END JOINTS IN LUMBER BY THREE TEST METHODS

U. S. FOREST SERVCE RESEARCH PAPER FPL 41 OCTOBER EVALUATON OF COMMERCALLY MADE END JONTS N LUMBER BY THREE TEST METHODS U. S. DEPARTMENT OF AGRCULTURE FOREST SERVCE FOREST PRODUCTS LABORATORY MADSON, WS.

SUMMARY Three different test methods were used to evaluate end-jointed material that was fabricated using commercial end-jointing techniques. Results. show that for end-jointed lumber the tensile strength as determined with a 3/32-inch-thickspecimen having uniform cross section throughout its length is approximately equal to the tensile strength as determined with a necked-down tension specimen. For scarf-jointed lumber, the modulus of rupture and the tensile strength are approximately equal, and for fingerjointed lumber the comparison of modulus of rupture and tensile strength apparently varies with joint configuration.

EVALUATON OF COMMERCALLY MADE END JONTS N LUMBER BY THREE TEST METHODS 1 By BLLY BOHANNAN, Engineer and M. L. SELBO, Chemical Engineer FOREST PRODUCTS LABORATORY 2 FOREST SERVCE U. S. DEPARTMENT OF AGRCULTURE NTRODUCTON An evaluation ofcommercially fabricated, endjointed laminations of various softwoods was made at the U.S. Forest Products Laboratory using three test methods to compare the tensile strength values as determined by each method and to investigate the correlation between tensile andbendingstrength, A tension test employing a necked-downspecimen of proper design has generally beenbelieved to provide the most critical evaluation of an end joint and is required for American nstitute of Timber Construction (ATC) plant qualification. Such a test has limitations, however, because of the difficulty and cost of preparing specimens and the need for adequate test facilities. Therefore, for day-to-day quality control tests, as required of laminating plants certified according to U.S. Commercial Standard CS 253-631 3 the necked-down tension test is often impractical. A strip-tension test using a thin rectangular specimen of uniform cross section throughout its length has been used in researchanddevelopment work on end joints at the Forest Products Laboratory and is now included in the ATC nspection Manual as an alternate test for day-today quality control. A bending test is also allowed by ATC as an alternate test on daily production. Both the strip-tension and bending tests are attractive for laminating plantsbecause specimens are easy to fabricate. The bending test is particularly attractive because it can be performed with relatively simple andinexpensive apparatus. The strength requirements of end-jointedlaminations as given in theu.s. CommercialStandard CS 253-63 are the same for each of the three types of tests, but it was believed that the strength might vary with the test method, type of 1 This research was done in cooperation with the American nstitute of Timber Construction, K Street, N.W., Washington 6, D.C. 2 Maintained at Madison, Wis., in cooperation with the University of Wisconsin. 3 U.S. Dept. Commerce. Structural glued laminated timber Comrnercial Standard CS 253-63. Apri, 1963.

joint, and possibly with species. A study of the effects of these variables on strengthwas needed. One study on end-jointed specimens prepared under laboratory control indicated thatthetensile strength determined from necked-downandstriptension specimens was about equal. 4 Results of a study conducted in the United Kingdom showed that the bending strength of single laminations containing end joints (both scarf and finger) was higher than the tensile strength of end-jointed, necked-downspecimens. 5 The purpose of the study, therefore, was to evaluate commercially made end joints in four different softwoods by the necked-downtension, strip-tension, and bending test methods and to compare strength values as determined by the different test methods, Evaluations of strength were made on scarf joints with a slope of in nominal 1-inch southern pine, Douglas-fir, West Coast hemlock, and western larch; on a finger joint in nominal 2-inchsouthern pine; and on a different type finger joint in nominal 2-inch Douglas-fir. This research was conducted at thelaboratory between April 1963 and January MATERAL AND FABRCATON Evaluation of end-jointed laminations was made on material furnished by members of the ATC. nstructions to the laminators regarding materials and method of fabricating end joints for this study included the following: Materials used will be coast-type Douglas-fir, southern pine, larch, and West Coast hemlock. Material shall be clear and straight grained, having a slope of 1 in 20 or flatter. Material shall be representative of the species and have a range of specific gravity, but the average specific gravity shall be about equal to the species average. Production equipment shall be used for cutting and gluing joints. ndustry will select and furnish all materials, end joint the laminations, and send the glued endjointed laminations to the Laboratory. The material requested was nominal 1- or 2-inch-thick(1 inch for scarf joints and2 inches for finger joints), 8-inch-wide, 8-foot-long laminations fabricated as follows: End joints shall be prepared at the opposite end of each 8-foot member, The member shall then be cut exactly 2 feet from one end and the end joint glued from the same pieces of each board. t is essential that this procedure be followed. After end gluing, the usual plant practice associated with the production of the end joint shall be followed, f the end-jointed piece is resurfaced in production before laminating, it shall be resurfaced by the laminator as is done in production; if not resurfaced in production, it shall not be resurfaced for these tests. Endjointed boards shall then be carefully packaged and shipped to the Laboratory. Material was furnished by commercial laminating plants and coded as fallows: Material designation B E D F A G H C J Species Surfaced after gluing PLAN SCARF JONTS Southern pine Southern pine Dougas-fir Douglas-fir Hemlock Hemlock Larch Larch Southern pine Douglas-fir Yes Yes Yes No No No No No FNGER JONTS Yes No Type of adhesive Phenol-resorcinol Casein Phenol-resorcinol Melamine-urea Melamine-urea Meamine-urea Phenol-resorcinol Casein Pheno-resorcinol Meamine-urea Twenty laminations were furnished of each of the scarf-jointed materials. The joints were made from nominal 1-inch lumber except for material D, which were fabricated from nominal 2-inchlumber and then dressed toafull1-inch thickness by the laminator. They were further dressed to a 25/32-inchthickness after they were received at the Laboratory, All scarf joints had an average slope of 1 in 10 except for joints 4 Sebo, M.L. Test for quality of glue bonds in end-jointed lumber. Symposium on Timber, ASTM Special Tech. Pub. No. 353. 1962. 5 Suney, J.G., and Dawe, R. S. Strength of finger joints. Wood, Vol. 28:9. 1963. FPL 41 2

of material E which had a slope of 1 in 11. Forty nominal 2- by 8-inch laminations were Materials D and had wood dowels and material furnished for each finger joint. The joints in E had a metal pin extending through the thickness material C were cut perpendicular and in mateof the lamination at the approximate midpoint of rial J parallel to the width of the boards. Details the joint. of each type finger joint are shown in figure 1. Figure.--Details of each finger Joint. Material C was cut perpendicular and material J parallel to the wide face of the board. SELECTON AND PREPARATON OF SPECMENS Necked-down, strip-tetnsion, and bending specimens were cut side by side across the width of the board, side matched. Since there were three different types of test specimens taken from each lamination, there were three combhinations which could be used in cutting the specimens. Starting from one edge of the laminations, the sequences by which specimens could be cut were strip tension, necked-down, bending; neckeddown, bending, strip tension: and bending, strip tension, necked-down. Each cutting sequence was followed for one-third of the laminations. Prior P#or to sorting and numbering of the laminations, the cutting sequence was randomly selected for each lamination to be tested. Fifteen scarf-jointed and 30 finger-jointed laminations of each material were selected for strength evaluation. During the selection, approximately two-thirds of the desired number passed the requirements set forth in the original study plan. The remaining one-third had to be slelected from less desirable material containing small localized strength-reducing characteristics such as cross grain or pitch pockets, These two-thirds 3

and one-third divisions seemed to be rather consistent for each of the 10 materials. After the desired number of laminations had been selected, they were randomly mixed so that the less acceptable laminations were intermixed with the acceptable ones. After the laminations had been numbered, they were then marked for cutting according to the previously discussed randomizing system. One end-jointed and one control specimen for the necked-down tension test and for the bending test and eight end-jointed and eight control specimens for the strip-tension test were prepared from each lamination. The midlength of the endjointed specimens was at the midlength of the end joints and the midlength of the control specimens was at a common unjointed cross section. The control specimens were end matched to the end-jointed specimens and were cut as close to the end-jointed specimen as possible. A 1-inchlong cross section was cut from between the ends of the end-jointed and control speciments to obtain a value of the specific gravity of each lamination. The thickness of rthe specimens was the thickness of the laminations from which specimens were prepared, except as indicated. Laminations that were not surfaced after gluing were not surfaced when specimens were prepared. The width and thickness of all specimens were measured to the nearest 0.0001 inch prior to testing, and the overall size of the specimens was as follows: Necked-Down Tension Specimens The necked-down tension specimens (fig. 2) for Figure 2.--A necked-down tension specimen ready for evluation. Specimen was anchored with wedge grips in the rigid heads of the test machine. M 125 146 FPL 41 4

the scarf joints were 2-1/2 inches wide and could not be tested without excessive crushing at 30 inches long, necked down on a 17-1/2-inch the grips when the specimens had a thickness radius to a width of 1/4 inch for a 10-inch length. equal to the thickness of the nominal 2-inch This gave a straight section along the full scarf laminations: therefore, these specimens were joint and for about 1 inch outside each scarf tip. ripped in half on a bandsaw, to produce two Specimens with finger joints from material J specimens approximately 3/4 inch thick. The were 2-1/2 inches wide and 26 inches long, sawn faces were surfaced. necked down on a 17-1/2-inch radius to a 1/4- Hardwood pads were not glued to the end inch width for a 6-inch length. Specimens with sections of the necked-down tension specimens. finger joints from material C were 2-1/2 inches wide and 26 inches long, necked down on a Strip-Tension Specimens 17-1/2-inch radius to an approximate width of 0.59 inch (one full pitch) for a 6-inch length. For The strip-tension specimens were cut on a the finger-jointed specimens the straight section table saw using a hollow-ground blade (fig. 3). extended across the joint area and for about All specimens were about 3/32 inch thick, Those 2 inches outside of each finger tip. cut from scarf-jointed 1-inch boards and from The control specimens for materials C and J 2-inch finger-jointed boards (from material J) and the end-jointed specimens from material C were ripped from the edges of the boards with- Figure 3.--Cutting strip-tension specimen with a hollow-ground saw. M 121 878 5

out any machining of the board surfaces; hence, Bending Specimens the width of the strip-tension specimens was equal to the thickness of the boards as received-- Bending specimens from scarf-jointed lamina- about 3/4 inch for the scarf joints and about tions were 2-1/2 inches wide and 30 inches long, 1-5/8 inches for the finger joints. The length except as indicated. As previously discussed, of all strip-tension specimens from scarf-jointed some of the laminations contained alinement laminations was 18 inches and from finger- dowels and by following the randomized sequences jointed laminations of material J, 11 inches. of cutting specimens it was impossible to elimi- Specimens fro m the finger- jointed, 2-inch nate these dowels from one-third of the bending southern pine laminations (material C) were cut specimens prepared from such laminations. Also, parallel to the face of the board and of a width one-third of the bending specimens from these equal to twice the pitch. On each edge of the speci- laminations with dowels were 2-1/4 inches wide. mens the outer finger feathered out at the mid- This change in specimen size was necessary to point (or midlength) of the joint (fig. 4). These provide sufficient material for strip-tension specimens were 12 inches in length. specimens. n summary, all scarf-jointed specimens were The finger-jointed specimens from material J about 3/32 by 3/4 by 18 inches; finger-jointed were 2-1/2 inches wide and 36 inches long and specimens from material J were 3/32 by 1-5/8 by those from material C were approximately 2.4 11 inches; and finger-jointed specimens from inches wide (four times the pitch) and 36 Figure 4.--Strip-tension specimen in wedge grips ready for testing. M 125 816 FPL 41 6

Conditioning of Specimens Upon receipt of the laminations at the Laboratory, they were marked with the assigned code designation and stored in an atmosphere maintained at 74 F. and 65 percent relative humidity until all materials were received--a period of about 3 months. After specimens were prepared, they were stickered and stored in the same controlled atmosphere for approximately 6 weeks, During this time a selected number of bending specimens were periodically weighed until nearly constant weights were reached to insure approximately 12 percent moisture content of the specimens at the time of test. No check of exact moisture content was made, TEST PROCEDURES Since the object of the study was to compare Necked-DownTension Test test methods, it was believed that the rate of stressing should be the same for each method. The necked-down t ens i on specimens were For the two tension tests the rate of loading was loaded with wedge grips in the rigid heads of the adjusted to produce a tensile stress of about test machine (fig. 4). The specimens weregripped 4,000 pounds per square inch per minute and for at each end over their full 2-1/2-inchwidth and the bending test to produce an outer fiber stress over an approximate 4-inchlength. A 200-pound of about 4,000 pounds per square inch per load was applied to the specimens to set the minutethroughout the elastic range. grips; then the specimens were loaded at a rate Figure 5.--Experimental apparatus with bending specimen ready for evaluation. Plywood with grain-of-face plies running parallel to was used under loading heads to prevent local crushing. M 125 147 FPL 41 7

of approximately 4,000 pounds per square inch per minute, Strip-Tension Test The strip-tension specimens were loaded with wedge grips 3 inches in length and of sufficient width to cover the full width of the specimens (fig. 3). The grips were anchored at top and bottom by ball and socket joints. Similar to the necked-down Specimens, the rate of loading was approximately 4,000 pounds per square inch per minute, The distance from the ends of the grips to the tips of the joints was from 1-1/2 to 2 inches for all specimens. Bending Test All bending specimens were evaluated under two-point loading (fig. 5). Specimens from the scarf-jointed laminations (nominal 1-inch material) were loaded over a 24-inch span with 12 inches between loading heads. The rate of vertical movement of the loading head was 0.39 inch per minute, to comply with an outer fiber-stress rate of about 4,000 pounds per square inch per minute. Specimens from the finger-jointed laminations (nominal 2-inch material) were loaded over a 32-inch span with 8 inches loading heads. The rate of vertical movement of the loading heads was 0.30 inch per minute, to comply with an outer fiber-stress rate of about 4,000 pounds per square inch per minute. With these distances between loading heads, the tips of the joints were at least 2 inches inside the heads. The loading apparatus (fig. 5) was chosen as one that might be practical in a laminating plant and still be similar to the apparatus recommended by the American Society for Testing and Materials Standard D 198-27. Roller nests were used at the supports, and the 1-5/8 -inch-diameter loading heads were fitted with Poller bearing ends to adjust for horizontal movement that might occur during test. One-eighth-inch-thickplywood with grain of face plies running parallel to the specimen was used under the loading heads to prevent local crushing. RESULTS AND DSCUSSON The results for each of the three test methods are given in table 1 and comparisons of results obtained by the three test methods are shown in figures 6 through 19. Reference lines having a slope of 1:1 are shown on each of the figures. Values given for the strip-tension specimens are the average of eight specimens from each lamination. Values for the necked-down tension specimens for control and jointed specimens of material C and for control specimens of material J are the average of two specimens produced by ripping a nominal 2-inch-thick specimen. The percentage of wood failure was estimated by visual observation of the bending specimens. Although such estimates may vary with individuals, they are, nevertheless, generally a reasonably good guide to the quality of glue joints. Failures of part of the bending specimens are shown in figures 20 to 25. Figure 26 shows some serious misalinement of joints during manufacture, which was noticeable, however, in only one group of material. n a study of this kind, which included over 3,000 specimens, it is difficult to avoid having some specimens prepared and evaluated which contain local irregularities that influenced their strength values. Also, a few of the bending specimens had dowels through their approximate center, although the dowels did not seem to greatly influence the strength of the joint. Furthermore, it seemed impractical to examine all the failed specimens in detail in an attempt to cull those whose strength was affected by the local irregularities. Such might have been desirable if exact strength values of joints in clear straight-grained boards had been the object of the study, but since a comparison of test methods was the essential object, it was believed that strength values for all specimens should be included. The specific gravity of a few of the laminations was either above or below the range that would be considered normal for the species in this study. Again, since test methods were being evaluated, the values for these few laminations were not culled. The relationships between necked-down and strip-tension strength values and between necked- FPL 41 8

Table.--Results of tension and bending tests on control and end-jointed specimens Piece No. Specific Tensile gravity 1 2 Strip Control Joint strength Modulus of rupture Necked-down Control Joint Control Joint Slope of scarf Percent wood failure MATERAL B--SOUTHERN PNE 2 3 4 5 0.50.57.57.56.46 14,900 15,800 19,900 21,200 8,300 11,100 12,900 11,500 11,700 5,800 13,300 11,200 19,000 18,000 16,200 10,300 11,000 11,300 11,200 9,530 13,000 15,800 14,200 15,100 8,640 11,900 15,400 13,500 11,900 8,150 3 100 75 3 60 100 3 100 6 7 8 9 10.53.55.48.49.57 14,400 16,800 10,200 12,700 19,000 12,600 14,600 8,720 10,200 16,000 18,000 19,900 14,900 18,800 21,700 8,920 11,900 13,000 11,000 14,400 11,600 11,200 19,100 13,800 14,900 11,900 13,000 11,000 12,700 14,500 100 70 75 100 11 12 13 14 15.49.44.52.50.55 15,400 12,100 1,500 18,000 21,800 10,700 11,100 8,370 11,000 17,900 20,800 13,600 22,300 20,000 14,300 9,260 12,900 11,500 11,900 12,900 10,800 12,500 12,500 15,000 10,000 8,800 11,100 12,000 12,800 75 75 60 Av....52 15,500 11,600 17,500 11,700 13,100 11,900 83 MATERRAL E--SOUTHERN PNE 2 3 4 5.53.59.53.51.51 13,500 22,200 22,800 19,100 18,000 15,600 13,300 16,000 16,600 13,300 8,870 20,100 23,600 20,300 19,600 18,000 13,900 14,400 14,000 14,800 13,300 13,700 13,500 12,100 13,400 1,600 14,500 10,900 13,200 13,000 1:11 1:11 1:11 1:11 80 65 3 100 3 100 65 6 7 8 9 10.54.47.50.46.61 17,100 13,100 15,600 12,100 17,000 11,600 12,600 13,700 10,100 10,900 20,600 10,300 15,200 16,200 13,700 12,800 10,700 12,800 14,100 14,900 11,800 11,600 11,000 12,700 14,000 11,000 11,400 11,600 11,400 1:11 1:11 1:11 1:11 1:11 75 50 80 40 11 13 14 15.52.55.44.54.60 21,100 17,500 15,800 15,600 15,400 11,700 11,400 11,800 8,920 12,600 16,200 17,200 11,100 13,900 9,600 12,700 9,520 10,900 11,900 11,900 12,800 12,900 9,940 10,000 17,100 10,200 11,100 9,110 10,600 13,400 1:11 1:11 1:11 1:11 1:11 75 3 65 100 20 Av....53 17,000 12,700 15,800 13,200 12,700 11,600 1:11 72 9

Table.--Results of tension and bending tests on control and end-jointed specimens--continued Piece No. Specific gravity 1 Strip 2 Tensile strength Necked-down Modulus of rupture Slope of scarf Percent wood failure Control Joint Control Joint Control Joint MATERAL D--DOUGLAS-FR 3 1 0.47 12,500 13,000 12,500 13,400 11,500 10,500 100 2.52 15,800 13,700 20,100 11,800 16,400 12,600 95 3.55 17,000 13,700 14,000 12,900 9,300 13,500 3 60 4.46 12,000 11,600 14,000 9,540 12,600 9,720 100 5.54 20,200 15,500 27,300 22,300 15,000 16,600 90 6.50 16,400 16,100 17,400 17,400 15,200 13,400 7.46 12,300 13,200 16,500 13,300 12,500 13,600 25 8.40 12,700 12,500 12,700 13,000 11,400 9,520 3100 9.52 16,100 17,300 22,200 16,100 14,600 15,700 40 10.48 11,800 9,920 16,100 9,720 11,800 12,200 3 11.38 12,000 10,500 12,600 10,100 10,800 7,340 100 12.43 3 12,700 11,900 13,800 11,900 12,000 10,500 100 13.42 10,800 14,200 12,200 14,400 11,900 11,900 95 14.53 18,100 16,400 15,500 16,700 16,300 16,600 50 15.56 9,700 11,600 16,200 13,400 12,600 11,600 90 Av....48 14,000 13,400 16,200 13,700 12,900 12,400 81 MATERAL F--DOUGLAS-FR 1.44 12,400 9,460 12,100 13,000 11,100 9,520 90 2.48 13,900 4,960 15,400 11,800 12,800 9,870 95 3 3.47 12,600 8,510 9,120 7,800 10,400 8,500 100 4.46 13,200 9,760 17,400 12,600 14,400 11,200 95 5.44 12,600 11,100 14,100 10,200 13,100 9,090 100 6.39 9,990 8,000 9,910 9,0 9,540 9,010 100 7.38 8,710 7,110 11,100 7,120 9,460 9,160 80 8.40 7,170 6,420 7,650 6,700 7,750 7,500 100 9.44 10,700 8,800 14,500 8,180 11,800 10,000 3 90 10.52 9,370 8,940 10,500 6,650 9,520 8,440 100 11.43 10,200 6,300 13,100 6,780 10,600 8,310 100 12.47 9,760 8,120 18,000 9,490 10,000 10,300 90 13.44 14,900 11,700 14,600 9,800 11,800 9,560 60 14.51 17,900 12,000 15,200 11,200 12,800 11.900 15.40 12,000 9,300 7,930 12,100 9,460 9,440 70 3 100 Av....44 11,700 8,700 12,700 9,540 11,000 9,460 91 FPL 41 10

Table.--Results of tension and bending tests on control and end- jointed specimens--continued MATERAL G--HEMLOCK 9,720 6,880 7,230 9,050 8,830 5,750 10,300 7,680 8,440 9,320 8,770 7,240 6,730 8,120 11,000 8,330 Piece No. 14,400 12,200 19,200 15,000 7,060 14,700 9,780 16,400 10,600 12,800 14,500 14,100 14,200 13,200 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Av... 6 7 8 9 10 11 12 13 14 15 Av... Specific gravity 1 0.42.48.42.48.40.37.44.39.46.46.42.43.38.39.43.43.34.30.40.28.41.40.49.42.40.40.30.38.40.36.43.38 Tensile strength Necked-down Strip 2 Contro 2,700 5,400 2,600 2,000 8,270 2,600 9,530 6,300 2,200 3,100 3,300 10,700 10,100 14,800 12,600 8,060 0, 800 13,200 9,120 13,000 9,120 3,400 9,700 13,500 11,900 13,500 8,890 13,600 11,300 MATERAL A--HEMLOCK 6,940 8,240 10,600 7,480 8,070 8,500 8,180 8,440 8,000 7,640 7,450 8,350 8,680 7,450 6,050 8,010 8,890 10,600 8,720 11,800 13,200 15,300 11,000 14,400 14,400 11,600 9,740 11,200 15,500 12,900 10,200 12,000 8,380 6,100 5,260 8,600 9,720 4,920 9,280 7,420 7,400 12,100 8,460 5,020 9,780 6,370 8, 7,800 5,820 6,350 4,930 8,150 6,190 8,480 9,990 7,240 7,120 8,940 9,650 9,720 4,490 8,820 7,840 Modulus of rupture Contro 12,500 10,800 11,900 10,800 9,160 12,200 9,890 12,800 10,900 11,200 10,900 10,700 10,500 12,400 9,280 8,380 6,940 9,100 10,900 9,800 12,600 10,900 8,340 10,000 10,900 10,700 10,000 Joint 9,250 5,740 8,480 10,800 9,640 5,170 7,540 10,600 8,760 8,600 8,940 8,260 9,830 10,200 8,890 7,720 6,980 8,520 5,810 9, 270 4,540 7,300 10,000 9,270 8,980 7,160 9,600 9,620 6,830 7,160 7,920 Slope of scarf 1: 10 1:11 1:11 1:11 1:9.5 1:9.5 1:9.5 1:9 1: 10 1:09 Percent wood failure 90 80 50 90 70 70 65 95 70 65 95 60 77 70 70 65 50 20 35 60 65 50 70 65 64 Joint Control Joint 1 2 3 4 5 11

Table.--Results of tension and bending tests on control and end-jointed specimens--continued MATERAL --LARCH Joint Piece No. Control Joint 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Av... 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Av... Specific gravity 1 0.50.59.38.57.50.42.48.56.54.49.51.54.49.46.50.50.50.57.48.47.45.60.49.45.53.54.50.43.48.47.42.49 Tensile strength Strip 2 Necked-down Contro 8,540 13,600 9,110 11,600 9,240 10,600 12,800 14,100 14,800 12,900 9,630 12,800 12,600 13,200 12,100 11,800 10,200 15,400 8,730 O,500 10,300 16,400 2,800 4,200 5,450 9,550 MATERAL H--LARCH 7,830 9,570 6,760 11,000 8,950 10,000 8,980 8,630 9,940 8,810 9,510 7,200 8,410 9,250 5,560 6,610 9,190 7,800 8,130 O,800 4,790 7,720 7,320 8,960 6, 10 7,840 5,100 5,030 6,520 7,190 13,000 13,800 8,400 15,700 13,800 9,650 13,600 14,200 16,100 12,600 7,470 21,900 12,200 13,000 9,580 12,400 11,000 16,100 15,700 24,500 12,400 7,670 13,100 13,100 14,800 10,100 16,300 4,660 8,600 12,700 10,000 9,550 5,730 7,380 12,000 6,980 9,600 8,250 13,300 7,440 10,800 8,160 12,400 9,400 8,450 6,420 4,080 8,500 7,460 7,570 4,300 3,770 7,320 5,770 7,960 4,180 3,840 9,000 6,690 Modulus of rupture Contro 10,400 11,700 7,650 16,400 12,100 10,400 10,800 16,000 15,000 11,900 14,200 13,200 12,300 10,400 13,600 1,000 11,800 12,200 15,800 8,810 8,550 9,620 12,300 11,100 11,800 9,100 10,700 Joint 0,700 8,660 7,430 9,420 9,990 1,200 8,910 0,400 9,940 9,030 9,880 8,010 9,700 1,400 9,760 6,940 9,540 7,290 6,380 9,020 9,260 4,640 4,100 5,670 10,300 6,860 7,200 7,420 5,920 4,690 7,020 Slope of scarf 1:9 1:9 1:9.5 1:09 Percent wood failure 3 100 3 100 95 70 90 100 70 3 100 80 65 75 95 86 5 15 40 5 45 5 15 30 30 10 5 5 15 3 100 30 24 FPL 41 12

5 10 15 20 25 30 -- Table 1. Results of tension and bending tests on control and end- jointed specimens--continued Piece No. 9,200 6,300 8,120 7,040 6,710 7,390 10,300 6,600 7,400 5,870 8,340 10,200 9,250 6,950 7,780 9,170 9,440 9,220 6,690 7,300 2 3 4 6 7 8 9 11 12 13 14 16 17 18 19 21 22 23 24 26 27 28 29 Av... 29,200 16,900 10,300 19,600 25,700 23,500 29,400 17,000 22,800 17,700 22,400 21,600 16,500 17,700 21,100 24,400 25,300 22,200 22,400 21,600.50.58.61.55.55 Specific Tensile, strength Modulus of gravity 1 rupture Strip 2 Necked-down 4 14,100 14,200 13,800 14,900 18,000 Control 6,160 8,540 9,020 0.62. 53.56.50.68 22,100 26,300 19,200.54.65.47.52.57 7,600 8,770 9,579 6,300 8,430 7,540 7,870 7,970.53.62.54.58.53.68.58.50.57.51 23,000 22,200 23,500 15,500 21,900 23,300 24,200 21,600 23,800 18,800 12,800 12,800 16,300 18,700 25,000 9,880 15,000 14,800 15,800 17,200 19,600 17,500 18,700 22,300 18,100 15,100 14,900 19,500.62.56.63.49.57 21,200 18,100 19,600 13,900 16,900.56 17,000 MATERAL C--SOUTHERN PNE 10,400 7,920 8,460 7,820 8,210 8,000 9,320 7,490 8,680 8,840 8,540 9,640 7,140 8,260 9,060 8,740 8,090 8,520 8,610 9,360 6,550 9,220 8,340 9,830 9,900 9,180 6,960 8,700 9,550 6,860 8,540 18,600 11,600 14,600 13,700 19,300 14,900 17,600 1,800 12,500 12,800 15,900 14,500 16,000 15,400 16,800 14,300 16,700 15,700 13,000 16,200 12,500 12,700 15,400 15,900 4,200 3,100 5,100 1,600 11,900 16,200 12,900 15,600 14,800 Joint 0,900 1,000 0,000 9,130 0,200 0,100 9,180 0,800 0,800 10,800 10,300 11,800 9,870 10,100 10,700 9,760 10,100 10,200 9,800 9,700 11,800 9,820 9,200 9,540 7,730 9,170 9,340 0,000 0,200 Slope of scarf Percent wood faiure 70 100 65 60 90 90 70 95 65 60 60 95 90 45 90 95 90 80 75 70 100 95 95 75 65 80 Joint Control Joint Control 13

Table 1.--Results of tension and bending tests on control and end-jointed specimens--continued Piece No. Specific gravity 1 Tensile 2 Strip strength Necked-down Modulus of rupture Slope of scarf Percent wood failure Contro Joint 4 Control- Joint Control Joint P.s.i MATER AL J--DOUGLAS-FR 2 3 4 5 0.62.62.46.43.49 18,700 20,600 16,400 15,300 9,980 7,280 12,300 9,340 8,190 7,900 17,600 21,400 14,200 13,600 14,400 8,640 14,600 12,600 14,900 9,340 12,200 7,540 10,400 9,100 11,600 5,680 12,300 9,300 8,650 9,160 75 70 90 95 6 7 8 9 10.42 11,200.53 17,600.53 18,200.49 12,600.52 17,800 8,920 8,660 8,980 8,160 11,400 12,400 17,500 13,200 13,700 19,300 9,750 9,060 9,440 6,540 11,400 11,200 13,800 13,700 11,900 13,900 8,430 12,300 10,600 7,260 12,300 95 100 80 90 75 11 13 14 15.52.47.52.52.55 13,400 9,240 16,900 9,910 20,800 9,620 6,640 12,500 7,530 8,650 18,600 11,500 22,100 12,800 22,900 9,030 7,960 9,620 7,660 10,400 12,600 12,100 13,100 10,100 14,600 9,480 8,420 11,700 6,880 7,730 75 90 55 100 16 17 18 19 20.40 14,200.53 19,200.47 10,800.44 16,200.64 16,400 8,470 10,100 8,460 9,090 10,600 14,200 21,600 14,900 15,300 19,800 8,470 10,900 10,200 9,680 9,630 10,700 13,200 11,700 11,700 14,200 9,440 11,700 10,100 10,100 11,300 80 75 70 95 21 22 23 24 25.46.49.47.44.49 13,200 19,300 14,300 11,200 14,600 9,780 9,770 9,310 6,0 8,940 14,400 14,200 15,300 10,500 16,000 8,0 10,600 9,720 7,300 8,170 11,100 12,200 11,900 11,100 12,000 9,780 11,200 10,700 8,400 8,580 90 95 100 100 26.60 27.49 28.56 29.62 30.50 15,100 15,600 16,500 18,900 11,700 9,540 10,800 11,000 11,500 8,970 14,500 21,600 19,400 20,400 16,900 11,000 12,900 10,400 12,700 11,300 11,800 12,100 14,400 8,840 12,400 11,300 9,800 8,400 13,100 9,670 80 90 90 75 90 Av....51 15,200 9,310 16,500 9,510 12,500 9,790 Based on volume at time of test and ovendry weight. 2 Values for strip tension test are average of 8 specimens. 3 Bending specimens that did not fail in joint. 4 Values are average of 2 specimens. FPL 41 14

Figure 6.--Comparison of tensile strength of necked-down and strip-tension, scarf-jointed specimens having an average slope of scarf of in. The reference line has a slope of 1 to 1. M 125 838 15

FPL 16

Figure 8.--Comparison of tensile strength (necked-down) and modulus of rupture of scarfjointed specimens having an average slope of scarf of 1 in 10. The reference line has a slope of 1 to 1. M 125 835 17

FPL 41 18

Figure 10.--Comparison of average tensile strength of necked-down and strip-tension, scarf and finger-jointed specimens. Materials C and J were finger joints and all others were scarf joints. Complete data are shown in figures 6 and 13. The reference ine has a slope of io. M 125 842 19

Figure 11.--Comparison of average values of tensile strength (necked-down) and modulus of rupture of scarf and finger-jointed specimens. Materials C and J were finger joints and all others were scarf joints. Complete data are shown in figures 8 and 14. The reference line has a slope of 1 to 1. M 125 844 FPL 41 20

Figure 12.--Comparison of average values of tensile strength (strip-tension) and modulus of rupture of scarf and finger-jointed specimens. Materials C and J were finger joints and all others were scarf joints. The reference ine has a slope of 1 to 1. M 125 847 21

Figure 13.--Comparison of tensile strength of necked-down and strip-tension, finger jointed specimens. The reference line has a slope of 1 to 1. M 125 846 FPL 41 22

Figure 14.--Comparison of tensile strength (necked-down) and modulus of rupture of fingerjointed specimens. The solid reference line has a slope of 1 to 1 and the broken lines go through the average value of each group of data. M 125 843 23

Figure 15.--Comparison of tensile strength of necked-down and strip-tension control The reference line has a slope of 1 to 1. M 125 839 FPL 41 24

Figure 16.--Comparison of average values of tensile strength of necked-down and strip- tension control specimens. The complete data are shown in figure 15. The reference ine has a slope of 1 to 1. M 125 845 25

Figure 17.--Comparison of tensile strength (necked-down) and modulus of rupture of control specimens. The reference line has a slope of 1 to 1. M 125 834 FPL 41 26

Figure 18.--Comparison of tensile strength (necked-down) of end- jointed and control specimens. The reference line has a slope of 1 to 1. M 836 27

Figure 19.--Comparison of modulus of rupture of end- jointed and control specimens. The reference line has a slope of to. M 125 837 FPL 41 28 GPO 821-725-4

Figure 20.--Failures of finger- jointed bending specimens. The bottom or tension face is shown for material C, southern pine specimens, and the side view with arrow pointing to the tension face is shown for the material J, Douglas- fir specimens. M 125 324 29

Figure 21.--Failures of scarf- jointed, larch bending specimens. joints had very low percentages of wood failure. n general, these M 125 322 FPL 41 30

Figure 23. --Failures of scarf- jointed, southern pine bending specimens showing reasonably high percentages of wood failure. M 125 323 FPL 41 32

Figure 24. --Failures of scarf- jointed, Douglas- fir bending specimens showing reasonably high percentages of wood failure. M 125 326 33

Figure 25.--Failures of scarf- jointed, hemlock bending specimens showing generally high but also a few low percentages of wood failure. M 125 327 FPL 41 34

Figure 26.--Scarf- jointedlaminations which were poorly alined during manufacture, M 124 254 35

down tension strength and modulus of rupture for scarf-jointed specimens are shown in figures 6 and 8 respectively. The same relationships for the finger- jointed specimens are shown in figures 13 and The average values for each material are shown in figures 10, 11, and 12. From figures 6 and 8 it is nearly impossible to visually determine the scatter and trends of the data for each material, so these figures were expanded to facilitate analysis. The expanded plots are shown in figures 7 and 9, with figure 7 showing the same data as figure 6 and figure 9 showing the same data as figure 8. Due to the variation of the data for each material, relationships between test methods are not readily apparent. Therefore, variation analyses were made on the data, and it was concluded that the variation of each set of data was fairly similar to that of the combined data for scarf joints; consequently, figures 6 and 8 of the combined data should be a reasonably good representation of the relationships between the test methods. Analyses using least squares methods were made on the combined data of figures 6 and 13, which show the relationship between the two tension-test methods for both finger- and scarfjointed specimens, and on the data of figure 8, which show the relationship between necked-down tension strength and modulus of rupture for the scarf- jointed specimens. The calculated slope of the line in these figures varied from 1:1 by a minimum of 0 percent to a maximum of 4 percent, depending upon the assumptions made in performing the calculations. Therefore, for all practical purposes, the strip-tension test method gives as good an indication of tensile strength of end- jointed specimens as the necked-down tension test. Also, the modulus of rupture is approximately equivalent to the tensile strength of scarf-jointed laminations having scarfs with a 1: 10 slope. These statements appear to be correct for strength values within the normal range of strength obtained from end joints in these tests (less than 13,000 pounds per square inch). For stronger material, the necked-down tension test gave higher strength values. This is shown in figures 15, 16, and 17, which compare strength values for the control specimens for the three test methods. The strength values for control specimens for the two tension tests represent values for one size specimens for the necked-down test (approximately 1/4- by 3/4-inch cross section) and two sizes for the strip-tension test (approximately 3/32- by 3/4-inch and 3/32- by 1-5/8-inch cross sections). The greatest difference in strength values occurred when 1-5/8-inch-wide strip specimens were used. The corresponding values for necked-down specimens were the highest obtained. t could not be determined from these results whether the specimen size caused the greater difference in values from the two test methods or whether the difference becomes greater as the strength increases. Previous work with the striptension test at the Laboratory indicated that smaller cross sections gave higher strength 4 values. A comparison of tensile strength as determined by necked-down tension test and modulus of rupture for the finger-jointed laminations is shown in figure 14. This figure shows that the relationship between these test methods varies with the of the finger joint. Figures 18 and 19 show the relationship between the strength of control and end-jointed specimens and are not directly pertinent to this study except to indicate the range of variation in values. At the start of this study it was not known how many specimens would be needed from each joint for the strip-tension test to give an accurate estimate of average strength of the joint. From previous studies it was estimated that eight specimens per lamination were about the minimum number required to determine the average strength. An analysis of the data from eight specimens per lamination gave an average withinboard coefficient of variation of 15 percent. The data from both finger and scarf joints are included in this value. Based on this 15-percentaverage coefficient of variation, the sample size needed to predict different percentages of accuracy was calculated to be as follows: Precision (Percent of mean) Specimens per joint ±5 36 ±10 9 ±15 4 ±20 3 An investigation was made of the necked-down tension data to determine if the strength of finger FPL 41 36 GPO 821-725-3

Table 2.--Variance of tensile strength of necked-down tension specimens Material Species Variance Number of specimens PLAN SCARF JONTS B E D F A Southern pine Southern pine Douglas-fir Douglas-fir Hemlock 5,6,476 4,388,223 11,272,110 5,021,557 4,097,909 15 15 15 15 15 G H Hemlock Larch Larch 4,191,2 4,562,532 5,491,500 15 15 15 FNGER JONTS C J Southern pine Douglas-fir 883,824 2,060,2 30 30 Pooled variance of scarf joints 5,610,198 Pooled variance of finger joints 1,472,338 60 joints was less variable than the strength of scarf joints. Values of the variance for each group of laminations are given in table 2. As expected, there was a difference in the variability between materials, since the joints were made by several manufacturers: however, except for one, this difference in variability was not great for the scarf-jointed laminations. The pooled variance for the finger joints was 1,472,338 and for the scarf joints was 5,610,198. For this set of data, therefore, the strength values of finger joints are less variable than the strength values of scarf joints. Comments on Fabrication of Joints n general, the joints appeared reasonably well made. This is further substantiated by the strength and wood failure values given in table 1. n one instance, however, there were indications that the glue used was either over-aged or possibly improperly formulated. This was borne out by generally low wood failures (fig. 21) and wide variations in strength values. Serious misalinement of joints was also in evidence (fig. 26). 37

CONCLUSONS This study evaluated the strength of end joints by three different test methods. The material tested was end glued by several laminators using commercial end- jointing techniques. The following conclusions were drawn from this study: 1. For all practical purposes, the tensile strength of end- jointed laminations as determined by the strip-tension test method, and also the modulus of rupture of scarf- jointed laminations having scarf slopes of 1 in 10, are equivalent to the tensile strength as determined by the neckeddown tension test method, This is valid provided the strength of the end joint does not exceed the normal range of strengths obtained in this study; that is, strengths of less than 13,000 pounds per square inch. 2. For control specimens and end-jointed specimens having higher strengths, the necked-down tension test gives the best indication of strength. The relationship between tensile strength and modulus of rupture of finger- jointed specimens apparently varies with the configuration of the finger joint. 4. The strength of the finger joints was less variable than the strength of the scarf joints having slopes of 1 in 10. FPL 41 38

PUBLCATON LSTS SSUED BY THE FOREST PRODUCTS LABORATORY The following lists of publications deal with investigative projects of the Forest Products Laboratory or relate to special interest groups and are available upon request: Architects, Builders, Engineers, and Retail Lumbermen Box, Crate, and Packaging Data Chemistry of Wood Drying of Wood Fire Protection Fungus and nsect Defects in Forest Products Furniture Manufacturers, Woodworkers, Teachers of Woodshop Practice Growth, Structure, and dentification of Wood Logging, Milling, and Utilization of Timber Products Mechanical Properties of Timber Structural Sandwich, Plastic Laminates, and Wood-Base Components Thermal Properties of Wood Wood Fiber Products Wood Finishing Subjects Glue and Plywood Wood Preservation Note: Since Forest Products Laboratory publications are so varied in subject matter, no single catalog of titles is issued. nstead, a listing is made for each area of Laboratory research. Twice a year, January 1 and July 1, a list is compiled showing new reports for the previous 6 months. This is the only item sent regularly to the Laboratory s mailing roster, and it serves to keep current the various subject matter listings. Names may be added to the mailing roster upon request. 39

The FOREST SERVCE of the U. S. DEPARTMENT OF AGRCULTURE is dedicated to the principle of multiple use management of the Nation s forest resources for sustained yields of wood, water, forage, wildlife, and recreation. Through forestry research, cooperation with the States and private forest owners, and management of the National Forests and National Grasslands, it strives - as directed by Congress - to provide increasingly greater service to a growing Nation. FPL 41 40 2.-41 GPO 821-725-2

Forest Service regional experiment stations and Forest Products Laboratory GPO 821-725-1

FOREST PRODUCTS LABORATORY U.S. DEPARTMENT OF AGRCULTURE FOREST SERVCE - - - MADSON, WS. n Cooperation with the University of Wisconsin