Increase in fines content and adhesion behavior in the pneumatic conveying of CB.

This is the second of a two-part series. Part one appeared in the July issue.

Test evaluation and discussion of the results

Increase in fines content

During conveying of beaded carbon black, the fines content is known to depend on the conveying conditions and the distance conveyed. As described at the outset, up to 10 circulations were performed in the conveyor and the fines content was determined after each round. Figure 5 shows the result of the increase in fines content as a function of the conveyed circulations, the type of carbon black, the pipe material and the conveying conditions. For better comparability, figure 5 shows the result up to the fourth circulation, which corresponds to a conveyed distance of about 270 m.

[ILLUSTRATION OMITTED]

Figure 5 shows that substantially higher fines contents are obtained in dilute phase conveying than in dense phase conveying. This effect is known in principle and was described in earlier publications (refs. 1 and 2). In these, the carbon blacks N 330, N 539 and N 660 were analyzed as a function of the conveying conditions and the circulations conveyed. The reason for the high fines content is the higher air velocity during conveying. The carbon black pellets are accelerated to 50-90% of the velocity of the conveying air, depending on the conveying method. If, for example, the air velocity is tripled, the kinetic energy is nine times as high because of the quadratic relation. Particularly in bends, the carbon black pellets may be destroyed.

Furthermore, figure 5 shows, for N 234 and N 650 exemplarily, that slightly higher fine contents are established for an identical conveying state in the rubber pipe. This effect can be explained by the different coefficients of friction between the carbon black pellets and the various pipeline materials. What appears here to be a disadvantage of the rubber pipe is, however, compensated, among other things, by other advantages, as will be shown.

In the state of dilute phase conveying in the rubber pipe, all three carbon blacks were investigated. It is found that N 772 undergoes the greatest destruction and N 650 the smallest. The difference in the fines between these two types after four circulations is more than 20%.

If the dilute phase conveying is compared with the dense phase conveying in the steel pipe in numerical terms, the fines for N 650 is about a factor of ten higher after the fourth circulation in dilute phase conveying. On the other hand, the carbon black N 772 shows completely different behavior even in dense phase conveying with steel pipes. The fines content is higher even in the state of delivery, increases after the first circulation and then appears to decrease again. This curve indicates that the resulting fines adhere in the pipeline.

Overall, figure 5 shows that the increase in fines content is the greatest after the first conveyed circulation (67 m). The gradient then becomes much less steep.

In this context, figure 6 shows the fines plotted against the surface characteristic CTAB after conveying three circulations for carbon blacks for this and a further investigation (ref. 1). The carbon blacks have been investigated under approximately the same conditions in the same pneumatic conveyor. The results were included in figures 6 and 7 in order to permit a better interpretation using a larger database. In the following graphs, a distinction is made only between dense phase conveying, represented by the solid circles, and dilute phase conveying, represented by the open circles.

[ILLUSTRATIONS OMITTED]

Here, the lines added for the two conveying conditions show that the fines increase with the carbon black surface areas. The higher the CTAB adsorption of the carbon blacks used, the higher the fines content is to be in the pneumatic conveying of carbon black.

Using the same nomenclature, figure 7 shows the fines of the carbon blacks represented in figure 6 after three conveyed circulations, in this case plotted against the DBP absorption.

No substantial dependence of the increase in fines content on the DBP absorption is evident from figure 7. The scatter of the values is substantially larger. If a line is fitted to the data points, a nonsignificant increase in the fines content with higher DBP absorption numbers is found only in the case of dilute phase conveying.

Adhesion behavior

The adhesion behavior of carbon blacks on the inner surface of the conveyor pipe is a further undesired side effect in addition to the increase in the fines content. As already presumed in the discussion of figure 5, the resulting fines of N 772 adhere to the pipeline resulting in a decrease instead of increase in the carbon black fines in the course of the further conveying. The test results are presented separately according to conveying condition/pipeline material.

Dense phase, steel pipe

In dense phase conveying using the steel pipe as conveying pipe (test 1), the carbon black N 772 tends to adhere strongly in the pipeline. Figure 8 shows a photograph of the interior of the pipe after five circulations on the left and one after ten circulations on the right.

[ILLUSTRATION OMITTED]

A substantial increase in the layer thickness on the inner surface of the pipe between the fifth and tenth circulations is evident. In the right photograph, the thickness of the carbon black layer is up to 20 mm at some points.

The photographs shown in figure 8 and the curves shown in figure 5 demonstrate that a considerable amount of the resulting fines has adhered to the inner surface of the pipe and can no longer appear as fines in figure 5. In spite of the strong adhesion in figure 8 after 10 conveyed circulations, stable conveying is possible until then. This is because very little or no fine material is present in the flowing air that conveying can be maintained at all over these distances.

In dense phase conveying in the steel pipe (test 2) with the material N 650, substantially less caking is observed than in figure 8 for N 772. Both the increase in the fines content per conveyed circulation and the adhesion behavior are reduced compared with test 1. It is to be expected that N 234 leads to even less adhesion in this conveying condition.

Dilute phase, steel pipe

In dilute phase conveying in a steel pipe (test 3) with the carbon black N 650, even less adhesion is observed, but this adhesion no longer has any substantial influence on the air mass flow.

With the same configuration, the carbon black N 234 (test 4) does not adhere at all to the pipeline, as can be seen in figure 9.

[ILLUSTRATION OMITTED]

Since the carbon black has a strong light-absorbing effect, the interior of the pipe has been illuminated by an electric torch in the left photograph. At some points, the light is reflected by the bare metal of the pipeline. Thus, no significant coating has formed.

For the test with N 772, which was not performed in the test configuration, in comparison with N 650 and N 234, the most pronounced adhesion behavior is to be expected when the observations are based on dense phase conveying with the steel pipe.

Dilute phase conveying with rubber inliner

The carbon blacks N 772 (test 5), N 650 (test 6) and N 234 (test 7) show no significant adhesion at all in the pipeline. As an example, N 650 (test 6) is considered in more detail here. Figure 10 shows photographs of the pipeline after five (left) and ten (right) conveyed circulations.

[ILLUSTRATION OMITTED]

Neither of the two photographs show any extensive adhesion. At some points, small crumb-like structures are visible on the surface.

Owing to the resilient rubber skin in the interior of the pipe, however, adhesions of this type are only of short duration since the resilient inliner can result in a pump effect. The difference between the internal diameter of the steel pipe and the external diameter of the rubber inliner permits these movements. The inliner expands as the pressure in the pipe increases and accordingly the opposite occurs in the case of a pressure drop. As a result of these pulsation movements, any adhering material is detached again. Detachment of material adhering to the inner surface of the pipe by means of the pulsation results in a sort of self-cleaning. In all tests carried out with the rubber inliner as a conveying pipe, this behavior occurred. The advantage of the rubber pipe, mentioned at the outset, is found here and can entirely justify tolerating the slightly higher fines content. This effect is in any case virtually negligible in dense phase conveying with rubber inliner, owing to the low fines level.

Dense phase conveying with rubber inliner

In dense phase conveying with the rubber inliner (test 8), the state of dense phase conveying has been tested for the carbon black N 234. A comparison with the configuration of dilute phase conveying with the rubber inliner (tests 5 to 7) shows that the change in the method of conveying has no effect on the adhesion behavior. The state of the pipe very closely resembles the photographs shown in figure 9.

On the basis of this knowledge, no tests have been carded out with the carbon blacks N 772 and N 650 with this configuration. Here too, absolutely no adhesions were supposed to be expected on the inner surfaces of the pipes.

Discussion of adhesion behavior

The mean aggregate size of the carbon black adhering to the pipe walls was also used for clarifying the different adhesion behaviors. This determination was carded out using a BI-DCP disc-type centrifuge with red light diode from Brookhaven. The aim was to test whether the aggregates, too, are destroyed on destruction of the pellets during the conveying process. Table 3 shows a comparison of the mean aggregate sizes in the state of delivery and after five and ten conveyed circulations. Dense phase conveying in the steel pipe (test 1) served for this purpose for the material N 772 and the state of dilute phase conveying in the steel pipe (tests 3 and 4) for N 650 and N 234.

Table 3 - mean aggregate size in three different
conveying conditions

                          Aggregate
Material                     size      N    N 650   N 234

State of delivery            (nm)     772    216      76
After five circula-          (nm)     202    211      78
tions                        (nm)     195    204      82

The data shown in table 3 indicate that the mean aggregate size of the adhering carbon black remains virtually constant independently of the conveying conditions. The energy input into the material is accordingly not sufficiently high to destroy the aggregates. The increase in the fines content therefore results only from the destruction of the pellets.

Furthermore, table 3 shows that the mean aggregate size of the carbon black N 650 has shifted towards slightly higher values compared with N 772. This can be explained by the higher degree of branching (higher DBP absorption). The primary particle sizes of the two carbon blacks are likely to differ only insignificantly, owing to the very similar CTAB adsorptions (see table 1). This permits the conclusion that the carbon blacks N 650 and N 772 differ substantially only with regard to the degree of branching of their aggregates. It may therefore be stated that a high degree of branching positively counteracts adhesion to the pipe walls during conveying. In addition, it may be concluded that carbon blacks having low structure and a low specific surface area (N 772) have a very strong tendency to cake.

The following interpretation is primarily of a speculative nature and would have to be backed up by a suitable choice of experimental carbon blacks and exact basic conditions (e.g., conveying of only one specific sieve fraction, as well as exactly set amounts of bead assistants). Such a procedure is, however, scarcely possible if it is intended to use conditions close to practice as in this investigation, since the required amounts of carbon black are very large.

As can easily be imagined, the aggregates of a low-structure carbon black, i.e., a carbon black which is composed of aggregates with a low degree of branching (virtually linear structures), can agglomerate more readily and thus have more contact points than a highly structured carbon black with highly branched aggregates. The higher the degree of branching of an aggregate, the fewer the contact points between one aggregate and another. In other words, the number of contact points with a neighboring aggregate, as well as with a pipe wall, is smaller for highly structured carbon blacks than for a low-structure carbon black. The number of contact points, i.e., possible interactions per aggregate, is on the one hand critical with respect to the possibility of pellet formation by a carbon black. On the other hand, the tendency to adhere to a wall is also increased. This is also confirmed in practice: An N 772 very readily undergoes dry beading, whereas this method of beading leads to considerable problems in the case of an N 650. In this case, the fact that them is little structure or a higher level of interaction has a positive effect on the pellet properties. In the case of adhesion to a pipe wall, the opposite behavior is observed. Here, the higher level of interaction of the low-structure carbon blacks with the pipe wall leads to greater caking.

The interaction between the carbon black aggregates or carbon black agglomerates and the pipe wall takes place through Van der Waals forces, since carbon black is known to be nonpolar.

The situation is similar for the bulk density. The more unbranched the mutual arrangement of the carbon black aggregates, the higher the density of the aggregates in a pellet, and thus also the bulk density of the carbon black. If the carbon blacks have a high degree of structure, the branching results in different packing densities in the pellets. The bulk density of carbon blacks with a higher degree of structure is accordingly lower than that of carbon blacks with less structure.

The dependency of the adhesion behavior on the primary particle diameter can be assessed only to a limited extent. N 650 exhibits very little adhesion, i.e., there is scarcely any caking on the inner surface of the pipe, and it is therefore possible reasonably to assess only any adverse effect as a result of a larger surface area. However, the carbon black N 234 with a substantially larger specific area shows absolutely no signs of caking. This indicates that a large surface area likewise appears to have a positive effect on the adhesion behavior.

Shift of the mass distribution curves

The destruction of the carbon black pellets results not only in an increase in the fines content and in the adhesion of material to the inner surfaces of the pipes. But it also results in a shift of the particle size distribution curves, since the fractions of the remaining particle sizes likewise change relative to the population.

Figure 11 shows the cumulative undersize plot for dense phase conveying in a steel pipe with a carbon black N 650 (test 2) after the first four circulations and after the tenth circulation. The plot deviates only slightly from the original curve. The fines, too, remain virtually constant at 6%. In comparison, figure 12 shows the cumulative undersize plot for N 650 in the case of dilute phase conveying (test 3). The fines increase for the same material from 4% in the state of delivery to over 50% after four conveyed circulations.

[GRAPHS OMITTED]

Figures 11 and 12 show the contrast between dense phase and dilute phase conveying in the case of the carbon black N 650 and with the use of the steel pipe as a transport pipe. The test sieving after individual circulations shows no significant changes compared with the state of delivery in the case of dense phase conveying (test 2). In contrast, dilute phase conveying (test 3) shows a considerable increase for the medium and small sieve mesh sizes, the largest jump taking place between the state of delivery and the first circulation.

Figure 12 shows the shift in the cumulative undersize plots towards smaller particles. The plot does not provide any exact information as to whether specific pellet fractions in particular are destroyed. However, the increase in the middle fractions in combination with a decrease in the upper fraction indicates that the larger and harder pellets are at least as much affected by the destruction, as are the smaller and softer ones. The problem of bead destruction in pneumatic conveying thus cannot be solved merely by increasing the mean pellet hardness. If, for example in figure 12, the particle fraction 250-500 [micro] m in dilute phase conveying is considered, this fraction being present in an amount only about 13% by mass in the as-delivered state, this fraction increases considerably after a single conveying circulation, which must originate from the larger particle fractions. The relationship between the particle size fraction of the carbon black and the pellet hardness related to this is discussed for N exemplarily.

A comparison of figures 11 and 12 shows that, in the case of dilute phase conveying, the shift towards smaller pellets is much more pronounced than in the case of dense phase conveying, in which case the effect plays virtually no role. This is confirmed by the lower stress on the material in the last-mentioned method of conveying.

Figure 13 shows exemplarily the pellet hardness on the individual pellet size fractions of the carbon black N 550 which are obtained after sieving, as a function of the mean pellet hardness which is stated, for example, in specifications.

[GRAPH OMITTED]

It is clearly evident that the carbon black pellets of larger diameters are harder than those of smaller diameters. For example, for a mean pellet hardness of 45 g, the individual pellet hardness of the fraction in the diameter range of 1.7-2.0 mm is 60 g, whereas this value is only about 16 g for a pellet in the fraction of diameter 0.7-1.0 mm. It is also evident that the mean pellet hardness stated in specifications is a representative cross-section. The bar graph in figure 13 shows in principle the mass fractions of the individual pellet fractions.

Figure 14 shows the average pellet hardness for N 234 after each circulation, both for dilute phase and dense phase conveying with the robber inliner (tests 7 and 8) and for dilute phase conveying with the steel pipe (test 4). The pellet hardness relates to a mean pellet size range of from 0.7 to 1.0 mm. The pellet hardness has been determined on the basis of this fraction since the standard fraction (1.4-1.7 mm) was not present in a sufficient amount owing to the conveying process. For this reason, the pellet hardnesses shown are also lower (figure 13).

[GRAPH OMITTED]

For the state of dilute phase conveying with the rubber inliner, an increase in pellet hardness from 9 g in the state of delivery to about 15 g after three conveyed circulations is evident. For dilute phase conveying with the steel pipe, an increase in the average pellet hardness from 9.5 to 14.5 g is observed. Compared with conveying in the pipe with rubber inliner, this is a quantitatively similar value, but a spectrum of seven circulations is considered in the case of the steel pipe compared with three in the case of the rubber inliner. This indicates that, when the principle of dilute phase conveying with rubber inliners as conveyor pipes are applied, slightly faster pellet destruction takes place than in steel pipes. In the case of dense phase conveying, the mean pellet hardness increases from 8.4 g after four circulations to only 10.9 g, which may be regarded as negligible. Here too, the gentler treatment of the carbon black pellets during dense phase conveying is once again recognizable.

On the one hand, this indicates that, in the pneumatic conveying of carbon black pellets, the softer ones are destroyed first. On the other hand, another reason for this presumably lies in the destruction of the large and hard pellets, which disintegrate into smaller but still hard fragments and may also be measured in the automatic pellet hardness determination. Both effects result in an increase in the mean pellet hardness on average.

For industrial applications, this knowledge means that, when large fines contents occur during carbon black conveying, it is not sufficient simply to use higher mean pellet hardnesses. The maximum individual pellet hardnesses in the large particle size classes increase considerably and hence very hard fragments also occur in the conveyed material. Particularly in low-viscosity mixtures, high individual pellet hardnesses can lead to dispersion problems.

Summary

In the pneumatic conveying of carbon black, problems of various kinds can occur. On the one hand, the carbon black fines increase to different extents as a function of the conveying system and of the carbon black type, and, on the other hand, certain carbon blacks tend to adhere in the conveying pipe, leading to blockages and hence to considerable cleaning work.

Three carbon black types which differ essentially in the specific surface area, characterized by the CTAB adsorption, and in the aggregate structure, characterized by the DBP absorption, were chosen for the investigations carded out in this work.

It was possible to show that the increase of fines depends substantially on the specific surface area (CTAB adsorption) of the carbon black and only to a nonsignificant extent on the aggregate structure (DBP absorption) in the case of dilute phase conveying.

In dilute phase conveying in the rubber pipe, slightly higher levels of fines occur compared with the steel pipe. The effect is probably due to different coefficients of friction between the carbon black pellets and the different pipe materials. The pellet destruction is furthermore accompanied by a shift in the particle size distribution curve. The particle size spectrum is shifted towards smaller particles. This may affect the mixing process, since the incorporation of smaller particles leads to lower power consumptions in the black incorporation time (BIT). Constantly changing conditions, as may occur, for example, through separation according to particle sizes, are then particularly problematic. Moreover, it was possible to confirm that the above mentioned effects occur primarily in the case of dilute phase conveying, whereas they play virtually no role in dense phase conveying.

If the tendency of the different carbon black types to adhere to the pipeline is considered, the greatest adhesion is found in the case of low-structure carbon blacks having a small specific surface area. It is possible to show that the tendency to adhesion depends essentially on the carbon black structure, which is characterized by the DBP absorption. The danger of pipe blockage decreases with increasing carbon black structure.

In principle, adhesion to conveying pipes occurs only when the pipe is made of steel. When rubber inliners are used, the resilience of the material ensures a pump effect, through which possible accumulations are detached again very rapidly by the action of mechanical force.

Summarizing the results presented in this investigation, it may be said that dense phase systems with flexible pipelines or inliners are to be recommended in the case of universally applicable conveyor systems for all carbon black types. Only then is it reliably possible to avoid adhesions and at the same time to minimize the increase of fines during conveying. It has also been found that the use of carbon black with higher pellet hardnesses does not inevitably eliminate conveying problems. As a result of this, high absolute individual pellet hardnesses may additionally give rise to dispersion problems.

References

(1.) Freitag, K-.H., "Economic aspects in the pneumatic conveying of carbon black," Schuttgut (Bulk material), 1,1995, pp. 115-117.

(2.) Heep D., Winkhardt G. and Nijman G., "Use and optimization of a dense-stream carbon black conveying system in a tire factory," Schuttgut (Bulk material), 1, 1995, pp. 103-114.

(3.) "What is carbon black?" Information publication from Degussa-Huls AG.

(4.) Hauser, G., "Pneumatic conveying in the Department for Mechanical Engineering and Apparatus Technology in Weihenstephan," Schuttgut (Bulk material), 1, 1995, pp. 145-147.

(5.) Nowak, R.H.T., "Quality assurance of bulk materials," Schuttgut (Bulk material), 1, 1995, pp. 135-138.

(6.) Rumpf, H., "Principles and methods of agglomeration," Chemie-Ingenieur-Technik (Chemistry-Engineering-Technology), No. 3, 1958, pp. 144-158.