Powder Pointers Winter 2009 Volume 3 No D

Brought to you by:

rd
Material Flow Solutions, Inc.
7010 NW 23 Way, Suite A Gainesville, FL 32653 Ph: 352-303-9123 E-mail: [email protected]

Preventing Segregation by In This Issue

Feature Article:

Proper Product Design Preventing Segregation by Proper

Product Design
Upcoming Events: Dr. Johanson to Teach
1
1
The Segregation Problem. Segregation is responsible for a significant Powder Pointers Preview 2
The SSSpinTester: Coming Soon 4
amount of lost product due to poor quality. Predicting and controlling Regular Feature: Learning the Trade 4
segregation is critical to optimize product design or mitigate quality issues
with bulk powders and granules. Segregation can be caused by several
effects. Fine particles sift down through a matrix of coarse particles. Dr. Kerry Johanson
Differences in repose angle can cause separation during pile formation.
Air currents may carry fine particles, depositing them preferentially in Will Teach at
different parts of the bin. In general, segregation is caused by a difference Upcoming Events
in a particle scale property. Thus, the segregation driving force is some
function of these key flow properties. Cohesion can often be used to April 26-29, 2010
mitigate segregation. However, the effect of cohesion is different World Congress on Powder
depending on the segregation mechanism. Technology – WCPT6
The Key. To solve a segregation problem we need a reliable means of Nuremberg, Germany
measuring segregation tendencies in bulk materials. Engineers use Cutting-Edge Lecture / Discussion
segregation data to optimize product design, creating a product with “Using Segregation Test Data to
minimal segregation, or to modify the process and minimize the effect of Predict Quality of Mixtures in
segregation in their plant packaging process or handling facility. In either Handling Systems”
case, the segregation pattern, segregation mechanism, and magnitude of May 4-6, 2010
segregation are key parameters in process or product design. Process Technology for Industry:
Any measurement of segregation must relate to the process. Segregation International Powder & Bulk
occurs due to differences in key particle scale properties and almost all Solids Conference – PTXi
materials will segregate if exposed to an external stimulus that induces Chicago, Illinois
different behaviors based on a key property. For example, fines can be Thursday, May 6:
carried by air currents. However, if the process has minimal air current Full Day Workshop on
events, then segregation will not be prominent. The real question is: will Segregation and Blending
the mixture segregate when exposed to a feed behavior similar to that (9:00 am-12:00 pm and
present in the process? Therefore, any measurement of segregation 2:00 pm-4:00 pm)
tendency should have three key elements. First, feed must be controlled
to allow the measured segregation to be scalable to process conditions.
June 7-8, 2010
ASME Continuing Education
Second, the segregation pattern must be included as part of the
Houston, Texas
measurement to predict the expected concentration leaving process
equipment. Finally, the magnitude of segregation must be quantified. We * * AND * *
can form a pile at conditions similar to what might exist in the process and October 14-15, 2010
then measure the concentration of key components down the pile using a ASME Continuing Education
slice box similar to that shown in Figure1. The back of this slice model is Chicago, Illinois
made of glass, allowing access to a view of the segregation pattern in the 2-Day Course entitled “Producing
bulk material. Observation view ports are chosen and reflectance Quality Powder Products”
spectroscopic methods employed to measure subtle (Continued on Page 2)
Preventing Segregation by Proper Product Design 2
Continued from page 1
color differences in a user-defined view port. Since segregation measurement is a scale issue, the view port size
must be large enough to contain a representative number of particles, yet small enough that local composition
differences are not lost in the averaging scheme.
The SPECTester Solution. If the
spectra of the pure components are
known, and the spectra of the
mixture in different view port
boxes along the pile are known,
then the pure component spectra
can be used to compute the
concentration of components along
the length of the pile. A new
segregation testing device called
Figure 1. SPECTester used in experiment showing the variable speed vibratory feed SPECTester measures segregation
system and the pile formed in the tester for analysis (note right door is potential of a bulk material using
open to expose the segregation hopper) both visible and NIR reflectance
spectroscopic techniques. Up to
Rt six components can be measured.
Experiments indicate that this
Define the size of the view port and measure the spectroscopic measurement
spectra along the top of the pile. Adjacent viewports
can overlap and tester can measure concentration at
method corresponds very well to
r
up to 50 locations along pile. actual concentration segregation
profiles measured on piles. This
new tool allows the user to
evaluate the effect of changing
particle scale properties on
segregation behavior, making it the
Figure 2. Measurement zone along pile top surface ideal product design tool. We
have applied this technique to
hundreds of materials, including pharmaceuticals, foods, chemical mixtures, and even the same material at different
particle sizes.
The experiment. Consider the effect of cohesion caused by liquid bridge between particles on the segregation of
bulk materials. Three differently colored sands – blue 70%, yellow 15%, and white 15% were – were used to
generate segregation data. Figure 3 shows the averaged spectra of the three sand components used in the mixture.
Note the different peaks at various wavelengths. (Continued on Page 3)

Powder Pointers Preview Future Topics

Coming Next Quarter – PAT Implementation To put you at the cutting-edge

The pharmaceutical industry is requiring that critical product properties be • Successful agglomeration
monitored and key process variables changed in an attempt to create • Process simulation and
quality product all the time. Other industries are implementing this type predicting behavior
of control with looser constraints. The bulk of PAT work has focused on • Milling – new techniques
inline monitoring methods. However, a portion of PAT methodology
relates key process and product variables to process behavior. Robust We encourage and welcome your
processes must be developed with enough variables to adjust process suggestions and special requests
behavior and relate this to sound models that predict process behavior for powder flow topics which you
from knowledge of key material flow properties. Our next Newsletter will would like to see included in future
examine how to do this for typical handling systems and solids processing editions of Powder Pointers.
unit operations. Contact: Susan at 352-379-8879
Preventing Segregation by Proper Product Design 3
Continued from page 2
How it works. These pure component spectra include 600

the effect of surface texture, color and particle size. The 500

mixture spectra also are affected by particle orientation. 400

If the particles were spherical and all the same size, then

Intensity
300
there would be no optical difference between the
particles in the mixture and pure particles placed in the 200

component trays. In this case, the mixture spectral 100

intensity Fmixj(λ) would be a simple linear combination 0

of the spectral intensity of pure components (Fi(λ)) 400 500 600 700 800 900
Wavelength (nm)
based on the local fraction (xi,j) of each component.
White Sand Blue Sand Yellow Sand Mixture
However, smaller particles fill the voids between coarse
particles, creating a shadow effect for the coarse Figure 3. Spectra of three colored sand components and sample
particles. The fine particles within the voids occupy a mixture spectra resulting in a composition of 12.5%
white, 79.3% blue, and 8.2% yellow sand
proportionally greater percent of area than the volume
fraction would suggest. This effect can be modeled by adding a weighting
Fmix j (λ ) = ∑ Wi ⋅ xi , j ⋅ Fi (λ )
factor (Wi) to the linear combination of pure spectra (equation 1). The tester
i compares the computed intensity curve Fmixj(λ) with the actual measured
mixture curve Factj(λ) and adjusts the weighting factors Wi and local
Equation 1
component fractions down the pile xi,j to minimize the error between the two
curves using a least squares approach. The result is a radial concentration profile. Now consider the effect of
adding cohesion to the bulk material by comparing the segregation profiles given in Figure 4. A dimensionless
radius value of 0.0 corresponds to the top of the pile; a value of 1.0 corresponds to the pile bottom. Consider the
concentration profiles for the sand mixture without oil. The blue sand accumulates at the bottom of the pile while
the yellow and white sands accumulate at the top of the pile. When oil is added to the mixture then the cohesive
forces between the particles prevent segregation resulting in a nearly uniform concentration as a function of radial
distance. The standard deviation relative to the mean of each component concentration, or segregation intensity
(SIi), is used to quantify the overall segregation potential of the mixtures (equation 2).

Segregation of Sand Mixture 0.00% Oil Segregation of Sand Mixture 1.00% Oil

0.9 0.9

0.8 0.8

0.7 0.7

0.6 0.6
Concentration

White White
Concentration

0.5 0.5

0.4 0.4
Blue Blue
0.3 0.3

0.2 0.2

0.1 Yellow 0.1 Yellow

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Dimensionless Radius (r/Rt) Dimensionless Radius (r/Rt)

Figure 4. Segregation Test results for sand mixture 0% and 1.00 % oil

Mitigating segregation through product design. Segregation intensity varies 2

between 0 for uniform mixtures to 1.0 for completely segregated mixtures. Thus, a  xi , j 
simple change in product properties can prevent segregation from occurring. ∑j 1 − xavg 
SI i =  i 
Likewise, one could have changed the particle size or shape of key components and
n −1
mitigated segregation. The trick, however, is having the ability to easily measure
segregation potential quickly and accurately. The new SPECTester allows both Equation 2
accuracy and speed, making product design a very simple task. Since the tester
requires only 10 to 15 minutes to run, many potential mixtures could be evaluated (Continued on Page 4)
Preventing Segregation by Proper Product Design 4
Continued from page 3
in just one day’s time, providing the necessary data to The SSSpinTester (patent pending) Will
truly optimize product design. Figure 5 shows the
reduction in segregation intensity as the oil content is Revolutionize the Pharmaceutical
increased, demonstrating how a small change in and Chemical Industries
formulation can mitigate segregation.
Spotlighted in the March issue of Tablets & Capsules
0.7
magazine. Planned release for this revolutionary tester
0.6 is Spring of 2010.
0.5 In its final stages of development, the SSSpinTester
Segregation Intensity

0.4
measures the strength of fine powders using a sample
as small as 0.05 gram. Current methods of measuring
0.3
the unconfined yield strength of a powdered material
0.2 require at least one liter of sample – usually hard to
0.1
come by in the pharmaceutical and chemical industries.
0
Able to quantify the strength of fine powders in as little
0 0.2 0.4 0.6 0.8 1 1.2 as 15 minutes, this novel tester takes its user to the
Miosture Content (%)
cutting-edge of productivity. Its 14x16 inch footprint
Blue Sand Yellow Sand White Sand makes it easy to accommodate in any testing
laboratory.
Figure 5. Segregation Intensity Index as a function of oil
content for a 3-color sand mixture The SSSpinTester will arrive at your facility, complete
with bonus testing cells, a preprogrammed laptop, and
Consider using this technique to help solve your next instruction manual with demonstration video on CD.
segregation problem, or to characterize your current
materials. For more information: Contact Kerry at 352-303-9123

Important Announcement: Growth Results in a Move

Material Flow Solutions is pleased to announce that we have outgrown our previous facility and have moved across
town to a new location. Our new contact information is:
7010 NW 23rd Way, Suite A Gainesville, FL 32653 Ph: 352-303-9123 Fax: 352-379-8878

Learning the Trade

Knowing and understanding key material properties is power to characterize bulk material flow behavior. We will
empower you quarterly as we discuss one of these fundamental flow properties and its industrial application.
Surface Tension. Surface characteristics of particles dictate the magnitude of
inter-particle forces between adjacent particles. These adhesive bonds between
particles result in bulk cohesion. Liquid bonds between particles can cause
significant forces to hold particles together, resulting in a material that will arch
over outlets and form stable ratholes in process equipment. We have developed
models relating inter-particle forces to cohesive behavior. These can be used to
estimate the cohesion in a process during the product design phase of the project.
Thus, the design engineer provides a shorter path to market as well as a more robust process design with limited
bulk property information. Practical application of surface tension data include, but are not limited to:
• Segregation prevention • Process design during formulation development
• Hang-up prevention • Process control
• Risk analysis • Flow rate prediction
• Blending analysis • Drying efficiencies