Made from Visually Graded Hem-Fir Lumber

Flexural Tests of Large Glued-Laminated Beams Made from Visually Graded Hem-Fir Lumber by James W. Johnson April 1973 Research Paper 18 Forest Research Laboratory School of Forestry Oregon State University Corvallis, Oregon

Flexural Tests of Large Glued-Laminated Beams Made from Visually Graded Hem-Fir Lumber James W. Johnson Associate Professor of Forest Products Research Paper 18 April 1973 Paper 876 Forest Research Laboratory School of Forestry Oregon State University Corvallis, Oregon, 97331

CONTENTS Page Acknowledgements..................... iv Summary......................... iv Introduction........................ 1 Procedure.............. Procurement of lumber, grading, and data taken............ Beam design and assembly.................. 2 General design.................... 2 Face laminations.................... 2 End joints...................... 2 Manufacure of beams.................... 2 Testing...................... 2 Results and Discussion..................... 4 Modulus of elasticity (MOE)................ 4 Modulus of rupture (MOR).................. 4 Beam failures...................... 8 Conclusions............ 8 Appendix......................... 11 W

ACKNOWLEDGMENTS This study was a cooperative effort of individuals from several organizations and was performed under the general auspices of the American Institute of Timber Construction (AITC), which largely financed the project. Appreciation is expressed to everyone concerned and especially to AITC, the Steering Committee (see APPENDIX), Rosboro Lumber Company, and co-workers in this study from Oregon State University. SUMMARY Fifteen large glued-laminated beams made from visually graded hem-fir lumber were designed, fabricated, and tested to failure in static bending. When tested, the beams had an average moisture content of 7.5 percent. The beams were 40 feet long, 5 t/s inches wide and 24 inches deep. Before laminating, individual pieces of lumber were graded by specifications similar to those for the L-grades and the AITC recommendations for face laminations for Douglas fir. Three combinations (designs) were included, with 5 beams in each. Performances of the beams during testing are tabulated and graphed. Guidelines were established for the grades and locations of material needed for different levels of modulus of elasticity (MOE) and modulus of rupture (MOR). Least values of MOE and MOR obtained from beams of each design were: combination 1, 1,788,000 and 5,350: combination 2, 1,660,000 and 4,580; and combination 3, 1,724,000 and 3,950 psi. Results of the tests will provide a basis for the development of industry laminating specifications for visually graded hem-fir lumber. it,

1 Flexural Tests of Large Glued-Laminated Beams Made from Visually Graded Hem-Fir Lumber James W. Johnson INTRODUCTION The objectives of this study were to design large glued-laminated beams fabricated from visually graded hem-fir lumber, test the beams in static bending, and measure their performance. Results of the tests would provide a basis for the development of industry laminating specifications. Beams of three combinations were included: the combinations were determined by grade of lumber and location within the beam, Although each combination was not to be designed for a specified stress value, the three designs were expected to provide allowable design stress values in bending ranging from 1.600 to 2,400 pounds per square inch (psi). When this study was started in 1972. there were no industry specifications for the use of visually graded hem-fir lumber in the production of glued-laminated timbers. Results obtained from the previous study of E-rated hem-fir ]timber indicated that beams with high bending strength also could he made by using visually graded hem-fir material, provided the lumber was sorted and properly used (Design and Test of Large Glued-Laminated Beams Made of Nondestructively Tested Lumber, by James W. Johnson, Forest Research Laboratory, Oregon State Univ.. Report T-27. November 1971 ). The use of visually graded item-fir lumber would provide laminators with an excellent, but long underused group of species for laminating stock. This is of special importance during periods of high demand for Douglas fir and Southern pine lumber to be used for purposes other than laminating. Inclusion of hem-fir as a source of laminating stock could help eliminate potential shortages as well as have a favorable effect on the economy of manufacturing laminated timber. This study, and the one cited above, were guided by a steering committee composed of industry and university personnel. Members of the steering committee for this study are listed in the APPENDIX. In this study, 15 beams 5 1/8 inches wide, 24 inches deep, and 40 feet long were tested. Five beams (replications) were included in each of the three combinations. The beams were fabricated at Rosboro Lumber Company, Springfield, Oregon. PROCEDURE Procurement of lumber, grading, and data taken Initially, about 16,000 board feet of unseasoned 2- by 6-inch hem-fir lumber of 12, 14, and 16-foot lengths and of Structural Joist and Plank grades (25 percent No. 3, 25 percent No. 2. balance No. I and Better) was obtained from Simpson Timber Company in Washington. Willamette Industries in Oregon, and Sierra Pacific in California; each supplied about one-third. The lumber from Washington and Oregon was western hemlock (Tsuga helerophylla [Raf.j Sarg.) primarily: most of that from California was white fir (Abies eoncolor (cord. & Glen.I Lindl.). Later, after checking grades, an additional 1,000 board feet of white Cu and 1,600 board feet of hemlock (all laminating grade L3}, was obtained from two Oregon mills D. R. Johnson Lumber Company and Frank Lumber Company. After kiln-drying at a conventional schedule (maximum dry-bulb temperature. 180 F) and planing the pieces to I V-inch thickness at Clear Fir Products, Springfield, Oregon, the lumber was graded by graders of Western Wood Products Association (WWPA) and West Coast Lumber Inspection Bureau (WCLIB). Descriptions for laminating grades of hem-fir were not available, so the lumber was graded on the same basis as the structural lamination grades (L-grades) as contained in the WWPA Special Products Rule 500.00. Structural Laminations, or paragraph 154. WCLIB Grading Rules No. 16 for Douglas fir. No dense. or dose-grain grades were used, however; all lumber was graded as medium grain.

Lumber was sorted and tallied. Data recorded for each piece included the following: identification. Joist and Plank grade. Laminating grade, length. weight. moisture content measured at three locations along the pieces of lumber: and average modulus of elasticity (MOE) measured by the Irvington E computer. The information obtained is not part of this report, but is retained by the American Institute of Timber Construction (AITC). The E computer was used for convenience in measuring weights that, in turn, were used in determining specific gravity. Values for MOE were recorded because of the added opportunity to obtain MOE data on a sample of hem-fir: they were not used in designing or assembling the beams. A knot survey was made on about 1,200 board feet of the L3-grade lumber (core material of the beams) selected at random. This was done on individual pieces of lumber. Later. before final gluing and after end jointing and dry assembly of the beams, a knot survey was made on the central 20 feet of each beam, but this survey was done only for the outer four compression and four tension laminations, which consisted of lumber of grades and LI, plus the special lumber of the face laminations (Figure I ). All knot-survey data were taken for possible use by AITC at a later date. These data are not included in this report, nor did they influence design or assembly of the beams. Beam design and assembly General design. Beams were composed of 16 laminations, each of 11/2 inch thickness. Three combinations of beams were designed, and five beams were fabricated for each combination. Combinations of hem-fir were similar to combinations of Douglas fir as follows: Combination 1, hem-fir Douglas fir, 26 F Combination 2, hem-fir Douglas fir, 24 F Combination 3, hem-fir Douglas fir, 20 F Design of the three combinations is shown in Figure 1. The pieces of lumber of LI grade used in the tension zone and the face (tension) laminations were of near average or above specific gravity of hem-fir (0.39 or above, based on oven-dry weight and volume at 12 percent moisture content). Face laminations, Special attention was given to selection of the face (tension) laminations because of their importance in determining beam strengths. Lumber of the 301 grades (AITC Recommendations) was used in the different hem-fir beam combinations as follows: 301-26, 301-22, and 301 20 in combination I, 2, and 3 respectively (Figure I). Tension laminations were of low-line to average quality: no high-line face pieces were used. Tension laminations, as well as those of other grades, were manufactured on a continuous basis, and the pieces of lumber were placed randomly within laminations. End joints. All end joints in the beams were finger joints. End joints in the face (tension) lamination were allowed to occur naturally, as in production. In each combiantion of 5, however, at least one beam had a joint in the face lamination that would be located between load points (under maximum bending moment) during test. Locations of end joints in the remaining area of each beam were in accordance with Commercial Standard CS 253-63. Manufacture of beams All beams were straight (without camber) and were manufactured in accordance with applicable provisions of manufacture and quality control as contained in CS 253-63. Melamine-urea adhesive was used in end joints; a phenol-resorcinol for fact bonding. Caul pieces of lumber were used between the laminating frame and the tension laminations. Each beam was surfaced on two sides and end trimmed at the laminating plant. Testing Beams were tested at the Engineering Laboratory of Oregon State University on a universal testing machine by standard procedures of the American Society for Testing and Materials (ASTM). A sketch of the arrangement is shown in Figure 2. 2

11 Combination 1 Slope of grain Combination 2 Combination 3 Slope of Slope of grain grain LI L3 1:14 1:12 1:12 1:8 L3 L3.Neutral axis L3 L1' 301-261 L3 301-221 C7 L3 1:8 1:12 1:12 L1' 1:14 301-201 1:12 1These laminations were of near average or above specific gravity for hem-fir (0.39 or above, oven-dry weight and volume at 12 percent moisture content). Figure 1. Combinations of glued-laminated beams made from visually graded hem-fir lumber and tested in bending. Grades of lumber were laminating grades similar to L-grades and AITC Tension Lamination Recommendations for Douglas fir; they were graded as medium grain. STEEL BEAM Figure 2. Arrangement for testing large glued laminated beams. Measurements are in feet. 3

The 40-foot beams were tested in static bending over a 38-foot span with 8 feet between the two load points (15, 8, and 15-foot spacing). Deflection at the centerline between supports was measured to the nearest one-hundredth inch with a taut wire and scale; centerline deflection over a 6-foot span between supports was measured to the nearest one-thousandth inch by an Ames dial mounted on an aluminum yoke. Both yoke and wire were supported from nail-type pins at the neutral axis. The scale was read with a telescope. Lateral restraint was provided at each end of the beam at support points (not shown in Figure 2) and at two other locations 33 inches each side of the midlength of the beam. Loading of the beams was continuous to failure with head speed of the machineset at I inch per minute. Deflections were read at 2,000-pound increments. After failure of each beam, moisture content was measured for each lamination, near the failure, with a resistance-type moisture meter. Also, photographs and notes were obtained for each beam. RESULTS AND DISCUSSION Individual and average test values for the beams are listed in Table 1, and load-deflection data in Table 2. Average moisture contents of the beams at time of test were about 7.5 percent. Modulus of elasticity (MOE) The two types of MOE values obtained in this study and their abbreviations are defined as follows: E measured over the total span of 38 feet, includes shear deflection; E,, measured over a 6-foot span between load points where shear is assumed to be nonexistent. The two types of MOE values correlated quite well (correlation coefficient = 0.839), considering the limited range of the values obtained (from 1,660,000 to 1,882,000 psi for E,) and that only 15 pairs of values are included (Figure 3). Values of Ens were about 6.5 percent greater than Er (Table 1). Differences were small among MOE values for the three combinations of beams, especially differences among the values obtained over total span (Er). However, MOE was consistently greater for beams of combination 1. Small differences among the three combinations might be expected because of small differences in visual grades in design of the beams (Figure t), and because MOE values for lumber of different visual grades do overlap. MOE values for pieces of lumber were not used in placing the pieces within the beams. Modulus of rupture (MOR) Values of MOR for beams of the three combinations are listed in Table 1, along with moisture content and shear. Also listed (column 4) is the allowable design stress for each beam, which was calculated by adjusting the value of MOR by factors for duration of load, safety, and size of beam (formula in footnote to Table I ). Each combination was not designed for a specified stress value, but the beams were tested to determine the stress values that would be obtained. The three designs were expected to provide allowable design stress values ranging from 2,400 to 1,600 psi. The allowable design stress values obtained ranged from 3,790 psi (beam 1-2) to 2,040 (beam 3-5). Least design stress values obtained from beams of combination 1, 2, and 3 were 2,770; 2,370; and 2,040 psi, respectively (Table 1, column 4). Based on averages, the MOR values from beams of combination 1 were greatest, but little difference was found between beams of combinations 2 and 3. Relations among individual and average values of MOR and MOE are graphed in Figure 4. Least variation of MOR values (and most variation for MOE) occurred among beams of combination 2. 4

Table 1. Results of Bending Tests on 40-Foot Beams, 2 Feet in Depth, Laminated from Visually Graded Hem-Fir Lumber modulus Modulus of of Allowable elasticity (MOE) Moisture rupture design Total Between Shear Beam content' (lior)z stress3 span load points stress 1 2 3 4 5 6 7 S Psi Psi if psi. at psi Psi Beams with AITC 301-26 as tension lamination 1-1 7.9 6810 3520 1788 1922 221 1-2 7.7 7330 3790 1802 1925 238 1-3 6.9 7060 3650 1860 1933 229 1-4 7.3 5350 2770 1808 2049 173 1-5 7.2 7200 3730 1805 1982 235 Avg 7.4 6750 3490 1813 1962 219 with AITC 301-22 as tension lamination Beams 2-1 7.6 4900 2540 1660 1670 158 2-2 7.0 5310 2750 1882 2037 172 2-3 7.3 4740 2450 1743 1780 153 2-4 7.3 4580 2370 1769 1970 147 2-5 7.4 5500 2850 1722 1756 178 Avg 7.3 5010 2590 1755 1841 162 Beams with A1TC 301-20 as tension lamination 3-1 7.8 5860 3030 1855 1987 190 3-2 7.9 5300 2740 1839 1911 171 3-3 7.7 6300 3260 1728 1815 204 3-4 7.4 5910 3060 1724 1859 193 3-5 7.8 3950 2040 1737 1862 127 Avg 7.7 5460 2830 1777 1887 177 'Average of all laminations, taken near break, with resistance-type moisture meter. 2 Average weight of beams used in calculating M.OR. 31!0R (1:2.1)(1:0.92), where 1:2.1 is factor for safety and load duration and 1:0.92 is factor for adjustment from 24 to 12-inch size, compatible with the standard size factor of (12/d)'"9. 4Based on maximum load.

Table 2. Loads and Deflections over Total Span, in Inches, from Bending Tests of 40-Foot Reams, 2 Feet in Depth, Laminated from Visually Graded Hem-Fir Lumber Load, lb x Beam number 103 1-1 1-2 1-3 1-4 1-5 2-1 2-2 2-3 2-4 2-S 3-1 3-2 3-3 3-4 3-5 2 0.35 0.39 0.36 0.37 0.31 0.39 0.38 0.36 0.35 0.38 0.36 0.38 0.38 0.36 0.38 4 0.71 0.75 0.71 0.73 0.66 0.79 0.72 0.73 0.72 0.76 0.71 0.72 0.75 0.74 0.74 6 1.07 1.10 1.05 1.09 1.02 1.16 1.06 1.09 1.07 1.13 1.06 1.06 1.12 1.10 1.11 8 1.43 1.46 1.40 1.45 1.38 1.55 1.40 1.47 1,46 1.51 1.42 1.42 1.49 1.48 1.47 10 1.80 1.82 1.74 1.81 1.73 1.95 1.76 1.85 1.84 1.88 1.78 1.77 1.87 1.85 1.86 12 2.17 2.18 2.10 2.17 2.09 2.37 2.10 2.22 2.22 2.27 2.14 2.13 2.24 2.23 2.25 14 2.54 2.55 2.45 2.54 2.44 2.77 2.44 2.59 2.58 2.66 2.50 2.48 2.63 2.60 2.62 16 2.90 2.90 2.79 2.90 2.80 3.17 2.80 2.97 2.96 3.03 2.87 2.84 3.01 2.97 2.99 18 3.27 3.25 3.15 3.25 3.16 3.54 3.15 3.33 3.33 3.42 3.23 3.19 3.38 3.33 3.36 20 3.65 3.62 3.51 3.63 3.53 3.97 3.50 3.73 3.74 3.82 3.60 3.56 3.78 3.73 3.76 22 4.03 4.00 3.87 4.01 3.90 4.39 3.87 4.16 4.12 4.19 3.98 3.93 4.17 4.12 -- 24 4.41 4.36 4.24 4.38 4.27 4.80 4.22 4.57 -- 4.59 4.34 4.27 4.61 4.52 -- 26 4.78 4.74 4.60 4.74 4.63 -- 4.58 -- 4.99 4.72 4.66 5.07 4.90 -- 28 5.16 5.11 4.95 -- 4.99 -- -- -- -- 5.40 5.09 -- 5.50 5.29 -- 30 5.53 5.48 5.30 -- 5.35 -- -- -- -- -- 5.48 -- 5.91 5.68 32 5.93 5.87 5.66 -- 5.71 6.33 -- -- 34 6.37 6.32 6.03 -- 6.10 36 -- 6.75 6.42 -- 6.48 38 -- 7.21 -- 6.89 Max Defl 6.74 7.21 6.58 5.09 6.89 5.12 4.82 4.67 4.34 5.40 5.48 4.98 6.58 5.90 3.76 Max. Load 3.58 38.5 37.0 27.9 38.1 25.6 27.8 24.8 23.8 28.8 30.2 27.8 33.1 31.3 20.7 Beams symmetrically loaded over 38-foot span, with two loads 8 feet apart (15, 8, 15 feet spacing).

106 x 2.1 T-_I COMBINATION a 2 3 0 1.9 LO 1.6 w w 1.7 0 2 1.6 Y I.438x -664(103) C.C. 0.839 N- 15 1.51 rl 1 1 1L 1.5 1.6 1.7 1,8 1.9 x IO MOE (Er ), PSI (x) Figure 3. Relation between modulus of elasticity (MOE) obtained over total span (Er) and MOE obtained over a span of 6 feet between load points (E5). 7 6 4 3,. 1.6 1.7 1.8 106 x 1.9 MOE (Er), PSI Figure 4. Relations among individual and average values of modulus of elasticity (MOE) over total span (Er) and modulus of rupture (MOR) of five each glued-laminated beams for three combinations of hem-fir. The average of all five beams of a combination is indicated by the larger, solid symbol.

Beam failures Remarks concerning failures of each beam are given in Table 3. To determine contributing causes of failure or select the initial break (trigger point) was not always possible. With several of the beams, however, initial and final failures were seen. Also, the start of compression under the load points or in the top lamination was seldom possible to determine. The remarks in Table 3, which are a consensus of several observers at time of test, plus a recheck after test of the fractured beams, indicate whether the particular statement was a result of direct observation or of expert opinion. Typical failures of the beams arc shown in Figure 5. In most beams, tension and shear failures were extensive throughout the middle 20-foot section, and some shear failures extended to one or both ends. Some beams broke abruptly, but most gave warning by cracking sounds during test. With several beams, combinations of strength-reducing characteristics seemed to influence strength considerably. CONCLUSIONS Glued-laminated beams of different combinations of lumber similar to visual grades L1,, and L3 of Douglas fir can be constructed of visually graded hem-fir lumber to give satisfactory values for modulus of rupture (MOR); but material used in highly stressed areas must be selected carefully and well-made joints are needed. Guidelines were established regarding visual grades and locations of hem-fir lumber needed for different levels of MOE and MOR. Least values of MOE and MOR obtained from testing 5 beams (average moisture contents, about 7.5 percent) of each design were: combination 1, 1,788,000 and 5,350; combination 2, 1,660,000 and 4,580; and combination 3, 1,724,000 and 3,950 psi. Table 3. Remarks Concerning Failures of Hem-Fir Glued-Laminated Beams. Combination-1 beams, with A.TC 301-26 as tension lamination (lam) 1-1 first cracked at 30,000 pounds in the tension lamination near centerline of the beam. There was local slope of 1:12 and a bark pocket. Final failure was in the same area at a load of 35,800 pounds. MOR was 6,810 psi. 1-2 first cracked at about 30,000 pounds, but area was unknown. The beam failed at 38,500 pounds at a joint in the tension lamination 30 inches outside a load point. The break extended to a joint in the second lamination 20 inches from centerline of the beam. A knot was near the joint. After test, compression was found under and between both load points; some extended into the fourth lamination from the top. MOR was 7,330 psi. 1-3 failed at a pin edge knot 15 inches from centerline of the beam and a finger joint 12 inches outside of a load point, both in the tension lamination. Apparently, the knot failed first and the joint soon after. Compression was found under one load point, in the top lamination. MOR was 7,060 psi. 1-4 failed at a 1/2-inch corner knot in the tension lamination under one load point, with the break extending diagonally across the board to 24 inches past the centerline of the beam. The second lamination had a brash break 15 inches away from the knot, toward the centerline. There was juvenile wood in the central portion of the third lamination. MOR was 5,350 psi. 1-5 possibly failed in shear between the third and fourth laminations on one end of the beam and near mid-depth at the other end. But laminations 4-11 failed as in bending near the centerline of the beam. The first three laminations were intact (no failures). Compression was found in the top lamination under one load point. MOR was 7,200 psi. 8

Combination-2 beams, with AITC 301-22 as tension lamination 2-1 probably failed at a 11/4-inch edge knot in the tension lamination 11 inches inside one load point. Joint in the second lamination failed 20 inches from the centerline of the beam on the opposite side of the centerline from the edge knot. MOR was 4,900 psi. 2-2 failed at a 11/4-inch edge knot in the tension lamination 12 inches inside one load point. The break extended across the board to a /-inch edge knot 12 inches away. There was a failure in a joint in the second lamination between the two knots. MOR was 5,310 psi. 2-3 first cracked (20,000 pounds) near the centerline of the beam at a 3/4-inch spike knot in the tension lamination. Final failure, at 24,800 pounds, was at a joint 8 inches, and a 7/8-inch edge knot 14 inches, from centerline of the beam, both in the tension lamination. In the same cross section, the second lamination failed at two knots, each about 11/4 inches in size. MOR was 4,740 psi. 2-4 failure was somewhat questionable. In the tension lamination, there was a partial joint failure 4 inches inside one load point and a failure at a l'/4-inch knot 68 inches out from the load point. The second lamination had a brash break 19 inches out from the load point. The third had a failure at a knot and the fourth had a joint failure, both near the load point. Probably, there was compression wood in the second and fourth laminations. Several knots in the tension lamination between load points did not fail. MOR was 4,580 psi. 2-5 failed first (28,000 pounds) and finally (28,800) at two spike knots on opposite edges of the tension lamination about 17 inches inside one load point. The two knots were 5/8 and %2 inch in size. The break extended to a 3/-inch knot 35 inches outside the load point. The second lamination had a joint failure just above the two knots. MOR was 5,500 psi. Combination-3 beams, with AITC 301-20 as tension lamination 3-I failed at a cluster of three '/-inch spike knots and juvenile wood in the tension lamination 8 inches inside one load point and at a joint 10 inches from the knots toward the centerline of the beam. The break extended along the tension lamination to a 1'-inch edge knot 9 inches on the other side of the centerline. MOR was 5,860 psi. 3-2 broke suddenly at a joint in the tension lamination 16 inches outside a load point. The break went through a I-inch knot 12 inches on the other side of centerline. The beam may have failed first in the second, third, fourth, or fifth laminations, from 6 to 20 inches on the opposite side on the centerline from the joint. MOR was 5,300 psi. 3-3 cracked first (20,000 pounds) at a 1 118 -inch edge knot in the tension lamination 14 inches inside one load point. Final break (33,100) was at the same knot, but also at a I 118 -inch edge knot in the tension lamination 25 inches from the first break, toward centerline of the beam. This beam was very tenacious from the first to the final failure. MOR was 6,300 psi. 3-4 first cracked (20,000 pounds) at a 13/8-inch edge knot in the tension lamination under one load point. The final break (31,300 pounds) was between centerline of the beam and the opposite load point in the tension lamination. The break went through or by three knots: a 11/16-inch knot 6 inches, a 11/16-inch knot 16 inches, and a '/z-inch edge knot 18 inches distant from the centerline. At end of the test, compression was found in the top lamination under both load points. This beam was very tenacious, with continual cracking noises from first to final failure. MOR was 5,910 psi. 3-S failed at a 15/8-inch edge knot in the tension lamination 20 inches from centerline of the beam. Six inches away from the edge knot, there was a failure in the second lamination at a 1'h-inch edge knot. MOR was 3,950 psi. 9

Figure S. Typical failures of hem-fir beams, side views. The combination number is indicated. 10

APPENDIX Steering Committee The following were members of the Steering Committee for the study in 1972 on the design, fabrication, and testing of large glued-laminated timber beams made from visually graded hem-fir lumber: Thomas E. Brassell, Chairman.............. AITC Wayne Bauder............. Riddle Laminators Bob Boehm........... Woodlam, Incorporated Richard Calletti......... Standard Structures, Incorporated Bob Halford............. Duco-Lam, Incorporated Clyde Hughes............ Rosboro Lumber Company James W. Johnson........... Oregon State University Cal Luding.............. Weyerhaeuser Company Neal Pinson......... Western Wood Products Association W. P. Smith.............Tumac Lumber Company Jim Spencer............ Boise Cascade Corporation Ray Todd.......... West Coast Lumber Inspection Bureau II