Лако-красочные материалы — производство

A. Adhesive Consumption and Glue Spread in the Production of Particleboards

Several aspects regarding the proportion of adhesive to be used in the production of particleboards must be evaluated to obtain good results:

proportion of adhesive on individual particles proportion of adhesive in particle mixtures and fractions proportion of adhesive in the total particle mix

distribution of the adhesive on the surface of the particles, and proportion of the particles’ surface area covered with adhesive.

The resin load content on wood as a measure of the consumption of adhesive is one of the more important parameters to consider during the production of particleboards. From a technological standpoint a certain minimum amount of resin is necessary to obtain the desired properties of the boards resulting in sufficient bonding of the individual par­ticles. However, an excessive resin load imparts some technological disadvantages, such as high moisture content and hence possible problems with high vapor pressure during hot pressing. Furthermore, for economical reasons, the consumption of adhesive should be as low as possible as the resin contributes considerably to the costs of the finished boards. The resin load, however, is only an overall average on the total mixture of particles, without considering differences in particle size distribution and the shape of the individual particles. Moreover, the resin load gives no direct indication of the area-specific consump­tion of the adhesive, which is the amount of resin solids content based on the surface area of the particles. The expression ‘‘resin-robbing by the fines’’ is well known and describes the exceedingly high consumption of adhesive based on mass of particles owing to the great surface area of the fine particles [429,430].

The resin load on wood chips can be described in the following two ways:

mass resin load (percent or grams of resin solids content per 100 g dry particles) and surface-specific resin load (grams of resin solids content per square meter of surface area).

If one of these two terms is known, the other can be calculated assuming a uniform distribution of the resin on the particle surfaces and estimating the total surface area of the particles.

In the production of particleboards mixtures of particles are always used as raw material, and thus the particles differ in size and shape. A size grading of the particles can be performed by sieving, where two of the three dimensions of the particle must be smaller than the standard measure of the actual sieve mesh to be passed. An exact sieving of the particles according to their size, therefore, is only possible for particles of rather similar shapes. Particles can differ widely in shape. A simplification to describe their shape is to assume that they are squared, flat with length l, width b, and thickness d for medium and coarse particles and rather cubic for the fines. Since the sieve mesh is usually graduated according to a logarithmic scale, for the theoretical calculations of the particle size

logarithm of particle length (mm)

Figure 9 Example of a particle size distribution, the calculated mass resin load (gluing factor), and the distribution of the resin solids content. The overall adhesive resin consumption was assumed to be 8% resin solids content/dry wood. (After ref. 429.) distribution this was also assumed to be logarithmic and similar to a gaussian distribution. Distributions on an industrial scale might differ from this model.

Each particle fraction has a certain relation to its resin load according to the size of the particles. Because of the great surface area of the fine particles their resin load increases strongly (linearly with the term d_1). Even if there is only a small proportion of a mass fraction of very fine particles in the mixture, the high consumption of resin solids content of this fraction has a negative impact on the resin load of the coarse particles. Figure 9 shows an example of a particle size distribution with the calculated mass resin loads and the distribution of the resin solids content on the different fractions of the particle size distribution. Particle length was assumed to vary from 25 mm for the coarsest particles to 0.6 mm for wood dust, according to experience with industrial particle mixtures.

Because of the reasons discussed above, usually core layers and face layers are glued separately. In the core layer rather coarse particles predominate and in the face layer rather fine particles predominate. This separate gluing enables the use of different com­positions of the glue mixes (e. g., different addition of water and hardener) and different resin loads (gluing factors) for the two layer types. An example of separate gluing is shown in Fig. 10, with separate gluing of the core layer CL (6.5% mass gluing factor) and in the face layer FL (11.0% mass gluing factor). The mass ratio of the layers CL:FL is 60:40. Figure 10 shows the particle size distributions and the mass gluing factors of the individual particle size fractions for this example of separate gluing.

Samples of industrial core layer and face layer particles, before and after gluing, can be fractionated by sieving, and thus sampling has to be done at the same time before and after blending. In the case of aminoplastic adhesives each particle fraction, glued or not, can be investigated for its nitrogen content. By knowing (i) the content of nitrogen as well as the resin solids content in the glue mix and (ii) the moisture content of the particles glued or not in the various fractions, the mass gluing factor of each glued particle size fraction can be calculated. Figure 11 shows the results of one such calculation for glued core layer and face layer particles. Even if the absolute values may differ from the calcu­lated ones, the resin load (by weight) and the particle size show that the same shape of distribution curve is obtained.

Assuming that the gluing of particles of different sizes is performed randomly with their surface area as the decisive parameter, for various homogeneous particle size frac­tions and for different particle size mixtures the theoretical mass gluing factors and the distribution of the resin solids content can be calculated and correlated with the same values obtained experimentally, by analysis. There are some indications [431-433], how­ever, that glue distribution is not exclusively influenced by the surface area of the particles, but has a certain preference for coarser particles. This may be due to the effectiveness of the adhesive application, thus to the separation and distribution of resin droplets, or to the mixing action in the blender after application of the resin on the wood particles (wiping effect). The concept that the particle surface area exclusively influences gluing is quite clearly invalid, if glue droplets and the surface to be glued have similar size. Meinecke and Klauditz [431] mentioned diameters of glue droplets of 8 to 110 pm, depending on the type of spraying and Lehmann [434] mentioned up to 200 pm. The latter values are of the same order of magnitude as the size of the finest particles used for the calculations above.

Besides the surface area of the particles several other parameters also have some influence on the necessary resin consumption, e. g., type of boards, thickness of the sanding zone, type and capacity of the blenders, separation and spraying of the resin (depending on if only the wiping/spreading effect occurs during blending or if instead spraying of the resin is used), shape of the particles for the same particle sizes, dependence of the slender­ness ratio on particle length, concentration and viscosity of the glue resin, or a partial size degradation of the coarser particles in the blender.

New strategies in blending take into account the reality of the higher resin con­sumption by the finer particles, e. g., by removing the dust and the finest particles from the particle mix before blending. Also an exact screening and classifying of the particles before blending can improve the distribution of the resin on the particle surfaces and can help to spare some resin. A lower consumption of resin not only means lower costs for the raw materials, but also helps to avoid various technological disadvantages. With the resin, water is also applied to the particles; as long as this amount of water is low enough, especially in the core layer, no problem should occur with a too high vapor pressure during hot pressing. Often, however, the moisture content of the glued core particles is too high, due to an excessive gluing factor. The high vapor pressure in the board at the end of the press cycle tends to expand the fresh board; if venting is not done very carefully, blistering of the boards at the end of the continuous press or after the opening of the press might occur. Additionally, the heat transfer by the steam shock can be delayed if the vapor pressure difference betweeen the face layer and the core layer is small. If the moisture content of the glued core layer particles is high, the moisture in the glued face layer particles must be reduced. Also spraying water onto the belt before the forming station and onto the surface of the formed mat cannot be done due to the problems with the too high moisture content in the mat and hence with the too high vapor pressure.

Gluing of particles is usually done in quickly rotating blenders by spraying the resin mix into the blender. Due to the rotation of the blender a partial degradation in the size of the particles can occur. While blending OSB strands this degradation must be avoided; this is done by using slowly rotating big blender drums with a diameter of approximately 3 m. The liquid adhesive is distributed by several atomizers in this blender drum.

Gluing of fibers in MDF production is usually done in the so-called blowline between the refiner and the dryer. The advantage of this method is that it avoids resin spots at the surface of the board. The disadvantage, however, is the fact that the resin passes the dryer and can suffer part precuring. This causes some loss of usable resin (approximately 0.5 to 2% in absolute figures); therefore the glue consumption in blowline blending is higher than in the mechanical blending. Due to this fact mechanical blenders have lately been installed again in a few factories. The theory of turbulent flow blowline­gluing is not yet clearly defined [435,436]. However, some equations attempting to describe it have been recently presented [436].