RESISTANCE OF SEVERAL TYPES OF GLUE IN WOOD JOINTS TO FATIGUE STRESSING

Size: px

Start display at page:

Download "RESISTANCE OF SEVERAL TYPES OF GLUE IN WOOD JOINTS TO FATIGUE STRESSING"

Jeremy Gibbs
5 years ago
Views:

1 FREST PRDUCTS LIBRARY FREST RESEARCH LABRATRY REGN STATE UNIVERSITY RESISTANCE F SEVERAL TYPES F GLUE IN WD JINTS T FATIGUE STRESSING Information Reviewed and Reaffirmed June 1955 No UNITED STATES DEPARTMENT F AGRICULTURE FREST SERVICE FREST PRDUCTS LABRATRY Madison 5, Wisconsin In Cooperation with the University of Wisconsin

RESISTANCE F SEVERAL TYPES F GLUE IN WD JINTS T FATIGUE STRESSING- By W. Z. LSN, Technologist, D. W. BENSEND, Technologist, and H. D. BRUCE, Chemist Forest Products Laboratoryr 2 Forest Service U. S. Department of Agriculture MI AN.

2 RESISTANCE F SEVERAL TYPES F GLUE IN WD JINTS T FATIGUE STRESSING- By W. Z. LSN, Technologist, D. W. BENSEND, Technologist, and H. D. BRUCE, Chemist Forest Products Laboratoryr 2 Forest Service U. S. Department of Agriculture MI AN. nn Introduction Glued joints in wood aircraft structures are subjected to repeated stressing. Failures in the glue bonds are usually attributed to poor gluing technique or to deterioration of the glue by exposure, but so far as is known have never been traced to fatigue due to repeated stressing. Nevertheless, it is possible that some failure may be the result of fatigue. It is, therefore, important to know how glues resist the weakening influence of vibration and how different types of glues vary in this respect. Several types of glue have been tested at the Forest Products Laboratory by repeated stressing of the glue joint of a specially designed specimen. The specimens were made with either thick or thin glue lines. In addition, matched specimens of the same type, and also lap joints of plywood, were tested in shear to detect any relation that might exist between resistance to fatigue stressing, and shear strength. 1 This is one of a series of progress reports prepared by the Forest Products Laboratory relating to the use of wood in aircraft. Results here reported are preliminary and may be revised as additional data become available. -Maintained at Madison, Wis., in cooperation with the University of Wisconsin. Report No Agriculture-Madison

Preparation of Test Specimens The eight glues included in this study were of seven types as listed in table 1. They were all commercial glues currently on the market.

Assemblies were constructed by gluing thin plywood (1/1-inch) to thick plywood (1/2-inch), both of high quality yellow birch veneer bonded with phenolic film glue.

3 Preparation of Test Specimens The eight glues included in this study were of seven types as listed in table 1. They were all commercial glues currently on the market. They were mixed and used according to the manufacturers' directions. Gluing details are given in table 1. The type of specimen used in the fatigue tests is shown in figures 1 and 4. Assemblies were constructed by gluing thin plywood (1/1-inch) to thick plywood (1/2-inch), both of high quality yellow birch veneer bonded with phenolic film glue. The thin plywood consisted of 1/32- inch face plies on a 1/24-inch core. The thick plywood was made of eight plies of 1/16-inch veneer with the grain of the two center plies in the same direction to provide balanced construction. The three-ply material was cut into 5- by 12-inch pieces and the eight-ply material into 2-1/2- by 12-inch pieces, with the grain of the face plies in both at right angles to the long dimension. After conditioning to 12 percent moisture content the surfaces to be glued were lightly standed. These pieces of three- and eight-ply plywood were glued together, with the adhesive being investigated, to make a 1/2-inch lap joint. This was accomplished as shown in figure 2 using a piece of the eight-ply plywood, waxed and taped to prevent sticking, as a filler block. After removal from the press, the filler blocks were discarded and the assemblies were conditioned for at least 1 week at 8 F. and 65 percent relative. humidity, after which a groove was sawed to the glue line at a in figure 2 to leave a lap joint of exactly one-half inch. The assembly was then sawed into specimens 1 inch wide, as shown in figure 1, and a 3/8-inch hole was bored through the thin plywood, seven-eighths inch from the end to accomodate a fastening in the fatigue machine. The specimens with thick glue lines were prepared in much the'same way, except that 1/2-inch veneer shims were used, one under the filler block, the other, 1-7/8 inches wide, between the thin and the thick plywood, and extending from the left edge of the specimen as shown in figure 2, to the point where notch a was later cut. This left a trough, 1/2 inch deep, 5/8 inch wide, and 12 inches long, which was filled with glue and covered by the piece of thin plywood to be glued. As the glue cured, the glue-line thickness became less than 1/2 inch. The shrinkage amounted to 12 to 6o percent. Average thicknesses of the glue lines when tested are given in table 2, column 3. Specimens were cut from this assembly in the same manner as those with the thin glue lines. Four assemblies, glued as shown in figure 2, were prepared with each glue, two with thick glue lines and two with thin glue lines, except for the alkaline hot-press phenol-formaldehyde glue, C, with which Report No

only thin glue lines were prepared. From each assembly, eight specimens were obtained of which four were tested in fatigue and four in shear, alternately selected.

For a further comparison of shear and fatigue-test results, specimens were prepared for shearing by gluing two pieces of three-ply 1/1-inch birch plywood together and cutting the assembly into

4 only thin glue lines were prepared. From each assembly, eight specimens were obtained of which four were tested in fatigue and four in shear, alternately selected. The shear test specimens were cut to a length of 3-1/4 inches and tested in the standard plywood shear-testing machine. For a further comparison of shear and fatigue-test results, specimens were prepared for shearing by gluing two pieces of three-ply 1/1-inch birch plywood together and cutting the assembly into lap-joint specimens with, 1/2-square-inch shear area. Test Procedures Fatigue-testing machines of the constant-deflection type (fig. 3) operating at 1,79 revolutions per minute were used. The machines were provided with automatic counters and a precision switch which shut the current from the motor whenever a specimen broke. The specimen was so adjusted in the machine that the motion was the repeated-stress cycle illustrated by figure 4A, that is, the deflection was always to the tension side of the neutral point of zero stress on the glue line and never to the side at which the glue line would have been under compression. Previously, exploratory trials had been made of the reversing-stress cycle, represented by figure 4B, with which it was found that the threeply birch plywood invariably broke through the face plies at the point corresponding to a in figure 4A before there were any signs of failure in the glue line. Hence, for a test of the glue joint rather than of the plywood, the repeated-stress cycle was preferable. As a first step, exploratory experiments were made to find a load that would result in failure within a reasonable length of time. An average static load of 6.7 pounds, applied 3 inches from the point a in figure 4A so as to put the glue in tension, caused failure when fatigue specimens glued with urea-formaldehyde glue were tested as cantilever beams, but a 4.-pound initial load induced failure at about 3, cycles of repeated stress. Because the thin plywood varied in stiffness, the deflections resulting from a particular load varied somewhat between specimens,but the initial deflection produced by a load of 4. pounds was arbitrarily chosen for all the fatigue tests subsequently made in this study. Whenever a specimen withstood 1,, cycles without breaking, it was removed from the machine. Report No

Glues B and G withstood means of 36,9 to 397,3 cycles before failure. al specimens failed. 3. Glue H withstood a mean of 31, cycles before failure. All specimens failed.

5 Results The results of the fatigue tests are given in table 2. The data on the thin glue lines fell into three groups as follows: 1. Glues A, C, D, E, and F withstood means of 4,431,1+ to 6,627,5+ cycles before failure, with the means calculated on the assumption that unfailed specimens broke at 1,, cycles. 2. Glues B and G withstood means of 36,9 to 397,3 cycles before failure. al specimens failed. 3. Glue H withstood a mean of 31, cycles before failure. All specimens failed. The data of the fatigue tests on the thick glue lines fell into five groups as follows: 1. Glues E and D were not broken by 1,, cycles. 2. Glues A and F withstood means of 2,273,5+ to 3,152,4+ cycles before failure, with the means calculated on the assumption that unfailed specimens broke at 1,, cycles. 3. Glue B withstood a mean of 628,4 cycles before failure. All specimens failed. 4. Glue H withstood a mean of 75,1 cycles before failure. All specimens failed. 5. Glue G withstood a mean of 2,2 cycles before failure. All specimens failed. The differences between the groups is highly significant, but there is no statistical significance between the means for the glues within the groups. These results indicate considerable differences between types of adhesives in their ability to withstand repeated stressing. The resorcinol, melamine, and alkaline phenol resins were superior to the urea resins or to the casein. The failure of the urea- and casein-glued joints in the fatigue tests was largely in the glue, whereas in most others, the failure was largely in the wood. The wood adjacent to the acid phenol glue, B, was clearly Report No

6 injured by the acid of the glue and failed with shallow wood failure. There was no indication that the wood was injured by the less concentrated acid of the urea and melamine glues, F and G, or by the strong alkali of glues C and H. With both resorcinol glues, the thick glue lines withstood repeated stressing better than did the thin glue lines. This appeared to be true also for the casein glue but the difference was not important. Likewise, the difference may not be significant for the acid phenol glue. The thin glue lines with urea glue were definitely superior to the thick glue lines. Table 3 gives the results of shear tests on both fatigue-type specimens and the birch plywood lap-joint specimens. In a comparison of the results of the fatigue and shear tests, disregarding the data of the acid-damaged specimens of glue B, there is some indication that the lower the wood failure in the sheared specimens, the lower the number of repeated stress cycles to cause failure. The evidence, however, is not definite but there appears to be no relation between the shear strength values and the number of repeated stress cycles to cause failure. The conclusion is that a shear test of a glued joint is not a satisfactory criterion as to the behavior of the joint in fatigue stressing. Report No

$71181 co 1: al CD C rl c 114. c.).9 P4 r I I vn in if\?) od N-- tc-n ca to 4-, (1) a.) III dr, c%),c) ca c.) 4-, o u-n H C 111 H H H CU I rg Ci-1 Pl cf o c Co H H N- %.$ ) I I P.+ I H -P o o - n-4 2 -,-1 a:' -P En (1) rn -.-1 -P 4-) I C -P co as CLT CD ca ki 4) I -9 4-1 '.P4 It.14) o cor co pi pi..-i Pi 4-) ri) H 1 a Ts.d-i 1E1 -i a.

) I I P.+ I H -P o o - n-4 2 -,-1 a:' -P En (1) rn -.-1 -P 4-) I C -P co as CLT CD ca ki 4) I -9 4-1 '.P4 It.14) o cor co pi pi..-i Pi 4-) ri) H 1 a Ts.d-i 1E1 -i a.

7 71181 co 1: al CD C rl c 114. c.).9 P4 r I I vn in if\?) od N-- tc-n ca to 4-, (1) a.) III dr, c%),c) ca c.) 4-, o u-n H C 111 H H H CU I rg Ci-1 Pl cf o c Co H H N- %. ran b.. I I 1-1 r-i I g 4, q q -,-1 tll -1-) g,--i b o U gf, FA 4-,I t-i I al o Pi C +' U Pi 1.13 o..4 l-i ca [cif s-i 6 bp g p., ,-1 a.).,-1 Ei 4-, 4-, 4-, -P 4-, ci) a> 4, -P al 4-) cd 4-) P -H P-1 1 to ) =- ) C.) I I P.+ I H -P o o - n-4 2 -,-1 a:' -P En (1) rn P 4-) I C -P co as CLT CD ca ki 4) I '.P4 It.14) o cor co pi pi..-i Pi 4-) ri) H 1 a Ts.d-i 1E1 -i a.) g rq,-1 -P,-1 4-,.1-I H 43 I M I al co rd. m U,i Pi 3q $.4 F-1 +3 q o al.4.r:a I:4 14 I-I M Report No. 1539

$H ta.34 4-) 4 4-3 -4-) 43 ca s.-1 4.) nci tll ri r 4-1 $4 :1 S +, 4 C 1-1 d g o i Fi rp 1-1 i b -.-A k M co CV N nc N r r $4 LC\ H c co cc), \ MMPr \ I> af < 41 P4 rta 4-1 9 1 *I IA 1-4 (1). 4) U ) L.$

8 H ta.34 4-) ) 43 ca s.-1 4.) nci tll ri r 4-1 $4 :1 S +, 4 C 1-1 d g o i Fi rp 1-1 i b -.-A k M co CV N nc N r r $4 LC\ H c co cc), \ MMPr \ I> af < 41 P4 rta *I IA 1-4 (1). 4) U ) L. r-1 H. rd) H. V fel 4-3 r>i ra.. e 141 npen1 N hc )n M KN rc\ Nal te\q MMK\KN II tcnrc\oo n.ont<nnnoitc-n Hnrcn-tmmonoo_I-o-NaloH s es es es es es es es es es es es es es es Hnts-CHC-WN\DHlr\ 88 N-K\N-1-N-,41HC.-- ut.5 n *5 *5 *5.5.5 'HnN\HMMlW\LC\ tr\ M,,c-o-Nwainoafo-Invoitc\t-...l'altr\\HHtc\ es es es es es ft. ws nono.npn H m nhmhocoo-to\orrn Hcont NH H H 1 e t f-i I g tel H ) I r t re;p P1 4.2 HC) HH14-V-IM H1-NC---t`oorc\Nt-nreNn \c w.5 N r-1 H_H 4c KNIKAWNIKVANN-WN MK\MC---NC-K\H\DNN HHHHHHHHr4HHHrAHH \ID1/4\\CAWNIA4C\W\n-1.\ HHHHHHHHHHHH HH H c4-1.9 o to n A.44P1MoAAWFAr=4C=41: C1) fai H k tf) rt-a g r-i to H i. d ri-i. d d H I cn 4-I n;-. 1:1 ri H a pc H 3 H co -1 rg o d H C H ' I-1. I -P :-> o +3 P.1 4-, H I g g,-1 T s o -.-1 H 11.1PLIPH,-1 P dh d H 1,--4 CH s4 -I-, N-I ei-i H o -P cd H CA C.) Ill, A ati r 4-) 1::1,-I 4-1 g tir n--1 lr'n 1E1 ',I.d k 1 ai H 4-1 P.4 cil HI afil N-Nr Report No

: C : thin : 751 : 95 639 : 96 D : thin.94 68 : 88 : thick 612 : 94 E : thin ; 758 : 87 685 : 92 thick 538.

9 Table 3:--Averaged results of shear teats on glued srciméns Glue : Glue-line : Fatigue-type specimens : Lap joints of three-ply symbol : thickness : : 3/32-inch birch plywood : Shear strength : Wood failure : : :Shear strength:wood failure. P.s.i. Percent : P.s.i. : Percent : A : thin : : 716 : 92 : thick : 421 : 36 : : B : thin : : 59 : 1 thick : : C : thin : 751 : : 96 D : thin : 88 : thick 612 : 94 E : thin ; 758 : : 92 thick : : : F : thin 647 : : 79 : thick 413 : 6o G : thin 672 : : thick 34 : 1 H : thin 689 : : thick 424 : 3 i f-port No. 1539

10 Figure 1.--Fatigue test specimens with thin and thick glue lines. Z M F :4

11 t

) r1 -P P- n.! tij 3 cn N.1-1 - P- C.). C6 o E -)J 6).

12 C.) 6) C.) n-i U U ri ch E. zri b U1 CD 1::) F.-. 6) r - E r-i C.) r1 -P P- n.! tij 3 cn N P- C.). C6 o E -)J 6).Z CI) 6) d bk) 47 -P -P I R /.C1 C7.4

13 W 1 C.) 1 1/2 ;14 (1) cl) S.. -o) ti X

EFFECT Of THICKNESS Of GLUE LINE ON STRENGTH AND DURABILITY. Of GLUED WOOD JOINTS. Information Reviewed and Reaffirmed. June No.

EFFECT Of THICKNESS Of GLUE LINE ON STRENGTH AND DURABILITY Of GLUED WOOD JOINTS Information Reviewed and Reaffirmed June 1960 No. 1616 FOREST PRODUCTS LABORATORY UNITED STATES DEPARTMENT OF AGRICULTURE