Mobile Edition
Print
Get the Mag
Weekly UpdatesConverting wood into carbon has been done for centuries, and the products have been used for fuels, adsorbents, and industrial raw materials. Recent studies (Byrne and Nagle 1997a, 1997b) on the production of crack-free monolithic porous carbon from wood for use as precursors of structural ceramics and composites has demonstrated new uses for wood which has attracted interest from a number of research groups. Carbon produced through the controlled thermal decomposition of wood in an inert atmosphere has been shown to retain the ultrastructural integrity of the parent wood (Blankenhorn et al. 1972). Because of its high reactivity and excellent machinability, wood-derived carbon has been used as a template in the synthesis of high performance biomorphous carbide ceramics (Klingner et al. 2003, Zollfrank and Sieber 2004, Rambo et al. 2005). Carbonized wood also has shown promise as a precursor for the production of carbon/carbon and carbon/polymer composites (Byrne and Nagle 1997c). To date, however, no comprehensive study has been conducted to investigate the mechanical properties of such composites.
Carbonization of natural solid wood has several limitations including: long carbonization processing times for large samples, retention of structural defects (knots), marginal reproducibility, and dimensional limitation of products due to the size of trees and the solid wood cut from those trees. Therefore, to expand the breadth of products that can be produced from carbonized wood, wood-based composites, such as medium density fiberboard (MDF), have been used to make porous carbon (Kercher and Nagle 2002). When MDF is used as the raw material, the carbonization of large samples can be completed within 1 day (Kercher and Nagle 2002) without crack formation, and some of the properties of the final products can be engineered through the control of density and particle size of the MDF panels (Treusch et al. 2004).
The primary objective of this study was to conduct a preliminary investigation to determine the potential for use of carbon preforms made from wood-based materials in the production of high-value composites by evaluating some of the basic physical and mechanical properties of carbonized medium density fiberboard (CMDF) infused with epoxy and phenolic resins.
Materials and methods
Carbonization of MDF
Hardwood MDF with a nominal thickness of 19 mm, purchased from Home Depot (Catonsville, Maryland), was cut into 305 mm long and 203 mm wide pieces. The boards were carbonized in argon with a flow rate of 0.5 L [min.sup.-1] with the entire process conducted in a retort furnace (CM 1200 Rapid-Tem, CM Furnace Inc., Bloomfield, New Jersey). The thermal schedule was modified from the prior work of Nagle and coworkers (Byrne and Nagle 1997a, Kercher and Nagle 2002) and is outlined as follows:
50[degrees]C/h to 110[degrees]C: maintain for 3 hours
15[degrees]C/h to 200[degrees]C
30[degrees]C/h to 400[degrees]C
15[degrees]C/h to 600[degrees]C
50[degrees]C/h to 1000[degrees]C: maintain for 1 hour
300[degrees]C/h to 25[degrees]C
Resin infusion (specimen shaped before resin infusion)
After carbonization, all of the CMDF boards had a uniform thickness of 11 mm. The CMDF boards were machined into test coupons prior to resin infusion, with the length direction of the coupons parallel to the panel formation direction, and thickness being equal to the MDF. For apparent density, 50 by 50 mm square blocks were used. For flexural properties tests, bar-shaped coupons were used (178 mm long and 23 mm wide). For tensile property evaluation, dog-bone shaped specimens were used with an overall length, overall width, length of the narrow section, and width of the narrow section of 165 mm, 22 mm, 57 mm and 13 mm, respectively. The dimensions of the fracture toughness specimens were 102 mm long and 23 mm wide. For each property test and each type of material, five test coupons were prepared, producing a total of 60 test coupons.
Infusion of epoxy resin.--The epoxy resin system, ProSet 125 (resin) and 229 (hardener) by Gougeon Brothers Inc. (Bay City, Michigan), was used in this study. The resin and hardener were mixed at a weight ratio of 10:3 at room temperature according to the manufacturer's manual. Before infusion, the specimens were submerged and weighted down in the resin mixture. A vacuum of -25 in. Hg (gauge) was applied and maintained for 10 minutes before release; this process was repeated an additional time. Specimen surfaces were then cleaned of excess resin by wiping, and the cleaned specimens were stored at room temperature for 15 hours followed by post-curing at 83[degrees]C for 8 hours. After the resins were cured, all of the test coupons were examined using standard radiographs (x-ray imaging at 80 kV) to ensure uniform resin infiltration.
Infusion of phenolic resin.--The phenolic resin, Durite SC 1008 by Hexion Specialty Chemicals (Columbus, Ohio), was selected as it is specially designed for laminate infusion. Although this resin is cured by heating to 100[degrees]C for 30 minutes, slight heating will decrease the viscosity and facilitate infusion. Preliminary studies showed that when the resin was heated to the temperature range of 60[degrees]C to 70[degrees]C, viscosity was reduced for a period long enough to permit complete resin infusion of the samples. The same vacuum procedure described for epoxy resin infusion was used for phenolic resin infusion, except that the process was conducted at 70[degrees]C. After cooling of the submerged samples to room temperature, the samples were cleaned and then stored at 50[degrees]C for 24 hours, followed by curing under the following temperature regime:
150[degrees]C/h to 80[degrees]C
80[degrees]C/h to 94[degrees]C and remain for 20 minutes
150[degrees]C/h to 150[degrees]C and remain for 60 minutes
300[degrees]C/h to room temperature
All of the test coupons were imaged by x-ray after resin curing as specified in the section on epoxy resins.
Mercury porosimetry measurement
Total porosity and pore size distribution of the samples before and after the polymer infusions were measured using a mercury porosimeter (AutoPore IV 9500, Micromeritics Instrument Corporation, Norcross, Georgia). A contact angle of 130[degrees] and a surface tension value of 473 mN/m for mercury were used in the calculation of pore sizes (Blankenhorn et al. 1978). A total of about 2 g particulate samples were collected from different parts of five specimens of each type of composite to obtain an average evaluation of the porosity for that type of material.
Mechanical tests and calculations
A screw-driven ATS universal test machine (Series 910, Applied Test Systems Inc. Butler, Pennsylvania) was used for all of the mechanical tests. Dog-bone tensile tests, the four-point bending test, and the single-edge-notch fracture test were conducted according to ASTM D638-03, ASTM C651-91, and ASTM D5045-99, respectively. A constant crosshead speed of 1.27 mm/min was used in all of the tests. All displacements were measured using a MTS extensometer (Model: 632.118-20, MTS, Eden Prairie, Minnesota), and the data were collected using a HP data acquisition unit (Model: 34970 A, Hewlett-Packard Company, Loveland, Colorado). Sample orientation and load direction in the mechanical tests are as illustrated in Figure I.
During the mechanical tests, all failures took place at the expected locations in the test coupons, and all data from the fracture toughness tests followed the criteria specified in the test standard. Therefore, data from all of the tests were included in the calculation of mechanical properties. Statistic analysis of the data were conducted using Microsoft Office Excel 2003 at a confidence level of [alpha] = 0.05.
Results and discussion
Porosity and pore distribution
The original CMDF had a porosity of 58.9 percent, but after resin infusion, the porosity decreased to 4.8 percent for epoxy-infused samples and to 29.1 percent for the phenolic-infused samples (Fig. 2). Most of the pores in the original CMDF had a diameter smaller than 10 [micro]m. After the carbonized material was infused with epoxy resin, most of the pores were filled with resin and only a small amount of the pores, in the size range between 2.5 [micro]m and 9.0 [micro]m, remained. For the samples infused with phenolic resin, pores smaller than 1.5 [micro]m were filled with resin, but there were a relatively large number of unfilled pores in the size range between 1.5 [micro]m and 9.0 [micro]m (Fig. 2). The high porosity in the phenolic-infused samples is primarily due to high shrinkage of the resin during polymerization and resin bleeding in the curing process which is discussed in the next section.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
Apparent density
Samples infused with epoxy displayed a greater apparent density than those infused with phenolic resin (Table l) with the former having a weight gain of 96.7 percent compared to 66.7 percent for the latter. The epoxy resin used in this study was designed to cure at room temperature, and during the infusion process the epoxy resin began to polymerize shortly after infusion into the porous carbon preforms. The heat released from the polymerization reaction appears to have facilitated the curing process permitting more resin to be retained in the samples. The smaller apparent density associated with the phenolic resin infused samples was primarily associated with bleeding of the resin during the curing process at an elevated temperature. During the infusion process, as the temperature dropped from 70[degrees]C to room temperature, the viscosity of the resin increased and the resin stayed temporarily within the structure of the porous carbon preform. But, because the resin was not cured to the B stage during the 50[degrees]C-24 h storage period, when the temperature increased during the curing cycle, the viscosity of the resin decreased, resulting in resin bleed from the samples. Vaporization of the resin solvent and the water produced from the condensation reaction also could have prompted resin bleeding via the formation of bubbles. It is conceivable that postcuring of the phenolic resin while maintaining air pressure on the samples would have reduced resin bleed to produce a greater density sample.
FEED