Evaluation of black locust (R. pseudoacacia) as raw material for wet-process hardboard

Transcription

1 Evaluation of black locust (R. pseudoacacia) as raw material for wet-process hardboard C. Carll W. Eslyn G. Myers W. Brewer D. Staton Abstract Wet-process hardboards were made from fiber from small, whole black locust (R. pseudoacacia) stems (including bark) and from pure locust heartwood. Heartwood extractives were apparently stem-extracted and lost to cyclone condensate during digesting. Despite this apparent loss of extractives, the hardboards showed good resistance to weight and strength losses caused by the decay fungi C. versicolor and G. trabeum. Dry mechanical properties of all locust boards were acceptable. Dimensional stability properties of boards from pure heartwood were good, but the boards from whole stems showed considerable moisture absorption in exposure to high humidity, and corresponding excessive thickness swelling. These boards from whole stems showed poor exterior durability. The heartwood of black locust (R. pseudoacacia L.) is one of the most decay-resistant woods native to the United States. The wood is dense and therefore not very suitable for use in flakeboard, but its high density is not a drawback to its use for hardboard or medium-density fiberboard, suggesting it might be a suitable material for producing decay-resistant hardboard. The objectives of this study were to evaluate the suitability of locust heartwood and of whole locust stems for production of wet-process hardboard. Because black locust logs are often crooked and therefore difficult to debark, industrial use of locust for hardboard would require chipping of nondebarked or poorly debarked stems. Hence, commercial hardboard made from black locust would invariably contain sapwood fiber and some bark. This is why we evaluated boards from whole locust stems. We examined decay resistance of hardboards made of locust heartwood to see if the decay resistance of the heartwood could be imparted to wet-process hardboards. We also evaluated mechanical properties of hardboards made of whole locust stems, and dimensional stability properties of hardboards made from heartwood and from whole stems. In addition, hardboard from locust was cursorily evaluated for exterior durability. Procedure Raw material Locust heartwood lumber was sawn from a locust bolt of approximately 7-inch (18-cm) top diameter (inside bark). The lumber was chipped, and the chips screened on square mesh screen. Pieces retained on 1.25-inch (32-mm) mesh were discarded as oversized, and pieces falling through 0.25-inch (6.4-mm) mesh were discarded as fines. Whole locust stems were obtained from what we believe was a clone of trees approximately 18years old. Diameter (inside bark) of trees in the clone ranged from 4 inches (10 cm) to 6 inches (15 cm) at ground level. Sticks were cut from these trees to a 2.25-inch (5.7-cm) diameter (inside bark); in aggregate, these sticks were, by volume, approximately half heartwood/half sapwood. They were chipped without debarking and the chips screened as were those of locust heartwood. In the The authors are, respectively, Forest Products Technologist, Research Plant Pathologist, Research Forest Products Technologist, Physical Science Technician, and Biological Lab. Technician, USDA Forest Serv., Forest Prod. Lab., P.O. Box 5130, Madison, WI Material for this study was donated by the Univ. of Wisconsin Arboretum and Masonite Corp. James W. Evans, Statistical Mathematician, of the Forest Prod. Lab., advised the lead author regarding statistical analyses and performed some of the analyses. This paper was received for publication in May Forest Products Research Society Forest Prod. J. 35(3): FOREST PRODUCTS JOURNAL Vol. 35, No. 3 11

2 screening operation, much of the bark passed the 0.25-inch (6.4-mm) mesh screen and was therefore discarded as fines. Refining, board manufacture, and heat treating Chips from whole locust stems and from heartwood were fiberized (in separate batches) in an 18-inchdiameter (46-cm), single-disk pressurized refiner equipped with coarse plates. Because the digester and refiner were of the batch type. steaming time of chips in the digester ranged from 10 to 30 minutes. After being heated with 302 F (150 C) saturated steam, chips were refined at the same temperature, blown from the refiner, and dropped into a polyethylene bag. Although steaming time was as long as 30 minutes, steam temperature was low to moderate, insofar as 130 C is generally recognized as the minimum temperature at which essential softening and weakening of lignin in the middle lamellae occurs. We therefore believe that this was a moderate steaming treatment. Wet-formed hardboard disks (8.5-in. [22-cm] diameter and 1/8 in. [3.2 mm] thick) were manufactured from each of the two types of pulp. Board density was 60 pcf (0.962 g/cm). All hardboards from whole locust stems were manufactured first, and hardboards from locust heartwood were manufactured later. Enough fiber to make a disk was placed in a container and diluted with soft, cold water. While stirring the pulp slurry, 1/2 percent phenolic resin and 3/4 percent wax emulsion (percentage of ovendry fiber) were added and allowed to mix thoroughly before adding 1 percent alum and enough sulfuric acid to lower the slurry ph to between 3.5 and 4.0. Stirring was discontinued and the slurry poured into an 8.5-inch-diameter forming cylinder, where the water was drained away and a fiber mat formed. Water from the forming cylinder was not recirculated but sewered. All fiber mats were pressed for about 1 minute at 100 psi (689 kpa) mat pressure in an unheated hydraulic press to compact the fiber and expel excess water. Mats were compacted, dried, and the resin cured in a 14 by 14-inch steam-heated platen press at 374 F (190 C). The press cycle was 6 minutes long, consisting of 1 minute at 500 psi (3.45 MPa), 2 minutes at 100 psi, and then 3 minutes at 500 psi mat pressure. All hardboards were subsequently heat-treated for 1 hour in a 320 F (160 C) convection oven. Commercially manufactured wet-process hardboards of mixed southern hardwood species, not including locust, were used in this study for comparison. These commercial boards were neither oil-tempered nor heat-treated as shipped to our Laboratory; they received the same heat treatment as the laboratory boards. Measurement of physical and mechanical properties Measuring sound wave travel in commercial boards with a sonic velocity meter confirmed that they had a "machine direction." For this reason, mechanical and dimensional stability test specimens were cut from these panels such that there was an equal number of test specimens cut in each direction. Dimensional stability (linear expansion and thickness swelling) of both types of laboratory boards and of commercial boards was measured. Dimensions were measured at the following conditions: 50 percent relative humidity (RH) at 73.5 F (23 C), 90 percent RH at 80 F (27 C), and ovendry. Static bending and lateral nail resistance (LNR) tests were performed on laboratory boards from whole locust stems and on commercial boards. These tests were performed as prescribed in American Society for Testing and Materials (ASTM) standard method D 1037 (1) except that LNR test specimens were cut 2-3/4 inches (7.0 cm) wide rather than 3 inches (7.6 cm) wide. Tests were performed on specimens stored in a climate of 80 F/65 percent RH for 27 months, and on specimens stored in a climate of 80 F/90 percent RH for 23 months, ovendried, and then conditioned to equilibrium in a climate of 80 F/65 percent RH. We included storage at 80 F/90 percent RH in this study to evaluate the effect of exposure to high humidity and to mold and mildew fungi (with which high humidity conditioning rooms are always or almost always infested) on mechanical properties of the hardboards. Decay resistance testing Laboratory-manufactured hardboard from black locust heartwood and commercial hardboard were tested for decay resistance based upon changes in weight, static bending properties, and lateral nail resistance. Exposure to decay fungi and weight loss determinations were based on the ASTM D 2017 method for testing natural decay resistance of wood (2). In order to perform mechanical tests following weight loss determinations, however, we used larger test specimens and, hence, larger soil bottles than are customarily used for this test. In addition, an alternate method was used to sterilize the test specimens. Test specimens from both types of hardboard were 2 by 5 inches (static bending specimens) and 2-3/4 by 6 inches (lateral nail resistance specimens). The latter were 1/4 inch narrower than specified by ASTM D 1037 so that they might fit into the 2-quart (1.9-1) widemouth mason jars that were used as decay test containers. Test fungi included the white-rot fungus Coriolus versicolor L. ex Fr. (Mad-697) and the brown-rot fungus Gloeophyllum trabeum (Pers.) Murr. (Mad-617). The test procedure was as follows: The mason jars were filled with 400 grams (ovendry weight) of soil, laid on their sides, then enough water was added to bring the soil up to a moisture content of 36 percent. Paper feeder blocks (used with C. versicolor) or southern pine feeders (used with G. trabeum) were inserted in the jars. Each jar was sealed with a 6-inch-diameter (15-cm) milk filter and autoclaved for 90 minutes with saturated steam at 15 psi (103 kpa) pressure. One-third of the jars were then inoculated with C. versicolor, one-third with G. trabeum, and the 12 MARCH 1985

3 remaining third were noninoculated (controls). These were then incubated for 3 weeks at 80 F and 70 percent RH. Test specimens were ovendried, weighed, then soaked in sterile distilled water for 3 hours followed by a 3-second dip in boiling water to sterilize their surfaces. This procedure brought the moisture content of the test specimens up to 11 to 15 percent. This is between the 20 percent used by Behr (3) and the 8 percent recommended by Hong et al. (4) for initial moisture content of test hardboard specimens. The test specimens were then aseptically placed into the mason jars and incubated for 12 weeks in the case of those subjected to C. versicolor, and 15 weeks for those subjected to G. trabeum (Fig. 1). Half the noninoculated controls were removed at 12 weeks and the remaining half at 15 weeks. After 8 weeks of incubation, moisture contents of the soil had dropped to only 15 percent, hence 130 milliliters of sterile distilled water was added to each jar. In addition, to slow down water loss and prevent contamination (which had occurred in a few jars), the jars were now sealed with two, rather than one, sterile milk filters, and canning lids were loosely attached. Following incubation, the test specimens were removed, ovendried, weighed, and their weight losses calculated. The specimens were then conditioned to equilibrium at 80 F and 65 percent RH after which mechanical testing was conducted. The procedure described above permitted calculation of 1) Weight loss as a result of water/temperature exposure and of water/temperature/fungal exposure. 2) Moisture content of specimens upon removal from test jars. 3) Marginal strength loss as a result of exposure to decay fungi, beyond that attributable to exposure to moisture and temperature. Exterior durability Twelve 6- by 6-inch (15- by 15-cm) specimens each of commercial hardboard and of hardboard from whole locust stems were placed in a test frame, painted on one face with two coats of white exterior acrylic latex paint (housepaint), and placed outdoors facing north at Madison, Wis. None of the specimen edges were painted. The specimens were inspected periodically for substrate failure, warp, mildew growth, and paint failure. Results and discussion Refining properties and board manufacture The chips from whole locust stems and those from locust heartwood each refined to fine fluffy fiber with modest amperage draw on the refiner motor. Bauer- McNett screen analysis and Asplund drainage rate of fiber from either type of chip are shown in Table 1. During the early stages of heating the chips, the cyclone was cooler than it was during refining. During these early stages, steam exiting the refiner condensed in the cyclone. This condensed steam draining from the cyclone was dark brown to black and foamy. We suspected the dark color of the condensate was from steamextractable heartwood extractives. Analysis of the condensate from steaming of heartwood chips indicated that the color was from nonpolar or weakly polar molecules. Liquid chromatography indicated the presence of tiny amounts of five wood sugars (on the order of mg/ml concentration) but no furfural or hydroxy furfural. Boards of 60 pcf (0.962 g/cm 3 ) nominal density were manufactured without difficulty from each type of locust fiber. These hardboards were slightly denser than the commercial hardboards from mixed southern TABLE 1. - Bauer-McNett screen analyses a and Asplund drainage rates b of fiber from R. pseudoacacia heartwood and from whole R. pseudoacacia stems. Figure 1. - Appearance of decay resistance test setup after 12 weeks' incubation. Left to right: Jar with noninoculated specimens; jar with C. versicolor-inoculated specimen on a paper feeder block: and jar with a G. trabeum-inoculated specimen resting on a southern pine feeder block. FOREST PRODUCTS JOURNAL Vol 35, No 3 13

4 TABLE 2. - Dimensional stability properties of laboratory-manufactured boards from R. pseudoacacia and of industrially manufactured boards, species, making direct comparison of properties difficult. Furthermore, steaming of chips for the commercial boards in this study was not identical with our laboratory steaming; commercial steaming used higher steam pressures (hence temperatures) but for shorter times. This difference in steaming may have influenced respective board hygroscopicities. This possible difference in influence on board hygroscopicity should be kept in mind when comparing locust boards to commercial boards. Dimensional stability Descriptive statistics of linear expansion, thickness swelling, and moisture content of the two types of locust hardboard and of the commercial hardboard are given in Table 2. Note that boards made from whole locust stems reached higher equilibrium moisture contents (EMCs) than did locust heartwood boards or commercial boards. This higher EMC of boards from whole-stem locust was consistent, both between and within humidity conditions (Fig. 2). At either humidity condition, the EMC differences between whole-stem boards and either of the other two board types were statistically significant (0.01 alpha level) as determined by analysis of variance and Duncan's multiple range test (DMRT). Figure 2. - Equilibrium moisture contents of commercial and laboratory hardboards at 50 percent RH (at 73.5 F) and at 90 percent RH (at 80 F). Differences in dimensional stability properties were often difficult or impossible to identify with certainty insofar as they usually were not striking; and where specimens of a particular board type showed greater dimensional movement, they often showed higher average density than specimens of the other board types. The notable exception to this statement was for thickness swelling between 50 percent and 90 percent RH. Despite the fact that board specimens from whole locust stems had densities between those of heartwood boards and those of commercial boards, they showed appreciably and significantly (0.01 alpha level) more thickness swelling than either of these board types (as determined by analysis of variance and DMRT). This. reflects the fact that they went through significantly greater (0.01 alpha level as determined by analysis of variance and DMRT) moisture content change between 50 percent and 90 percent RH than did specimens of either of the other two board types. This greater thickness swell of boards from whole locust stems also showed up later in test fence exposure and is probably due to the inclusion of some bark in these boards. Mechanical properties Descriptive statistics of bending properties and lateral nail resistance of hardboard from whole locust stems and of commercial hardboard are given in Tables 3 and 4. As with dimensional stability properties, differences in mechanical properties are difficult or impossible to identify with certainty, as the specimens from boards from whole locust stems were denser as well as stronger. No elegant comparison is attempted, insofar as we cannot say that the sample of commercial board is representative of the population of wet-process hardboards that are commercially manufactured in the United States. No great difference exists between property values of boards stored at 65 percent RH and those of boards stored in humid (and mold/mildew infested) conditions. For either type of hardboard, comparatively higher mechanical property values were associated with comparatively higher specimen ovendry weights rather than with storage in lower humidity conditions. If storage in high-humidity conditions degrades mechanical properties of hardboards of these types, the damage is so subtle that we could not detect it. 14 MARCH 1985

5 TABLE 3. - Bending properties of laboratory-manufactured boards from whole locust stems and ofcommercial boards. TABLE 4. -Lateral nail resistance oflaboratory-manufactured boards from whole locust stems and ofcommercial boards. Decay resistance: weight losses None of the specimens incubated in noninoculated soil jars lost weight; obviously weight loss, had it occurred, would have been due to leaching. However, all specimens exposed to decay fungi lost weight (Fig. 3). Weight loss was appreciably less in the black locust hardboard than in the commercial board, irrespective of the decay fungus involved. We expected this considering the excellent decay resistance of black locust heartwood. The lower percent weight loss of specimens from locust heartwood was statistically significant ( alpha level) as determined by group t-test and by factorial analysis of variance. G. trabeum caused roughly twice as much weight loss as C. versicolor in both types of hardboard. The greater percent weight loss caused by G. trabeum was statistically significant ( alpha level) as determined by factorial analysis of variance. In addition, there was a statistically significant ( alpha level) specimen size effect as determined by factorial analysis of variance. The smaller (bending) specimens consistently lost a greater weight percentage than did the larger (lateral nail Figure 3. - Percent weight loss of hardboard specimens as a result of incubation in soil jars in the presence of decay fungi. resistance) specimens. This can probably be attributed to the greater ratio of edge surface to specimen volume of the smaller specimens. Fungus invasion of specimens occurs most readily from the edges. Hence the fungus can completely inhabit smaller specimens more rapidly. Decay resistance: mechanical property values Mechanical property loss as a result of exposure to either decay fungus was calculated as the difference between mean property value of specimens exposed in sterile jars and mean property value of specimens exposed to the fungus, divided by mean property value of specimens in sterile bottles. As with weight loss, specimens of locust heartwood board showed appreciably less loss of mechanical properties as a result of exposure to either decay fungus than did specimens of commercial boards (Table 5 and Fig. 3). Exterior durability The heat-treated (but not oil-tempered) hardboards from whole locust showed poor exterior performance. This poor performance is not due solely to a lack of oil tempering, as commercial boards that received the same heat treatment and no oil tempering performed much better. The locust boards showed much greater thickness swelling at 11, 21, and 25 months of exterior exposure than did the commercial boards. At 21 and 25 months, the swelling of locust boards was severe; the nails appeared sunken below the board surface (Fig. 4). Half of the test specimens were nailed to the test frame with three nails each, such that the specimen could cup, concave side out, along its bottom edge. Specimens of boards from whole locust stems showed considerable cupping at 11, 21, and 25 months of exposure, whereas specimens of commercial boards did not (Fig. 4). We examined the specimens at 25 months of exposure the day after a heavy rain. At that time all locust boards showed a great deal of fiber raising, whereas FOREST PRODUCTS JOURNAL Vol. 35, No. 3 15

6 TABLE 5. -Mechanical property losses from 12 weeks incubation in soil jars. Figure 4. - Appearance of laboratory-manufactured hardboard from whole locust stems (left) and of commercial hardboard (right) after 25 months' weathering at Madison, Wis. Cupping and thickness swelling are evident in the laboratory board. commercial boards did not (Fig. 5). One of the locust board specimens had undergone severe substrate failure by 25 months of exterior exposure. This specimen had a rough surface with soft (delaminated or decayed) areas. Some of these areas were so soft that a finger could be pushed into them, breaking the paint film on the surface of the specimen (Fig. 6). Specimens of locust board showed slight mildew growth at 21 months of exposure, whereas specimens of commercial board did not. We believe that the boards produced from fiber from whole locust stems performed poorly because they were too absorptive of liquid water. It seems likely that greater absorptivity would correlate with greater hygroscopicity (which whole-stem locust boards exhibited in dimensional stability tests). Such behavior is probably caused by inclusion of some bark in these boards. Conclusions 1. R. pseudoacacia heartwood and whole R. pseudoacacia stems were each refined into highquality fiber by pressure refining with what we believe to be a moderate steaming treatment and modest energy consumption for refining. 2. Nonpolar or weakly polar heartwood extractives were steam-extracted from locust heartwood chips during digesting and refining with saturated steam at 302 F (150 C). 3. Fiber from small, whole (nondebarked) R. pseudoacacia stems can be manufactured into wet-process hardboards with reasonably good dry mechanical properties. 4. Storage for 2 years in a climate of 80 F (26.7 C) and 90 percent RH did not result in any consistent loss of mechanical properties in either commercial hardboards or whole-stem locust hardboards. Figure 5. - Appearance of commercial hardboard (left) and laboratory-manufacturedhardboardfrom whole locust stems (right) after 25 months' weathering at Madison, Wis. Fiber raising evident in the laboratory board. Figure 6. - Severe substrate failure of a specimen of laboratory-manufactured hardboard from whole locust stems after 25 months of weathering at Madison, Wis. Note soft area where paint film has been broken by finger pressure. 16 MARCH 1985

7 5. Hardboards from whole locust stems showed greater hygroscopicity than commercial boards or than laboratory boards from locust heartwood. 6. Hardboards from locust heartwood exhibited reasonably good dimensional stability. Hardboards from whole locust stems exhibited a moderate amount of linear expansion in dimensional stability tests, but excessive thickness swelling. This is likely caused by inclusion of some bark in these boards. 7. Despite the loss of some extractives, which occurred in steaming of chips, wet-process hardboard from locust heartwood fiber showed good resistance to weight and strength loss caused by the decay fungi C. versicolor or G. trabeum. 8. Hardboards from whole locust stem fiber showed poor exterior durability. Shortcomings in performance were cupping, thickness swelling, fiber raising, substrate failure, and slight mildew growth. This poor performance may be due to inclusion of (some) bark in these boards, and correlates with the high hygroscopicity exhibited by these boards. In addition, locust sapwood may be very hygroscopic. Literature cited FOREST PRODUCTS JOURNAL Vol 35, No 3 17