Anderson, Browne, Brouse, Hopkins, Lee, McBain, and Robertson.  - 65 areas*  The r o l e of stress concentrations i n premature f a i l u r e has been  reviewed*  The most u s e f u l stress d i s t r i b u t i o n i s one i n which the maximum  stress occurs  i n a (single) known portion of the glue l i n e *  A l i s t has  been included of the d i f f i c u l t i e s t o be overcome i n any e f f o r t t o measure the stresses i n a p a r t i c u l a r specimen*  The hope of successfully using stress  analyses t o reduce v a r i a b i l i t y ( i n spite of the above-mentioned d i f f i c u l t i e s ) has been expressed* Methods of reducing the v a r i a b i l i t y between breaking loads of specimens have been considered.  These began with the recommendations of  the Adhesive Researoh Committee^ ), proceeded t o those of various other 4  a u t h o r i t i e s , and f i n a l l y incorporated some of the author's views.  The r o l e  of s t a t i s t i c a l techniques, controlled stress concentrations, and controlled lathe oheok orientation i n reducing v a r i a b i l i t y has been noted. niques i n c l u d e i  Other tech-  f i r s t l y , transforming breaking loads of glue joints t o per-  centages of wood strength (for comparative  purposes); seoondly, the use o f  several adhesive q u a l i t i e s to avoid the dilemma created by attempting t o measure the strength of strong glue with weak woodj and f i n a l l y , employing the rate of reduotion i n strength induced by a series of accelerated weathering  cycles o f increasing severity* Theories regarding a few s p e c i f i c sources of exoessive v a r i a b i l i t y  i n test r e s u l t s have been formed.  The premature f a i l u r e of Tension Normal  specimens due t o stress concentrations a t undesirable locations i n the speoimen i s the f i r s t of these*  Others include the effeot of sawout depth, and  the varying pattern of springwood and summerwood i n introducing excessive v a r i a b i l i t y into t e s t data* Theories have been expressed regarding the e s s e n t i a l of the i d e a l mechanical t e s t f o r glue bonds*  requirements  The major requirements whioh  - 66 have been, set a r e i  f i r s t , that t e s t . r e s u l t s be reproduoible; second, that  r e s u l t s be expressed i n a u n i v e r s a l l y reoognized measure of force or s t r e s s ; t h i r d , that specimens be of a design simple t o prepare t o a high standard of aoouraoy; fourth, that the maximum number of speoimens be obtainable per unit area of plywood; f i f t h , that the machinery required be inexpensive; and s i x t h , that the method be adaptable t o both researoh and production testing. Methods of reducing-the v a r i a b i l i t y due t o c e r t a i n suspected sources have been studied. These sources include the d i f f e r i n g springwoodsummerwood pattern between speoimens (and w i t h i n d i f f e r e n t areas of a single speoimen), the o r i e n t a t i o n of the wood c e l l s t o the surfaoe of the veneers, the roughness o f veneer surfaces before gluing, and the necessity for standa r d i z i n g every d e t a i l of machine and specimen design. The above-mentioned study resulted i n the development of a new method of t e s t which has been named the Glueline-Cleavage Test.  The method  i s t o measure the force required t o oleave or s p l i t a specimen by means of the wedge.  The a c t i o n i s s i m i l a r t o s p l i t t i n g wood with a wedge except that  the knife i s placed along the glue l i n e instead of on a simple (or s o l i d ) block of wood.  Figures 1 and IA, Appendix A, i l l u s t r a t e the maohine with a  specimen i n p o s i t i o n f o r t e s t i n g .  For produotion t e s t i n g of plywood, one-  inoh square speoimens are cut from the sheets a t an angle of 45° to the grain.  For researoh purposes, on the other hand, where aoouraoy i s of the  utmost importance, speoial specimens have been designed.  In these speci-  mens the wood has been out so that the c e l l s i n t e r s e c t the surface of the veneers a t an angle of 10°.  The method makes use of t h i s small angle, plus  the r e l a t i v e weakness of the wood i n tension perpendicular t o the grain, t o concentrate the stress i n a small area of glue l i n e immediately below the  -  knife edge*  67  -  Care must be taken to see that the knife i s applied t o the  correct edge of the glue l i n e , otherwise the rupture w i l l tend t o occur along the grain of the wood instead of down the glue l i n e *  When t h i s i s  done every specimen s p l i t s through the adhesive or a t the wood-glue i n t e r face*  Where cross-banded edge-grain veneers are used i n test specimens the  v a r i a b i l i t y i s reduced beoause of the more equal d i s t r i b u t i o n of the springwood-summerwood w i t h i n each specimen and between specimens* In the exploratory t r i a l , using lathe-out veneer samples, the Glueline-Cleavage Test gave more reproduoible r e s u l t s than the Tension Shear or  Tension Normal methods.  A second comparison was made, t h i s time between  the Tension Shear, Tension Normal, Blook Shear (modified specimen s i z e ) , and the Glueline-Cleavage Test Methods*  EGxlO Specimens with veneers cross0  banded a t an angle of 90° were employed i a t h i s t r i a l .  The l a t t e r t r i a l was  conducted with conditions designed t o y i e l d minimum v a r i a b i l i t y between the breaking loads of d i f f e r e n t specimens*  Again the Glueline-Cleavage Test  gave the highest aoouraoy* I  A t h i r d t r i a l was conducted t o determine whether the Glueline-  Cleavage Method would maintain t h i s superiority over a wider range of bond strengths and treatments, and t o compare the accuracies of a larger number of specimen designs.  For t h i s purpose nine Glueline-Cleavage and four  Tension Shear designs were included. "Treatments" included four strengths i of hot-press phenolic r e s i n adhesive and s i x s e v e r i t i e s of accelerated weathering.  R e p l i c a t i o n was provided by i n c l u d i n g eight trees i n the sample.  One speoimen from each t r e e was subjeoted t o every treatment p r i o r to t e s t . quality»  Eaoh t e s t speoimen yielded two independent  combination  estimates of bond  the breaking load of the meohanioal t e s t , and the percentage of  wood f a i l u r e .  I n e f f e c t twenty-six estimates of the various bond q u a l i t i e s  - 68 -  were a v a i l a b l e f o r comparison, two f o r each t e s t specimen design*  Although  estimates of bond q u a l i t y based upon wood f a i l u r e had been ruled out previously on t h e o r e t i c a l grounds, i t was thought worthwhile t o compare them mathematically with mechanical t e s t r e s u l t s * Meohanioal estimates of bond strength proved superior i n accuracy to those based upon per oent wood f a i l u r e *  The Glueline-Cleavage and  Tension Shear methods proved equally aoourate when used with specimens of the type a v a i l a b l e f o r "production-testing" or "trouble-shooting" purposes* The machine used f o r Glueline-Cleavage t e s t s was very inexpensive by comparison with that required for the Tension Shear t e s t s .  This oould overcame  a r e s t r i c t i v e f a c t o r i n promoting q u a l i t y control of the gluing process i n plywood manufacture.  A method has been demonstrated f o r insuring that the  r e l a t i v e strengths of wood and glue are known, information whioh may  lead  to more acourate conclusions regarding the nature of the bond f a i l u r e s . The superior accuracy of the Glueline-Cleavage Test, when used with crossbanded EGxlO Specimens, was 0  the method are*  confirmed.  Other worthwhile advantages of  (1) the glue l i n e of every specimen i s exposed f o r inspec-  t i o n , a most desirable feature, (2) test speoimens are easy to manufacture and consequently r e l a t i v e l y inexpensive, (3) less t e s t i n g time i s required than with the other methods compared, and (4) a maximum number of specimens may be prepared from a given sheet of plywood.  This l a t t e r feature i s of  value when i t i s necessary t o use experimental designs involving large numbers of matohed specimens* A l l the above-mentioned advantages of the Glueline-Cleavage Test, with one exoeption, confirmed the findings of e a r l i e r t r i a l s . '  This  exception r e l a t e d t o the accuracy of the method when used for productiontype t e s t i n g .  The f i r s t comparison indicated the Glueline-Cleavage Test  - 69 to be of superior accuracy, whereas the t h i r d t r i a l indicated only equality i n t h i s respect.  The data and analyses  of the f i r s t and t h i r d t r i a l s are  not d i r e c t l y comparable, nevertheless i t was a n t i c i p a t e d that they would indicate the same trend.  Further experimental work seems t o be indicated  to resolve thiB d i f f e r e n c e . In spite of the improved acouracy demonstrated f o r the GluelineCleavage Test the v a r i a b i l i t y between the breaking  loads of matched speci-  mens i s s t i l l disconcertingly large and a few observations experimentation might be appropriate.  on further  Research, by i t s very nature, un-  oovers new problems during i t s progress.  One which appears worthy of  further i n v e s t i g a t i o n i s the stress d i s t r i b u t i o n i n a t e s t specimen.  Suf-  f i c i e n t work(45) has been done t o indicate that t h i s problem may be solved by using a mathematical approach.  A more p r a c t i c a l and r e l a t e d problem i s  to develop a method that makes use of present knowledge regarding the stress d i s t r i b u t i o n t o reduoe the v a r i a b i l i t y between specimen breaking  loads.  This i s amenable t o s o l u t i o n since the region of maximum stress i s known. Bond strength i s frequently correlated with s p e c i f i c gravity of the pieces bonded.  A correction based upon the s p e c i f i c gravity of the wood a t the  most highly stressed portion of -the bond appears t o offer a means of reducing  v a r i a b i l i t y between the breaking  loads of otherwise matohed specimens.  Probably a s i m i l a r correction could be determined f o r v a r i a t i o n s i n the angle a t which the wood o e l l s i n t e r s e c t the highly stressed portion of the glue lines i n d i f f e r e n t specimens. Another point deserving a t t e n t i o n i s the conversion of breaking loads t o a common basis so that comparisons may be made between specimens whioh are not matohed for veneer thickness, density, and species.  The lack  of such a system has been one of the l i m i t i n g factors i n the use of meoh-  a n i c a l test methods.  For oertain researoh purposes i t i s quite f e a s i b l e  to use a speoimen of standard dimensions, f o r oertain other purposes, however, t h i s may be impraotioal. Plywood-bond q u a l i t y oontrol frequently requires that tests be made on speoimens with more than one veneer t h i c k ness.  Conversion of a l l data t o a oommon basis f o r oomparison would be  desirable.  The author has demonstrated, by means of tests not recorded  here, that breaking loads are correlated with veneer thicknesses.  Once  t h i s c o r r e l a t i o n has been worked out t o the necessary degree of accuracy i t should be possible t o oonvert the measured breaking load t o that expeoted with veneer of the ohosen standard thiokness.  Likewise, i t should be  possible t o e s t a b l i s h a s i m i l a r system t o compare bond strengths between d i f f e r e n t s p e c i f i c g r a v i t i e s or species. The author i s convinoed that researoh to further improve the accuracy and the adaptability of the Glueline-Cleavage Test, along the lines indicated above, would be a worthwhile contribution t o further development i n the f i e l d of adhesives and adhesion.  71 BIBLIOGRAPHY American Society f o r the Testing of Materials (A.S.T.M.) Standards, Philadelphia, A.S.T.M., 1953. American Society f o r the Testing of Materials, Symposium on Adhesives, Philadelphia, A.S.T.M., 1945. Bensend, D.W., and Preston, R.J., Some Causes of V a r i a b i l i t y i n the Results of Plywood Shear Tests, Madison, Wisconsin, U.S. Forest Products Laboratory Report No. R1615, 1946. Bergin, E.G., The Significance of Wood Failure i n Glued J o i n t s , Canadian Woodworker, March, 1953, ( r e p r i n t ) . Bergin, E.G., and W.E. Wakefield, A Comparison of Test Methods for the Evaluation of Cold-Press Urea-Formaldehyde Resin Glues, Ottawa, Canada, Forest Products Laboratories of Canada, 1946. Bethel, J.S., and J.B. Huffman, Influence of Lathe Cheok Orientation on Plywood Shear Test Results, Sohool of Forestry, North Carolina State University, 1950. B l i s s , C.I., and D.W. Calhoun, An Outline of Biometry, New Haven, Conneotiout, Yale University, 1951, Mimeo. B r i t i s h Standards I n s t i t u t e S p e c i f i c a t i o n B.S.1203, Synthetic Adhesives for Plywood (Phenolic and Aminoplastio), 1945. Brouse, Don, Factors A f f e c t i n g the Test Value of Casein Water-resistant Plywood, Purdue University Master's Thesis, 1927, (unpublished). Chelvarajan, B.K., Comparative Study of Indian and American Plywood Shear Test Standards, Indian Forester, V o l . 80, No. 1, 1954. Commercial Standard CS 35-49, Hardwood Plywood, U.S. Department of Commerce, National Bureau of Standards, 1949. Cousins, F.W., Determining Stress i n Glued Joints by Photo-elastio Method, Timber News, V o l . 58, No. 2127, January, 1950. Department of S c i e n t i f i c and I n d u s t r i a l Researoh, F i r s t Report of the Adhesives Research Committee, London, His Majesty's Stationery O f f i c e , 1922. Department of S c i e n t i f i c and I n d u s t r i a l Researoh, Second Report of the Adhesives Research Committee, London, His Majesty's Stationery Offioe, 1926. Department of S c i e n t i f i c and I n d u s t r i a l Researoh, Third and F i n a l -Report of the Adhesives Researoh Committee, London, His Majesty's Stationery Offioe, 1932.  - 72 (16)  Department of S o i e n t i f i o and I n d u s t r i a l Research, Report of the Committee on the Mechanical Testing of Timber, London, His Majesty's Stationery O f f i c e , 1934.  (17)  Elmendorf, A.» Basio Tests f o r Plywood, Forest Products Research Sooi e t y Proceedings, V o l . 2, 1948.  (18)  Elmendorf, A., Methods of Testing Wood Adhesion, Wood and Wood Products, May, 1952.  (19)  Frooht, M.M., Photo E l a s t i c i t y , New York, N.Y., John Wiley and Sons, 1949, 2 volumes.  (20)  Hopkins, R.P., Evaluation of Resin Adhesives, Forest Produots Research Society Proceedings, V o l . 3, 1949.  (21)  Kline, G.M. and F.W. Reinhart, The Fundamentals of Adhesion, Paper Trade Journal, V o l . 129, No. 26, December, 1949.  (22)  Knight, R.A.G., Requirements and Properties of Adhesives f o r Wood, Forest Products Researoh B u l l e t i n No. 20, London, His Majesty's Stationery Offioe, 1950.  (23)  Knight, R.A.G., Doman, L.S., and Newall, R.J., D u r a b i l i t y Tests on Plywood Adhesives - Series,II, F i f t h Year's A n a l y s i s , Investigations i n t o Glues and Gluing, Progress Report Sixty, Prinoes Risborough, England, Department of S c i e n t i f i c and Industrial Research, Forest Produots Laboratory, 1951.  (24)  Knight, R.A.G., Adhesives for Wood, New York, N.Y., Chemical Publ i s h i n g Co. I n e , 1952.  (25)  Knight, R.A.G., D u r a b i l i t y and Performance of Adhesives f o r Wood, P l a s t i c s Progress, 1953, ( r e p r i n t ) .  (26)  Knight, R.A.G., The Assessment of Bond Quality i n Glued Joints, Part 1, Analysis of Series I I Plywood Experiments, Princes Risborough, England, Department of S o i e n t i f i e and I n d u s t r i a l Research, Forest Products Research Laboratory, 1953.  (27)  Laoey, P.M.C. and H.A. Howe, The Testing of Glues by the Glueline Method, Prinoes Risborough, England, Department of S c i e n t i f i c and I n d u s t r i a l Research, Forest Produots Laboratory, 1949 and 1950.  (28)  Levon, M., Proposal f o r the Establishment of Rules f o r the Standardi z a t i o n of Test Methods and Test Pieoes for the Strength Testing of Plywood, Paper t o the Timber Researoh Committee of the International Union of Forest Researoh Organizations, 1939, (N.V.M.).  (29)  Marra, A.A., Glue Line Doctor, Southern Lumberman, V o l . 183, No. 2290, September, 1951.  (30)  Marra, G.G., and J.W. Wilson, Preliminary Report on Test Method for Evaluating G l u a b i l i t y of Hardboard, Pullman, Washington, D i v i s i o n of I n d u s t r i a l Research, Washington I n s t i t u t e of Technology, Washington State College, 1952.  - 73 Maxwell, J.W., Shear Strength of Glue Joints as Affected by Wood Surfaces and Pressures, Syracuse University, Teohnical Publication, No, 64, New York State College of Forestry, 1944. Newall, R.J., Tests Connected with the Proposed Revision of B.S.1203, Progress Report 63, Investigations into Glues and Gluing, Princes Risborough, England, Department of S o i e n t i f i o and I n d u s t r i a l Research, Forest Products Laboratory, 1953. Northoott, P.L., Analysis of the Stress Pattern Induces i n a Plate of I n f i n i t e Width by a 90° Wedge Acting upon a Matching Groove i n the Plate, M.E.561, Vancouver, Canada, Mechanical Engineering Department, University of B.C., 1953* (unpublished). Northoott, P.L., The Development of the Glueline-Cleavage Test, Journal of the Forest Products Research Society, V o l . I I , No. 5, Deoember, 1952. Orth, Otto G., J r . , Radio-frequenoy G l u i n g — A Research Project, P a c i f i o P l a s t i c s , October, 1947, (N.V.M.). Perkins, N.S., Predicting Exterior Plywood Performances, Forest Products Research Society, Proceedings No. 4, 1950. Royal Canadian A i r Force S p e c i f i c a t i o n C-22-2 (A s p e c i f i c a t i o n f o r the c e r t i f i c a t i o n of oertain types of adhesives), Ottawa, Department of National Defenoe for A i r , 1942, Rudkin, A.W., A Simple Method of Testing Glue Lines i n Tension, Reprint No. 102, A u s t r a l i a , Council for S c i e n t i f i c and I n d u s t r i a l Researoh, D i v i s i o n of Forest Products, 1947. Selbo, M.L., and W.Z. Olseh, D u r a b i l i t y of Woodworking Glues i n D i f ferent Types of Assembly J o i n t s , Journal of the Forest Produots Researoh Society, V o l . I l l , No. 5, 1953. Truax, T.R., The Gluing of Wood, U.S. Dept. of Agr. Bui. 1500, 78 pp., Washington, D.C, U.S. Government P r i n t i n g Offioe, 1929. Truax, T.R., F.L. Browne, and Don Brouse, Significance of Meohanioal Wood-joint Tests f o r the Selection of Woodworking Glue3, Madison, Wisconsin, U.S. Dept. Agr. Forest Products Laboratory, 1929. Truax, T.R., Development of Wood Adhesives and Gluing Teohnie, Transactions of the A.S.M.E., No. 54, February, 1932, United States Forest Produots Laboratory, Veneer and Plywood, U.S. Dept. Agr., Forest Products Laboratory, Madison, Wisconsin, 1919. United States Forest Produots Laboratory, Madison, Wisoonsin, E f f e c t of Heat on the Properties and S e r v i c e a b i l i t y of Wood, Experiments on Thin Wood Specimens, 1945. Wakefield, W.E., The Tension Normal t o the Glue Line Plywood Test, Mimeo. No. 121, Ottawa, Canada, Forest Produots Laboratories of Canada, 1947.  74 -  APPENDIX A  .5  - 75 -  TABLE IA, APPENDIX A  Tension Shear Tests of Douglas P i r Springwood Bonded t o Springwood and Summerwood t o Summerwood - Using RCAF S p e c i f i c a t i o n C - 2 2 - 2  Summerwood t o Summerwood  Wood Failure  %  Breaking Load Lbs.  915  20  415  100  945 800  5 100  520  100  460  100  930 850  100 100  620 485  100 100  950  95 20  545 390  100 100  90 70  510  50  Breaking Load Lbs.  Wood Failure*  1125 1135 1175 960  Mean Breaking Loads $ Standard Deviations! Coefficients of V a r i a t i o n :  Springwood t o Springwood  988.5 142.2  100  lbs. n  1 4 . 4 per oent  %  570  90  645 565  100 100  520.4 80.0  lbs. n  1 5 . 4 per cent  Note - The C o e f f i c i e n t of V a r i a t i o n of a l l the above specimens combined i s 3 5 . 3 per cent. •Percentage of wood f a i l u r e was estimated using the Douglas F i r Plywood Association standard.  - 76 -  TABLE IB, APPENDIX A  Blook (Compression) Shear Tests of Douglas F i r Springwood Bonded t o Springwood and Summerwood t o Summerwood - Using a Modified Speoimen Size l Square w  Summerwood t o Summerwo od Breaking Load Lbs.  Wood Failure*  3550  90 95 65 100 50 65; 30 95 85  •  2830 3700 3870 3480 3650  2240 4020 4020 Mean Breaking Loads» Standard Deviationsi Coefficients of V a r i a t i o n i  % •  3484.4 l b s . 589.0 16.9 per oent  Springwood t o Springwood Breaking Load Lbs.  Wood Failure*  2540 2330  100 15  -  2710 2820 2670  2330 1710 2160 2810  2453.3 l b s .  %  30 95 95  1Q0 10 100 70  362.6 n 14.8 per oent  Note - The Coefficient of V a r i a t i o n of a l l the above speoimens combined i s 23.9 per cent. •Percentage of wood f a i l u r e was estimated using the Douglas F i r Plywood Association standard.  - 77 -  TABLE 1C, APPENDIX A  Tension Normal t o the Glue Line Plywood Tests o f Douglas F i r Springwood Bonded t o Springwood and Summerwood to Summerwood Summerwood t o Summerwood  Mean Breaking Loadst Standard Deviations! Coefficients of V a r i a t i o n :  Breaking Load Lbs.  Wood Failure*  570 690 745 490 490 395 590 710  100 100 90 100 100 100 100 100  *.  585.0 l b s . 123.3 n 21.1 per oent  Springwood t o Springwood Breaking Load Lbs.  Wood Failure*  645 770 610 790 740 87© 635 525  100 100 100 100 100 100 100 100  %  698.1 l b s . n 113.1 16.2 per oent  1f  Percentage of wood f a i l u r e was estimated using the Douglas F i r Plywood Association standard.  - 78 -  TABLE 2, APPENDIX A  Coefficients of V a r i a t i o n of Breaking Loads Exploratory Comparison of the GluelineCleavage Test with Tension Shear and Tension Normal to the Glue Line Plywood Test  Coefficients of V a r i a t i o n - per oent Sample No.  Tension Normal  1 2 3 4 5 6 7 8 9 10 11 12  38.0 41.1 31.4 31.8 32.4 30.0 28.0 26.6 42.5 60.1 40.1 48.4  Means t Notei  37.5#  Tension Shear  GluelineCleavage  29.1 28.2 27.5 28.1 25.5 25.6 22.4 19.2 24.0 26.2 • 27.6 24.5  13.4 23.5 16.6 18.9 18.6 16.4 16.9 15.4 14.3 16.5 20.3 17.6  25.6%  17.4??  Eaoh sample (for eaoh method of t e s t ) oonsists of 26 speoimens. Total number of speoimens tested 26 x 3 x 12 s 936. :  Glueline Cleavage EG x10° Glueline Cleavage EG x10° Specimens. 90° Knife Speoimens. 90° Knife Not Grea oed Greased Before Eaoh Test Breaking % M.C. S.G. Breaking % M.C. S.G. Breaking % % M.C. S.G. Breaking % % M.C. S.G. Breaking % M.C. S.G. % % % Load Wood at Oven Load Wood at Oven Load Wood at Oven Load Wood at Oven Load Wood at Oven (lbs.) Failure Test Dry (lbs.) Failure Test Dry (lbs.) Failure Test Dry (lbs.) Failure Test Dry (lbs.) Failure Test Dry  Tension Normal tothe Glue line  440 505 387 455 450 150 425 385 300 270  345 345  415 370 285 440 380  430  414 267  100  267 352 335 264 265 213 389 280 380 337 185 365 355 245  80 90  100 100 100 100 90  95 ioo j 95 " 85 100 '  100 ' 100  100 100  100 100 100 100  100 100 100 100 100 100 100  100 \ 100 \ ioo !  100 95 95 95  100  .70 .68 .68 .68 .67 .68 .67 .67 5.0 .68 4.5 .69 4.1 .65 6.7 .68 5.1 .64 4.7 .64 4.9 .65 4.6 .66 4.7 .66 4.8 .65 4.7 .55 5.0 .64 7.2 .65 4.8 .64 5.6 .65 4.7 .63 5.4 .64 6.4 .64 4.1 .64 5.8 .66 4.5 .'65 6.0 .65 4.6 .65 4.8 .63 5.6 .64 6.9 .66 4.6 .68 4.3 5.5 5.5 5.1 5.5 4.e 5.1 4.9  97.7  5.16 .655  295 309 345 302 310 300 290 325 280 298 282 318 332  323 295 306 307 327 312  320  392 420 425 267 282 295 280 310 215 280 305 280 233 275  308.1  42.2  <T= 91.8  M.C. S.G.  348  100 100 85  95 100 100 100 100 95 85 90 80  100 100  100 100 100 90  100 100 60  100 100 95 90  100 85  50  80 95 90 60 40 70 95  6.2 5.7 5.3 5.6 5.7 5.5 5.0 5.8 5.1 5.6 5.4 6.1  .66 .66 .64 .64 .58 .69 .70 .66 .69 .68 .68 .69 5.0 .70 5.1 .68 5.4 .66 5.5 .72 4.8 .67 5.6 .71 5.4 .68 5.7 .69 5.3 .70 4.e .68 5.5 .68 4.5 .69 4.5 .67 6.1 .65 ' 7.0 .68 4.9 .64 6.0 .64 6.1 .65 5.7 .68 5.0 .69 5.1 .69 5.2 .69 5.5 .69  Moisture Content Speoifio Gravity  -  Blook Shear  1873 2038 ! 1593 1127 1242  90  1203 2035  100 100 95 100 100 100  !  1748 1962 1648  100 100  1\ 1'  1797 1648  !  1  1995  2015 1775 1915 1495 | ,1572 1420 1981 1648 1315 1863 1360 1805  1055  2125 1985 2515 2215 1325 2105 1540 2185 2290  95 95  85 85 95 90 95 75 95  100 85  100 95 80 95 90  100 85 75 85 95 90 95 85 95 90  6,e  .74  6.0 .70  .67 .73 .72 .74 .69 .69 .69 .62 .68 .60 .59 .62 .62 .66 .60 8.C .57 5.4 .69 5.1 .66 4.3 .68 4.4 .70 7.5 .71 6.5 .70 5.8 .69 5.0 .69 5.2 .69 8.5 .67 3.9 .68 5.1 .68 5.7 .66 6.2 .67 5.2 .68 6.5 .68 5.0 .68  5.1 5.8 6.2 5.1 7.3 6.4 7.4 5.6 5.1 5.2 4.6 5.4 5.0 5.2 5.5  89.4  5.45 .674  1754.6 356.2  209 209 260 222 228 267 306 260 241 267 319 196 228  254  196 280 274  222  209  202 202  5 5 5 5  20 20 15 5 5 5 5 5 15 15  10  15 0  10 0 10  10  248 280 260 293 254  5 25 5 5 5  1 • 260 228  20  215 222 260 241 267 306 254  20 20  5  10 5 5  10 10  5.9 6.4  5.2  5.4 3.9 3.4 3.6 4.9 4.8 5.4 4.9 6.5  . 5.0  5.2  4.8 5.1 6.7 5.7 5.1  5.4  6.6 6.4 6.7  5.0 5.2  6.0 4.5 3.7  4.1 3.8 5.1 4.6  5.0 7.0 5.3  .69 .67 .69 .66 .66 .66 .69 .70 .66 .64 .64 .67 ! .69 .65 .65 .69 .68 .68 .69 .70 .66 .68 .69 ' .69 | .66 .69 .70 ..68 (.69 .66 .68 .69 .70 j .68 .65  92.4  5.74 .672  246.8 33.1  267 254 254 267 246 222  235 235 190  215  246 280 241 22e 241  15 10 20 15 10 5 5 5 5 5 20 5 10 10 15  5.5 5.5 5.7  5.1 5.1  5.2 4.7 5.5 5.5  4.9  5.5 6.1 5.1 5.4 5.9  .69 .66 .67 .69 .69 .70 .71 .68 .67 .68 .68 .69 .66 .66 .68  |  I  15  35  35  35  N= 35 x = 337.4  Tension Shear  9.6  5.21 .676  241.6  10.3  5-38 .681  23.0  T A B L E 3,  A  - 80 -  TABLE 4, APPENDIX A Statistics Mean Percentage of Wood Failure  No. of Speoimens Tested "N"  Mean Breaking Load (Pounds)  Standard Deviation (Pounds)  Coefficient of Variation (Per oent)  Tension Normal t o the Glue Line  97.7  35  337.4  91.8  27.2  Tension Shear  89.4  35  308.1  42.2  13.7  Block Shear  92.4  35  1754.6  356.2  20.3  9.6  35  246.8  33.1  13.4  10.3  15  241.6  23.0  9.6  Glueline-Cleavage EGxlO Speoimens (knife not greased) 0  Glueline-Cleavage EGxlO Specimens (knife greased) 0  --82 - _  F \"G. 2, APPENTMX A s  SPRINGWOOB  S U KIM ETC WOOD  LUEUNE  DIRECTION OF WOOD F I B R E S SQUARES S A W N l'* \" (FINISHED SIZE)  f  FOR  G L U E U H F JANUARY  . ARROWS, INDICATING DIRECTION IN WHICH KNIFE IS TO PENETRATE GLUELINE, TO B E M A R K E D ON E A C H SQUARE BEFORE SAWING C O M M E N C E S  1  CLEAVAGE:  %  E G A10  SPECIMENS  1951  FIG. 3 , APPEND!* A  TENSION  PATTERNS  SHEAR  FOR  TENSION NORMAL AND BLOCK SHEAR NUMBERING  AND  CUTTING  GLUELINE  TEST  CLEAVAGE  SPECIMENS F I G . A, A P P E N D I X A  86  - 87 TABLE 1, APPENDIX B  Press Load No. F i r s t Day (Glue "A ) S  Second Day (Glue «D > H  Third Day (Glue "C ") T  Fourth Day (Glue "B")  Glue Blank Numbers  1 2 3 4 5 6 7 8  6A-1 U-l 11A-2 12A-2 5A-2 13A-2 2A-1 4A-1  1QA-1 1A-2 13A-1R 2A-2 6A-2 11A-1R 3A-2 7A-1  10A-2 9A-2 1U-2R 5A-1 9A-1 13A-2R 8A-1  1 2 3 4 5 6 7 8  2D-1 6D-1 11D-1 1D-2 6D-2 12D-1 13D-2R 1D-1  4D-2 3D-2 11D-1R 10D-2 4D-1 7D-1 11D-2 9D-2  10D-1 3D-1 13D-1R 5D-1 12D-2 8D-2 13D-2  1 2 3 4 5 6 7 8  5D-2 13D-2R 1C-2 12C-2 13C-2 60-2 9C-2 2C-1  12C-1 11C-1 7C-1 8C-2 11C-IE 1C-1 5C-1 2C-2  6C-1 13C-1 10C-2 4C-2 13C-1R 8C-1 9C-1 —  3C-1 11C-2R 3C-2 10C-1 11C-2 70-2 4C-1 —  1 2 3 4 5 6 7 8  3B-2 11B-2R 10B-1 9B-1 5B-1 11B-1 6B-1 7B-2  8B-2 11B-2 13-1 6B-2 10B-2 13B-2R 3B-1 9B-2  1B-2 13B-1R 4B-2 12B-2 4B-1 13B-1 2B-2  12B-1 13B-2 7B-1 2B-1 8B-1 11B-1R 5B-2  mm  mm  -  4A-2 12A-1 13A-1 8A-2 7A-2 11A-1 3A-1 ™  5D-2 2D-2 11D-2R 8D-1 7D-2 9D-1 13D-1 —  -  1  GLUELmE-CLEAVAGE B3SAKIK3 LOADS  Adhnslve  A  D  C  D  iVoattinring Cycle *  2  a  10  11  12  13  14  15  2  a  10  11  15  a  8  10  11  12  13  L4  15  2  8  12  13  14  15  2  8  10  11  12  13  14  15  ' 1 II III IV V 71  235 135 110 120 120 20  210 155 no 110 70 45  260 195 195 130 110 45  345 175 175 105 31C 20  295 165 155 105 l?5 105  270 155 145 155 125 100  235 150 120 100 no 100  230 190 125 150 125 165  540 255 215 240 125 165  610 3» 215 290 230 K'5  440 275 215 165 195 185  400 415 230 185 105 155  260 175 215  240 -"15 190  360 320 305 300 230 2fO  3B5 340 210 295 190 170  455 250 250 305 270 150  425 345 250 205 220 170  540 340 270 230 190 185  470 300 255 230 220 210  370 295 255 275 215 ie5  425 340 270 235 190 175  370 345 260 270 240 230  220 130 155 125 150 105  240 155 105 85 S 0  UO 170 120 105 85 80  260 220 165 130 150 120  240 150 195 125 140 140  210 11D 105 110 100 85  195 70 120 90 70 55  230 175 L20 110 140 140  380 345 215 250 275 170  360 250 220 290 215 65  410 315 285 270 240 185  uo 335 270 270 215 190  390 300 205 205 210 190  335 285 240 240 135 210  315 270 235 230 215 205  315 355 275 220 750 230  I II 111 IV V VI  215 150 120 45 110 10  230 125 130 125 100 65  255 145 85 80  315 \X> 140 U5 125 125  235 140 125 105 85 100  215 1/.0 100 90 85 70  220 165 120 U5 125 125  515 190 160 270 140 100  610 290 210 230 155 120  430 30) 250 255 215 165  450 285 215 210 195 195  390 305 230 215 215 185  UO 345 230 175 215 170  435 340 2B0 270 275 230  400 360 195 235 16s 130  365 365 125 50 150 140  415 365 260 285 280 190  4*tt 300 210 235 190 175  435 315 240 305 215 175  360 280 260 250 210 175  400 295 255 250 150 190  3*5 335 235 270 205 205  215 145 120 125 110 0  175 155 85 20 0  300 185 130 140 140 110  295 175 150 145 125 125  260 UO 130 130 120 110  205 145 110 90 35 30  190 150 105 105 IDG 65  150 155 130 120 130 105  340 235 255 215 240 175  410 305 203 140 200 5  455 425 305 340 255 220  500 415 295 210 235 185  415 240 240 240 310 175  340 305 255 205 190 175  355 285 215 220 215 215  285 320 240 770 255 270  I II III IV V VI  210 30 100 15 35 0  195 50 20 0 0 0  250 90 65 90 3 0  285 20 10 10 0 0  150 30 100 35 5 0  210 45 70 5 0 0  190 130 90 80 13 0  330 175 40 2 8 31  325 250 110 22 8 5  385 335 150 165 145 30  320 215 5 0 0 0  335 335 155 7» uo 10  360 175 120 23 0 0  325 270 175 85 110 10  290 125 20 0 0 0  300 200 25 0 0 0  400 285 150 165 125 10  390 250 100 45 10 0  380 240 185 15 20 0  285 185 150 35 0 0  340 240 140 165 125 0  345 220 205 150 65 20  215 120 90 LOO 110 15  195 140 40 5 0 0  295 155 no 65 18 0  285 45 110 90 , 0 0  275 170 125 105 30 10  170 110 LOO 45 0 0  155 60 90 80 45 20  185 155 120 85 35 40  365 240 185 170 215 15  430 190 UO 0 0 0  440 340 305 215 210 20  430 305 240 140 15 20  510 345 250 205 205 20  340 270 215 65 0 10  325 320 185 140 130 70  335 305 150 120 55 30  I 11 III IV V VI  160 5 12 0 0 0  175 90 0 0 0 0  20 3 0 0 0 0  170 5 0 0 0 0  125  165  15"  305  270  J05  270  295  240  185 45 0 0 0 0  270 185 0 0 0 0  230 120 0 0 0 0  300 2 0 0 0 0  235 42 0 0 0 0  195 7 0 0 0 0  250 80 0 0 0 0  235 155 7 0 0 0  230 175 0 0 0 0  175 25 10 0 0 0  28 1 0 0 0 0  210 2 0 0 0 0  230 18 0 0 0 0  220 10 22 0 0 0  175 33 20 0 0 0  H5 110 12 0 0 0  165 10 6 0 0 0  345 165 5 2 0 0  295 155 8 0 0 0  365 205 60 0 0 0  325 185 10 10 0 0  255 175 110 0 0 0  250 155 IB 0 0 0  250 155 140 25 0 0  275 145 8 0 0 0  8  10  Tro. » tio. 11 12  13  14  15  2  e  10  9 Troe No. U 12  us ICO  Accelerated Weathering Cycle  A  B  175  255  6 10  Tree No. 11 12  2  8  I II III IV V VI  380 280 195 235 215 230  300 360 260 190 305 195  410 300 240 240 235 165  550 385 260 215 195 190  410 280  1 :i 111 IV V VI  280 360 200 260 170 215  330 335 200 190 20 2  585 325 340 280 255 210  1 11 111 VI  365 385 140 88 35 5  345 350 175 210 10 2  1 11 111 IV V VI  235 185 5 0 0 0  330 230 6 0 C 0  nV  230 215 195  1  1  11  12  13  .14  15  2  13  14  15  215 145 165 110 90 103  255 170 175 175 L40 125  320 170 135 175 US 45  305 195 165 U5 U.0 140  145 45 100 90 70 100  175 170 125 150 140 150  230 175 125 150 145 145  385 300 320 200 300 230  350 270 260 200 270 215  365 400 270 255 250 240  365 360 320 295 260 260  410 280 305 260 250 250  410 300 235 280 260 280  320 320 260 240 305 280  355 335 275 240 280 335  305 260 280 250 270 295  380 410 25i 270 370 105  345 295 365 280 270 295  425 355 385 340 275 300  410 220 250 175 230 210  240 220 175 255 185 190  285 240 185 240 215 230  295 360 250 315 355 305  215 150 140 120 110 145  190 150 no 70 90 15  270 205 145 130 130 20  215 185 190 150 UO 130  260 195 145 130 105 120  165 45 120 110 100 70  185 130 125 no 125 125  175 170 110 no 130 110  360 240 260 210 275 200  360 200 210 150 175 145  415 360 315 270 270 250  490 320 285 295 240 150  385 320 275 260 240 280  355 240 240 235 250 275  300 255 250 175 215 205  365 345 285 250 275 305  350 320 230 235 240 215  320 185 170 210 190 150  365 260 255 290 190 215  390 315 240 360 210 220  385 250 275 305 260 250  255 305 205 230 235 190  325 270 230 230 230 130  315 305 240 230 305 280  300 300 175 155 120 45  185 125 105 30 90 125  125 85 80 20 0 0  280 140 120 105 90 10  300 3C 105 120 40 0  275 190 125 120 45 10  170 10 90 55 05 70  190 80 100 80 125 35  165 no 110 100 100 100  350 195 200 190 190 70  325 175 130 185 165 5  520 320 260 280 205 220  320 295 205 155 80 20  390 335 270 190 195 205  300 235 215 185 210 165  190 235 175 140 140 165  280 275 205 195 210 170  345 295 200 210 230 125  280 275 145 175 110 110  435 305 295 295 195 190  335 260 250 365 85 10  435 230 305 255 130 10  295 175 215 165 175 155  325 210 195 145 UO 140  315 285 210 210 150 205  •260 175 7 0 0 0  150 5 85 5 0 0  165 105 29 0 0 0  165 30 3 0 0 0  205 10 3 0 0 0  215 20 45 28 0 0  105 10 40 0 0 a  145 10 45 15 0 0  155 60 90 70 55 20  295 165 100 1 0 0  270 45 13 30 0 0  255 55 110 10 0 0  305 10 90 25 0 0  305 215 U.5 110 5 0  260 145 130 85 0 0  210 145 100 45 0 0  280 170 125 125  345 280 125 5 2 0  240 110 9 0 0 0  345 15 no 70 15 0  315 140 15 0 0 0  275 130 70 15 20 0  255 155 100 55 0 C  260 145 130 120 25 25  230 140 120 UO 60 20  14  15  220  340 255 240 235 230 1?5  305 280 205 220 215 205,  675 365 280 275 165 175  355 295 210 235 255 175  385 250 250 215 175 170  405 340 195 185 55 0  415 380 260 140 60 30  345 365 215 205 120 240  315 20 1 0 0 0  305 30 15 . 0 a 0  285 185 35 25 0 0  iSr  u  10  13  <l -30  10  2  3  280 320 205 215 255 220  195 165 135 110 120 100  335 275 215 210 230 210  345 285 215 220 280 215  435 240 185 105 20 0  295 220 205 170 80 0  295 175 13 0 0 0  275 185 145 5 0 0  no  105  T A B L E 2 , APPENDIX  CD CD  B  ton:. 3 u  Adhesive A  Accelerated 'iJealharin^j Cycle I II III IV V  n i  B  II  i nIV V VI  C  D  I II III IV V VI I II III IV V VI  10 2 586 4.50 350  (.00 UO 373 370 350  330 320 313 357  a uo 400 332  310 300 281  503 450 350 330 313  10  693 359 234 273  353 348  7ia 470  545  JOO  251 462 421  592  585  Tr«>• No. 11 12 429 6U 421 3B9 365 382 349 329 306 282 289  283  522 403 302 396  305  513 465  614 372 324 277 290  13 567 390 310  392 220  15 439 305  232 200 112 337 246 241 311 225 208 382 502 392  285 245 303 438  2J5 298  2V8  208  115 174 248 245 260 335 237 290 559 472 317 358 272 245 304 295 220 326 245 172 345 244 272  564 562 450 355 311 415 217 373 302 304 210 325 396 274 301 3W 200 301 330 313 360 80 302 196 5 280 428 390 472 443 495 5U0 350 350 161 260 231 238 250 76 100 142 0 515 0 0 0 0 29 0 0 0 0 0 0 0 0 0 0 0 0 0  472 400 350  14  292  242  198  523 127 155  383 249 271 0 0 0  125 0 0  2 205 149  205 104  157 125  8 250 182 >6 71  37 0  318  266  114 145  12b  208  24e 124  13-r 74 50  119 161 50  0 0 216 62  0 0 0 0  194 4 0 0  0 0 0 0  30  0 0 0 0 0  .1 1/ , M i ' . -K.vt  11 Tr«« Mo. 10 304 229 211 102 148  107  336 208 193 140 13-! 58  11  12  236 U3 156  240 175 187 147 112 82  152 183  72 a6 243 254 165 133 106  339 - 291 214 219 85 148 25 77 0  0 214 0 0 0  0 0  0 0  217 0 0 0 0 0  309  149 171  205 132 93  219 126  174  128  64  27 226 33 0 0 0 0  12  13 177 100 167  uo 115 83  279 172 125 123 62 65 224 155 39 0 0 0  128 18 0  0 0 0  14 254  105 106 167 83  84  274 178 121 123 104  15 159  192 108 173 155  123 329 145 181 110 158 120  52 133 241 110 122 85 0 0  136  I06  186 8 0 0 0 0  18 0 0 0 0  76 27 18 0  2  8  10  U2 4J6 400 400 400 367  601 500 410 410  '635 U8 390 396  400  500 432  450 460  450 503 450 480 450 400 400  410  483  677  541  516 434 437 431 337  430  537 498 577 467 427  513 435 492 502 512  683  729  534 314 442' 268  385  404 250 146 100  440  553  577  400 377  ICO 0 0  505 498  4-.'-! 362 325  435 340 422 418 356  400  300 196 ICO 0 0  12  411 394 U6 385 404  450 40J 350 286 410  500  13  1Y.>•* No.  11  275 143  157  0 0  227 410  13 481  337 UO 354 379 358 412 311  370 372 248  425 342 482 413 432  543 492  175 236 326 331 478 257 ICO 346 235  494 236 118 3 0 0  533  322 216 45 0  0  369 130 84 40 0 0  14 4t4  15 391  320 335 322 302 353 261 357  308 450 376  404  10  232  274  286  224 214  i6e  180 140  340  268 219 200  367  317 276  435 359  247 335  323 333 285 370 396 218 136 53 38 38  8  369  353  320 J28 414 362 348 4J2 286 247 227 337  2  269 287  215 77 77  259 192  295 228 164 212 78  222 228 156  217  223  180 168 142  240 19S 242  328 34 U  258 178  4 0  184  337 202 183 130 38  166 137  0 0 0 0 100 145 48 0 32 . 0 0 0 0 0 0 0  165 164  152 95 35 74  175 21  0 0 0 0  Tro<1 Uo. 12 11 306' 209  237 227 158  13  14 220 220  172 235  230 147 212 130 215 219  193  221  231 202  176 165  226 259 251  242 241 238 139 151 266  196 207 107  269  215  231  159 136 194 222 • 219 213  0 105 0 90 301 221 47 55 0 0 0 0 0 0 0 0  173 163 168 97  191 175 147 75  113 0 173 0 0 0 0 0  205 213  150 203  196 221  15 234 226 192 215 156 lea  121 190 110  323 199 203 165 155  122 142 142 82 8  209 187 170 147 124 85  168 237  185  51 25 0 0 0  192  219 37 0  0 0 0  cc PER CffiT :«XD FAILURE0? TENDION ZHEAR TE3T  Adhesive A  fi  Accelerated H'SJIUI wing Cycl« I II III IV V VI I II  i nIV V  VI  c  D  I II III IV V VI I II III IV V VI  10 True Ko. 12 100 95 ICO 100 100 100 100 100 ICO 100 100 100 100 100 ICO 100 ICO 100 100 100 85 100 70 ICO ICO 100 85 95 100 100 100 100 ICO 100 ICO 40 103 100 100 95 10 as 100 ICO 100 0 70 0 70 ICO 100 70 0 60 0 5 0 55 0 100 0 ICO 25 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  8 2 100 ' 100 ICO 100 ICO 100 ICO 100 ICO 100 100 100 110 100 100 100 90 100 50 100 10 ICO 10 70 70 100  10 100 1JO 100 100  11  11  U ICO  100 100 1ICO O0 100  ICO  100 100 100 AO 100 100 70  95  80 25  0 100 0 0 0 0 0  15 ICO 100 100 100 ICO 100 100 100 ICO 100 100 100 ICO 100 100 ICO 100 ICO 100 100 100 100 100 100 100 100 95 100 50 100 BO 100 100 100 0 100 100 100 0 0 0 0 0 0 0 0 0 0 14  2  8  100 100 100 90 10 100 0 95 10 85 10 80 85  50  10 60 100  35 ICO 40 30  100  0 100 100 0 0 100 0 0 300 0 5 25 0 0 0 • 90 100 100 0 0 50 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 • 0 0 0 0 0 0 0 0 0 0 0 0 0 0  Tre» No.  11  12  100 UO  100 100  100  ICO 101 ICO ICO  HO 100 90  100 100  20 50 15 100 100  ICO ICO  100 0 0  100 100 0 10  0  I  Design No.  100 100 100  0 0  100  0 0 0 0 0 0  50  0  0' 0 0 0 0  13  14  15  100  80 100 100 80 20 70  100 60 100  95 10 5 0 5 100  10 100 10 100 0 100  100  25  90 ICO 60 100 100  100 20 0 100 0  15 20 0 0 0 0 0  0 0 0 0  0 0  0 0 0 0 0  0 0 0 0 0 0  ICO 100 ICO  0 0 0 0 0  0 0 0 0 0 0  12  2  6  ICO  100 100  100  ICO 100 100  ICO 100 90 90 ao 10  10 100  100 ICO ICO  ICO ICO ICO ICO  100  90 90  50  90 85 60 40 100  BO 100 100 70 70 15 50 0 30 0 0  0 5  60 0 0 0 0 0  95 0 0 0 0 0  85  Tr-e No. 11  12  13  100  ICO  ICO ICO  ICO 100  ICO  95 95  ICO  100 100  100 100  95 75  100 90 85  90 85 90 100  100 80 75  ICO ICO  65  10  100 0 0 0 0 0  0 0 0 0 0 0  0 0 0 0 0 0  100 0 0 0 0 0  eo 95 10  15  60  60  95 100 90  ICO 100  ICO ICO  14 15 lOO 100 100 ICO 100 90 ICO 100 75  ICO ICO 15 • 100  100 95  90 95 75 85  100 95 90 100  ICO  ICO  • 90  0  100  25  5 0  U. IC 5 30 0 0  0 0  95 95 100  25  90 100 100 100 100  PO  30  95 110 100 20  100 0 0 0 0 5  ICO  0 15 0  0 0  0  0 0  2  8  10  100 0 10  100  ICO  1(JU  25  20 10  ICO  50 5 5 5  0 100 50  95 95 10 10 5  100 5  60  0 0 0 0  100 60 c 0  0  95 0  100 0  0 0 0  75 0 0 0 0  0 0 0 0 0  0 0 0 0 0 0  0  0  Tre. e Ho. LI 95  12  13  100  100  eo 90 50 30 100 20 95 ICO 20 75 ICO 50 50 25 75 100 ICO 100 10 15 95 10 80 5 5 70 0  14  ICO 30 80 65 85 to  10  60  15 100  50  95 20 100 30 90 10  20  50 0  100 0  50 30  50 10 95 00  100  100  0  0 0 0 0  0 0 0  100 0 0 0 0 0  100 0  0 0 0  100 0 0 5 0 0  0 0 0 0 0 0  0 0 0 0 0 0  0 0 0  60 0  40 0 0 0  85 35 5  0  0  0  5  0 0  0  0 0  0 0 0  c 0 0 0  60 10  5  0  0 0 0  0 0  T A B L E 2 , APPENTMX.  B  = 2°m 353222 43553 4  ,  -  9 0  nr&i  =. So3S4 222222 S|222S 555522  45555 25334 :45511  - zm3 .3323 -ran  43333 ogss • 5 g:s|| 5 3 S | | | 555333 °55532 3223JI  9  0  *9  => 253522 3.533s ~4n& mm  l  - .3338  24553 o2g235 22522 • §S5|1| - 525355 43535 „ 04332 523532 42551 mm  9  - $m  555535 s ^ | H  * 033355 553555 33333 •  mm  mm  • 24255 323333 222222  322111  022233  2232-2  4  .  23 223:232 332233 33331?  02343 23432  a  223333 33333^  -•oSSSI 4335| 5535|I Kgiii  " 4 225332 2222J 223322 22223 •  » °5SSriS 4 5 3 3 | 45553 333111  * 333223 £52222 43332 2-32||  04323  9  553233  43333  4 .§355S -21535 135551 552111 *  m*& 45511  sssgas  i  0  SSiiH  ' . "33533 45333 433S3.333211  .  . 023323 223333 233333  f  ^ 4332S 333333 333333 333323 * 253533 223533 43333 222331  ^ ogggig * °§3S3i 45535  ^ | | | |  55435 mm mm - "5 s 2 .lass g;g§§ S  Si5i  S  - 55453  gS  3  032233 24333  SSiiii  SasS^l  Sillll  =« -35355  » 335335 222223 023333 333311 ^  323353 02334  3 222233 ^35333 o35333 535111 * o33333 335333  £833  «54233  9 o£55SS 45553 355353 S § | i i i  - =23S32  = 355553 233335  s?s||| . f ~  45555  " *  I I ^ I I 5J|I!i  35314  323532 533135  2322-^  323323 33343  555551 S i l l l !  s 535352 052255 523551 355111  035335 °5SSSS §2SS2| § S l | | g  • 523555 3o5555 54553 535111  9  " ^ 03453  223333  nnu nun  43331 s l i m  *  23222 •  - =aSSS3  o3535S 32S1II s s i i l i  - 033232 43332 34332. 2^111 •ntwtw. pai>j>x»3T  023332  232322 023335  l  o35522 5232  032332 332533 332113  - 43333 o22222 222231  - o25sss 03335s 553323 s a i m - ossaas 45355 533523  2333H 533111  - 233533 o33333 323222 323311  5SH11  " ^ -SS35S 45355 SSSHI  43353 333333 43333 333311  " ^ 45353 353533 43331 353111  » o555S5 555333 535511 521111 P  335231  ^iiii  sss-sgl  " - « s > «  s  -as£>s -*  - S=a»S -  "==S*B "  - 223233  if g  "nsE=-B  z. MX  PP. <  n UJ  -33333 24333  » =55232 332322 333333 33343  45353 -SSS|1 5311H  s 335525 o2S2S3 2§522| § 5 | | | !  f  mm  ^li '421% 4433  9  - 2 £ K  * ^  222252  „ 343-2 .2^5 223S2- 4 3551  » ^ I S a S ^25252 55^11 23SII1 f  mm  4542 S E E  r  " * §52SS§ oSSSSS SgSSSS 555J||  » -S5Sgg  min  2253:51  -354  -3«s-s  -ass--:  535111 -ags-s  %  PERCWTACIE lUDUCTiaiO  Adhaiiva  Accelerated Weathering Cycle  2  6  10  11  12  14  15  2  0 23.2 1.0.3 31.7 30.0 36.3  25.7 32.t t3-9 47.6 t9.3 52.5  3.5 50.0 67.4 62.0 50.e 51.5  23.9 31.0 35.3 38.1 45.7 43.8  0 34.6 40.7 48.9 56.2 56.0  0 . 31.2 t5.3 64.7 tO.6 t5.1  25.0 57.9 55.1 43-0 53.0 57.0  0 30.5 47.2 74.5 45.1 34.4  35.5 53.1 35.5 67.3 50.6 60.7  36.9 tO.3 t3.7 45.4 t6.6 39-1  14.2 2i,.0 tO.9 44.2 47.1 49.3  0 34.5 24.1 65.0 35.6 41.4  7.4 28.5 32.3 29.4 9.0 17.6  t.7 52.6 t2.2 49.7 57.0 55.0  32.6 49.7 56.8 46,6 22.6 40.9  4.0 43.0 76.0 44.2 53-2 54.7  10.7 52.6 60.4 43.5 40.6 33.9  0 58.5 64.2 54.4 76.5 64.3  27.1 52.6 98.5 100.0 100.0  19.t 31.7 to.3 U.5 48.l 38.6  0 2t.O 63.3 6t.5 66.2 99.2  18.5 50.6 48.0  0 tt.e 1.6.4 51.4 41.5 85.8  12.7 35.6 52.8 53.3 51.4 53.1  l.t 36.9 46.4 42.5 39.2 65.4  9.8 46.0 43.6 53.2 53.3 53.7  27.6 U.2 49.9 60.8 38.0 54.9  34.5 62.6 49.4 84.3 100.0 100.0  6.6 53.4 100.0 100.0 100.0 100.0  27.0 tO .3 57.3 100.0 100.0 100.0  34.1 40.9 83.1 100.0 100.0 99.2  34.3 77.6 80.2 100.0 100.0 100.0  21.4 53.9 100.0 100.0 100.0 100.0  23.1 64.1 20.0 95.5 100,0 100.0  11.8 58.0 86.6 100.0 100.0 100.0  12.8 43.3 38.3 100.0 100.0 100.0  32.1 80.5 100.0 100.0 100.0 100.0  100.0 100.0 100.0 100.0 100.0  58.1 44. 61.0  e  13  0 75.7 70.4 76.1 100.0 100.0  31.6 63.9 73.3  40.4 45.0 73.4 61.4 72.1  64.8 65.7 58.7 83.7  12.5 45.6 48.4 63.5 65.6 84.9  51.2 45.1 42.7 58.2 70.0 76.1  11.7 U.3 77.9 93.5 100.0 100.0  34.3 50.6 66.6 82.6 100.0 100.0  U.3 100.0 1CO.0  51.0 100.0 100.0 100.0 100.0 100.0  tco.o 1C0.0 103.0  III DJIEAJ\D!G L0AD3 OF  22.3 43.4 39.5  36.6 35.5 40.1  51.6 U.7 33.6 57.3 69.9 29.1 59.2 43.7 58.6  100.0 100.0 100.0 100.0  TEKJILN  61.7 61.3 39.0 69.7 69.3  67.2 47.4 52.9 62.6  38.4 55.2 55.9 70.6 76.7  55.1 62.0 61.0  55.9 45.0 66.6 52.0 63.5  86.0 100.0 100.0 100.0  51.4 59.8 55.5 69.0 100.0 100.0  54.1 93.5 100.0 100.0 100.0 100.0  39.4 93.4 100.0 100.0 100.0 • 100.0  26.7 56.7 76.9  43.5 97.6 100.0 100.0 100.0 100.0  JlffiAA  12.1 19.3 20.5 20.5 20.5 23.1  TE^T.i  26.8 40.0 40.0 40.0 41.4 29.4 34.1 41.4  4.6 10.5 20.5 20.5 20.5  12.5 20.5 25.0  40.8 63.4 78.6 85.4 19.0 56.1 71.3 85.4 100.0 100.0  U.9 36.5 45.4 45.7 41.0 33.7  17.6 28.6 25-3  26.3 31-7 20.6 35.9 41.4  40.3 53.4 42.1  76.5 100.0 100.0  42.0 IB.3' 46.8  99.4 100.0 100.0  31.2 33.5 36.5 50.1 46.6 58.4  16.5 10.1 30.5 50.0  20.5 35.0 40.7 23.1 43.0 45.4  33.e 45.1 48.6 56.7  11.4 37.9 24.5 34.8  31.0 30.6 23.9  16.6 18.0 19.1 36.6 20.3  24.1 42.7 31.9 31.5 54.3 37.0  3.0 19.0 31.0 10.8 25.0  11.0 27.1 24.5 16.8 36.1  9.6 22.5 18.9 55.7 10.3 35.1  67.8 40.0 52.7 56.7  51.1 27.4 30.4  17.5 43.2 23.0 23.4 34.5  40.0 45.7 61.4 88.7 100.0  38.2 34.0 50.6 82.3 62.3  57.0 65.8 90.5 100.0 100.0 100.0  20.8 32.1 39.0 57.4 23.1  39.6 59.1 91.6 100.0 100.0  32.0 66.6 64.5 92.6 100.0 100.0  20.2 53.0 70.7  22.4 20.3 45.4  31.7 22.5 25.8  22.0 13.3 30.8 15.4 42.3 42.6  22.2 54.6 50.6  86.6 100.0 100.0 100.0  66.6 87.6 74.1  5.9 35.9 32.4 65.0 100.0 100.0  69.5 100.0 100.0 100.0 100.0 100.0  38.8 92.6 100.0 100.0 100.0 100.0  1.6 84.6 100.0 100.0 100.0 100.0  27.6 30.0 40.6 33.4 51.7  14.5 45.4 21.2 51.7 20.1 18.6  13.5 10.1 36.7 15.6  35.7 39.4 37.5 63.9  46.9 19.8 53.6 29.1  38.4 37.2 42.7 52.0 40.6  22.7 11.6 12.7 15.1 58.2 64.1  72.1 58.0 100.0  40.1 40.1 66.4 96.6  35.3 42.1 47.4 54.5 61.6 73.7  12.0 78.1 100.0 100.0 100.0 100.0  35.7 100.0 100.0 100.0 100.0 100.0  100.0 100.0 100.0  100.0 100.0 1CO.0 100.0  19.5 29.9 34.3 31.5  H PERCLKTACE REDUCTION., Hi FEU CENT ,*A.D F A I L U ^ CF THijIDH SHEAR T E J T Design l.'u.  Adh»9iT«  Ac col or* ted thering Cycle  T A B L E 3, A P P E N D I X f;  s  I  B  - 92 TABLE 4A, APPENDIX B Comparisons of the Differences i n Predicted Percentage Reductions ( i n breaking load or % wood f a i l u r e ) Design 11 - Design 11% WF Adhesive  A  B  C  D  Weathering Cyole I II III IV V 71  2  8  35.5 53.1 -54.5  6.0 21.6 63.9 68.3 74.8 80.0  -32.7 -39.4 -29.3  -50.0  10  -40.0  Tree No. 11 12  13  40.4 30.0 73.4 1.4 2.1  46.7 0 64.8 65.7 58.7 83.7  22.3 43.4 39.5 52.4 53.8 73.5  36.6 25.5 40.1 9.8 58.8 70.2  14  15  -12.7 61.7 61.3 39.0 69.7 39.3  51.7 1.6 ' 67.2 47.4 52.9 62.6  I II III IV V VI  -15.0 -41.5 -35.8 -45.6 -23.5 -15.7  27.1 -47.4 -1.5 5.0 0  12.5 45.8 48.4 63.5 -9.4 -15.1  51.2 -34.9 -7.3 -26.8 70.0 76.1  0 51.8 44.7 33.6 57.3 69.9  0 -41.6 -44.8 55.9 -29.4 -8.3  -5.0 -55.0 -39.2 -44.9 -13.0 81.0  -90.0 55.9 -45.0 66.6 -48.0 63.5  I II III IV V VI  24.5 -37.4 -50.6 -15.8 0 0  6.8 -46.6 0 0 0 0  11.7 -5.7 -22.1 -6.5 0 0  34.3 -49.4 -33.4 -17.4 0 0  29.1 59.2 -56.3 -31.4 72.8 -8.7  -60.3 -55.6 -14.0 0 0 e  51.4 -40.2 -44.5 -31.0 0 0  -48.3 -41.3 -23.1 —8.2 -5.5 0  I II III IV V VI  -67.9 -19.5 0 0 0 0  -6.3 0 0 0 0 0  -55.7 0 0 0 0 0  -49.0 0 0 0 0 0  -23.1 -10.7 0 0 0 0  -45.9 -6.5 0 0 0 0  -60.6 -6.6 0 0 0 0  -56.5 -2.4 0 0 0 0  Mean of the differences = *5*5% Standard Deviation of the differences s ±37.36#  -  93  TABLE 4B, APPENDIX B « Comparisons of the Differences i n Predioted Peroentage Reductions i n Breaking Load Design 11 - Design 4 Adhesive  Weathering Cycle  2  8  10  Tree No. 12 11  13  14  15  I II III IV V VI  35.5 12.2 6.0 24.1 18.8 8.4  6.0 -3.8 7.7 8.7 -6.9 0  -63.3 -2.9 -15.0 8.4 -10.3 -1.2  34.8 -25.4 20.7 9.8 9.5 24.4  9.6 -2.0 10.4 -2.1 14.7 24.4  36.6 -12.1 -9.9 2.2 6.4 10.7  7.3 -2.4 22.8 -14.8 5.6 -2.5  51.7 17.7 19.4 -4.8 13.8 23.5  I II III IV V VI  -2.3 24.4 18.8 11.2 26.5 -15.7  -27.1 -8.3 -12.0 6.9 0.8 0  12.5 7.5 -8.3 10.2 12.3 21.6  51.2 4.4 -6.5 7.4 12.4 18.5  -5.4 2.7 -8.0 -19.1 0.9 9.9  -2.4 7.4 7.6 -1.2 11.1 -9.0  -2.6 11.9 9.6 8.9 13.3 14.3  -34.8 23.3 1.5 18.8 8.5 9.2  I II III IV V VI  32.2 17.2 -9.7 29.8 50.0 6.8  -12.0 11.7 16.7 2.1 0 0  10.0 -4.0 14.6 15.2 6.0 0  30.9 -34.1 3.9 13.1 0 0  29.1 21.0 -10.8 -3.2 -16.3 -5.1  0.7 -3.2 33.6 21.4 0 0  30.9 -9.4 1.7 10.0 23.1 10.3  7.1 26.2 29.1 28.8 9.7 17.4  I II III IV V VI  11.7 -8.1 5.6 0 0 0  0.4 0.4 0 0 0 0  14.3 0.7 0 0 0 0  29.0 6.1 0 0 0 0  6.9 -7il 8.0 0 0 0  37.4 9.2 9.5 0 0 0  13.8 49.8 6.2 0 0 0  15.2 2.0 2.6 0 0 0  Mean of the differences s *6.5?S Standard Deviation of the differences =  +15,42%  - 9U TABLE 5A, APPENDIX B Stannary of Slope Eatios Weathering Cycles  1  2  3  I  9.16  11.83  13.29  3.05  4.23  4.17  II  11.41  9.08  9.88  7.61  6.60  III  14.22  11,82  12.68  12.02  IV  11.08  10.47  11.25  V  11.77  9.42  VI  8.23  Totals 65.87 Means  Designs 5  11  13  6.38  3.85  3.37  59.33  4.27  7.85  10.68  10.05  77.43  9.18  9.33  10.54  9.75  11.24  100.78  10.16  12.44  11.01  12.21  8.72  9.42  96.76  12.12  7.85  10.86  9.80  9.10  8.97  11.39  91.28  10.95  11.40  7.38  6.54  8.49  7.31  7.83  9.57  77.70  63.57  70.62  48.07  49.85  47.07  53.39  49.80  55.04  503.28  4  10.978 10.595 11.770  8.012  6  8.308  7  8.898  7.845  8.300  Totals  9.173  TABLE 5B, APPENDIX B  Analysis of Variance of Slope Ratios Souroe of V a r i a t i o n  DP  S.S.  Totals  53  368.8970  Designs  8  101.0362  12.6295  3.772**  2.99  Weathering Treatments  5  133.9482  26.7896  8.002**  2.99  40  133.9126  3.3478  Error  M.S.  Calculated V.R.  Indicates significance a t Pa.01 l e v e l  Tabled V.R. P=.01  TABLE 6, APPENDIX B Single Degree Comparis (Single Degree) Comparisons of Designs  Totals  d  of Slope Ratios  d2  Div. w«  k  MSad /ok  V.R.  2  c  400.12  303.22  96.90  9389.6100  18  6  86.9408  25.969**  1 vs. i i  65.87  49.80  16.07  258.2449  2  6  21.5204  6.428*  3 v s . 11  70.62  49.80  20.82  433.4724  2  6  36.1227  10.789**  1 vs. 13  65.87  55.04  10.83  117.2889  2  6  9.7741  3 vs. 13  70.62  55.04  15.58  242.7364  2  6  20.2280  6.042*  1*2*3 v s . 13  200.06  165.12  34.94  1220.8036  12  6  16.9556  5.065*  1*2*3 v s . 4*5*6  200.06  144.99  55.07  3032.7049  6  6  84.2418  25.163**  4*7 v s . 11*13  101.46  104.84  3.38  11.4244  4  6  0.4760  1*2*3 v s . 4*5*6*7*11*13  7.31  s  *4.08 s  Variance r a t i o f o r  2.920  0.142  significance a t P - .01 l e v e l ,  Variance r a t i o f o r significance at P r .05 l e v e l .  NOTE.- B O A R D S ;  2 THICK, B Y A  WIDE ARE. TO B E  •RIPPED F R O M T H E G R E E N A"*A A S SHOWN IN THE DIAGRAM; KI l_N -fcRI ED TO 5l MOISTURE C O N T E N T ; AND P L A N E D T O M A K E " V E N E E R S * 0 . 2 0 0 " t 0 . 0 0 5 " THICK.  1 NO  CUTTING PLAN FOR BOARDS FLAT-GRAIN, —EBG El-GRAIN, K. EGX10°(! ) E D G E . -  IWAKINC  .THE.  DECEMBER  1952  G R A I N  W I T H  A N (VNtiLE.  OF  THE. CELLS' IO"  W I T H  S U R F A C E . .  FIG. 1 APPEMDIX.  B  9 -  97 -  (NO' 1TQ9) s  GLUELINE-CLEAVAGE DESIGNS Ha4oo"h—  ^VENEERS EDGE-GRAIN ' 5 U T W I T H Tne. C E L L S A T lO° T O H E S U R F A C E . "SPRINGWOOD - S U M M e R . w o o D B A N D S o F" O P P O S I N G V E N E E R S LAID U P ^RALLEL.(LAMIMftTCIi) r  S P R I N G W O O D A N D SUMMI£RWOOC G L U E D AT 9 0 ° A N G L E . ( C R O S S -  EiANDE©).  DESIGN No.4-.  OPPOSIHG AT  THE. S EDGE.CELLS S U R F A  A m c A s N o . 2 . , E>UT U S I N G G R A I N V E N E E R S ( ie. W I T H ALI<3<^£D P ^ K . A L L C L T O T H E C E S ) .  D E S I G N No.7  D E S I G N No.S.  T n e S A M E A S No.A., B O T USIHG FLAT-GRAIN VEMEEKS,  A  10"  V £ N t £ R S  CT*.OSSE.TJ  A n G L E - .  D E S I G N No.£.  D E S I G N No.5.  T H E S A M E . A S No.i., B U T U S I N G E B S £ - G H A , I « VEneeRS ( W I T H THE. C C L L S A L I G N E D " P A R A L L E L To T H E . S U R F A C E S } .  .  EfietHS E D G E - G K A I M B U T W I T H T H EC E L L S A T I O °T O T H E S U R F A C E . SP^INCWOOD -  T  T H E . S A i v \ e A sN o . 5., B U T U S I N ^ F L A T - G R A I N V E M E E R S .  T H E S A M E A S NO.3., B U T USING EDCE A IN V E N E E R S ( ije. W I T M THE WOOD C E L L S A L I G N E D P A R A L L E L To T H E S U R F A C E S ) .  D E S I G N No.9.  T H E SA(*\E A s N o . ,G., B U T USING FLAT - GRAIN. VE.N E E R S .  TENSION-SHEAR DES\GNS tNo' I0T )3^ 5  O  DESIGN  No. 10.  DESIGN  No. 12.  V E N E E R S A M D uv A R L I D E N T I C A L WITH N o . 5 . T w o SETS A R E KE^UIRED T o "PFIOVHSE-G T E S T SPECIMEHS,  V G . N B . E H S A M O L A V U P Ane I D E N T I C A L WITH NO. Q . T W O SE.TS A R E . "REQUIRED To P R O V I U L Q> T E S T SPECIMENS.  ' D E S IGN  No.\\.  EDL-E-G^AIN. V E N E E R S , C R O S S - E I A N I X I D . T W O SETS A R E . R C C ^ U I R £ D TO "PROVIDE. G TEST SPECIMENS.  D E S I G N S OF T E S T  SPECIMENS  FOR  OF T H E I R  COMPARISON  DESIGN  SENSITIVITY*, T H E  VARIABILITY  OF  RESULTS.  THEIR  NOTE/.-  T E S T  F O R £>PEcir<\E.N N U M B E R I N G  CuTTmc  P M T E R H S  No 13.  F L A T - G R M N \/F.MF.ERS, C R O S S - E»ANDEO, OTHERWISE T n u S A M E A s N o . II.  AWD ~  SEE.F I G .  A-.  FIG, g.  APPEMBIX B  -  MARKING S Y S T E M r>ELCErv,E,fLR-  195*2  98  F O R  -  GLUL-BLANKS  PIG 3  APPEMDIX  B  TYPICAL  L O G A R I T H M I C  D E S I G N T E S T E D  lOOl  T R A H S F O K M A H O H S  i  A F T E R  D E S I G M I  T E S T E D  F O R  L I H E A R  C O R R E I A T I O H  7  A F T E R  D E S I & N H I  T E S T E D  li  A N A L Y S E S .  A F T E K  2 1  r-KX)  OZ  r  UJ  U  8  UJ P - 8 0  2 Q  6 0  o o < UJ (^40  u  OZ  D  UJ  U  m -  (Q <  © m o a M fft  .Soups  I ?.  in  o  m  o NURCH1954  F l G. 5 , APPENDIX E>  REDUCTION  IN  BKEAKING  LOAD,  101 -  (PER  CENT)  HUGHES OWENS 11 * B  F I G . 6, A P P E N D I X  B  100  AO  HUGHIS OWENS  ] K i  F I G . 7,  APPENDIX  B  - 104 -  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0106855/manifest

Comment

Related Items