5.2. Effects of screen plate design on fractionation

5.2.1. Presentation of results

To test hypothesis 3, the passage ratios of the Bauer-McNett fractions to total pulp were used to explain the fractionation effects of the various screen plate designs. The passage ratio (Eq. 15), which combines the mass and volumetric reject rates, was employed as a quantity for expressing the fractionation effect uniformly, in accordance with hypothesis 4. The normalised freeness (Eq. 24) was used as an alternative to the passage ratio (Eq. 16) to represent the fractionation effect.

5.2.2. Bauer-McNett fractions

The passage ratios of the Bauer-McNett fractions R16 and R30, calculated according to Eq. 16, for the various screen plates in the first screening stage are presented as functions of the passage ratio in Fig. 22, and those for the second stage in Fig. 23. Correspondingly, the passage ratios of the fractions R50 and R200 and P200 are presented in Figs. 24 and 25. Unlike the situation with the other fractions, all the results for R50 are combined into one plot. Some variation appeared in the results due to variations in sampling, laboratory analyses, process conditions, feed pulp properties and the modelling of the fractions, but certain trends were perceivable.

The results clearly indicated that the greater the average fibre length of a fraction was, the lower its passage ratio. Comparison of the passage ratios of the fractions with the bold line which denotes the overall passage ratio of the pulp demonstrated a tendency for P200 and R200 to be enriched in the accept flow and R16 and R30 in the reject flow, while R50 tended not to be fractionated. The passage ratios of the intermediate fractions, R30, R50 and R200, were close to that of the total pulp, while only the coarse (R16) and fine fractions (P200) differed markedly from the overall passage ratio. There seemed to be no substantial differences in the passage ratios of the fractions between the first and second screening stages, even though the proportions of the fractions in the feed by weight were different. The passage ratios of the short fibre fractions, P200, R200 and R50, seemed to be slightly higher in the second stage than in the first, however, whereas those for the long fibre fractions (R30 and R16) were almost at the same in both stages. No great differences in fractionation appeared between the screen plates, operating conditions and screening stages, i.e. in the feed pulp properties, but some diversity did emerge.

At a constant pulp passage ratio, an increase in slot width raised the passage ratio of the R16 and R30 fractions. This was seen clearly in the first stage, where low-contoured screen plates were used. In the second stage, with high-contoured screen plates, the difference appeared to be smaller. There did not seem to be any differences in the passage ratio of the R50 fraction between the slot widths, but the passage ratios of the R200 and P200 fractions were slightly higher with the narrower slots.

Alongside the slot widths, the contour of the screen plate also had an effect on the passage ratio of the fractions. At a constant pulp passage ratio, a low-contoured screen plate produced a lower passage ratio for the R16 and R30 fractions and a higher one for the R200 and P200 fractions than either a medium or high-contoured screen plate. The differences between the latter two screen plates were small, with differences in passage ratio apparent only in the R16 and P200 fractions.

In order to deduce the fractionation efficiency of pressure screening, the fractionation index of Karnis (Eq. 17) for fibre length in the first screening stage (feed freeness 150 ml) is illustrated in Fig. 26. The fractionation index for low and medium-contoured screen plates having two slot widths was evaluated at pulp passage ratios of 0.45 and 0.60 on the basis of Figs. 22 and 24 and Table 5, assuming the volumetric reject rate to be 0.25 in both cases. The fractionation index (i.e. the efficiency of fractionation) increased linearly with increasing fibre length. In addition, the fractionation index increased as the passage ratio of the pulp, the slot width or the contour of the screen plate decreased. The fractionation index for the longest fibre length in Fig. 26 is equal to the Q-index for the R16 fraction (Eq. 20).

The fractionation index for the second screening stage is illustrated in Fig. 27 for identical situations to those in Fig. 26, except that the average feed fractions in question refer to a feed freeness of 340 ml in Table 5. Fractionation efficiency depended on the feed fractions. The efficiency remained somewhat lower than in the first stage when considering the same passage ratio and screen plate design.

5.2.3. Freeness

The results for the normalised accept freeness (Eq. 24) as a measure of fractionation are presented here. The correlation between the normalised accept freeness and the passage ratio in the first screening stage is illustrated in Fig. 28, which is based on freeness data for four screen plate designs and employs the same notations as in the previous section.

Fig. 28 contains quite similar information to that presented in the previous section, 5.2.2. At a constant passage ratio, the normalised accept freeness was highest on average in the case of the widest slot width, and lowest with the lowest contouring of the screen plate. There appeared to be no clear difference between medium and high-contoured screen plates. The normalised accept freeness showed greater variation than the passage ratios of the Bauer-McNett fractions, probably due more to alterations in feed pulp fibre properties than to inaccuracy in the freeness analysis.

As there is an obvious correlation between the two sets of data, the freeness change in screening should be presentable in terms of the alteration in Bauer-McNett fractions, and vice versa. Since the previous sections have shown that the intermediate fractions remained almost unchanged in screening, the alterations in the coarse and fine fractions should be enough to predict the change in freeness. Indeed, the following experimental relation between freeness and Bauer-McNett fractions was found:

Equation 32.

where

CSF_out is freeness in the accept or reject stream (ml),

CSF_F is freeness in the feed stream (ml),

w_c,out is the proportion of the coarse fraction (R16) in the accept or reject stream by weight,

w_c,F is the proportion of the coarse fraction (R16) in the feed stream by weight,

w_f,out is the proportion of the fine fraction (P200) in the accept or reject stream by weight,

w_f,F is the proportion of the fine fraction (P200) in the feed stream by weight.

The predictive power of Eq. 32 was shown to be good, as illustrated in Fig. 29, where the output freeness values for pressure screening are modelled with this equation.