Thick wood particle

It is not evident to us that the behaviour of a thermally thick wood particle, upon entering a hot reactor, is predictable firom standard TGA-studies because of the differences in heating processes. In a TGA-study one typically finds that a heating rate of a few to a few hundred degrees per minute has been applied. This is to be compared with the heating of a particle rapidly moved from 20 C to a reactor of 400 or even up to SOO °C. We have chosen wood, rather than cellulose or lignin, because it is unclear whether wood can be described by simple combination of the pyrolysis behaviour of its components. Antal [16] clearly states that this is not the case. 

A considerable amount of work has been done on the characterization and modelling of the pyrolysis of thermally thick panicles of wood, [13, 17, 25, 28, 29, 32, 34 - 47], These experiments have been conducted on single particles of wood (cylinder, sphere, parallelepiped). However, few experiments have been carried out on a bed of thermally thick particles of wood [48-51], and we have not identified in the literature any work on the modelling of the pyrolysis of a bed of thermally thick particles,... 

Figure 1. Temperature histories at the indicated locations in a 1.5 cm thick wood particle pyrolyzed with a constant heat flux of 1.6E-4 W/m. Initial particle moisture was 110% (dry basis).

Figure 5. Measured weight fraction tar produced from a 1.5 cm thick wood particle as a function of moisture content for 3 different heating rates.

Common grades of laminates tend to be thin materials ranging from 0.5—1.5 mm in thickness, therefore for most appHcations they must be supported. In the manufacture of furniture, cabinetry, and countertops the laminates are bonded to particle board or plywood. Since the laminates consist largely of ceUulosic paper, their dimensional stabiHty is similar to wood, particularly to particle board. 

Waferboard, a more recent wood constmction product, competes more with plywood than particle board. Waferboard and strand board are bonded with soHd, rather than Hquid, phenoHc resins. Both pulverized and spray-dried, rapid-curing resins have been successfully appHed. Wafers are dried, dusted with powdered resin and wax, and formed on a caul plate. A top caul plate is added and the wafers are bonded in a press at ca 180°C for 5—10 min. Physical properties such as flexural strength, modulus, and internal bond are similar to those of a plywood of equivalent thickness. 

After the wood particles are coated with resia, the particles are uniformly distributed iato a board by an air laid process. The art of the process is ia controlling and getting a uniform distribution of the wood particles by blowiag them out onto a collection chain. After forming the board shape it is moved to hot presses where the wood particles are consoHdated and the resia cured. From the hot presses the boards move to trim saws where the boards are cut square to their fiaal size. Ia some cases, the boards are sanded to final thickness and surface smoothness. 

Plastic laminated sheets produced in 1913 led to the formation of the Formica Products Company and the commercial introduction, in 1931, of decorative laminates consisting of a urea—formaldehyde surface on an unrefined (kraft) paper core impregnated with phenoHc resin and compressed and heated between poHshed steel platens (8,10). The decorative surface laminates are usually about 1.6 mm thick and bonded to wood (a natural composite), plywood (another laminate), or particle board (a particulate composite). Since 1937, the surface layer of most decorative laminates has been fabricated with melamine—formaldehyde, which can be prepared with mineral fiUers, thus offering improved heat and moisture resistance and allowing a wide range of decorative effects (10,11). 

As already indicated above, what one may consider a surface depends on the property under consideration. Adhesion is very much an outer atomic layer issue, unless one is dealing with materials like fibreboard in which the polymer resin may also be involved in mechanical anchoring onto the wood particles. Gloss and other optical properties are related to the penetration depth of optical radiation. The latter depends on the optical properties of the material, but in general involves more than a few micrometer thickness and therewith much more than the outer atomic layers only. It is thus the penetration depth of the probing technique that needs to be suitably selected with respect to the surface problem under investigation. Examples selected for various depths (< 10 nm, 10 s of nm, 100 nm, micrometer scale) have been presented in Chapter 10 of the book by Garton on Infrared Spectroscopy of Polymer Blends, Composites and Surfaces... 

Wood particles used for the outer layers were comprised of that fraction of Pallmann milled particles which passed a 10-mesh screen and were retained on a 16-mesh screen with random lengths to 1/2-inch. Wood flakes which remained on a 10-mesh screen and between 0.008 and 0.012-inch thick, with random lengths to 3/4-inch and random width to 1/4-inch were used as core in the three-layer board. 

Wood flakes have been reacted with BO/triethylamine [44]. Flakes modified to 20% WG gave a flakeboard that absorbed 25% less water and had reduced thickness swelling up to 50% as compared to an untreated flakeboard. Similarly, wood particles were treated with PO prior to board manufacture [45]. PO-treated boards showed excellent decay resistance. 

Recently, particleboards have been prepared from mixtures of acetylated and untreated wood chips [55]. Thickness swelling and water absorption after soaking in water for 24 h decrease as the number of acetylated chips increases. The specimens containing 100% of acetylated chips show no sign of decay. Further, particleboards from acetylated wood particles have been reported to have higher mechanical properties than those from PO-treated particles [56]. 

With wood waste, the feedstock characteristics (disposition in bulk, thick particles, low bulk density), lead to a very bulky bed. These conditions favour the rate of the pyrolysis process through a enhanced heat transfer and evacuation of the pyrolysis products. The loss of weight has been followed during the experiments. There are not much differences between the loss of weight for pyrolysis temperatures of 5S0 °C or 750 °C. In the fiiture, it would be interesting to perform such experiments with pyrolysis temperatures from 500 °C to 800 °C, in order to find the best appropriateness between the pyrolysis products quality and their heavy metals contents. 

An experimental facility for the low-temperature (T < 500 C) pyrolysis of CCA treated wood particles (between 2 and 135 mm long, between 2 and 17 mm wide and between 0.5 and 2 mm thick) was built with the aim of maximising the fraction of metals that is contained in the ash upon a maximal burnout of the wood. Two different configurations were tried updraft [15] and downdraft [11, 12, 14] fixed bed. 

Identifying the physical characteristics of wood that have a significant effect on the dynamics of carbonization of beds of thermally thick particles. 

---