Segregation data has two uses. Engineers may use segregation data to improve product design by developing a product with the least segregation tendency. They may also intend to customize processing methods to reduce the effect of segregation in their plant packaging process or handling facility. In both cases, the key parameters required for process or product design are the segregation mechanism, segregation pattern, and segregation magnitude.
Segregation takes place through a number of mechanisms. Determining the primary segregation cause and the segregation pattern created through handling is important to avoid de-mixing of the final detergent mixture while handling and packaging. Any difference in property between materials can result in the separation of vital material components. However, segregation issues in typical handling systems are caused by five common reasons, as described below.
At the time of handling, fine particles may sift through a matrix of coarse particles. This mechanism needs the void space between the adjacent particles to be sufficiently big to allow the fine particles to permeate. Usually, this needs a particle size difference of approximately 3:1. Furthermore, inter-particle motion is needed to offer an approach to expose blank void spaces to fine particles.
The fine particles must also be adequately free flowing to avoid arching between adjacent particles and the void spaces must be vacant enough to accept fine particles. Overall, this type of segregation generates a radial pattern as the material develops a pile in the process equipment. The fine particles build up near the pile charge point and reduce in concentration towards the edge of the pile.
Differences in Angle of Repose
Two materials may have different angles of repose. Therefore, as these two materials descend in a pile, they typically develop overlapping piles where the material with the sharpest repose angle builds up near the top of the pile, while the material with the flattest repose angle builds up near to the pile edge.
Usually, there is a distribution of these two materials along the pile’s surface. Repose angle differences of about 2° lead to considerable segregation. Materials of diverse particle sizes can have an adequate difference in repose angles to bring about this kind of segregation.
However, particle size difference is not a requirement for angle of repose segregation and materials of the same size can separate through this mechanism. Furthermore, the process must also produce piles during handling or processing to bring about this type of segregation.
The mixture may include fine particles that are tiny enough to be taken by air currents in the handling system. These fine particles leave the air stream when gas velocities fall below the entrainment velocity. This results in the separation of fine particles and coarse ones in handling systems. The fine particles usually accumulate close to the container walls.
A source of air currents in process equipment is needed for this type of segregation. This source of air can result from free fall of a compressible material. When the falling stream influences the material level, the trapped air is expelled from the interstitial pores and bears the fine particles in the resulting dust cloud. This segregation usually results in a radial pattern during pile formation, but the fine particles are at the bottom of the pile and not the top.
If the mixture is fine enough, then the entrained air in the interstitial voids can lead to fluidization of the material. Since a large particle falls into this fluidized layer, momentum causes the large particles to pass through this fluid layer, causing a top-to-bottom segregation of fine and coarse particles. This mechanism necessitates a source of air and the capability of the bulk material to hold onto entrained air for a long amount of time.
A fluidized layer of material can drop its entrained air if it remains immobile in a container that was just filled. Percolation drives air up through the bulk material. Usually, this process creates fissures in the bulk material where the gas escapes. The local velocity in these fissures is comparatively high and can carry fine particles in the process, leading to top-to-bottom segregation. This gives rise to the size separation of fluidizable material with broad particle size distributions.
It is important to identify the cause of segregation to prevent any processing that will cause the issue. It is also essential to know the pattern of segregation in order to offer a means of re-mixing material if needed. Gaining insights into the segregation mechanism will also help to establish what needs to be done to the material to develop a product that is less likely to segregate.
Segregation Testing Results: Background
Previously, scientists fed bulk materials onto a pile, sectioned a conical pile into annular sections, and then carried out chemical or size analysis on the material obtained in each annular section. This is a lengthy and tiresome process which causes only a few measurements to be collected along the pile. Plant personnel usually employ thieves to sample piles in process equipment.
This technique also produces only a few measurements next to the pile and disturbs the surface, thereby altering the segregation pattern. Other scientists measure the segregation pattern by positioning material in a funnel flow hopper and then discharging the hopper, gathering the output stream, and measuring the quality of the mixture by particle size or chemical composition.
This technique convolutes the segregation with the flow pattern from a specific hopper or bin. Unless the bin design in the tester is exactly similar to the bin design in the process, there is no assurance that the composition during discharge from the tester will be same as that which is found in the original process.
Moreover, the cohesive properties of the bulk material can hinder flow from the segregation tester hopper, although hang-ups may not take place in the process with the same material.
One issue with existing segregation testers is that they do not correlate well with the process. Any two or three materials can segregate if the properties are different and the mixture is exposed to adequate external stimuli. For instance, incorporating sufficient fluidization gas will separate materials of various densities, although such severe conditions may never take place in the real process.
The valid question to be answered is that if the mixture is subjected to a feed behavior the same as that which is present in the process, will it segregate? Therefore, any measurement of segregation tendency must have three main components. Firstly, the feed should be controlled so that it enables the measured segregation to be scalable to process conditions.
Secondly, the pattern of the segregation must be incorporated as part of the measurement in order to estimate the probable concentration leaving the process equipment. Finally, the magnitude of the segregation should be quantified to provide guidance in determining if the mechanism is, in fact, a probable issue in the process. Ideally, the concentrations based on chemical components in the mixture are preferred.
Relation of Test Results to Process
This article describes how to determine the key features that must be controlled or measured to render segregation measurements scalable to process conditions. For instance, consider that material just drops from a mechanical conveyor into the process vessel.
Material entering the process vessel usually falls freely from a certain fall height at a particular process velocity. In this case, note the dynamic effects resulting from particle rebound or by sliding down the pile scale linearly with the geometry of the process equipment (shown in Figure 1).
Figure 1. Typical rebound of free fall particles down the pile surface.
This is not the case for gas carried with the input stream. However, analogous models can be developed for this case and a scale law can be created. For the sake of this article, and as the segregation data is for a simple comparison, assume that there are little to no gas effects at the time of the filling of a typical pharmaceutical process.
The small scale segregation tester can subsequently be operated at a steady drop rate and fall height. The material was deposited in a slice model, forming half a pile, at a rate of about 1 liter per minute. The free fall dimension was configured to be 0.57 times the diameter or width of the receiving bin.
It must be noted that the complete diameter of the receiving bin in the process would correspond to twice the width of the slice mode bin in the tester as the slice model was filled on one side. For a typical bin with a width of 1.5 m, this would correspond to a drop of around 0.85 m.
The pattern of segregation needs the measurement of segregation to be done by analyzing spatial concentration data. This spatial analysis always includes a sample size or viewport related to it. It is essential to decide in advance the volume of material that represents a moderate segregation basis.
The viewport must include adequate particles to be statistically relevant. If that volume holds just one or two particles, then any segregation measurements are bound to fail because the segregation measurement volume cannot represent an average sample. If that sample volume is similar to the bin size, then every segregation measurement will signify the overall average concentration contained in the process vessel. The proper segregation volume must be selected at some point in between.
Of course, these spatial segregation measurements need access to some spatial view of the material after it has segregated because of initial process filling. One way to obtain access to the cross-section of a pile is by filling a slice model with material and observing the segregation pattern through the side of the slice model using optical methods (demonstrated in Figure 2).
Figure 2. Schematic of a segregation tester. Dump material into box and observe the change in color intensity along the pile as measured just below the top surface of the pile (rectangle section). These changes in color intensity are an indication of differences in either chemical composition or in particle size and can be used to estimate the segregation of key components in the system.
The observation of segregation through the sides of a slice model will be biased to the optical pattern that is present at the wall of the segregation tester. Consequently, one must be careful while loading the tester to ensure that a representative sample is visible through the side of the tester. Filling the tester across the width of the slice model will disperse the material to the tester wall, developing a uniform material across the tester.
The distribution of representative material to the tester wall is also facilitated by restricting the thickness of the slice model. However, it must be noted that very thin slice models are subject to banding owing to wall effects that would not exist in broader slice models. For typical materials, this reduces the slice model to the lowest thickness of around 25 mm.
Reflectance spectroscopic techniques can be used to measure slight differences in color. As this article addresses a discrete particle system, multiple measurements are needed to find out the average concentration within a specified viewport.
Segregation measurement also has a scale problem. The size of the selected viewport must be big enough to hold a representative number of particles, but small enough so that differences in local compositions are not lost in the averaging scheme (see Figure 3).
Figure 3. Measurement zone along the pile top surface.
The tester viewport area was defined as 1.27 cm2 and 36 sample readings were averaged within the viewport area. The viewport was moved to gather data at around 30 points along the length of the pile at a point that was approximately 1.27 cm beneath the top surface of the pile as represented in the schematic in Figure 2.
If the spectra of the pure components and the spectra of the mixture in different viewport boxes next to the pile are known, then the pure component spectra can be used to calculate the concentration of the components along the length of the pile.
One disadvantage of this method is that good data needs several spectral measurements obtained at multiple points along the pile. Obtaining this data manually is tiresome and lengthy. Thus, an automatic instrument was built to control the feed, collect the pure component spectra, and measure spatial concentration profiles for spectral measurements (see Figure 4).
Figure 4. SPECTester used in segregation analysis.
Results of SPECTester Analysis
Figures 5, 6, and 7 provide the segregation pattern and data. The concentrations are marked as a function of dimensionless radius. A radius of 1.0 is the bottom of the pile and a radius of 0 is the top of the pile.
This profile demonstrates the reasonable segregation of the API and a considerable segregation of active B and inert C components, as represented by the segregation intensity (shown in Figure 7). In the same figure, inert A and active A components exhibit moderate segregation.
Active A, active B, and inert C components tend to build up at the top of the pile, while the inert A components amass at the bottom of the pile (as seen in Figure 5). API is likely to build up at the top of the pile and near to the bottom of the pile. Measurements denote that active B and inert C components are the worst acting in terms of segregation.
Furthermore, the API shows that adequate segregation tendency is a factual issue in production. This kind of profile signifies angle of repose segregation. However, API may be segregating as a result of two mechanisms. Some of these inert components that represent major segregation do not need to be combined well with the mixture in order to maintain a good product.
However, the API segregation is large enough that dry formation processes will have to be observed carefully to guarantee good (uniform) product formation. This mixture must be kept in a well-designed handling system to ensure that it remains blended following the blending step prior to being tableted.
Figure 5. Radial segregation profile for heart-healthy medication.
Figure 6. Cumulative radial segregation profile for heart-healthy medication.
Figure 7. Segregation intensity for heart-healthy medication.
At times, the segregation trend can be easily viewed if observed from a cumulative concentration viewpoint. The fraction of any component along the pile can be integrated or summed and can be normalized in relation to the actual average concentration of that component along the pile (shown in Figure 6). When there is no segregation, this procedure would represent a straight line passing through the point (0,0) and (1,1) when plotted against the dimensionless radius.
The build up near the top of the pile is denoted by a positive deviation outside of this line. The accumulation towards the bottom of the pile is denoted by a negative deviation out of this line. The accumulation at both bottom and top are represented by an S-shaped curve. The percentage deviation from the mean concentration for any one component is indicated by the magnitude of the deviation from this line.
The segregation is plotted in this way to relate the segregation profile to the deviation from the average or mean concentration. These plots offer a rapid method to see how bad the segregation is from a mean concentration perspective, while providing a feel of the type of segregation taking place.
This information has been sourced, reviewed and adapted from materials provided by Particulate Systems.
For more information on this source, please visit Particulate Systems.