Forintek Canada Corp.
Western Division
2665 East Mall
Vancouver, B.C.
V6T 1W5
Machining, Laminating, Fastener Withdrawal and Finishing
Properties of Hybrid Poplar
by
Derek Williams
prepared for
Ministry of Forest Research Branch
Kalamalka Research Station
Vernon, B.C.
March 1998
For their contribution to the planing and sanding testing the author wishes to acknowledge the FRBC Valued-Added Skills Centre, Abbotsford, for use of their facilities, equipment, and support from Skills Centre staff, including Peter Vogt, Norbert Milota, and Stewart Buckman. For their work in the fastener withdrawal tests, strength tests and determination of density and ring counts the author wishes to acknowledge Jules Gardy, Rollie Wakeman and Roy Abbott of the Wood Engineering Department of Forintek Canada Corp. For their work with the laminating properties the author wishes to acknowledge Axel Andersen of the Composites Department and Elias Mucha of the Value Added Technology Transfer Program at Forintek Canada Corp. Finally, the author wishes to acknowledge Stu Eaton and Ed Santella of Cloverdale Paints Ltd. and Dave Lick, chief instructor at BCIT's Painting and Decorating Trades Department for their contribution to the paint and furniture finishing testing and evaluations.
Table of Contents
This study examines the utility of using hybrid poplar in typical woodworking processes. The following three processes were considered: machining (including fastener withdrawal), laminating and finishing. In addition the effects of kiln drying on shrinkage and warp and general strength properties of hybrid poplar were also evaluated.
The machining operations studied were those encountered in commercial secondary wood processing (value-added) plants, including planing, sanding, boring and shaping. The machining tests were conducted according to American Society for Testing and Materials (ASTM) D 1666-87: Standard Methods for Conducting Machining Tests of Wood and Wood-Based Materials. The fastener withdrawal tests determine the force necessary to withdraw two types of fasteners (nail and screw) from a number of samples. The fastener withdrawal tests were conducted according to ASTM D 1761-88: Standard Test Method for Mechanical Fasteners in Wood. Density (or specific gravity) and ring per inch were determined for approximately half of the boards used in the machining tests. Knowing these values may lead to insights on how variability in these parameters affects the machining, laminating, finishing and strength properties of hybrid poplar.
The laminating tests study the performance of hybrid poplar when four different types of adhesives are used in conjunction with three different types of glue press systems: cold set press, hot platen press and a radio frequency press used in laminating sample blocks. The resulting samples were tested for delamination and shear strength of the glued joint according to the ASTM D 1101: Standard Test Methods for Integrity of Glue Joints and D 905: Standard test Method for Strength Properties of Adhesive Bonds respectively.
The finishing properties of hybrid poplar were evaluated using three types of finishes: paint, furniture type of finishes and wax finishes. The paint tests are typical paint industry tests for the adhesion of the coating to a substrate and a gloss test. The following ASTM test procedures were used: D 4541 Test Method for Pull-Off Strength of Coatings Using Portable Adhesion-Tester, D 3359 Test Method for Measuring Adhesion by Tape Test and D 523 Test Method for Specular Gloss. The tests for furniture type finishes follow no formal test procedure as none exist; however, the applicability of using hybrid poplar in typical furniture finishing was subjectively evaluated by applying six different types of stains in combination with four types of sealers and six clear finish coats. The wax finish tests again follow no formal test procedure other than following the instructions provided by the wax finish manufacturer and use of subjective evaluation to judge the results.
The lumber used for the above and following tests had to be kiln dried at Forintek as a matter of course. Hence the opportunity presented itself to perform shrinkage and warp evaluations on the hybrid poplar lumber as supplied. Shrinkage and warp can affect the utility and of a wood species for particular products and depending on the grading rules used its grade determination and value. The evaluation was based on comparing the degree of shrinkage and warp prior to after kiln drying with the latter results being evaluated against standard warp tables found in NLGA Standard Grading Rules for Canadian Lumber.
The general strength properties evaluated were modulus of rupture (MOR) and modulus of elasticity (MOE). The test procedure followed to determine MOR and MOE was ASTM D 143 Standard Methods of Testing Small Clear Specimens of Timber .
It was felt that the results of these tests could be made more meaningful if hybrid poplar could be compared to other wood species or general standards if possible. As the author has recently completed a wood machining study of 15 B.C. wood species, comparing hybrid poplar with five similar species provides a sense of the relative properties of hybrid poplar. The five species selected were trembling aspen, red alder, black cottonwood, western white pine and lodgepole pine. In the laminating tests the same five species were tested along with hybrid poplar for the block shear strength values. No comparison could be made for the various finishing tests due to the lack of comparative data.
The lumber to be tested was supplied directly from Rouck Bros. Sawmill in Lumby as green lumber in two thickness sizes (1 and 2 inch), in widths ranging from 4 to 8 inches and in lengths ranging from 4 to 9 feet, the majority being in eight to nine foot length. The lumber was marked with identifying colours on the ends: red, white and blue to signify the size of log from which it was sawn. Refer to Appendix I for a list of the sizes of lumber in each colour category received. Of note was the fact that a lot of lumber (with green ends) was sent to Forintek inadvertently as it came from pulp grade logs. However not to miss out on an opportunity to test this lumber, it was agreed that if funding permitted both MOE and MOR tests would be performed on this lot of lumber as a comparison with lumber from saw logs. All the lumber once received was immediately stickered and stored indoors for a period of one month to await kiln drying.
As in the case for the hardwood tested in the previous machining study the moisture content targeted for the hybrid poplar lumber was 9%. The kiln schedule used can be characterized as being very gentle (low air velocity, low dry bulb final temperature and long duration). This schedule was used to limit drying defects such as checking, as only a limited amount of lumber was available for the other tests.
Once dried lumber from saw logs was marked for the cuttings to be made, stamped with a four-digit number for identification and sawn into pieces. Every attempt was made to obtain cuttings from each of the three coloured board ends though this was not always possible especially for the blue end boards, as they were too narrow and short.
The cuttings were stored in a conditioning chamber set at relative humidity of 65% and a temperature of 20° C to maintain a moisture content of 9%. Samples were kept in the conditioning chambers until a machining test was ready to commence. Refer to Appendix II for a cross-reference list of board number to part number. Small sections were cut from each of the 50 boards to determine specific gravity and number of rings per inch. Oven-dry weight over oven-dry volume as measured by water displacement was used to determine the specific gravity for each board.
The planer samples were further sanded to make them ready for the finishing test, which require a finer finish. The laminate samples were taken from boards used in the machining tests and from previously unused boards though no sanded surface could be used in the tests.
The mean and standard deviation of the specific gravity values of 27 hybrid poplar samples was calculated to determine how it compared to the specific gravity for the two species from which this hybrid poplar is derived, namely western black cottonwood (Populus Trichocarpa) and eastern cottonwood (Populus deltoides). The mean value of specific gravity obtained for hybrid poplar was 0.375 with a standard deviation of 0.043. The published means for both types of cottonwood are 0.334 and 0.386 respectively. The hybrid poplar samples tested clearly fell between these two values with a tendency of having a specific gravity closer to that for eastern cottonwood. Higher specific gravity values are always superior to lower values for strength considerations alone.
The mean and standard deviation for the number of rings per inch were also determined. Refer to Appendix III for a table showing specific gravity (or relative density) and rings per inch for various part numbers. Similar data from the previous machining study show western black cottonwood, red alder and trembling aspen having ring per inch mean values in the 10 to 15 range. It is evident that this hybrid poplar is a very fast growing tree with 1.48 rings per inch.
The test procedures followed the tool and machine settings as prescribed in ASTM D-1666 whenever possible. However, some settings were modified in the interest of obtaining the best possible surface quality. These modifications are discussed separately under the "Procedure" heading for each test.
The wood machining tests were performed on two different samples cut from each board, as specified in the ASTM standard though this became difficult to accomplish with narrow boards. The planer test samples were each 36 in. long and 4 in. wide. The shaping test samples were each 12 in. long and 3 in. wide. The boring tests were performed on the same piece of wood as the shaper test.
The tests examine the surface quality of the machining operations and this was done both visually and by touch. The ASTM D1666 standard uses five quality grades based on the amount and severity of defect present, as follows:
Table 1 lists each machining test and the quality grades used in determining overall performance.
Table 1 Quality Grades Used in Determining Overall Performance for each Machining Test
|
Machining Test |
Performance Criteria |
|
Planing |
Grade 1 |
|
Sanding |
Grade 1 |
|
Boring |
Grades 1 and 2 |
|
Shaping |
Grades 1 and 2 |
The practice of using different grades or combination of grades for each test may appear to be inconsistent. However, the grade combinations are those that are considered suitable for most purposes found in secondary wood processing. Samples of each grade and type of defect were kept for reference.
Planing is second to sawing as the most important machining operation in a wood processing plant, since all lumber has to be dressed to size and/or surfaced prior to further use. Accordingly, this study has placed more emphasis on the planing test than the other machining tests. Planing provides an excellent opportunity for adding value to a product. This machining process can be performed by a planer or the more versatile moulder.
The planing test was conducted on a Weinig through-feed moulder with five spindles. Only the top spindle was used in the planing tests. The machine had a variable feed rate with a spindle rotation speed of 6000 rpm.
Two hydraulically clamping cutter heads were used: one with a 12° hook angle (also known as the rake or cutting angle) and the other with a 20° hook angle. The corrugated back knives were cut from individual bar stock, balanced to within 0.2 grams, and aligned within the cutter head. All knives were the industry standard high speed steel (HSS), M-3 type. Each knife was carefully ground using an aluminum-oxide wheel to a consistent cutting circle on a profile grinder. The knives were ground to an extremely keen edge through several lapping passes. The planing tests compared two hook angles and four numbers of knife marks per inch (KMPI).
4.1.2 Measurement of Surface Quality
The number of KMPI is often used as a measure of surface quality in planing. It is determined by the feed speed, spindle rpm and number of knives making the final cut. As the latter two parameters were not varied during the tests, altering the feed speed was the only way of changing the number of knife marks per inch. Table 2 shows the number of knife marks per inch typically found in various wood products:
Table 2 Comparison of Knife Marks per Inch by End Use
| Product | KMPI |
| Lumber | 4 to 8 |
| Exterior Wood Products | 8 to 12 |
| Millwork | 12 to 16 |
| Furniture | 16 to 20 |
4.1.3 Procedure
Five passes were made with each sample having a 0.1 inch (2.5 mm) depth of cut under different sets of conditions. The first four runs used a 20° hook angle cutter with four feed speeds that yielded 8, 12, 16 and 20 knife marks per inch. The fifth run used a 12° hook angle cutter that produced 20 knife marks per inch. The number of knife marks per inch was determined by running a few test pieces through the moulder, and adjusting the feed rate until the desired number was repeatedly measured. Table 3 lists the machining parameters for each planing test. All the samples were run butt to butt to eliminate the occurrence of snipe and related feeding problems that can result in burn marks and subsequent overheating of the knife edges. All specimens of a common thickness were run one after another, and then the machine height setting was changed to accommodate thinner stock.
Samples were graded for the presence of fuzzy grain, raised grain, torn grain and chip marks. The sense of touch was found to be an efficient method for determining the presence and severity of raised and fuzzy grain.
Table 3 Machining Parameters for the Planing Tests
|
Run |
Hook Angle |
KMPI |
Feed Speed (Ft/min)* |
|
1 |
20° |
8 |
63 |
|
2 |
20° |
12 |
42 |
|
3 |
20° |
16 |
31 |
|
4 |
20° |
20 |
25 |
|
5 |
12° |
20 |
25 |
* Note: Non-jointed cutter block -- only one of the knives produced the final cut
Comparison of the planing properties was based on the percentage of defect-free samples in each species. Each of the five runs was evaluated separately. Table 4 provides a summary of the results for each planing test runs and an average for all five runs.
The best results for planing hybrid poplar occurred with a 20° hook angle at a feed speed of approximately 31 feet per minute which yielded 16 knife marks per inch. The most prevalent defect at all feed speeds was fuzzy grain though the severity was in nearly all the cases, grade 2 (good) and it could be considered as being light fuzzy grain. However, the fuzzy grain was completely removed by sanding. If the criteria for evaluation hybrid poplar was to include good to excellent samples (defect free), the average number of samples pieces in each run meeting the standard would have been 93%. These results would indicate that hybrid poplar can be planed successfully if the tooling is kept sharp and it has a hook angle between 12º and 20º and that light sanding is required afterwards to remove minor fuzzy grain.
Table 4 Summary of Planing Tests - Percentage of Defect-Free Samples
Species |
Run 1 |
Run 2 |
Run 3 |
Run 4 |
Run 5 |
Average |
|
% |
% |
% |
% |
% |
% |
|
Lodgepole Pine |
100 |
100 |
100 |
92 |
94 |
97 |
W. White Pine |
90 |
74 |
86 |
94 |
96 |
88 |
Black Cottonwood |
72 |
49 |
32 |
26 |
58 |
47 |
Red Alder |
80 |
96 |
92 |
96 |
88 |
90 |
Trembling Aspen |
66 |
90 |
46 |
96 |
100 |
80 |
Hybrid Poplar |
69 |
53 |
72 |
57 |
62 |
63 |
Run 1: 20° Hook Angle, 8 KMPI
Run 2: 20° Hook Angle, 12 KMPI
Run 3: 20° Hook Angle, 16 KMPI
Run 4: 20° Hook Angle, 20 KMPI
Run 5: 12° Hook Angle, 20 KMPI
4.2 Sanding
Sanding is a method of preparing the wood surface for the application of a finish coating (e.g., sealer, stain, oil, and lacquer). Sanding wood properly is the first step in applying a first class finish. Mistakes made in sanding often show up after the finish is applied.
Sanding is an abrasive action that leaves a scratch pattern on the wood surface, and is directly influenced by the grit size of the sandpaper. To remove excessive scratch patterns, progressively finer grit sandpaper is used until the desired surface smoothness is reached.
This test was conducted on a CEMCO wide belt sander with two heads. The belt sequence was a 80 grit, cloth-backed aluminum oxide belt on the first head and a 120 grit, cloth-backed aluminum oxide belt on the second head. The feed rate was adjusted to 20 ft/min (6.1 m/min).
The planing samples were used for the sanding tests though they were not cut down to a 12 inch length as recommended in the ASTM standard but left in the 3 foot length as they were to be used in the finishing tests as well. In order to use the whole width of the belt, the sanding specimens were staggered across the belt.
The basis of comparison for sanding properties was the percentage of fuzzy grain present in each sample. The percentage of defect free-pieces samples was 96 % with the remaining 4 % of the samples being in grade 2 or good category. Figure 1 shows how well hybrid poplar did when compared to the other five species. It is evident that hybrid poplar sands well.
Figure 1. Comparative sanding performance
Shaping is similar to planing with respect to cutting action and type of tooling. The major difference is the shaper's ability to profile curved pieces of wood, whereas the moulder/planer can only profile straight lumber. A shaper is a versatile machine and can produce a variety of cuts (grooves, rebates, profiles, dados, etc.).
The shaping test was conducted on a SCM T120C single spindle shaper operated at a spindle speed of 7200 rpm. The tooling used was a six pocket (wing) cutter head with removable HSS knives though only three of the pockets contained a knife (this was a specially ordered cutter head with three 10° and three 20° pocket angles). Each pocket in the cutter head was numbered as were the knives and corresponding gibs. Shaping tests were conducted using the 10° knives on the hardwoods and 20° knives on the softwoods. The knives were ground to the same profile shown in the ASTM standard.
A jig was designed to hold the sample in place while shaping. The jig was similar to that described in the ASTM standard, with the addition of toggle clamps to help secure the workpiece. The workpiece was fed by guiding the jig against a bearing mounted underneath the cutter head. This set-up produced duplicate curved patterns with every pass.
Prior to being shaped, each sample was bandsawn to a pattern using a template. Two passes were made on the shaper: a preliminary pass that removed a lot of material followed by a second pass that removed the required 1/16
inch (1.6 mm) of material. This was accomplished by placing a 1/16th inch strip of veneer between the sample and the jig's base. All the cutting action was against the grain as is the usual way of operating a shaper safely.
All samples were examined visually and by touch. Samples were graded for the presence of fuzzy grain, raised grain, torn grain and end grain tear out. As in the evaluation of planing pieces, the sense of touch was found to be a more efficient method of determining the presence and severity of raised and fuzzy grain.
Shaping properties were based on the percentage of good to excellent samples present. For hybrid poplar, the value was 96% which indicates that it is a good species to shape however, this must be qualified to some extent. It was observed that no appreciable shaping defects occurs in the straight portion of the cut but as the cut started to go across the grain in the curved portion rough end grain appeared. This defect was categorized as being to a slight degree (grade 2) but it occurred in 60% of the samples. Fuzzy grain was the next most prominent defect occurring in just 12% of the samples again to a slight degree. The percentage of defect-free samples was 17%. Figure 2 shows how hybrid poplar compared to the other five species. It is on par with alder and greatly exceeds the performance of black cottonwood.
Figure 2. Comparative shaping performance
4.4 Boring
Boring like shaping is performed extensively in furniture and other woodworking plants. Boring is performed to create holes for the insertion of dowels or attachment of metal hardware
Boring holes into wood can serve two purposes: (1) to receive dowels and (2) to pre-drill "pilot" holes for insertion of screws or other type of fastener. The bored hole must be round without any noticeable distortion and have an inner surface conducive for good glue bonding (free of torn and crushed grain). Dowels serve to (1) reinforce the joint and (2) accurately position adjacent parts, thereby allowing fast assembly. Boring is extensively used in the manufacture of RTA furniture.
Two bits were used in the boring tests: (1) a single twist, solid-centre point bit (referred to as single twist bit from here on) as stated in the ASTM standard, and (2) a brad and lip point bit (referred to as brad point bit from here on). The single twist bit was similar to an auger style bit. The brad point bit differed from the single twist bit in that it had two flutes (twists) instead of one, and two lips ground into the ends of its flutes. The lips in the brad point bit act as spurs severing wood fibres before they are cut. Both styles are common in use however, the brad point bit is normally regarded as producing a more accurate and clean hole. The brad point bit was added to the test, as it was more representative of current industrial practices in B.C.
Adaptations to the ASTM recommended test procedure were necessary, since the recommended spindle speed of 3600 rpm was found to cause burning of the wood and overheating of the boring bits. Furthermore, the recommended single twist bit was found to be too aggressive in cutting (its threaded centre-point pulled the wood into the bit too fast). After a number of trials, a spindle speed of 1200 rpm was found to be satisfactory for both bit types. The removal of threads by grinding reduced the pulling tendency of the single-twist bit.
The boring tests were conducted after the shaping test using the same samples.
The boring machine used was an Ashina bench drill press with a ½ hp motor. The feed rate was approximately 1 inch in 6 seconds.
The brad point bit was used to bore the first hole into each of the wood samples. This hole was placed at the pointed end of each sample. A jig was made so that the holes were placed in approximately the same location for each specimen. A removable wooden base board was used under the test pieces, which was replaced when worn.
The same procedure was replicated using the single twist bit. The two holes were spaced at least 2 inches (50 mm) apart and away from any edges.
All samples were examined with the aid of a magnifying glass. The bored holes were graded for the presence of crushed, fuzzy, or torn grain, and general smoothness of cut.
Comparisons of boring properties using the two types of boring bits were based on the percentages of excellent and good samples present in each species. The brad point bit outperformed the single twist bit by a wide margin, 68% to 4% respectively for defect-free samples and 80% to 8% respectively for defect-free and good samples. This performance improvement is due to the cutting action of the brad point bit's two lips that sever the wood fibres with a shearing cut before the cutting edges remove the chip.
The most common defect for the brad point bit was tearing of the fibres followed closely by fibres being crushing up against the inside of the hole. For the single twist bit the most common defect by far was grain crushing followed by grain tearing. Crushing of the fibres can present problems if glue is to be inserted into the hole as the glue will adhere to fibres that are not firmly attached, hence failure is likely. Figure 3 shows how hybrid poplar did when compared to other species. Hybrid poplar performed the best when the brad point bit was used though with the single twist bit its performance was the poorest.
Figure 3. Comparative boring performance of single twist and brad point bits
4.5 Comparative Summary of Machining Properties
Table 5 presents a comparative summary of the results of the four machining tests performed with hybrid poplar and the other five species. It can be seen that hybrid poplar exceeds only black cottonwood in the planing test however, it must be pointed out that when the good to excellent results are considered hybrid poplar scores in the high 90's. The sanding properties of hybrid poplar are on par with the other wood species. The single twist boring results can be considered insignificant as the type of boring bit is seldom used by industry. However, it can be seen that hybrid poplar responds well to an improvement in tooling with the brad point bit. Here it scored the highest. In the shaping test hybrid poplar scored in the high 90's and it is comparable with some of the better wood shaping species from B.C.
Table 5 Summary of Comparative Machining Properties
| Species | Specific Gravity |
Planing Good to Excellent Ave. Best |
Sanding Defect- free |
Boring Good to Excellent Single Twist Brad Point |
Shaping Good to Excellent |
|---|---|---|---|---|---|
| % | % | % | % | ||
| Hybrid Poplar | 0.38 | 69 72 | 96 | 8 80 | 96 |
| Black Cottonwood | 0.39 | 47 72 | 80 | 10 29 | 33 |
| Red Alder | 0.49 | 90 96 | 100 | 25 67 | 94 |
| Trembling Aspen | 0.47 | 80 100 | 96 | 55 73 | 98 |
| Lodgepole Pine | 0.46 | 97 100 | 98 | 19 44 | 86 |
| W. White Pine | 0.43 | 88 96 | 96 | 57 65 | 100 |
The fastener withdrawal tests provide quantitative data on the force required to withdraw nails and screws. The screw test measured the maximum withdrawal force required to pull two screws from either flat or vertical grain samples (chosen at random), whereas the nail test measured the maximum withdrawal force required to pull two nails from each of the three grain orientations (tangential, radial, or end grain).
All tests were conducted on a Tinius Olsen Universal Testing Machine. This machine had an accuracy of ± 1%. A specially designed gripping device shaped to fit the base of the two types of fasteners was used to permit accurate sample placement and true axial loading. A clamping assembly was used to hold the wood sample to one platen of the machine.
The screw withdrawal test determined the maximum force required to withdraw a screw fastener driven in at right angles to the wood surface. Screws were randomly driven into either the tangential or radial face.
Fifty samples were tested, each measuring 2 in. by 6 in. with their depth at least equal to the length of the screw. The tests involved standard one inch, No. 10 gauge, flathead, low-carbon steel wood screws. A screwdriver was used to insert the screw into a pre-drilled hole bored on a drill press at right angles to the surface. Each screw was only used once.
The samples were pre-drilled to a depth of ½ inch using a #39 drill bit (0.1 inch diameter). The location of the holes was within an area ¾ inch from the edge and 1-½ inch from the ends and at least 2-½ inch apart. Samples were tested within the prescribed one hour period after driving of the fasteners. The screws were driven by hand using a screwdriver to a depth where the threads were no longer visible. This permitted the gripper to firmly secure the screw head.
Each sample was placed in the clamping assembly, the gripper secured around one of the screw heads and aligned axially. A uniform platen withdrawal rate of 0.1 inch per minute was set for the universal testing machine. The maximum force in pounds was measured for each screw.
5.1.3 Screw Withdrawal Properties
Table 6 shows the wood specific gravity and screw withdrawal force for hybrid poplar and the other five species. There is a reasonably good relationship between increasing specific gravity and required force for screw withdrawal.
Table 6 Relationship between Specific Gravity and Screw Withdrawal Force
|
Species |
Specific Gravity |
Average Force (lb) |
Standard Deviation |
|
Hybrid Poplar |
0.38 |
395 |
73.1 |
|
Black Cottonwood |
0.39 |
302 |
52.9 |
|
White Spruce |
0.40 |
347 |
49.8 |
|
Lodgepole Pine |
0.46 |
435 |
58.8 |
|
Trembling Aspen |
0.47 |
482 |
79.9 |
|
Red Alder |
0.49 |
518 |
54.5 |
Figure 4 shows a comparison of the average force required to withdraw a screw from hybrid poplar and five other species. It is clearly on par with the other hardwood species in this group and it even surpasses one of the pine species.
Figure 4. Comparative Screw Withdrawal Force
The nail withdrawal test determined the maximum force required to withdraw a nail driven in at right
The nails tested were the 6d plain-shank, diamond-point, round-wire, low-carbon-steel type. Each nail was only used once. Fifty samples of hybrid poplar were tested. Each sample measured 6 in. long and 2 in. wide and high.
Six nails were driven by hand with a hammer to a total penetration of 1 ¼ in.; two nails in the tangential face, two in the radial face, and one in each end. A jig was designed to ensure that ½ in. of nail shank remained above the surface. Nails were driven least ¾ in. from the edge, 1-½ in. from the ends, and no closer than 2 in. apart. Nails driven into the end grain were not placed in line with each other. Testing was completed within one hour after the nails were driven. The testing machine along with the gripping device and clamping assembly were the same as used in the screw withdrawal test. The maximum withdrawal force in pounds was measured for each nail.
5.2.3 Nail Withdrawal Properties
Table 7 shows the wood specific gravity and nail withdrawal force for hybrid poplar and the other five species. The highest results were obtained from nails driven into the tangential and radial faces (tangential being slightly higher on average). As expected, the end grain face produced the lowest values though the end grain face values for hybrid poplar were only about 15% lower whereas for the other species the end grain values were about 1/3 lower. Figure 5 shows a comparison of the average force required to withdraw a nail from the tangential, radial and the end grain faces for hybrid poplar and five other species.
Table 7 Relationship between Specific Gravity and Nail Withdrawal Force
|
Species |
Specific Gravity |
Average Force (lb.) |
||
|
Tang. |
Radial |
End |
||
|
Hybrid Poplar |
0.38 |
98.3 |
94.9 |
82.2 |
|
Black Cottonwood |
0.39 |
108.5 |
100.7 |
70.8 |
|
White Spruce |
0.40 |
110.7 |
100.8 |
69.1 |
|
Lodgepole Pine |
0.46 |
131.6 |
116.1 |
85.2 |
|
Trembling Aspen |
0.47 |
165.8 |
169.8 |
102.4 |
|
Red Alder |
0.49 |
198.7 |
190.5 |
153.1 |
Figure 5 shows a comparison of the average force required to withdraw a nail from the tangential, radial and the end grain faces for hybrid poplar and five other species.
Figure 5. Comparative Nail Withdrawal Force
The laminating tests determine the applicability of using hybrid poplar in glued wood products when delamination and shear strength are taken into consideration in the evaluation of a glued joint. In order to given a broader answer to the question of how well hybrid poplar performs as a glued wood product, four adhesives typically used in the secondary wood processing industry were tested in conjunction with three different types of glue presses.
The four types of adhesives used in the test were as follows: polyvinyl acetate (PVA), polyvinyl acetate with a crosslinking agent added (PVAc), urea formaldehyde (UF) and phenol resorcinol formaldehyde (PRF).
PVA is white glue typically found in woodworking applications like furniture making. PVAc and UF represent adhesives that provide better moisture resistance than PVA. They are used for furniture and non-structural building components and require mixing of two components prior to use. These adhesives can be cured at room temperature but are commonly used with heated platens or radio frequency curing to increase production rates. PRF is a high end adhesive that would probably not be used with this wood species but it is an adhesive that can provide exterior exposure durability with the capacity for room temperature curing. Like the PVAc and UF this adhesive requires mixing of two components and has to be used within a certain time period.
The three types of glue presses used were as follows: a hot platen, a cold set press and high frequency press. The hot platen press uses heat to cure the glue joint by transferring heat from the metal platen (top and bottom) to the glue joint. This type of press is well suited for plywood and veneer and thin lumber stock but it is limited in use for very thick stock as wood is a good insulator. The second type of press uses the normal room temperature to cure the glue joint providing that the temperature is in the 18 to 20° range. A clamp carrier or glue reel is an example of this type of press. This type of press is the most inexpensive and simplest to use and operate. The last type of press uses electrical energy in a wave form to cure the glue joint in a relatively short period of time depending on the size of the generator and surface area of glue line to be cured. Though this type of press promises quick curing, tight control of wood moisture content is critical if arcing problems are to be avoided. Ideally, the moisture content should not go above the 12 to 13 % range.
The two ASTM standards used to test the glued joints are D 905 and D 1101. The first test determines the shear strength of the glued joint and the second test determine the extent of delamination of the glue joint when subjected to wet and dry cycling. The shear tests determine the maximum force necessary to fracture the glue bond when a shear force is applied in compression. An examination was also made of the bonded surface and specifically how much of the joint failure was a result of wood failure and conversely, glue failure.
The force necessary to break the glue bond is influenced by three factors working in conjunction with each other. These factors are the inherent strength of the adhesive, how well the adhesive interacts with the surrounding wood fibre and the strength of the surrounding wood fibre. To give an indication of the relative shear strength of hybrid poplar in compressive shear loading, it was decided to perform a secondary test in which samples of hybrid poplar would be compared to the five selected species mentioned above. Though the shear test would be done with solid wood and not along a glue line.
The delamination test on hybrid poplar unlike the shear test was not comparative. This test determines the accumulated percentage of delamination along the glue line for each batch of samples. To be considered a good glue joint the percentage of delamination should be no more that 10% of the overall length of all the glue joints hence comparing to other wood species was not necessary.
Glued wood samples were prepared in five lots with each lot comprised of five glued blocks. The wood that made up the glue blocks were previously planed on both wide faces, then sawn in the middle so that each block would in effect be glued to itself. The first lot used a PVAc cured in a hot platen, the second lot a PRF cured by cold set, the third lot a PVA cured by cold set, the fourth lot a UF cured by radio frequency and the final lot, a PVAc cured by radio frequency. Each glued block was further divided into five smaller pieces¾ three to be used for the shear test, one to be used in the delamination test and one spare piece. A total of 75 shear samples and 25 delamination samples were tested.
The shear test as stated in the ASTM standard is intended primarily as an evaluation of adhesives for wood; however, it is often used to evaluate the strength of a glued wood joint as in this case. In a shear test the true indication of a good bond is if there is wood failure present in the fracture area. The amount of wood failure was recorded by judging the percentage of the surface area of the joint that exhibited fibres torn away from the opposing surface.
A Tinius Olsen Universal Testing Machine was used to apply a shear force in compression. Fitted to the machine was a shear tool with a self-aligning mechanism to ensure uniform lateral distribution of the load.
To test the shear strength of the hybrid poplar glue line each of the glue blocks were cut into five pieces measuring exactly 2 by 1½ inches. Each smaller sample was accurately knotched by a bandsaw in such a way that a ¼ inch segment was removed from diametrically opposite ends of each lamination. This way the shear force could be concentrated along the glue joint.
All sample pieces were numbered to keep a track of which adhesive and glue curing system was used in their construction.
The shear load was applied with a rate of load of 0.15 inch/minute to failure. The shear stress at failure was calculated in pounds-force per square inch of glue line area between the two laminations.
To test the shear strength of a solid piece of hybrid poplar relative to the five other wood species, the samples were cut in exactly the same manner. Samples from the five other species were conveniently available with planed surfaces ready for gluing. Only the shear stress at failure was recorded.
The results for block shear for hybrid poplar with glue joints appear in Table 8. Due to the low shear values obtained for UF adhesive using the RF glue press it was decided to use a urea with a thickening agent or filler to increase its viscosity thereby reducing the flow of urea into the wood fibre before curing can occur. Unfortunately testing a UF with filler with a RF press was not possible however, the thickened form of UF was tested using the hot platen and the cold set method and the results can also be seen in the last two columns in Table 8.
Table 8 Comparative Shear Results Using Different Glue/Press Combinations
|
Hot Platen |
Cold Set |
Cold Set |
RF |
RF |
Hot Platen |
Cold Set |
|
|
Part # |
PVAc |
PRF |
PVA |
Urea |
PVAc |
Urea |
Urea |
|
B23 |
1708 |
|
|
|
|
|
|
|
W15 |
1649 |
|
|
|
|
|
|
|
W16 |
1776 |
|
|
|
|
|
|
|
W17 |
1570 |
|
|
|
|
|
|
|
W18 |
1220 |
|
|
|
|
|
|
|
Average |
1585 |
||||||
|
B25 |
|
2046 |
|
|
|
|
|
|
B21 |
|
1751 |
|
|
|
|
|
|
R5 |
|
1377 |
|
|
|
|
|
|
R21 |
|
1598 |
|
|
|
|
|
|
W2 |
|
1245 |
|
|
|
|
|
|
Average |
1603 |
||||||
|
B22 |
|
|
1543 |
|
|
|
|
|
B24 |
|
|
1875 |
|
|
|
|
|
R10 |
|
|
1803 |
|
|
|
|
|
R11 |
|
|
1862 |
|
|
|
|
|
W4 |
|
|
1569 |
|
|
|
|
|
Average |
1731 |
||||||
|
R6 |
|
|
|
859 |
|
|
|
|
W19 |
|
|
|
682 |
|
|
|
|
R12 |
|
|
|
1493 |
|
|
|
|
R7 |
|
|
|
1115 |
|
|
|
|
R9 |
|
|
|
1145 |
|
|
|
|
Average |
1059 |
||||||
|
R14 |
|
|
|
|
2210 |
|
|
|
R13 |
|
|
|
|
2032 |
|
|
|
W3 |
|
|
|
|
1443 |
|
|
|
W8 |
|
|
|
|
1490 |
|
|
|
W20 |
|
|
|
|
1824 |
|
|
|
Average |
1800 |
||||||
|
B6 |
|
|
|
|
|
1572 |
|
|
W10 |
|
|
|
|
|
1622 |
|
|
R11 |
|
|
|
|
|
1806 |
|
|
R24 |
|
|
|
|
|
1457 |
|
|
R24B |
|
|
|
|
|
1894 |
|
|
Average |
1670 |
||||||
|
B6 |
|
|
|
|
|
|
1608 |
|
W10 |
|
|
|
|
|
|
1591 |
|
R11 |
|
|
|
|
|
|
1877 |
|
R24 |
|
|
|
|
|
|
1400 |
|
R31 |
|
|
|
|
|
|
1786 |
|
W34 |
|
|
|
|
|
|
1965 |
|
Average |
1705 |
*Note: Sample number denotes source board number and hence colour of board end marking.
Excluding the UF values using the RF press it can be seen from the above table that all of the adhesives used were capable of providing a satisfactory shear strength no matter what type of glue press was used. Filled UF provided good shear strength whereas un-filled UF, which provides a colorless glue line, could not provide an adequate shear load. The tests conducted with un-filled UF clearly show that hybrid poplar does not work well with low viscosity adhesives.
The shear test results for solid wood blocks (with no glue joint) appear in Table 9.
Table 9 Block Shear Results for Solid Wood Samples
|
Species |
Shear Force (lb.) |
|
Black Cottonwood |
1350 |
|
Western White Pine |
1359 |
|
Trembling Aspen |
1675 |
|
Lodgepole Pine |
1716 |
|
Hybrid Poplar |
1876 |
|
Red Alder |
2037 |
The results in Table 9 show that hybrid poplar is only exceeded by red alder in compressive shear loading. It must be emphasized that the number of samples used in this test was small hence the data may not be that representative.
A comparison can also be drawn between the value for the solid wood shear test for hybrid poplar and the average of the glue joint shear tests in Table 8. It can be seen that the glued joint shear tests (except for the UF radio frequency samples) range from being 16% to 4% below the strength of solid hybrid poplar which is an indication of how well hybrid poplar glues.
This test method determines the resistance to delamination of wood laminations typically used in exterior applications. It is, in effect, an accelerated means of measuring the effects of exterior exposure. Similar test procedures are employed in other countries to test glued wood bonds though, in practice, the number of wet and dry cycles is reduced to one rather than the three stipulated in the ASTM standard. This test requires the evaluations be done in two stages: after one complete wet-dry cycle and then following a second wet-dry cycle.
Equipment requires includes a pressure vessel similar to the ones used in paint finishing lines where both a vacuum and elevated atmospheric pressure can be applied, a water aspirator system to create a vacuum of between 25-30 inches of mercury, and a convection oven capable of heating air up to 153 ° F and capable of circulating air at 500 feet per minute.
Test sample pieces were cut from the larger glue blocks mentioned above to a size of approximately 2 by 2 inches and then weighed. With weighed down samples in the pressure vessel warm water was added to cover the samples and a vacuum of between 20 to 25 inches of mercury was drawn for five minutes which was followed by a pressurization period of one hour at 75 psi. Upon completion of wet cycling, the samples were dried in an oven for 10 hours at 160 ° F at an air velocity of 500 feet per minute. The end grain glue joints were placed directly parallel to the air stream. Upon reaching 115% (i.e. at 15% moisture content) of the initial dry weight the samples were examined for delamination of the end grain joints only.
All delamination lengths were totaled and expressed as a percentage of the total length of glue joint in all the samples of a lot. A second wet-dry cycle was then completed as before with the delamination recorded as before for each lot.
6.2.3 Delamination Test Results
The first wet-dry cycle yielded the following results as found in Table 9:
Table 10 Delamination Test Results after first wet-dry cycle
|
Hot Platen |
Cold Set |
Cold Set |
RF |
RF |
|
|
Part # |
PVAc |
PRF |
PVA |
Urea |
PVAc |
|
B23 |
0.0% |
||||
|
W15 |
56.0% |
||||
|
W16 |
13.0% |
||||
|
W17 |
0.0% |
||||
|
W18 |
5.0% |
||||
|
Average |
14.8% |
||||
|
B25 |
0.0% |
||||
|
B21 |
0.0% |
||||
|
R5 |
0.0% |
||||
|
R21 |
72.0% |
||||
|
W2 |
0.0% |
||||
|
Average |
14.4% |
||||
|
B22 |
0.0% |
||||
|
B24 |
0.0% |
||||
|
R10 |
26.0% |
||||
|
R11 |
0.0% |
||||
|
W4 |
13.0% |
||||
|
Average |
7.8% |
||||
|
R6 |
53.0% |
||||
|
W19 |
65.7% |
||||
|
R12 |
100.0% |
||||
|
R7 |
100.0% |
||||
|
R9 |
62.0% |
||||
|
Average |
76.1% |
||||
|
R14 |
4.0% |
||||
|
R13 |
28.0% |
||||
|
W3 |
7.0% |
||||
|
W8 |
0.0% |
||||
|
W20 |
52.0% |
||||
|
Average |
18.2% |
*Note: Sample number denotes source board number and hence colour of board end marking.
Excluding the UF results, it can be seen that every test batch had at least one sample that had high delamination values though 65% of the samples had values below 10%. The very high values account for 25% of the values which is significant. Only the cold set PVA batch had delamination results that were below the 10% level. The PRF had the one 'wild' result but other than that it faired well as would be expected with this waterproof, structural adhesive.
The second wet-dry cycle yielded the following results as found in Table 10:
Table 11 Delamination Test Results after second wet-dry cycle
|
Hot Platen |
Cold Set |
Cold Set |
RF |
RF |
|
|
Part # * |
PVAc |
PRF |
PVA |
Urea |
PVAc |
|
B23 |
0.0% |
||||
|
W15 |
29.0% |