Screw thread shape and fastener type effects on load capacities of screw-based particleboard joints in case construction.
In this study, the suitability of several mostly screw-based fasteners for connecting 5/8-inch- (15.87-mm-) thick furniture grade particleboard was assessed. Separate statistical tests were made for the effects of screw shank diameter, screw thread pitch, screw thread design, and fastener type (screw, screw + PVC anchor, or dowel) on edge screw withdrawal resistance (SWR), bending moment resistance of corner joints, and lateral edge load of butt-jointed shelf units. A thicker shank diameter (0.252 in) gave significantly improved edge SWR and bending moment resistance but thinner screws gave the highest edge load of shelf joints. A thread pitch of 9 threads per inch (tpi) gave the highest edge SWR, and screws with a fully threaded shank performed best in the lateral edge load tests. Beveled threads made little difference to performance compared with the same screw brand and type with plain threads. Opposing trends for edge SWR and lateral edge load of shelves with increasing screw shank diameter are interpreted in terms of fracture mechanics and contrasting load distribution patterns on the substrate surrounding the fastener. Combining screws with a large expandable PVC anchor or replacing with an unglued dowel connector reduced the fastener performance in all three tests.
Particleboard is well suited for the construction of modular furniture because of its uniformity, moderate density, ease of cutting, and low cost. For these reasons it has become the single most widely used reconstituted wood composite panel in the furniture manufacturing industry (Wu and Vlosky 2000). There are four furniture grades of particleboard: M1, MS, M2, and M3. The M2 grade is currently the most commonly used grade for "flat-pack" ready-to-assemble (RTA) shelving and cabinet case work. The RTA industry is keen to adopt the cheaper and lighter weight MS or even M1 grade panels more widely to reduce product weight and cost. However, a recent comparative survey by Semple et al. (2005) of MS and M2 grade particleboard made by two companies found that the MS grade failed to meet American National Standards Institute (ANSI) standard A208.1 (ANSI 1999) for minimum edge screw withdrawal resistance (SWR). M1 grade is not used in furniture at present and there are no ANSI-recommended minimum SWR values for this grade (CPA 2003).
In North America, screws with or without dowels are the most commonly used fastener systems for assembling RTA furniture shelving and cabinets with simple unglued butt joint configurations. For cost and versatility, dowels are commonly used as the primary connector system in applications such as kitchen cabinets (Zhang and Eckelman 1993) and frames for upholstered furniture (Eckelman et al. 2002). Joint bending moment resistance for dowels of various thickness, embedment depths, and joint configurations for furniture case work has been extensively studied (Eckelman 1969, 1971; Zhang and Eckelman 1993; Zhang et al. 2001, 2002). The research shows load capacity of butt joints connected with dowels is largely a function of dowel diameter, the presence or absence of adhesive, and the embedment depth into the face member. Butt joints connected with screws and/or dowels have become very common with the trend for commodity furniture to be sold to the consumer in the flat pack or RTA format, which greatly reduces shipping costs (Efe et al. 2004). The most common problem associated with using such joints in particleboard is delamination damage in the core material that holds the fastener, which is also the weakest zone of the particleboard (Kelly 1977).
Previous research has already established that tapered wood screws are unsuitable for use in composite panels such as particleboard because of the wedge effect and the reduced delamination resistance of particleboard (Eckelman 2003, Norbord 2004, WPIF 2005). According to Rajak and Eckelman (1996) untapered sheet metal screws appear to offer a low-cost solution for connecting particleboard furniture components. Technical data sheets for medium density fiberboard and particleboard (Norbord 2004) recommend a small core diameter and a widely spaced "aggressive" thread pattern. This implies a relatively high shank to root diameter ratio, but no specific value is given. Edge SWR generally increases with screw shank diameter (Rajak and Eckelman 1993), while other work (Eckelman 1973, 1975; Broker and Krause 1991) has determined the critical effects of embedment depth and ratio of pilot hole to shank diameter on edge SWR in particleboard. Embedment depth has a greater effect on SWR than shank diameter (Eckelman 1973, 1975; Bachman and Hassler 1975), but the practical length of screws used to connect furniture components is limited by cost considerations (Eckelman 2003).
This study focuses on appropriate fasteners for lower-grade (MS) commodity particleboard. A range of short, low-cost, mostly screw fasteners is tested for edge SWR, corner joint bending moment, and lateral edge load capacity of butt-jointed shelf units in which the screw embedment depth into the edge of particleboard components is limited to 1 inch in all cases. No information has been found enabling comparison of the performance of plain vs. specialty screws with modified threads or anchor systems. Based on findings of previous research, the principal hypothesis tested in this study is that thicker screws, specialty screws, or combinations of screws and anchors could improve the SWR, bending moment resistance, and lateral edge load of corner joints made from particleboard. The objectives were to:
1. Determine the effects of shank diameter, thread pitch, thread shape, and fastener type (screw, screw + anchor, or dowel) on edge SWR, bending moment, and lateral edge load of corner joints constructed from lower grade (MS) particleboard, and recommend the most appropriate fastener/s for shelf construction from MS particleboard.
2. Compare fastener performance across different loading configurations and comment on whether edge SWR tests alone provide an adequate indication of the suitability of a fastener for butt joints constructed from particleboard components.
Materials and methods
Five 4- by 8-foot panels (1.22 by 2.44 m) of 5/8-inch-(15.87-mm-) thick furniture grade particleboard were purchased from a local building supply retailer and cut into 2- by 4-foot (0.61 by 1.22 m) subpanels. The subpanels were cut into components for use in edge SWR tests, corner joint bending moment tests, and shelf loading tests. All precut components were stored in a conditioning chamber at 20 [+ or -] 1[degrees]C and 65 [+ or -] 5 percent relative humidity for 3 weeks prior to drilling and assembly of test samples. The following physical and mechanical properties of the panels were measured prior to the fastener tests: thickness, ovendry density, moisture content (MC), modulus of rupture (MOR), modulus of elasticity (MOE), and internal bond strength (IB). For each property, two specimens were cut from random locations on each panel. The specimen sizes and tests for physical and mechanical properties were made in accordance with ASTM D 1037 (ASTM 2000).
[FIGURE 1 OMITTED]
The fasteners used in this study were obtained from commercial suppliers. A total of 13 screw fasteners and 1 dowel were evaluated. All of the screws are untapered "sheet metal" screws but vary in specifications including shank diameter, thread pitch, and thread design (Fig. 1). Screws A, and H to K are fully threaded, while screws B to G are threaded on the bottom 1 inch of the shank only. Screws D and G are specialty screws with modified thread edges. The bottom four threads on screw D are approximately square-shaped while on screw G the threads are crenulated or flower shaped; micrographs of these are shown at x20 magnification in Figure 1. The motivation for crenulated or square thread designs is to allow for easier insertion, but the effect on edge SWR in particleboard is unknown. Screw H was tested with two different anchors (fasteners L and M). For comparison with the PVC anchors, a 5/16-inch multi-grooved dowel (N) was tested in the bending moment and lateral edge load tests. The general hardness of the wood and inspection of the cross grain of the dowels suggests they may have been made from birch (Betula spp.) (Ellis 2005).
Measurement of edge SWR
For each fastener, a total often 3- by 6-inch (76.2- by 152.4-mm) edge SWR replicates (2 per panel) were cut and a pilot hole (between 85% and 95% of shank diameter) was drilled into the core midway along the long edge of each specimen. While the machine direction of the store-bought panels was unknown, the orientation of the SWR samples was kept consistent relative to the long edge of the panels, i.e., the long edge of the SWR samples was perpendicular to the long edge of the panel. Pilot hole diameter and the ratio of fastener shank diameter to pilot hole diameters, in percent, for each fastener are listed in Table 1. Each fastener was embedded 1 inch (25.4 mm) into the edge of specimens regardless of total fastener length. The 1-inch embedment depth corresponds to the depth of embedment into the edge-drilled members of the bending moment and the shelf lateral edge load tests. The pilot hole extended a distance of approximately 0.08 inch (2 mm) beyond the tip of the screw as recommended for composite panel edges (Norbord 2004). The test procedure for edge SWR was done in accordance with ASTM D 1037 with a loading rate of 0.6 inch (15.2 mm) per minute.
[FIGURE 2 OMITTED]
Construction and testing of corner joint bending moment resistance
Corner joints for testing bending moment resistance were composed of a face-drilled member and an edge-drilled member. These were 7 inches (17.8 cm) wide; the length of the edge-drilled member was 0.625 inch less than the face member to ensure joint symmetry, i.e., the distance from both outer edges to the apex of the joint was kept constant at 4 inches. The joint and loading configuration for bending moment specimens is shown in Figure 2a. The members were connected by two fasteners spaced 4 inches (101.6 cm) apart and 1.5 inch (38.1 mm) in from each short edge. This is based on the optimum spacing for fasteners in the edge of particleboard, which should not be less than 1.5 to 2 inches (Rajak 1989, Ho 1991, Zhang 1991). These previous studies found that during testing of edge-drilled particleboard, the components split and the resulting delamination zone extended a little over 1.5 inches on either side of screw or dowel. For larger screws, the average length of the delamination zone in the edge member was 3.5 inches, suggesting that when joining 3/4-inch-thick particleboard, the optimum distance between two screws should be 4 inches (Rajak and Eckelman 1996). Fasteners were embedded through the face of the face-drilled member to a recommended depth of 1 inch (Zhang and Eckelman 1993) into the edge of the edge-drilled member.
Bending moment resistance tests were performed on a Syntech 30 D Universal Testing Machine. Load (F) was applied to the apex of the corner joint specimen at a rate of 0.25 in/min (6.35 mm/min). The ends of the specimen rested on roller plates to permit unrestricted movement during loading (Eren and Eckelman 1998) as shown in Figure 2a. Peak or "ultimate" breaking load ([F.sub.t]) was recorded for each specimen and these were then converted to corresponding bending moment resistance values using the following expression from Zhang and Eckelman (1993):
[M.sub.t] = 1.181[F.sub.t] 
where [M.sub.t] = ultimate bending moment resistance, in-lb; [F.sub.t] = peak load, lb. The coefficient 1.181 is the distance between the two inner edges divided by 4, i.e., d/4.
Bending moment resistance of a corner joint can be tested in two configurations; where the edges are forced outwards by loading the apex, or where compression loading is applied to the ends. The apex loading configuration was used here since it is slightly stronger and less prone to non-uniform load distribution (Zhang and Eckelman 1993).
Construction and testing lateral edge load capacity of butt-jointed shelves
For each fastener, 10 simple butt-jointed shelf configurations measuring 18 by 7 by 4 inches (457 by 177.8 by 101.6 mm) were constructed. These consisted of a 7- by 18-inch top shelf supported by two 7- by 4-inch face-drilled sides as shown in Figure 2b. The long edges of the shelves were cut perpendicular to the long edge of the panel. The appropriate diameter pilot holes for each fastener were predrilled before assembling the shelf units. The fasteners were inserted through the face of the side supports and penetrated into the edges of the shelf to a distance of 1 inch. Embedment depth was restricted to 1 inch so as to reduce the likelihood of the shelf tops failing in bending. This is because the embedment depth of the screws used in a butt joint has a critical effect on its strength and mode of failure (Rajak and Eckelman 1996). They showed that simply increasing the length of screws from 2 to 2.5 inches resulted in a significant jump in butt joint strength as the mode of failure shifted from delamination to bending.
The fasteners were spaced 4 inches apart and 1.5 inches in from the front and back edges of the shelf, the same spacing used for the bending moment samples in Figure 2a. This was adapted from the fastener spacing found in commercial flat pack shelving kits containing 5/8-inch-thick particleboard components and 1-5/8-inch-long screws and also from the recommendations of Rajak and Eckelman (1996).
Load (F) was applied to the shelf samples at a rate of 0.25 in/min by two loading noses located 4 inches in from each joint as shown in Figure 2b. This loading configuration was used by Zhang et al. (2002) to maximize delamination at the joint, rather than failure in bending at the midspan of the shelf. The peak load ([F.sub.t], in-lb) for the specimen was recorded and also the mode of failure (either delamination at the joint or shelf failure in bending). The average load value for each fastener is inclusive of all test specimens.
Experimental designs and statistical analyses
Factors considered likely to affect edge SWR, bending moment, and lateral edge load were assessed in four separate one-factor analyses of variance using four subsets of data. This approach was used in place of a multi-factorial design because each factor involved a different subset of fasters. Note that the levels, such as tread design or thread pitch, within each factor are discrete variables. The four contrasts are as follows:
1. Shank diameter: screws A, E, and K are plain-threaded and of the same thread pitch, i.e., 9 threads per inch (tpi), with three levels of shank diameter (0.146, 0.170, and 0.252 in).
2. Thread pitch: screws B, C, E, and H are all very similar in shank diameter (0.165 to 0.170 in), and represent four levels of thread pitch (8, 9, 14, and 16 tpi).
3. Thread design: screws D, E, and G from the same manufacturer are compared. All are no. 8 gauge screws with 9 tpi, vary little in shank diameter (from 0.166 to 0.178 in), and represent three discrete levels of thread design (plain, squared, or crenulated). Obtaining the screws with different thread designs but with exactly the same shank diameter was not possible. Note that for the lateral edge load tests, screw G was omitted due to insufficient supplies.
4. Fastener type: fasteners H, K, L, M, and N, representing five levels (thin screw, thick screw, large anchor, small anchor, and dowel). The thicker fasteners K, L, M, and N vary in shank diameter from 0.25 to 0.35 inch. It was not possible to obtain the different types of fastener in exactly the same diameter. The no. 8 screw (screw H) with the two anchor types, L and M, are compared with the screw on its own. Note that SWR tests do not apply for the dowel.
Single-factor analyses of variance for each of the above four contrasts were performed using JMP Statistical Software. A check for normality in population distributions showed that all were close to normal. A summary of the analysis of variance (ANOVA) results for four factors is shown in Table 2. All of the average values for each test for each fastener (listed alphabetically) are given in Table 3; only those contrasts with effects that were significant for p [less than or equal to] 0.05 will be shown graphically. Statistical comparison of means is made using the least significant difference (LSD) bar.
Physical properties of test material
The basic physical and mechanical properties of the particleboard sheets used to fabricate the fastener loading test specimens are given in Table 4. The thickness, density, and MC were fairly consistent; coefficient of variation (COV) = 2 percent for density and <1 percent for thickness and MC. Variation in IB was higher with a COV of 25 percent while COV values for MOR and MOE were lower at 11 and 14 percent, respectively. Average density of the particleboard was 39 pcf (625 kg/[m.sup.3]) and IB was 74 psi (0.51 MPa); however the MOR values of the panels were below the ANSI A208.1 minimum of 1,813 psi (12.5 MPa) for MS grade particleboard. The particleboard did exceed the standard for MOE of 276 ksi (19 GPa). This standard gives no recommended density range for MS grade panels; however, a survey of particleboard properties across grades (Semple et al. 2005) found an average density for the MS grade of around 40 pcf (640 kg/[m.sup.3]), indicating that the particleboard used here is most likely MS grade with relatively high bond strength.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Effects of shank diameter, thread pitch, thread design, and fastener type
From the summary (Table 2) of the results from the ANOVA tests for the shank diameter, thread pitch, thread design, and fastener type factors, it can be seen that all factors had significant effects on the performance of fasteners. Exceptions to this were thread design, which had no effect on edge SWR or lateral edge load capacity, and thread pitch which had no effect on bending moment.
Note from Table 2 that shank diameter ranging from 0.146 inch to 0.252 inch had no statistically significant effect on lateral edge load capacity of butt-jointed shelf units, but there was nevertheless a slight downward trend (which is examined further later) with increasing shank diameter. The significant effects of screw shank diameter on edge SWR and bending moment are shown graphically in Figure 3. Using a thicker screw (shank dia. = 0.252, screw K) resulted in highest edge SWR (285.6 lb) and bending moment (164 in-lb). The results for edge SWR were in agreement with earlier work by Eckelman (1973, 1975) whereby edge SWR tends to increase with screw diameter. Work by Rajak and Eckelman (1993) also shows a regular increase in edge SWR from particleboard with increasing screw diameter, but only if pilot hole diameter is between about 79 and 86 percent of shank diameter.
The effects of screw thread pitch are shown for edge SWR in Figure 4a, and lateral edge load in Figure 4b. Screws with 9 tpi were significantly (p < 0.001) higher in edge SWR than those with 8, 14, or 16 tpi (Fig. 4a), while lateral edge load was maximum in the case of screw H with 14 tpi (Fig. 4b). Screw thread pitch was not expected to have affected lateral edge load capacity, and this was supported by there being no significant differences between screws B (8 tpi), C (16 tpi), and E (9 tpi). It is important to note that screw H differed from the others in that the shank was fully threaded. Unlike screws B, C, and E, where the threaded portion of the shank is only present in the edge-drilled member, the fully threaded shank of screw H passes through both the face- and edge-drilled members of the shelf unit. This may help explain the higher lateral edge load capacity for this screw.
Although not statistically significant, screws with crenulated threads (screw G) were higher in bending moment (139.5 in-lb) than the same size screws with plain threads (screw E; 127.2 in-lb) or with square threads (screw D; 127.9 in-lb). The reason for this is unclear, since a similar trend might be expected for edge SWR as well, but this was not observed. Further testing would be required to confirm whether the observed crenulated thread effect on bending moment is consistent.
Edge SWR, bending moment, and lateral edge load capacity of shelf units were all significantly affected by fastener type (screw, screw with anchor, or wood dowel), as can be seen in Figure 5. The performance of the no. 8 screw (H) with PVC anchors (fasteners L and M) in edge SWR and bending moment was significantly lower than if using this screw alone or a thicker screw (screw K). However, using a small anchor (fastener M) was not significantly different from using either a thin or a thick screw to connect butt-jointed shelf units (Fig. 5c). Using a larger anchor (fastener L) or a wood dowel resulted in significantly lower lateral edge load capacity of butt-jointed shelves compared to using a thin screw (screw H) on its own. The poor performance of the anchors was contrary to expectations since screws with PVC anchors are commonly used for connecting particleboard furniture components in countries such as Korea (Lee and Park 1991). However, it has been recommended by the Wood Panel Industries Federation (WPIF 2005) that such fixings, particularly those that involve the expansion of a component such as a PVC anchor, not be inserted into the edges of particleboard. The reduced edge SWR of the anchors may have been caused by two factors: the taper of the PVC anchor and the expansion of the anchor in the pilot hole could have caused damage to the panel core, and/or the outer thread configuration of the anchor provided less mechanical interlocking with the substrate.
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Failure modes during loading tests
In the edge SWR and the bending moment tests, the only mode of failure was fastener pull out from the edge-drilled member once the threads ruptured the surrounding substrate in the core of the board, confirming the expectation that the core is the weakest zone in the board. This was different from the failure modes reported by Zhang and Eckelman (1993) in bending moment joints connected with glued dowels, where the face member broke or the dowel itself snapped. In the case of the lateral edge load specimens, there were two failure models that are shown in Figure 6. In most cases, the specimens failed in delamination of the edge-drilled member at the sites of fastener insertion (Fig. 6a), and a minority failed in bending at the midspan of the shelf (Fig. 6b). The frequency of failure type for each fastener is given in Table 5. Up to 30 percent of the test specimens failed at midspan in those cases where only screws were used, suggesting stronger joints. Among these, failure at midspan was highest for screws A, C, E, and F. Note that while midspan failure was high (30%) for the dowel joint, the overall lateral edge load capacity for shelves with dowels was low (354 lb or 1575.3 N). This difference is reflected in the high variation in breaking load of dowel-connected shelves, with a COV of 20 percent (Table 3). It is expected that both joint edge load capacity and its variability would improve if glued dowels were used. Glued dowels were not used in this study as it was focused on removable fasteners.
Screw gauge effects for different loading configurations
A clearer picture of how screw performance varies with the different loading configuration can be seen in Figure 7, which compares the effects of increasing screw gauge of the same brand of fully threaded screws (0.146 in for no. 8 screws up to 0.252 in for no. 14 screws) on edge SWR, bending moment resistance, and lateral edge load. These averages are indicative for screw gauge only and no specific statistical comparisons can be made in this group for shank diameter due to the confounding effect of different thread pitches. Edge SWR and bending moment follow a similar trend with increasing screw gauge up to no. 12 (0.22 in, screw J), after which edge SWR continues to increase with screw gauge while bending moment decreases (Fig. 7a).
The trend for lateral edge load capacity with increasing screw gauge is opposite to that for edge SWR, i.e., while edge SWR increases with increasing screw gauge, the edge loading capacity of the butt-jointed shelves decreases (Fig. 7b). The pilot hole in the edge-drilled member acts as a crack or flaw in the panel. Since the pilot hole is only 85 to 90 percent of the root diameter of the screw (Fig. 1), the insertion of the screw will exert an internal mode I opening force on the surrounding substrate, which is compounded by loading perpendicular to the hole as in the shelf tests. As the corner joints of the shelves are loaded in compression or bending, the screws effectively pry the upper and lower faces of the shelf apart, adding to the internal stress already acting on the material surrounding the screw. With the thicker screws, the depth of material between the particleboard surface and the screw is reduced and a greater volume of material is disrupted by the insertion of the screw.
[FIGURE 7 OMITTED]
Thinner screws, including H, A, F, and C, gave the highest lateral edge load capacity (Fig. 7b, Table 4). Plain sheet metal screws with a fully threaded shank (screws H or A) could be recommended as good overall fasteners for particleboard shelving purposes. Screws H and A were also among the cheapest of the fasteners tested (42% of the cost of the "particleboard screw" B). Other work (Madriz 1997) has found no direct relationship between cost of fasteners and performance in joints, indicating that strong, cost-effective joints can be obtained through rational design procedures without the need for expensive fasteners.
1. Screws on their own performed markedly better than either dowels or screws with PVC anchors in all three tests. The use of thicker screws or screws with anchors is not recommended for butt-jointed shelving applications using particleboard.
2. Screws with specialized thread configurations such as beveled edges or alternating thread heights performed well, but were not significantly different from plain-threaded screws of the same dimensions in any of the loading configurations.
3. Screw length was restricted to 1.5 inches or less for cost reasons and to minimize the possibility of fracture in the middle of the particleboard shelf during the lateral edge load tests. Most of the lateral edge load specimens failed in delamination of the edge member at the joint, although up to 30 percent of the shelves connected with small-diameter screws failed at the midspan, indicating higher joint loading capacity.
4. If a screw has high edge SWR, it cannot be presumed suitable for connecting modular furniture components from particleboard because there is little correlation between the performance of a screw in the edge SWR test and its performance in the bending moment resistance or lateral edge load tests. The bending moment results by themselves indicate that screw I (no. 10) was best, with optimum screw diameter falling between 0.200 and 0.225 inch (5.08 and 5.72 mm). Thicker gauge screws (no. 12 and no. 14) performed best in edge SWR tests, whereas the thinner gauge screws (no. 6 and no. 8) performed best in the lateral edge load test.
5. Plain, fully threaded sheet metal screws with around 14 tpi and a screw gauge of 8 to 10, shank diameter ranging from 0.165 to 0.195 inch (4.19 to 4.95 mm), were the best performing fasteners overall for general purpose top-loaded butt joints made from 5/8-inch- (15.87-mm-) thick particleboard.
American National Standards Institute (ANSI). 1999. Particleboard. ANSI A208.1. ANSI, Washington, DC. 11 pp.
American Society for Testing and Materials (ASTM). 2000. Annual book of ASTM standards; Section 4. ASTM D 1037, Construction (Wood). Vol. 04.01. ASTM, West Conshohocken, PA.
Bachman. G. and W. Hassler. 1975. The strength of various furniture constructions, their components and fasteners. Part 1. Holztechnologie 16(4):210-221.
Broker, F.W. and H.A. Krause. 1991. Preliminary investigations on screwings in dynamic tests. Holz roh-Werkstoff 49(6):381-384.
Composite Panel Association (CPA). 2003. Buyers and specifiers guide to North American particleboard, MDF and hardboard/fiberboard manufacturers and products. Composite Panel Assoc., Gaithersburg, MD. 15 pp.
Eckelman, C.A. 1969. Engineering concepts of single-pin dowel joint design. Forest Prod. J. 19(12):52-60.
______. 1971. Bending strength and moment rotation characteristics of two-pin moment-resisting dowel joints. Forest Prod. J. 21(3): 35-39.
______. 1973. Holding strength of screws in wood and wood based materials. Agri. Res. Bull. 85. Purdue Univ., West Lafayette, IN. 15 pp.
______. 1975. Screwholding performance in hardwoods and particleboard. Forest Prod. J. 25(6):30-35.
______. 2003. Chapter 6: Strength of screws in wood composites. In: Textbook of Product Engineering and Strength Design of Furniture. Purdue Univ. School of Agriculture, West Lafayette, IN. pp. 56-67.
______. Y.Z. Erdil, and J. Zhang. 2002. Withdrawal and bending strength of dowel joints constructed of plywood and oriented strand-board. Forest Prod. J. 52(9):66-74.
Efe, H., Y.Z. Erdil, A. Kasal, and H.O. Imirzi. 2004. Withdrawal strength and moment resistance of screwed T-type end-to-side grain furniture joints. Forest Prod. J. 54(11):91-97.
Ellis, S. 2005. Personal Communication, Assoc. Professor, Dept. of Wood Sci., Univ. of British Columbia, Vancouver, BC.
Eren, S. and C.A. Eckelman. 1998. Edge breaking strength of wood composites. Holz als Roh- und Werkstoff 56:115-129.
Ho, C.L. 1991. The use of performance tests in evaluating joint and fastener strength in case type furniture. MS thesis. Purdue Univ., West Lafayette, IN.
Kelly, M.W. 1977. Critical literature review of relationships between processing parameters and physical properties of particleboard. Gen. Tech. Rept. FPL 10. USDA Forest Serv., Forest Prod. Lab., Madison, WI. 64 pp.
Lee, P.W. and H.J. Park. 1991. Strength of furniture joints constructed with PVC anchor and screw. Mokchae Konghak 19(1):22-30.
Madriz, C. 1997. Cost and strength analysis of corner joints constructed with fasteners commonly used in kitchen cabinet industry. MS thesis. Purdue Univ., West Lafayette, IN. 83 pp.
Norbord. 2004. Medium density fiberboard: How to use MDF--screws. In: MDF from Start to Finish. Technical Data Sheet, Composite Panel Assoc., Gaithersburg, MD. www.norbord.com/mdf-ps-howto.htm. Accessed February 18, 2005.
Rajak, Z. 1989. Efficient use of screws in the construction of corner joints for case goods. MS thesis. Purdue Univ., West Lafayette, IN. 124 pp.
______ and C.A. Eckelman. 1993. Edge and face withdrawal strength of large screws in particleboard and medium density fiberboard. Forest Prod. J. 43(4):25-30.
______ and ______. 1996. Analysis of corner joints constructed with large screws. J. Tropical Forest Prod. 2(1):80-92.
Semple, K.E., E.K. Sackey, H.J. Park, and G.D. Smith. 2005. Properties survey of furniture grade particleboard Part 2--MS and M2 grade comparison and a practical in-situ test for internal bond strength. Forest Prod. J. 55(12):125-131.
Wood Panel Industries Federation (WPIF). 2005. Annex 2A: Particleboard (wood chipboard). In: Panel Guide. www.wpif.org.uk/pg/39_annex2av2.pdf. Accessed: February 18, 2005.
Wu, Q.L. and R.P. Vlosky. 2000. Panel products: A perspective from furniture and cabinet manufacturers in the southern United States. Forest Prod. J. 50(9):45-50.
Zhang, J. 1991. Rational design of dowel joints. MS thesis. Purdue Univ., West Lafayette, IN. 77 pp. Unpublished.
______ and C.A. Eckelman. 1993. The bending moment resistance of single-dowel corner joints in case construction. Forest Prod. J. 43(6):19-24.
______, Y.Z. Erdil, and C.A. Eckelman. 2002. Lateral holding strength of dowel joints constructed of plywood and oriented strand-board. Forest Prod. J. 52(7/8):83-89.
______, F. Quin, and B. Tackett. 2001. Bending strength and stiffness of two-pin dowel joints constructed of wood and wood composites. Forest Prod. J. 51(2):29-35.
He Jun Park
Gregory D. Smith*
The authors are, respectively, Visiting Scientist, Dept. of Forest Products, Iksan National College of Agriculture and Technology, Jeonbuk, Korea (firstname.lastname@example.org); Post Doctoral Fellow and Assistant Professor, Dept. of Wood Sci., Univ. of British Columbia, Vancouver, BC, Canada (email@example.com; firstname.lastname@example.org). The authors wish to thank Pan American Screw, Inc., Conover, NC, for providing some screw samples for tests and Natural Resources Canada for financial support of this work. This paper was received for publication in March 2005. Article No. 10020.
*Forest Products Society Member.
Table 1. -- Specifications for fasteners. Fastener dimensions Fastener Root Shank Pilot hole ID Name diameter diameter diameter (in) A No. 6 drywall 0.09 0.146 0.078 B No. 8 particleboard 0.102 0.165 0.092 C No. 8 floor screw 0.097 0.165 0.092 D No. 8 "wood-master" 0.104 0.178 0.093 E No. 8 "lo-root" 0.104 0.170 0.093 F No. 8 "type 17" 0.106 0.170 0.093 G No. 8 "Heco" 0.103 0.166 0.093 H No. 8 sheet metal 0.118 0.164 0.108 I No. 10 sheet metal 0.138 0.193 0.125 J No. 12 sheet metal 0.161 0.218 0.141 K No. 14 sheet metal 0.184 0.252 0.156 L Anchor fastener 0.180 0.270 0.248 M Anchor fastener 0.303 0.387 0.281 N 5/16-in wood dowel -- 0.312 0.313 Fastener dimensions Fastener Pilot hole/shank ID Name diameter Thread pitch (%) (tpi) A No. 6 drywall 86.7 9 B No. 8 particleboard 90.2 8 C No. 8 floor screw 94.8 16 D No. 8 "wood-master" 89.4 9 E No. 8 "lo-root" 89.4 9 F No. 8 "type 17" 87.7 9 G No. 8 "Heco" 90.3 9 H No. 8 sheet metal 91.5 14 I No. 10 sheet metal 90.6 12 J No. 12 sheet metal 87.6 11 K No. 14 sheet metal 84.8 9 L Anchor fastener 91.8 5 M Anchor fastener 72.6 5 N 5/16-in wood dowel -- -- Table 2. -- Summary of ANOVA results for shank diameter, thread pitch, thread design, and fastener type for edge SWR, bending moment, and lateral edge load capacity. Response variable Factor Edge SWR Bending moment Lateral edge load Shank diameter p = 0.001 p = 0.001 ns (a) Thread pitch p < 0.001 ns p = 0.007 Thread design ns p = 0.002 ns Fastener type p < 0.001 p < 0.001 p < 0.001 (a) ns = not significant at p [less than or equal to] 0.05. Table 3. -- Means and COV for edge SWR, bending moment, and lateral edge load. Edge SWR Bending moment Lateral edge load Fasteners Mean COV Mean COV Mean COV (lb) (%) (lb-in) (%) (lb) (%) A 218.1 11.9 132.4 21.6 423.0 12.5 B 181.8 10.6 130.4 9.1 376.7 13.9 C 192.1 12.4 132.3 11.6 409.0 14.1 D 265.5 5.9 129.0 5.0 392.5 13.5 E 257.8 8.8 127.2 6.7 389.4 7.1 F 259.8 7.7 140.9 8.0 412.9 13.6 G 261.6 8.0 139.5 7.3 -- -- H 223.4 11.9 138.9 8.5 458.2 13.8 I 256.3 13.6 171.3 18.9 396.9 14.8 J 257.6 16.9 169.4 15.4 385.5 13.3 K 285.6 18.1 164.0 10.9 383.6 21.0 L 151.7 9.6 89.3 28.0 312.6 15.9 M 143.0 8.5 127.6 13.8 418.1 17.3 N -- -- 88.4 19.5 354.2 20.9 Table 4. -- Mean physical and mechanical properties of the particleboard panels. Ovendry Thickness density MC MOR MOE IB (lb/ (x[10.sup.4] (in) [ft.sup.3]) (%) (psi) psi) (psi) 0.64 (0.2) (a) 39 (2.0) 8.9 (0.4) 1717.1 30.3 (11.2) 74.3 (14.0) (25.4) (a) Values in parentheses are COVs in percent. Table 5. -- Frequency of failure modes of butt-jointed shelves. (a) Mode A B C D E F H I J K L M N Bending 30 20 30 10 30 30 10 20 10 0 0 10 30 Delamination 70 80 70 90 70 70 90 80 90 100 100 90 70 (a) Bending failure occurs at midspan failure (breakage of shelf) as shown in Figure 6a; delamination occurs at the connection points as shown in Figure 6b.
|Printer friendly Cite/link Email Feedback|
|Author:||Park, He Jun; Semple, Kate; Smith, Gregory D.|
|Publication:||Forest Products Journal|
|Date:||Apr 1, 2006|
|Previous Article:||Performance tests of small barn frame constructed with round mortise and tenon joints.|
|Next Article:||The experimental particleboard manufacture from sunflower stalks (Helianthus annuus L.) and Calabrian pine (Pinus brutia Ten.).|