EVALUATION OF METHODS OF ASSEMBLING PALLETS U. S. D. A. FOREST SERVICE RESEARCH PAPER FPL 213 1973 U. S. DEPARTMENT OF AGRICULTURE FOREST SERVICE FOREST PRODUCTS LABORATORY MADISON,WIS.
ABSTRACT The performance of pallets assembled by various methods--with nails, staples, or synthetic elastomeric adhesives--was evaluated by testing complete pallets by the free- fall- on- corner test and by testing pallet comers statically and dynamically (impact). In addition to the assembly methods, other variables investigated were species and moisture content at time of assembly. The species were red oak, Douglas- fir, hickory, yellow-poplar, and aspen. Moisture contents at time of assembly ranged from 12 percent (dry) to more than 30 percent (green). Hardened pallet nails, slightly thinner and shorter than the standard helically threaded pallet nail, performed better than did the standard nail, especially in the impact tests of pallet comers. Results of tests of both the corners and the complete pallets confirmed that the slender nails were superior to the standard pallet nail. Pallets assembled with pneumatically driven plastic-coated staples (2-1/2-in. legs, 15-gage wire) performed adequately as did the pallets assembled with two of four synthetic elastomeric adhesives.
EVALUATION OF METHODS OF ASSEMBLING PALLETS 1 By R. S. KURTENACKER, Engineer Forest Products Laboratory,- 2 Forest Service U. S. Department of Agriculture -- -- INTRODUCTION Fabricators of pallets experience difficulties driving standard pallet nails into dense woods that are relatively dry (less than about 22 pct. moisture content). Nails with diameters of 0.120 inch split the ends of 1-inch-thick deck boards, and the standard pallet nails bend objectionably when driven into dry hardwoods. Nail manufacturers some years ago, recommended a hardened nail slightly thinner (0.110-in. diameter) and slightly shorter (2-1/4 in.) than the standard nail for these conditions. They claimed the new nail would minimize splitting and bending. During recent years, advancements have been made in driving fasteners pneumatically. Among these was the development of a 2-1/2-inch-longstaple from 15-gage wire that could be driven into dense species, especially at high moisture content. Interest has been shown in the possible use of synthetic elastomeric adhesives. A pallet assembled with this type of material and capable of adequate performance would eliminate some of the problems with nailed pallets such as splits at the nails and nail popping, which often occur if nailed pallets assembled with the wood at high moisture content dry out in service. To obtain information on these fastening systems that might be of benefit to pallet producers, the U.S. Forest Products Laboratory evaluated both complete pallets and pallet corners assembled with the different fastening methods. Five species of wood were included as were variations in moisture content. The pallet corners were evaluated both statically and dynamically. The complete pallets were subjected to rough handling by dropping vertically on the same corner from a height of 40 inches six times. Mechanical Fasteners MATERIALS The following four different mechanical fasteners were used: (1) A nail with a wire diameter of 0.120 inch, a length of 2-1/2 inches, and shanks helically threaded with five flutes at a pitch angle of about 69 to the nail axis (standard pallet nail). 1 Part of the work on nails was reported in Forest Prod. Lab. Rep. No. 2238, "Performance Comparison of Slender and Standard Spirally Grooved Pallet Nails," by T. B. Heebink. (Now out of print.) 2 Maintained at Madison, Wis., in cooperation with the University of Wisconsin. -1-
(2) A nail also with a wire diameter of 0.120 inch and a length of 2-1/2 inches, but shanks that were annularly threaded. (3) A nail, referred to as a "slender hardened nail," with a wire diameter of 0.110 inch and a length of 2-1/4 inches. The shanks were helically threaded with four flutes at a pitch angle of about 61 to the nail axis. (4) A mechanical fastener, a staple, with 2-1/2-inch-long chisel-pointed legs. The staples had a 7/16-inch crown, and were made from 15-gage galvanized steel wire. The lower 1-1/8 inches of the legs were coated with a plastic material. Synthetic Elastomeric Adhesives During recent years, a number of synthetic elastomeric adhesives have come to our attention. Of these, four (designated A, B, C, and D) were selected for evaluation. A and B contained different synthetic elastomers; both were formulated without any organic solvent vehicle. C and D also contained different synthetic elastomers; both were characteristic of conventional construction adhesives formulated with an organic solvent vehicle. All of the four exhibited good adhesion to wood, some degree of resiliency, and gap-filling characteristics. Adhesive A was easy to apply and use, but had a rather long cure time. B was a two-part material that required accurate mixing, and had a limited pot life. C was a neoprene structural adhesive. D was a conventional construction adhesive that conformed to American Plywood Association specification AFG-01 for field gluing plywood to wood framing. Species Wood species used covered a wide range of densities, and included the following: Red oak (Quercus rubra L.), Douglas-fir (Pseudotsuga menziesii var. menziesii), hickory (Carya Nutt.), yellow-poplar (Liriodendron L.), and aspen (Populus sp.). Representative of lightweight, low-density species were aspen and yellow-poplar, whereas red oak and hickory represented heavy high-density species. Between these was Douglas-fir, a popular pallet wood in the west coast area. The evaluations did not involve every combination of fasteners and species. Pallet Corners SPECIMENS To evaluate the different fastening systems, pallet corners for both static and impact tests were fabricated as shown in figure 1. When assembled with staples, six fasteners were used per joint instead of three nails (fig. 1). Nails and staples driven into opposite edges of the stringers were staggered so that there was no danger of the points meeting in the center of a 2 by 4. Corners were rounded to minimize compression of the wood at the bearing points and to insure that the deflections measured during and after loading more accurately represented deformation or rack in the joint. When synthetic elastomeric adhesives were used to assemble the specimens, small wood shims were used in the joints to control the various glueline thicknesses. Specimens were fabricated at various moisture conditions ranging from dry (12 pct. moisture content or less) to green (more than 30 pct. moisture content) (tables 1 through 10). FPL213-2-
Figure 1.--Pallet corner used to evaluate fastening systems. M 136 684-3-
Table 1.--Static compression performance of pallet corners--hickory FPL 213-4-
Table 2.--Static compression performance of pallet corners--red oak -5-
Table 3. --Static compression performance of pallet corners --Douglas-fir FPL 213-6-
Table 4. --Static compression performance of pallet corners--yellow-poplar -7-
Table 5.--Static compression performance of pallet corners--aspen FPL213-8-
Table 6.--Impact performance of pallet corners--hickory -9-
Table 7.--Impact performance of pallet corners--red oak FPL 213-10-
Table 8.--Impact performance of pallet corners--douglas- fir -11-
Table 9.--Impact performance of pallet corners--yellow- poplar FPL 213-12-
Table 10.--Impact performance of pallet corners--aspen -13-
Complete Pallets Reusable pallets--both reversible and nonreversible, three-stringer, 40- by 48-inch, twoway entry, double-faced, flush stringer--were assembled with the various fastening systems and species. All pallets were fabricated from green lumber except those assembled with adhesive B, which were fabricated from dry lumber. Pallet specimens did not include all combinations of species, fasteners, and pallet styles. The reversible pallets were similar to the pallet in figure 2, whereas the nonreversible were similar to that in figure 3. Nails and staples were spaced and staggered in the pattern used for the pallet corners. With staples, six fasteners were used for each joint in a leading edge deckboard and four staples in each joint for interior deckboards. Small wood shims were used in the joints to control the glueline thickness. Figure 2.--Reversible pallet used in free-fall Figure 3. --Nonreversible pallet used in free -on-cornertest. M 141 112 -fall-on-cornertest. M 141 113' Conditioning All specimens were subjected to a period of conditioning to simulate the changes in moisture content in wood that occur in pallets during use. Both the pallet corners and the complete pallets fabricated from wood at 22 percent moisture content or higher were placed either in an atmosphere of 80 F. and 30 percent relative humidity until reaching an equilibrium moisture content of about 7 percent or in an atmosphere of 68 F. and 65 percent relative humidity until reaching an equilibrium moisture content of about 12 percent. Some of the oak and the Douglas-fir specimens fabricated at 12 percent moisture content were first conditioned in an atmosphere of 80 F. and 90 percent relative humidity until reaching an equilibrium moisture content of about 20 percent. Then they were dried to about 7 percent moisture content before testing. The other specimens fabricated from dry material were stored at ambient room conditions for a time approximately equal to that of the conditioning of the other specimens. FPL 213-14-
TEST PROCEDURES Tests of Pallet Corners The pallet corners were subjected to either a static or an impact compressive force applied at the apex (fig. 1). The static compressive force was applied by a universal testing machine operating at a head speed of 0.3 inch per minute (fig. 4). A 45-pound hammer falling freely between guides from successively increasing height increments of 1 inch (fig. 5) generated the impact compressive force. Figure 4.--Setup for test of pallet corner in a universal testing machine. ZM 120 231 Complete Pallets The complete pallets were subjected to the free-fall-on-corner test. This test indicates the resistance of the various fastening systems to dynamic racking stresses in the plane of the pallet deck. Each pallet was given six falls on the same corner from a height of 40 inches (fig. 6). Before test and after each drop, measurements were made of each diagonal of both top and bottom decks. The average amount of racking was then calculated from the change in these four measurements. -15-
Figure 5.--A pallet corner tested by impact compressive force. The hammer weighs 45 pounds, is raised with an electromagnet, and when dropped impacts the top of the specimen with its flat bottom surface. FPL 213-16- ZM 120 233
Figure 6.--Setup for free-fall-on-corner test of a 40- by 48-inch wood pallet. Each specimen was dropped from a height of 40 inches six times on the same corner. ZM 120 230-17-
DISCUSSION AND RESULTS Pallet Corners The results of the static and the impact loading of the pallet corners are given in tables 1 through 10. For the most part, with the oak, the Douglas-fir, and the aspen corners, the slender hardened nail performed as good as or better than the standard pallet nail. In the static tests, the corners with the slender hardened nail generally exhibited less deformation at maximum load than did comparable corners with the standard pallet nail. In the hickory pallet corners, those with the annularly threaded nail outperformed those with the standard pallet nail. This was undoubtedly due in part to less initial splitting than that associated with the helically threaded standard pallet nails. Failures of specimens with annularly threaded nails often involved broken nails, whereas the standard pallet nails did not break. Specimens with standard pallet nails failed by splitting of deckboard or of stringer at the nails or by splitting of both and by the nails shearing from the deckboard, pulling out of the stringer, or pulling through the deckboard. No tests were made with the slender hardened nail on hickory comers. With the mechanical fasteners, the average maximum static load was generally influenced directly by wood density; as the density decreased so did the average maximum static load. This appears reasonable because density directly influences withdrawal resistance of nails and staples.- 3 This influence of density was not noticeable in the impact-loading test of pallet corners. The plastic-coated staples did not perform as well as nails in the static loading of oak pallet comers, but performed better than nails with the Douglas-fir and aspen corners. In the impact tests of corners, the plastic-coated staples performed as good as or better than the nailed comers. Although the comparisons that can be made with adhesive B are limited, its performance in the static corner tests was consistently good. It was generally as good as or better than mechanical fasteners except when used with hickory and possibly yellow-poplar. In impact tests, corners assembled with this material outperformed all other fastener systems, However, other research has shown that moisture content of the wood is more critical with synthetic elastomeric adhesive B than with some of the other materials used here. 4 - Adhesive A was used only in hickory and yellow-poplar corners. It performed erratically, and no marked influence could be attributed to variations in joint thickness. Its performance with dry wood was poorer than that with green. In the static tests, this synthetic elastomeric adhesive performed somewhat below the level of mechanical fasteners. In the impact tests, however, its level of performance was generally somewhat better than that of the mechanical fasteners. The type of failure that occurred with both synthetic elastomeric adhesives A and B was influenced by wood density. With the low-density species, such as Douglas-fir, yellow-poplar, and aspen, there was a considerable amount of wood failure (fig. 7), whereas with high-density species, such as oak and hickory, a cohesive failure was characteristic (fig. 8). 3 Scholten, John A. Strength of Wood Joints Made With Nails, Staples, or Screws. U.S. Forest Serv. Res. Note FPL-0100. Forest Prod. Lab., Madison, Wis. 1965. 4 Kurtenaeker, R. S. Appalachian Hardwoods for Pallets: Effect of Fabrication Variables and Lumber Characteristics on Performance. USDA Forest Serv. Res. Pap. FPL 112. Forest Prod. Lab., Madison, Wis. 1969. FPL 213-18-
Figure 7.--Wood failure typical of that associated with synthetic elastomeric adhesives A and B (contained no solvents) when used with low-density species. M 136 912 Figure 8.--Cohesive failure typical of the synthetic elastomeric adhesives A and B (contained no solvents) when used with high- density wood. M 133 045-19-
Limited evaluations were made of adhesives C and D in oak corners. The performance was generally inferior to the other fastener systems, especially in the impact loading of the pallet corners. These two adhesives contained solvents that evaporated during curing, and caused shrinkage of the adhesive that resulted in voids and honeycombing (fig. 9). This undoubtedly reduced the satisfactory performance of the adhesives. Because of the unsatisfactory performance of these two materials, especially in the impact-loading evaluation, they were not included in the tests of complete pallets. Although the pallet corners assembled with various systems were evaluated both statically and dynamically (impact), the impact evaluations are reasoned to be more meaningful. This is because pallets seldom fail during static loading such as occurs in storage or in a warehouse. While being used as devices for handling materials, pallets are rough-handled and subjected to impacts; thus failures are more often caused by dynamic rather than by static loading. Figure 9.--Honeycombing typical of that associated with both synthetic elastomeric adhesives C and D (contained organic solvents). M 139 677 FPL213-20-
Complete Pallets The results of the free-fall-on-comertests of complete pallets are shown in figures 10 through 16. The synthetic elastomeric adhesives A and B outperformed the mechanical fasteners regardless of whether the pallets were made from a high-density species such as hickory or lowdensity species such as yellow-poplar or aspen. These pallets showed no residual deformation after six falls on the same corner from 40 inches (figs. 10, 15, and 16). In both the reversible and the nonreversible oak pallets as well as the nonreversible aspen pallets, the slender hardened nail outperformed other mechanical fastener systems such as the standard pallet nail and the plastic-coated staples (figs. 11, 12, and 16). It is also apparent that the performance of pallets assembled with standard pallet nails differs little from those assembled with the annularly threaded nails (figs. 10 and 15). Although many comparisons can be made with the information presented here, comparing either species or reversible and nonreversible pallets should not be attempted because of the difference in average weights. These differences in weight undoubtedly influence the results of a test that involves gravity as does the free-fall-on-corner test. Energy or the ability to do work is proportional to mass (weight divided by acceleration due to gravity). Thus the heavier the pallet, the more work will be done on it in any one fall from a given drop height such as 40 inches. This can be seen by comparing figures 11 and 12 in which the same fastener methods are used to assemble similar pallets except one is reversible and has an average weight of 72 pounds, whereas the other is nonreversible with an average weight of only 55 pounds. The 55-pound pallets, being lighter, have less work done on them during any one fall. The curves for these nonreversible pallets indicate less distortion for the same drop increment than do those for the reversible pallets assembled by the same fastener system. In-Service Pallet Performance In recent years, it has Seen possible to evaluate in-service performance of adhesiveassembled pallets and staple-assembled pallets. In the one situation, conventionally nailed pallets and pallets assembled with synthetic elastomeric adhesive A were placed in service in a brewery. In another pallets were evaluated by a cement block and brick company. The pallets were assembled with plasticcoated staples and synthetic elastomeric adhesive A as well as standard pallet nails. After 2 years service in the brewery, there was no evidence of any pallet (nailed or glued) requiring repair or removal from service. Some of the pallets were examined, and all appeared serviceable although they did show signs of having been in service. The pallets at the cement block and brick company were used outdoors in all types of weather; at the end of 1 year there were no complaints about their performance regardless of the assembly method. Visual examination indicated that the pallets assembled with the synthetic elastomeric adhesive A needed more repair than did those assembled with staples or pallet nails. The general appearance of all of the pallets examined indicated they had been exposed to a severe rough-handling environment. -21-
Figure 10.--Results of free-fall-on-corner test of reusable hickory pallet--three stringer, 40- by 48-inch, two-way entry, double-face, nonreversible, flush stringer. (Average weight of pallet, 74 lb.; each point is average of three pallets. Adhesive A, synthetic elastomeric without solvent vehicle.) M 141 114 Figure 11.--Results of free-fall-on-corner test of reusable red oak pallet--three stringer, 40- by 48-inch, two-way entry, double-face, nonreversible, flush stringer. (Pallets assembled at 22 pct. moisture content, dried before test. Average weight of pallet, 55 lb., each point is average of two specimens.) M 141 115 FPL 213-22-
Figure 12.--Results of free-fall-on-corner test of reusable red oak pallet--three stringer, 40- by 48-inch, two-way entry, double-face, reversible, flush stringer. (Pallets assembled green, dried before test. Average weight of pallet, 72 lb.; each point is average of three specimens.) M 141 116 Figure 13.--Results of free-fall-on-comer test of reusable oak pallet--three stringer, 40- by 48-inch, two-way entry, double-face, nonreversible, flush stringer. (Pallets assembled green, dried before test. Six staples used per joint for edge deckboards, four per joint for intermediate deckboards, Average weight of pallet, 68 lb.; each point is average of five pallets.) M 141 117-23-
Figure 14.--Results of free-fall-on-corner test of reusable Douglas-fir pallet-- three stringer, 40- by 48-inch, two-way entry, double-face, nonreversible, flush stringer. (Pallet assembled and tested dry. Six staples used per joint for edge deckboards, four per joint for intermediate deckboards. Average weight of pallet, 56 lb.; each point is average of three specimens.) M 141 118 Figure 15.--Results of free-fall-on-corner test of reusable yellow-poplar pallet-- three stringer, 40- by 48-inch, two-way entry, double-face, nonreversible, flush stringer. (Average weight of pallet, 46 lb.; each point is average of three pallets. Adhesive B, synthetic elastomeric without solvent vehicle.) M 141 119 FPL 213-24-
Figure 16.--Results of free-fall-on-corner test of reusable aspen pallet--three stringer, 40- by 48-inch, two-way entry, double-face, nonreversible, flush stringer. (Pallets with metal fasteners assembled green, dried before test. Pallets with synthetic elastomeric adhesive B (without solvent vehicle) assembled and tested dry. Average weight of pallet, 44 lb.; each point is average of three specimens.) M 141 120-25-
CONCLUSIONS Although not all combinations of each variable were evaluated, the work summarized here yields a number of conclusions, some of them are listed in the following: (1) The 2-1/4- by 0.110-inch slender, hardened, helically threaded pallet nail performs better than does the 2-1/2- by 0.120-inch standard helically threaded pallet nail when used in pallets. (2) Pallets assembled with 2-1/2- by 0.120-inch helically threaded nails and those assembled with 2-1/2- by 0.120-inch annularly threaded nails can be expected to perform equally well. Annularly threaded nails tend to cause less splitting, but they also tend to break more frequently, especially with the more dense species, than does the helically threaded standard pallet nail. (3) In pallet corners assembled with mechanical fasteners, density of the wood has less influence on performance when subjected to impact loading than when subjected to static loading. (4) Performance characteristics of the synthetic elastomeric adhesive B warrants consideration, but its versatility is limited because of its inadequacy when used with green wood. (5) Although adhesive A apparently works with either green or dry wood, the performance of adhesive A was somewhat erratic. However, it may warrant consideration for use in certain pallet applications: For example, if a moderate-handling environment exists and if damage could result from protruding nailheads caused by subsequent shrinkage of pallet parts. However, in actual pallet use it may not be entirely suitable for extremely severe rough-handling environments. (6) Adhesives containing solvents, such as synthetic elastomeric adhesives C and D that evaporate during curing and cause shrinkage of the adhesive are not apparently suitable for pallet applications. (7) Plastic-coated staples of at least 15-gage wire with at least a 7116-inch crown and 2-112-inch-long legs apparently warrant further consideration for pallet applications. (8) In pallet tests involving the influence of gravity, wide differences in weight caused by variations in density or type of pallet have some influence on the results. Trade or proprietary names are included for the benefit of the reader and do not constitute endorsement by the Forest Service, U.S. Department of Agriculture. FPL213-26-
APPENDIX An exploratory series of tests was conducted to correlate a bending test with the performance of pallet nails in the impact test of simulated pallet corners. Six types of helically threaded pallet nails (fig. 17, A-F), about the same size, were included. Two types, A and F, corresponded to two of the nails evaluated in the body of this report. Nail A was the 2-1/2- by 0.120-inch helically threaded standard pallet nail; F, the 2-1/4- by 0.110-inch helically threaded slender hardened nail. Nail B was the same as A, but was produced by a different manufacturer. C was a hardened nail, similar to F, but its diameter (0.120 in.) was the larger of the two. Two twisted square stock nails were included; D had a minimum pitch angle, about 16 to the axis, and E was hardened and had a blunt chisel point. Figure 17.--Six pallet nails tested in bending, and compared for carbon content, hardness, and performance in pallet comers subjected to impact loading. ZM 120 614 Five pallet comers were constructed and subjected to an impact compressive force applied at the apex (figs. 1 and 5). The specimens were oak, fabricated green (more than 30 pct. moisture content) and dried to 7 percent moisture content at the time of test. The heights of drop to cause 1/2-inch deflection are presented in table 11. Samples identical to the nails used in the pallet comers were sent to the Joliet Works of the American Steel and Wire Division of the U.S. Steel Corporation to determine carbon content and hardness values (table 11). Three each of the six nail types were tested in bending, center-loaded on a 2-inch span. The machine was a universal hydraulic type that operated at a head speed of 0.07 inch per minute. The load-deflection characteristics of each nail are presented in figure 18 and the maximum loads to failure in table 11. -27-
Several conclusions can be drawn from the data in table 11, which may be significant to future research on pallet nails. (1) Although nail C performed best in the impact test and had the highest carbon content, this type of correlation is poor for the rest of the nails. The nail with the lowest carbon content is the second best in the impact test. It appears that carbon content is not a good indicator of performance. (2) The two best-performing nails in the impact tests, C and E, are hardened. Nail F is also hardened, but of a smaller diameter than C and E. Correlation is good between hardness and the impact performance of pallet corners. (3) The results of the bend test, as expressed by maximum loads, indicate good correlation with the impact performance of the simulated pallet comers. This means that because hardness tests are difficult to perform on pallet nails, perhaps the bend test is a more logical test to be considered in developing performance specifications of pallet nails. Table 11.--Properties of six types of pallet nails FPL 213-28- 3.5-29-8-73 U.S. Government Printing Office 754-546/35
Figure 18.--Load-deflection curves for six different types of helically threaded pallet nails (fig. 17), which were tested in bending, center-loaded on a 2-inch span. (Each curve represents average of data from three tests.) M 141 121